Localizing Websites: Shifting Focus onto the End-User

ŞcoalaăDoctoral ădeăStudiiăFilologice Facultatea de Litere Universitateaă„AlexandruăIoanăCuza”ădinăIaşi TEZ ăDEăDOCTORAT Localizing Websites: Shifting Focus onto the End-User Doctorand: asist. univ. Lakó Cristian Coordonator doctorat: Prof. univ. dr. Rodica Dimitriu Contents List of tables ............................................................................................................................... 6 List of Figures ............................................................................................................................ 6 List of abbreviations and acronyms ........................................................................................... 8 1. Introduction ........................................................................................................................ 9 1.1 2. Justification of the thesis ........................................................................................... 10 1.1.1 Why website localization? ................................................................................. 10 1.1.2 The localization industry in Romania ................................................................ 11 1.1.3 Teaching localization in Romania ..................................................................... 12 1.2 Main objectives ......................................................................................................... 14 1.3 Research material and method .................................................................................. 17 1.4 Structure and content ................................................................................................. 19 Key concepts in localizing websites ................................................................................. 21 2.1 Introduction ............................................................................................................... 21 2.2 Globalization (g11n) ................................................................................................. 23 2.3 Internationalization (I18n)......................................................................................... 27 2.3.1 Internationalization and translation strategies ................................................... 29 2.3.2 Internationalization and text length ................................................................... 34 2.4 Localization (L10n) ................................................................................................... 37 2.4.1 Prescriptive versus descriptive website localization .......................................... 40 2.4.2 Economic factors involved in the localization process ...................................... 42 2.4.3 The Media and the Internet in the localization mix ........................................... 43 2.4.4 Norms and conventions in the localization process ........................................... 44 2.4.5 Personalization ................................................................................................... 48 2.4.6 Culturability ....................................................................................................... 49 2.4.7 Grice’săMaximsăappliedătoălocalization ............................................................. 54 2.4.8 Content marketing – delimitations in the sphere of web localization ................ 60 2.4.9 Classification of localized websites ................................................................... 65 2.4.9.1 Classification by language ............................................................................. 65 2.4.9.2 Classification by region .................................................................................. 66 2.4.9.3 Classification by recipient profile .................................................................. 67 2.4.9.4 Classification by direction .............................................................................. 68 2.4.9.5 Classification by level (degree) ...................................................................... 69 2 2.4.10 Localization and diacritics usage: a case study.................................................. 71 2.4.11 (Quasi-)Automatic webpage localization software ............................................ 77 2.5 2.5.1 Translation, adaptation, copywriting ................................................................. 82 2.5.2 Translation and business discourse .................................................................... 87 2.5.3 Translation theories applicable to website localization ..................................... 91 2.5.4 Machine translation for website localization ..................................................... 95 2.5.5 Translation memory and machine translation .................................................... 98 2.5.6 Translation through parallel corpora and the problem of originality ............... 101 2.5.7 Transcreation in localization. ........................................................................... 102 2.5.8 Reader/user-oriented translations and search engines ..................................... 103 2.5.9 Semasiology and onomasiology ...................................................................... 104 2.5.10 Glocalization or "looking in both directions" .................................................. 109 2.5.11 General considerations on website translation ................................................. 116 2.6 Website localization and search engine optimization ............................................. 117 2.6.1 Search engines ................................................................................................. 117 2.6.2 Search engine optimization (SEO)................................................................... 119 2.6.3 Invisibility of the translator: when authorship matters .................................... 120 2.6.4 Web content translation and search engine optimization (SEO) ..................... 122 2.6.5 SEO website localization ................................................................................. 127 2.6.6 Keyword research and website localization ..................................................... 130 2.6.7 Long-tail keywords .......................................................................................... 132 2.6.8 Tools for keyword research ............................................................................. 133 2.7 3. Translation ................................................................................................................. 80 Conclusions ............................................................................................................. 135 Specific issues in translating electronic text ................................................................... 137 3.1 Introduction ............................................................................................................. 137 3.2 Text and E-text ........................................................................................................ 138 3.3 Classification of texts .............................................................................................. 143 3.4 The layers of the text ............................................................................................... 145 3.4.1 The textual layer .............................................................................................. 145 3.4.2 Contextual layer ............................................................................................... 148 3.4.2.1 Co-text .......................................................................................................... 148 3.4.2.2 Context ......................................................................................................... 148 3 3.4.2.3 Register......................................................................................................... 150 3.4.2.4 Genre ............................................................................................................ 150 3.4.2.5 Style .............................................................................................................. 152 3.4.3 The cultural layer ............................................................................................. 157 3.4.3.1 Cultural management ................................................................................... 160 3.4.3.2 Romania’săculturalăprofile ............................................................................ 163 3.4.4 Temporal layer ................................................................................................. 166 3.4.5 Intentionality layer ........................................................................................... 166 3.4.6 Intertextuality layer .......................................................................................... 167 3.4.7 Anticipation layer............................................................................................. 168 3.4.8 Call-to-action (CTA) and persuasive words in SERPs .................................... 172 3.4.9 The text as a communication process .............................................................. 187 3.4.10 Presupositions .................................................................................................. 188 3.4.11 Implicature ....................................................................................................... 189 3.5 3.5.1 The verbal component...................................................................................... 192 3.5.2 The non-verbal component (non-text) ............................................................. 195 3.5.3 The functional component ............................................................................... 199 3.5.4 The technical component ................................................................................. 200 3.6 The elements of a website and the localization process (a technical view) ............ 201 3.6.1 Front-end elements........................................................................................... 202 3.6.2 Back-end elements ........................................................................................... 203 3.6.3 The human factor ............................................................................................. 205 3.7 4. Text and non-text on commercial websites. ............................................................ 190 Conclusion............................................................................................................... 208 Website localization analysis and case studies ............................................................... 210 4.1 Introduction ............................................................................................................. 210 4.2 Benchmarking website localization, from an SEO perspective .............................. 211 4.3 Is it GILT all the way, when it comes to website localization? .............................. 218 – A case study .................................................................................................................... 218 4.4 Which Way Website Localization: Translation or Copywriting? ........................... 229 4.5 Booking.com: a comparative study from a website localization perspective ......... 237 4.6 Website localization for the Romanian market – a cultural approach .................... 247 4.7 Reverse localization: Romanian companies targeting foreign markets. ................. 252 4 4.8 5. Conclusions ............................................................................................................. 257 Final conclusions ............................................................................................................ 259 Bibliography .......................................................................................................................... 266 Appendix A ............................................................................................................................ 280 Appendix B ............................................................................................................................ 281 Appendix C ............................................................................................................................ 282 Appendix D ............................................................................................................................ 283 Appendix E ............................................................................................................................ 294 Appendix F............................................................................................................................. 305 Appendix G ............................................................................................................................ 316 5 Listăofătables Table 1: Analysis of article length in SERPs ........................................................................... 55 Table 2: Translation vs. copywriting ....................................................................................... 86 Table 3: GILT vs. glocalization ............................................................................................. 115 Table 4: Traditional versus electronic text features ............................................................... 142 ListăofăFigures Figure 1: Standard US and German keyboards........................................................................ 25 Figure 2: Chinese keyboard ..................................................................................................... 26 Figure 3: Google using prescriptive approach for search suggestions .................................... 41 Figure 4:Showing tooltip ......................................................................................................... 50 Figure 5:Office Word 2007 vs. Office Word 2003 .................................................................. 51 Figure 6: Example of directed-user approach .......................................................................... 51 Figure 7: Personalization as a custom type of localization ...................................................... 52 Figure 8: Stratifications of GILT processes ............................................................................. 53 Figure 9: Irrelevant sponsored results ...................................................................................... 59 Figure 10: Facebook inline advertising vs. right hand intrusive rotating banner .................... 62 Figure 11: The most popular channels ..................................................................................... 63 Figure 12: Channels of marketing. Source MarketingSherpa [36] .......................................... 63 Figure 13: Language options ................................................................................................... 66 Figure 14: Absence of diacritics in Google suggestion for Romanian .................................... 73 Figure 15: Google suggesting words with diacritics for German ............................................ 73 Figure 16: Prestashop's advanced localization package .......................................................... 78 Figureă17:ăMassion’sădiagram ................................................................................................. 80 Figure 18: Code free translation tool with preview feature ..................................................... 81 Figure 19: Adaptation methods related to localization ............................................................ 85 Figure 20: Services required by small companies. Source: www.elance.com ........................ 94 Figure 21: Website contextual menu for translating pages in Chrome and Internet Explorer 96 Figure 22: Statistical machine translation determined translation unit (TU) ......................... 100 Figure 23:Term usage adapted to target market - autumn ..................................................... 106 Figure 24: Term usage adapted to target market - fall ........................................................... 106 Figure 25: The term Voucher vs. cupon and tichet in Google Trends ................................... 108 Figure 26: Glocalization versus GILT ................................................................................... 114 Figure 27: Click-through rate based on position in SERP ..................................................... 123 Figure 28: Communication in the keyword based transation process ................................... 125 Figure 29: Google result page based on the IP of the device................................................. 129 Figure 30: Keyword length and how it determines CTR ....................................................... 132 Figureă31:ăMunday’săfunctionalăcharacteristics after Reiss ................................................... 144 Figure 32: Audiomedial supplement in operative texts ......................................................... 145 Figure 33: Authority-like figure............................................................................................. 157 Figure 34: Three Levels of Uniqueness in Mental Programming (Hofstede et al. 2010:6)... 158 6 Figure 35: Manifestation of culture (Hofstede et al. 2010: 8) ............................................... 159 Figureă36:ăHofstedețsă6ăculturalădimensionsăretrievedăforăHall’săHC-LC countries ............. 164 Figure 37: Involved production at communicative level ....................................................... 167 Figure 38: Google Trends chart for car insurance related keywords ..................................... 184 Figure 39: Google Trends chart showing the dominant compare car insurance keyword ..... 184 Figure 40: SERP result........................................................................................................... 189 Figureă41:ăExampleăofă"Ourăclients”ăwebpage ...................................................................... 191 Figure 42: Santini's genres(2010b:97) ................................................................................... 193 Figure 43: Intratextual and extratextual factors ..................................................................... 195 Figure 44: The number of Internet users by language ........................................................... 214 Figure 45: Internet content by languages[45] ........................................................................ 214 Figure 46: Cookie notification ............................................................................................... 219 Figure 47: Eurosport in localized instances ........................................................................... 231 Figure 48: Languages and countries in which localized Yahoo websites are available[95] . 233 Figure 49: Pay upon receipt ................................................................................................... 248 Figure 50: Formal vs. informal inconsitency ......................................................................... 250 Figure 51: Google interface translation error......................................................................... 251 7 Listăofăabbreviationsăandăacronyms B2B – business-to-business PDI – powerădistanceăindexă(Hofstede’să B2C – business-to-customer cultural dimension) C2B – customer-to-business ROI – return on investment C2C – customer-to-customer QA – quality assurance CAT – computer aided translation SEO – search engine optimization/search CMS – content management system engine optimizer CTA – Call-to-action SERP – search engine results page CTR – click-through rate ST – source text e-commerce – electronic commerce TLD – top level domain (.ro, .de, etc.) e-text – electronic text TS – translation studies GILT – acronym for globalization, TT – target text internationalization, localization, TM – translation memory translation TU – translation unit (see also UT) G11n – globalization UAI – uncertainty avoidance index HC – high context (Hofstede’săculturalădimension) I18n – internationalization UI – user interface IP – Internet protocol address URL – uniform resource locator (web L10n – localization address) LC – low context UT – unit of translation MOC – massive online collaboration UX – user experience MT – machine translation Notation conventions bold – used for key concepts italics – for exemplifications underlined italics – for highlighting chapter, subchapter or subdivision references 8 1. Introduction A person-centered approach to any text, language, or culture will always be more productive and effective than a focus on abstract linguistic structures or cultural conventions. DouglasăRobinsonă(2007:112) Interhuman communication under the influence of highly technological communication channels opens up for any company, small or big, the opportunity to communicate globally, namely through online presence . Consequently, nowadays, millions of offers abound on the Internet. Another important factor that has determined a plethora of webpages is that the cost for setting up a website is very affordable, online advertising is significantly more affordable thană traditională advertising,ă andă becomingă aă publisheră isă ată everyone’să fingertip.ă Thisă translates into billions of webpages worldwide. According to Google there were 56 billion webpages indexed in 2013. However, a significant part of the information found on the Internetăisăconsideredă“noise”,ăas the information provided is inaccurate, misleading, useless or too general. Such information generates faulty communication. This leads to frustrated web-users and companies. In the era of e-commerce, businesses survive if they make sales, i.e. if they can rise above “theă noise”. If brand and notoriety are of some importance in the buying decision-making, there have been numerous new companies that surmounted the inherent difficulties of start-up companies. Whereas a company may create the ultimate product, it will not survive if it is not duly promoted. In the age of globalization and democratization of the Internet, companies and their products or services should be considered from a global perspective. Offers should be made available to a vast population of the globe, by applying strategies specific to the electronic communication channels. For this reason, website localization should be part of the global expansion strategy of any company. 9 1.1 Justificationăofătheăthesis Communicating ideas, concepts and thoughts from a source language into one or several target languages involves translation. However, websites are a complex mix of multimodal communication requiring apart from translation, several other combined processes supporting the idea of making a product or service be perceived as if it had been developed and produced within the target culture. 1.1.1 Why website localization? Ină recentă years,ă companyă webă presenceă hasă convertedă intoă growthă ofă marketă shareă andă extensionă ofă collaterală services.ă Contraryă toă theă economică trendsă ină theă lastă fewă years,ă companiesă thată haveă extendedă theiră activitiesă beyondă theă traditională meansă ofă businessă toă consumerăcommunication,ăclaimăthatătheyăhaveăsurvivedădespiteărecentăeconomicăcrisesăandă evenăturnedăaăprofit.ăThisăunderlinesătheăimportanceăofăofferingăeasilyăaccessibleăinformationă toă theă existingă andă potentială clients.ă Providingă theă informationă ină ană appropriateă manneră isă similarătoăconveyingătheămessageă“weăareătheăperfectămatch”. Fromă theă perspectiveă ofă companies,ă makingă availableă variousă productsă toă broaderă marketsă translatesă intoă increasingă theă possibilityă ofă generatingă moreă profit.ă However,ă accordingă toă severalăstudies,ăpotentialăbuyersărequestătoăreachăandăreadăinformationăinătheirăownălanguage.ă 75șăofătheăonlineăusersă chooseătoădoăsoăwheneverăpossibleăandă56.2șăareăwillingătoăpayăaă higherăpriceăforătheăsameăproductăorăserviceăifăinformedăinătheirăownălanguageă(DePalmaăetă al.ă2006).ăEnglishămayăbeătheăfirstălanguageăforăaăwebsiteăthatătargetsăglobalăconsumersăbutăită isă insufficientă accordingă toă theă sameă article.ăApparently,ă aă companyă mayă reachă 80șă ofă theă potentială globală onlineă buyersă byă localizingă theiră websiteă ină onlyă 12ă languages,ă whileă foră reachingă 90șă 23ă languagesă areă required,ă accordingă toă anotheră commonsenseadvisory.comă reportăfromă2012.ă[1] Fromă theă perspectiveă ofă localizationă serviceă providers,ă theă localizationă industryă isă rapidlyă extending,ăfollowingătheătrendsăandăneedsăofătheăglobalăeconomy.ăWithătheă“democratization”ă ofă theă Internet,ă ană ever-increasingă numberă ofă usersă isă grantedă accessă toă onlineă information.ă Throughătheăinformationăprovidedăcompaniesăcanăpersuadeăaăbuyerătoămakeăaăpurchase,ăbută asă seenă previouslyă mostă usersă preferă toă readă theă informationă ină theiră ownălanguage.ăThisă isă 10 whereă translatorsă withă localizationă skillsă cană makeă aă difference.ă Accordingă toă commonsenseadvisory.comă(KellyăandăDePalmaă2012:6)ătheălanguageăservicesămarketăgrewă ină2012ăbyă12.17ș.ăInăabsoluteăfigures,ăUSȘ33.523ăbillionăwereăspentăină2012ăforălanguageă andă localizationă services.ă Theă sameă articleă (Kellyă andă DePalmaă 2012:7)ă showsă foră Easternă Europeăaămarketăshareăofă4.39ș. Thereă areă severală professională organizationsă thată supportă theă localizationă industry:ă Localizationă Industryă Standardsă Associationă (LISAă -ă www.lisa.unige.ch,ă discontinued),ă Localizationă Worldă (www.localizationworld.com),ă theă Localizationă Instituteă (www.localizationinstitute.com),ă theă Globalizationă andă Localizationă Associationă (GALAă -ă www.gala-global.org)ă andă theă Localisationă Researchă Centreă (LRCă -ă www.localisation.ie).ă Localizationă Worldă (www.localizationworld.com)ă isă aă conferenceă andă networkingă organization.ă Currently,ă numerousă LSPă (Languageă Serviceă Providers)ă offeră localizationă servicesăasăwell. 1.1.2 The localization industry in Romania Although a very high percentage of Romanian companies export their products and services to the EU and other countries around the world, with regard to offering localized information on them for each of the specific markets, their skills and knowledge show deficiency. Foreign companies also prove to be deficient in localization competences when analyzing their products localized for the Romanian market (For hundreds of products on the Romanian market product description is faulty or not adapted to the local rules and laws.); hence, the desirability and necessity to teach further skills to translation trainees. In order to be successful in localization specific tasks, apart from language knowledge students should also be trained in cultural communication, international commercial law, basic Internet technology, and search engine related skills (necessary for shifting focus onto the end-user). Unfortunately,ă theă localizationă industryă ină Romaniaă isă stillă ină aă veryă incipientă stateă (Lakóă 2011).ă Thisă claimă corroboratesă theă tableă ofă theă topă 100ă translationă servicesă vendorsă (Kellyă andă DePalmaă 2012).ăWhileă numerousă Romaniansă mayă workă asă translatorsă foră internatională languageăserviceăproviders,ăthereăareănoăRomanianăcompaniesăthatăplayăanăimportantăroleăonă theă localizationă servicesă market.ă However,ă thereă mayă beă subsidiariesă ofă internatională languageăserviceăprovidersăinăRomania. 11 Furthermore,ă theă directionă ofă theă localizationă processă appearsă toă beă ratheră fromă Englishă toă Romaniană asă aă multitudeă ofă newă brandsă enteră theă Romaniană marketă eachă year.ă Whileă theă import/exportădeficităhasăbeenădecreasingăinărecentăyearsă[2],ăthereăareăfewăRomanianăbrandsă thată directlyă targetă theă outeră markets.ă Thisă mayă beă oneă ofă theă majoră reasonsă thată ledă toă ană almostă inexistentă Romaniană localizationă industry.ă Anotheră reasonă mightă beă theă factă thată Romanianăcompaniesăchooseăinternationalăcompaniesăforălocalizationăasăităisămoreăefficientătoă hireă theă localizationă companyă thată isă employingă translatorsă whoseă nativeă languageă isă theă targetă language.ă Thereă areă severală localizationă industryă experimentsă thată supportă thisă idea.ă Reverse-localizationă(Schäleră2006)ăisăratherăinexistentăinătheăcaseăofăRomanianăcompanies.ă However,ăthereăareăsomeănotableăexceptionsăthatăIăwillădiscussăinătheăchapterădevotedătoăcaseă studies. Anotheră majoră impedimentă toă offeringă localizationă oră reverse-localizationă servicesă isă thată thereăisănoătrainingăinăRomaniaăforăthisăspecificăfield.ăTheăstateăofătheălocalizationăindustryăină Romaniaăisămirroredăinătheăalmostănon-existentălocalizationătrainingă programsăinăRomaniană universities.ă 1.1.3 Teaching localization in Romania Localizationă shouldă beă aă priorityă ină theă curriculaă ofă theă universities,ă toă reflectă theă everă growingă demandă ofă translatorsă withă addedă competences,ă especiallyă localizationă skills.ăWhyă localizationăandăwebsiteălocalization?ăDueătoătheăfactăthatăaccordingătoăstatistics,ăe-commerceă andătransnationalăe-commerceăproliferatesărapidly.ăNielsenăcommissionedăbyăPayPalăpredictsă thatăbyă2018,ăthereăwillăbeă130ămillionăcross-borderăshoppersă[3].ăMarketingărelatedăwebsites,ă socială mediaă basedă productă promotion,ă computer-basedă training,ă andă onlineă applicationsă makeăupătheăonlineăelectronicămedia.ăTheătranslationămarketăcannotăfunctionăindependentlyăofă theăeconomicăenvironment.ăLikeăcomputerătextăediting,ăwhichăisăalsoăaămust,ălocalizationăandă websiteălocalizationăareăparamountăskillsăthatătranslatorsăneedătoădevelop. Initially,ă Iă reviewedă theă European Master's in Translationă (EMT)ă [4]ă syllabus,ă whichă isă ană EUăsupportedăactivity,ătoăverifyăifălocalizationăskillsăareărequiredăinătheăcurriculaăinăorderăforă aă universityă MAă programă toă beă acceptedă asă ană EMTă member.ă Amongă theă addedă specifică competences,ălocalizationăisăsuggestedăasăanăexample,ăbutăităisănotălistedăasăaărequirementă[5].ă 12 Thisă demonstratesă thată localizationă foră MAă ină translationă studiesă isă notă compulsoryă ată EUă levelăeither.ăThisăisăfurtherăseenăonătheăTranslatorăprofileăpageă[6],ăwhereălocalizationăisănotă evenămentioned.ăTheăEMTăTranslatorăTrainerăProfileă[7]ăagainălistsălocalizationăbutăonlyăinăaă footnote.ă Fromă allă theă EMTă members,ă thereă isă onlyă oneă Romaniană MAă program:ă Masterat European de Traductologie-Terminologie (European Master’s in Translation Studies and Terminology),ă“Babes-Bolyai”ăUniversity,ăFacultyăofăLetters,ăCluj-Napoca.ăTheăcurriculumăofă thisăMAădoesăcoverăgeneralălocalizationăbutătheătimeăframeăallottedăappearsătoăbeălimited. Next,ă Iă searchedă foră translationă universityă coursesă ină Romaniaă throughă Google.ă Aă straightforwardă searchă ină Googleă foră “traductologie/traducereă localizare”ă (MAă translationă localization)ăproducesănoărelevantăresults.ăIăanalyzedăthenătheăMinistryăofăNationalăEducationă (edu.ro)ă approvedă BAă andă MAă translationă andă communicationă programsă toă seeă ifă issuesă ină localizationăareăcoveredăandătoăwhatăextentă[8]. ByăinspectingăeachăofătheăBAăandăMAăprogramsăinătranslation,ăcurrentlyătaughtăinăRomania,ăIă discoveredă thată oută ofă theă 22ă BAă andă 19ă MAă programsă onlyă 5ă programsă offeră coursesă thată includeălocalizationăwhileăoneăcontainsăbibliographicalăreferencesăthatăalsoăincludeătopicsăină localization.ăThisă isă aă percentageă ofă roughlyă 12ș.ăTimeă allottedă toă localizationă isă variable.ă Forăinstance,ătheăBAăprogramăfromăUniversitateaăPolitehnicaăTimisoaraăassignedăaătotalăofă4ă hoursă foră general,ă gameă andă websiteă localizationă issues.ă Theă MAă ină Translation and interpreting in economics,ă establishedă ată Academiaă deă Studiiă Economiceă dedicatedă oneă seminarăforălocalizationămattersăandătheăstudyăofătheă LISAăPrimeră(aă guideăonălocalization)ă [9]. However,ă thereă isă aă usefulă onlineă courseă thată cană assistă translatoră trainersă andă traineesă toă commenceăorăextendătheirăskillsăinălocalization.ăTheăprojectăisăcalledăeCoLoTraină[10]ăandăită isăaăLeonardoădaăVinciăpilotăproject,ăinăwhichă“AlexandruăIoanăCuza”ăUniversityăofăIasiăwasăaă partner.1ăTheăprojectăconsistsăofăseveralămodules,ăallăofăwhichăareăhighlyăhelpfulăinătrainingă skilled,ă marketă orientedă translators:ă ICȚT,ăTerminologyă Management,ăTranslationă Memory,ă Localization,ă Projectă Management,ă andă eLearning/Blendedă Learning.ă Theă localizationă 1 For further details regarding this project and other localization training projects see Rodica Dimitriu and Karl-Heinz Freigang (2008) (eds.) Translation Technology in Translation Classes Ia i:ăInstitutulăEuropean 13 moduleă refersă toă softwareă localization,ă bută theă greată majorityă ofă theă conceptsă cană beă successfullyăappliedătoăwebsiteălocalizationăasăwell. ThereăareăpromisingăstepsătowardsăintroducingălocalizationămattersăinătheăcurriculaăofătheăBAă andă MAă ină translationă studiesă ină Romania,ă evenă ifă theyă tendă toă originateă fromă universitiesă withă aă ratheră economică oră technicală profile.ă Therefore,ă specializingă ină localizationă dependsă veryămuchăonătheătranslators’ăwillătoăadaptătoătheărequirementsăofătheămarket.ă Onăaăglobalămarket,ătheăInternetăisătheă mostărapidăchannelătoă approachăaffluentăclientsăwithă highă purchasingă power.ă So,ă anyă Romaniană companyă shouldă beă interestedă ină offeringă itsă productsă and/oră servicesă ină theă languageă ofă thoseă potentială clients.ăAnyă economică activitiesă basedă onă exportingă goodsă andă servicesă shouldă beă accompaniedă byă effectivelyă tunedă communicationă towardsă potentială businessă partnersă andă clients.ă Unfortunately,ă althoughă Romaniaăcouldăre-launchăitsăeconomy,ăespeciallyăthroughădevelopingăgoodsăandăservicesăforă theă externală market,ă deficiencyă ofă skillsă ină (reversed-)ă localizationă isă aă majoră drawback.ă Currentlyă thereă areă noă universitiesă ină Romaniaă thată activelyă teachă theă skillsă ofă localization.ă However,ă Romanians’ă earlyă contactă withă foreignă languagesă makesă themă skilledă candidatesă forătranslatingăfromăRomanianăintoăEnglish.ăEverăsinceăchildhood,ăRomaniansăcanăwatchăTVă cartoonsă andă moviesă thată areă subtitledă andă notă dubbed.ă Ifă theyă hadă someă trainingă ină localizationă theyă couldă aidă toă aă higheră extentă Romaniană companiesă ină marketingă theiră productsăandăservicesăglobally. 1.2 Mainăobjectives Theă fundamentală hypothesisă ofă myă thesisă isă thată SEOă localizedă websitesă mirroră ină aă moreă suitableă manneră theă communicativeă needsă ofă web-users,ă onă theă oneă hand,ă andă thată ofă companiesăonătheăother.ăAnăSEOălocalizedăwebsiteăimpliesă“traditional”ăwebsiteălocalizationă withă ană addedă layeră ofă searchă engineă optimization.ă Contentă isă producedă accordingă toă theă onlineă locală marketă specificsă andă isă reachedă throughă locală searchă engines.ă Suchă websitesă benefită fromă higheră traffică andă potentiallyă higheră sales.ă Asă aă secondaryă hypothesis,ă Iă willă demonstrateă thată fromă aă linguistică perspectiveă SEOă localizedă websitesă tendă toă useă aă descriptiveătranslationăapproachăwhereasă“traditional”ăwebsiteălocalizationăaăprescriptiveăone. 14 Fromă aă conceptuală pointă ofă viewă Iă proposeă aă specifică typeă ofă websiteă localization,ă i.e.ă onomasiologicalălocalization.ăOnomasiologyăisătheăstudyăofăverbalizationsăforăextralinguistică concepts.ă Whileă onomasiologyă isă aă strictlyă linguistică subfield,ă fromă theă perspectiveă ofă theă localizationă industryă aă moreă appropriateă namingă wouldă beă keyword-focusedă localization.ă SEOălocalizationăisăaătermăthatăencompassesăboth. By comparing website localization in general to the more specific SEO localization I will demonstrate that the latter is more efficient in communicating with the potential customers through the usage of search engines. Furthermore, I anticipate that the number of visitors and thus, of potential buyers will increase significantly as compared to websites that strictly follow the GILT processes. Anotherăobjectiveăofămyăthesisăisătoădetermineăwhichăofătheătextăproductionăstrategiesăisămoreă lucrativeă(measuredăinăproductionătimeăandăreturnăonăinvestment):ătranslationăorăcopywriting.ă Iăexpectăthatăinătheăcaseăofăparticularătextătypesăcopywritingăisăsignificantlyămoreăefficient. The thesis can also be considered as a starting point in localization training for Romanian and otheră internatională “minor”ă languageă students enrolled in translation and/or communication BA and MA programs. By using a multidisciplinary approach, this research aims to observe and analyze how the Romanian online business community promotes its products and services to various international markets. The number of Romanian companies on outer market is limited to a few, yet important Romanian companies. The websites of these and other foreign successful companies are analyzed in order to synthesize a list of factors that are accountable for their success. The data collected will be used for creating website localization models to be used by other companies as well. Romanian companies will thus benefit from the means of increasing exposure on the global market by localizing their products and services to foreign markets. While considering well-known brands for benchmarking purposes, I will demonstrate that even national brands and small local companies may gain global exposure, by using affordable online marketing methods. Concurrently, while there are hundreds of specialized and professional agencies in the field of globalization and localization, very often, end-users are not taken into consideration. For this 15 reason, the thesis will highlight the importance of website localization from the perspective of the end-user. This is also the main contribution of this thesis. Any website localization process should start from the end-users. This translates into a non-invasive approach, as I support the idea that the most rewarding type of text is the one targeting users that are actually interested in having displayed on the screen information related to their interests. This is the so-called inbound marketing, through which a potential customer reaches information according to his or her own will. The thesis also observes the most lucrative online advertising methods, best practice general localization issues, and best-in-class content marketing strategies. Content marketing includes optimizing text performance both from the perspective of the end-user and of search engine algorithms. Keyword research and keyword selection constitute the core of the user-oriented approach to website localization. Another area of interest relates to online communication practices, i.e. analyzing and comparing call-to-action segments in both the source website and the target website pages. While the great majority of researchers consider translation as a legitimate component of the localization process, in many cases translation is actually either not required or the industry will alternatively decide on employing copywriting. The thesis will analyze the similarities and differences between translation and copywriting strategies. A further list of supporting objectives is: - integrateăTranslationăStudiesătheoriesăwithătheăpracticesăfromătheălocalizationăindustryă - reviseăexistingăTranslationăStudiesătheoriesăonălocalization o reviewăsomeăofătheăissuesăinăinternationalizationă o redefineălocalizationăfromăaăcommunicativeăandămarketingăpointăofăview o classifyăwebsiteălocalizationăfromăseveralăperspectives o reviseătranslationăunitsăfromăanăSEOăperspective - developăaămethodăforăwebsiteălocalizationăbenchmarking - produceăaătemplateăforăSEOăwebsiteălocalizationă - measureătheăgeneralăperformanceăofăSEOălocalizedăwebsites - synthesizeă currentă factorsă toă beă consideredă byă companiesă thată plană toă localizeă theiră websitesăforătheăRomanianămarket o deviseă aă methodă foră collectingă call-to-actionă expressionsă specifică foră theă Romanianăonlineămarket. 16 - developă ană analyticală methodă foră reverseă localizationă soă thată Romaniană companiesă canăgainăaccessătoăinternationalămarkets 1.3 Researchămaterialăandămethod The research material is based on existing website content as I believe it accurately mirrors reality in this field of study. The websites analyzed include both websites of multinationals, for the purpose of benchmarking but also of Romanian companies in order to observe how they perform on the global market. Affiliate websites are also analyzed as most often the marketing approach and thus website content differs from the content of a branded localized website. I will examine the differences between and the performance of search engines, as well as the extent to which they may apply to the inbound, non-intrusive marketing trend. The main research methods are both quantitative and qualitative. Quantitative methods consist of comparing websites, analyzing competing websites and benchmarking. Qualitative research is based on reviewing the recent literature in the field of localization in general and website localization in particular. In the case of quantitative methods I will analyze website content from both an intra- and interlinguistic perspectives. Regarding qualitative methods, I will apply to competing websites theories from translation studies (TS) and the localization industry. I will also analyze how SEO website localization performs in comparison with “traditional”ăwebsiteălocalization. Through benchmarking, as a primary deductive approach, I will collect and sample data on which I can synthesize the typical characteristics of a localized website. By using online specific tools, I will be able to determine the user profile and behavior, which are of primary importance in determining the target audience and context. The analysis of the sample websites will be accomplished through comparison with the findings originating from the benchmarking process. The analysis will consider characteristics such as localization conventions and norms, language and text issues, aesthetic and technical matters. I will also highlight the repetitive and the circumstantial factors involved in website localization. Another (deductive) research method will be the analysis of various websites from the perspective of the theoretical framework offered by Translation Studies (TS). The literature in the field will be integrated and applied to both authoritative websites (of multinational 17 companies) and Romanian websites. Through this approach I will observe the extent to which TS and Localization theories are mirrored in the actual website localization process. For some of the outcomes (originating from empirical methods) I will also provide theoretical evidence. The list of websites chosen for analysis includes the following websites: - reference websites: google.ro, emag.ro, dacia.ro, samsung.com, bitdefender.ro etc.; - target websites: websites of several Romanian winegrowers The comparative method will primarily be used for reviewing websites both diachronically and synchronically. The diachronic perspective will be adopted in comparing the specific localization features of a website in different periods of time. The synchronic method will be employed in the analysis of different websites from the same period. The analysis will include estimating the level of localization, both from the perspective of typical website components (domain name, peripheral texts, keywords, general design and layout, interface, textual content, contextual content such as graphical elements, etc.), and at a suprasegmental level (the website as perceived by users in general). The main hypotheses of the thesis will be corroborated by methodically analyzing the differences and similarities between the target and reference websites. I will make extensive use of widely known tools such as Google KeywordPlanner, Google Analytics, Bing Keyword Research etc.), as well as of various online tools specific to translation and localization processes (translation memories, CAT tools, automated machine translation, etc.). User behavior and profile will be analyzed by using Google Analytics, Google Suggestion Tool, Google Trends, etc. Additional empirical methods: Observation: Through methodical and purposeful observation of the localized websites I will collect evidence that shows that SEO localized websites outperform the traditionally localized ones. Adopting the point of view of the end-user, I will observe closely how webpages from reference or competing websites perform on search engines. The thesis also includes an experiment I made with my students which consisted in asking them to produce articles based on the SEO awareness level. Such articles can be of three types: translations (no localization or SEO skills are involved), articles produced through 18 localization or copywriting (localization skills only), and SEO plus localization aware texts. The texts produced were then analyzed from the perspective of their performance on search engines (positioning in the results pages) and how they were perceived by web-users (average time spent on the webpage) or, alternatively, by a test group. The statistical method includes the analysis of statistical data available in search engine reports with applicability in observing users’ăbehavior and keywords by which the websites were accessed, time spent on each webpage, navigation patterns, etc. This method is also used in the thesis for the analysis of text length, for extracting and comparing call-to-action words and in case studies. The findings resulted from using these research methods will provide support in demonstrating the importance of the end-users in the process of website localization. The outcome of the analyses will be generalized and synthesized after each of the research stages. As previously stated, the main objective of the research is to demonstrate the efficiency of SEO localized websites over traditionally localized websites, while suggesting methods of directing web-users towards company websites, of converting web-users into customers, highlighting the factors that determine the conversion, etc. 1.4 Structureăandăcontent The first part of the thesis covers issues such as globalization, internationalization, localization and their more specialized types: reversed-localization and e-localization or website localization. References are made to the mental and cognitive processes localization involves: translation, adaptation, transcoding, transfer, trans-creation versus copywriting in the target language. As the ultimate purpose of localization is to communicate in a given locale, matters such as interaction, Gricean maxims, semasiology and onomasiology, borrowings etc. are also considered. This part also discusses aspects of a more technical nature involved in the localization process in order to provide a complete view of what localization involves. Interpreting the data provided by search engines and various online tools for keyword research plays a major role in shifting focus onto the end-user. Keyword research is of outmost importance to link the companies’ă offersă withă whată theă usersă lookă foră onă searchă 19 engines. In website localization this means that translation must be made according to statistics on keywords. The following chapter investigates specific issues of electronic texts and how features of etext affect the comprehension of the source content and the production of the target content. Issues such as classification of online text, layers of text, text and non-text, and technical issues such as the components of the website mix are discussed and analyzed. Theăfinalăpartăofămyăthesisăconsistsăofăseveralăcaseăstudiesăinăwhichătheoreticalăclaimsăfromă theăpreviousăchaptersăareăappliedăandătested.ăThus,ăIăproposeăaămethodăforăbenchmarking,ăbyă analyzingă severală internatională brandsă fromă severală perspectives.ă Anotheră concernă isă toă demonstrateă thată localizationă doesă notă alwaysă involveă translation.ă Withă regardă toă textă production,ă companiesă employă translation,ă copywritingă oră aă mixă ofă translationă andă copywritingăstrategies.ă Anotherăcaseăstudyă encompassesăkeywordăresearchăandăon-siteăSEOă optimizationăappliedătoătheătwoătypesăofăstrategies.ă Fromăaăratheră“traditional”ăTranslationăStudiesăandălocalizationăperspective,ăoneăofătheăcaseă studiesăimpliesătheăanalysisăandăcomparisonăofălocalizedăwebpagesăwithinătheăsameăwebsite. InătheălightăofătheăSEOălocalizationăandăreverseăSEOălocalizationăIăproposeă(a)ăaămethodologyă forăforeignăcompaniesăonăhowătoălocalizeătheirăwebsitesăforătheăRomanianămarket,ăpinpointingă culturală aspectsă thată mayă beă applicableă toă thisă particulară market;ă (b)ă aă methodologyă foră Romanianăcompaniesăonăhowătoălocalizeătheirăwebsitesăforăforeignămarketsă(namelyătheăUK).ă 20 2 Keyăconceptsăinălocalizingăwebsites 2.1 Introduction Globalization, internationalization, localization and translation are processes of making the transfer of the message from one language/culture/context to one or more target languages/culture/contexts in a more efficient and cost-effective way. Most frequently these processes are directed from a major language source to another major language or a minor language. In Romania most products are imported from China, the EU (especially from Germany) and the US, so, products must be localized to the cultural, linguistic and legal system locale of Romanians. For example, Apple is one of the largest US companies with a well-established position in the world due to both the quality of its products and to its marketing strategies. The localization process for the Apple products is implemented at regional level, and they cover most of the countries, including Romania. On the other hand, these processes can haveă aă reversedă directionă asă well,ă fromă aă “minor”ă language-cultureăpairătoăaă“major”ămarketălocale.ăNokiaăisăoneăofătheămostăsuccessfulăFinnishă brands in mobile technology that sells their products and services around the world, with the US as its most important market. Their products and websites can be considered as studying material to confirm the principles of the processes mentioned previously, as there is a site for each country in the world. Such renowned companies are proof that localization should be considered by Romanian companies as well. Some of them are successful. For instance, the Romanian car maker Dacia has localized its websites to most of the European countries and languages, and all the elements of the localization process are more or less attended to. The official site of the EU, although very complex in structure and information abundant, is one of the outstanding examples of a localized multilingual website of an institution. It can also be used as an example for localizing websites, especially in the case of those businesses that plan to offer law related information or merely need to adjust their company policy to a target market. Information related to return policy, privacy issues, cookie law and so on can be applied by analyzing websites that have already complied with local laws and regulations. 21 Unfortunately, there are numerous Romanian companies that fail to acknowledge the necessity for localization practices. Most often the translator considers that the text only needs to be translated, while failing completely when considering other aspects of localization. Furthermore, even text translations are deficient in being transposed into content in the target language. For instance, http://www.eusecholding.eu, although it offers services in several countries, completely fails when considering localization. One can trace the source text and context (Hungarian) easily, in both the English and Romanian languages, because the skills of the translator(s) are inadequate (“Ină whată weă areă other”ă andă “Ină ceă suntemă diferiti”ă foră theă Hungariană “Amibenă másokă vagyunk”).ă Also,ă the texts on the pictures have not been translated at all. (Visited in May 2012). Currently the website is offered only in Hungarian (visited in February 2014). Inătheăviewăofătoday’sărealitiesă- the Internet is a business and an information tool as well as an educational resource, it is vital to reveal the importance of the end-user in the localization process of multilingual websites so that the gap in the business-to-client (B2C) and businessto-business (B2B) communication is covered as much as possible. By observing both sides, the offer related to a product or service, on the one hand, and that of the potential client on the other, the localizer in the role of the mediator can facilitate the communication. There is a common goal that the two sides share as the seller is interested in generating sales and the buyer is interested in finding quickly what s/he is interested in. The search engine aware localizer can be a catalyst in this process by adopting an appropriate translation and localization method, and helping in the achievement of non-invasive communication. Noninvasive refers to delivering to the search engine user real and relevant information, in the appropriate contexts, by using common signifiers (keywords). Ină thisă chapteră Iă willă analyzeă terminologicală issuesă thată areă relatedă toă theă phenomenonă ofă localizationăinăgeneralăandăwebsiteălocalizationăinăparticular,ăasăaăfurtherăresearchătoămyăinitială findings.ă (Lakóă 2012a)ă Theă keyă conceptsă ină websiteă localizationă areă oftenă referredă toă asă GILTă(Globalization,ăInternationalization,ăLocalization,ăandăTranslation).ăIăwillăanalyzeă howă theă industryă andă theă academicsă ină TSă perceiveă GILTă andă howă theă 4ă maină processesă interrelateă andă influenceă theă successă oră failureă ofă theă localizationă outcome.ă Asă fară asă 22 translationăisă concerned,ă Iă willă reviewăseveralămodernădirectionsăinăTranslationăStudiesăandă howă theyă applyă toă localization.ăAsă alreadyă mentionedă ină theă introductoryă chapter,ă theă maină focusă ofătheăthesisă isăonă websiteălocalizationăandăoptimizingălocală websitesătoă locală searchă enginesă (foră instanceă google.ro),ă asă thereă hasă beenă aă majoră shiftă ină termsă ofă howă potentială buyersăperceiveătheirărelationshipăwithătheăsellers.ăTheădirectionăisănowăfromăclientătoăselleră (C2Bă–ăcustomer-to-business).ăTheăpotentialăclientăisătheăoneăthatăinitiatesătheăcommunicationă process,ăbyăsearchingăforăinformationăonăsearchăengines.ăIfătheăinformationăisănotăreachableă byăsearchăenginesăităcannotăbeăfoundăbyăweb-usersăeither.ăTheămoreăinformationăandăcontentăaă companyă offersă onă itsă productă oră servicesă theă higheră theă possibilityă ofă beingă found.ă Asă aă consequence,ă “contentă marketing”ă hasă becomeă theă fashionableă termă toă referă toă thisă relationship. 2.2 Globalizationă(g11n) Globalization occurred as a natural consequence to the ever-growing need of companies to extend and enter new global markets. Thus, the term globalization was initially used in an economic context, at decision-making levels of powerful companies looking to offer their products and services on a greater scale, to various customers around the globe. Only later did it come to the attention of the scholars in Translation Studies, when the need to offer products and services in various languages became a marketing necessity. Globalization, abbreviated as g11n by the business community, became a compulsory step in the process of translating product description and other related content, i.e. making content internationally available. Globalization isă “[…] simply about spreading a thing to several different countries, and making it applicable and useable in those countries. We suggest therefore that our industry should follow the general meaning the word globalization already has in other domains, which is simply the dictionary meaning.” (original emphasis, Cadieux and Esselink 2004:3) Merriam-webster.comădefinesăglobalizationăasă“theăactăorăprocessăofăglobalizing : the state of being globalized; especially: the development of an increasingly integrated global economy marked especially by free trade, free flow of capital, and the tapping of cheaper foreign labor markets”[11] 23 Dictionary.cambridge.orgădefinesătheăsameătermăasă“theăincreaseăofătradeăaroundătheăworld,ă especially by large companies producing and tradingă goodsă ină manyă differentă countriesă •ă when available goods and services, or social and cultural influences, gradually become similarăinăallăpartsăofătheăworld”.ă[12] These definitions cover the social and economic aspects of globalization, but do not include any reference to the communication processes involved, namely communicating with the global market in the language of each of the distinguishable local markets. Thus, globalization involves translation as well. Multinationals and professional associations such as MultiLingual.com, which already have substantial experience in offering products and services to different language communities, provide more precise, translation studies oriented definitions. The December 2011 issue of the online MultiLingual Magazine defines globalization as "... the process that addresses business issues associated with launching a product globally, such as integrating localization throughout a company after proper internationalization and product design."[13:60] Globalization, in my view, is the very first step, a decision made in the marketing department, which initiates all the other subsequent processes: internationalization (l18n), localization (L10n) and translation. It primarily involves managerial aspects such as deciding which markets (locales) are targeted first (usually affluent nations) or where the return on investment (ROI) is more significant. During the Globalization stage a budget is set for the other subsequent processes, deadlines are established, etc. Although only a limited number of markets are initially selected, a carefully designed product or service will allow for the addition of any number of market prospects. Thus, for instance, a game console publisher may decide to launch a new title that would cover the U.S., Japan and parts of the EU first, but then through appropriate planning they can extend their offer to other markets as well: South Korea, Russia, Turkey, etc. The extent to which globalization may be adopted is restricted very often by objective external factors: competition on a certain local market, costs involved in the subsequent processes (internationalization, localization, translation) especially when the markets are rather different from one another in terms of language and culture, geo-political issues, environmental problems, etc. 24 For instance, keyboards on laptops are rather tricky to be set appropriately for various languages, not to mention that no one can explain the logic behind placing the keys the way they are on typing machines. The first computer keyboards were based on the 26 letters of the English alphabet, but numerous languages have more than 26: Romanian - 31, German - 30, Hungarian - 44, French - 26 + 6 diacritics and 2 ligatures. In the beginning, Microsoft managed the issues of the required extra characters with the aid of software (in Office Word various symbols from any language can still be added into the document). Only much later were special characters added to the keyboards, and this is because it added extra cost to development and production. They were first added only from the operating system, and it wasă quiteă awkwardă toă getă usedă toă findingă whereă specială Romaniană charactersă ( , ,Â, ,Î)ă were supposed to be located. Programming those keys to have several functionalities from the software and accessing those functions by pressing various combinations of keys involved more costs and implementation time. The extra characters were engraved onto existing keys only later, depending on the size of the potential market. Figure 1: Standard US and German keyboards 25 There are even more problematic issues with languages such as Chinese, arguably considered by many to use an ideographic alphabet. Figure 2: Chinese keyboard While there is an improvement in the keyboard availability for local languages, Non-English users will still need to know what various function keys refer to: Esc, Ctrl, Shift, Enter, Backspace, Insert, Alt. Most programmers will not consciously be aware of where these English terms originate from. Therefore, the learning curve for using these keyboards and related software is slightly more abrupt than in the case of native English speakers, as these function keys cannot be adapted to a locale. With modern gadgets, however, such as those using virtual keyboards on touchscreens, planning for globalization is now significantly easier, as operation systems can include virtually any number of keys to match any target language, while also benefiting from the option to zoom or move the keyboard around the screen. Thus, although the initial costs for developing a touchscreen controlled gadget is more expensive, the processes to internationalize, localize and translate related content are far more costly effective, because most of these processes are achieved from software, which can be maintained and upgraded significantly easier. 26 2.3 Internationalizationă(I18n) Subsequent to the process of globalization is internationalization (I18n). Schäler defines the term as “[...] the process of designing (or modifying) software so as to enable users to work in the language of their choice (even if the software is not localized) and to isolate the linguistically and culturally dependent parts of an application in preparation for localization.” (2009: 158) Software is not the only product that can be internationalized. Dacia Duster, or any other car maker, needs to have its driving wheel on the right-hand side, for instance, if it is to comply with UK driving regulations, and, although there are only two options of placing the driving wheel, it is still about designing the product while minding local rules and regulations. Therefore, internationalization is not only about preparing a product or service for language changes, but also preparing it for seamless integration with the target locales. Inappropriate or partial internationalization, especially if the target culture has already been in contact with similar products or services, can lead to readjustments and extra costs, or even rejection of the products or services. Previous user experience,ă asă partă ofă theă reader/user’să profile, is an important background factor that can determine how a product or service is perceived. A car with the driving wheel on the left will not sell well in the UK or Japan as it does not comply with what drivers are used to. For instance, by using keyboard shortcuts such as CTRL+S for saving a document, “S”ăforă “saving”ă worksă foră theă Romaniană “salveaz ” as well. The same happens for CTRL+C (“copy”-“copiaz ”) andă CTRL+Pă (“print”-“printeaz ”, increasingly used instead of “tip re te” when using a printer). On the other hand, even in English some of the shortcuts do not necessarily indicate the operation to be achieved.ă“Undo”ăandă“redo”ăcannotăuseătheă“U”ă andă“R”ăkeysăasătheyă areăusedă foră “underline”ăandă“align right”, respectively. On the other hand, the use of CTRL+X foră “cutting” is rather iconographic (a welcome tendency especially in the use of gadgets - aătriangleăusedăforă“play”,ăaăsquareăsymbolăusedăforă“stop”,ă etc.).ă “Open”ă andă theă Romaniană “Deschide” use the same combination - CTRL+O. Hence, preparing software for a new market would be rather difficult. In such cases, it is more appropriate to let the user learn the key combinations. The process of preparing a service, in this case keyboard software, can be taken even further. For non-English users you can even choose between using a standard keyboard layout with diacritics and a programmer keyboard layout with or without diacritics. Another argument is the usage of a slightly different keyboard in the UK - an extra £, as compared to the standard US keyboard. 27 Thus, internationalization is not only about developing a strategy for providing a product in several languages. It also involves preparing a product for local rules and regulations, habits and needs while generating and preserving common communication patterns. Offering various customers from different parts of the world the opportunity to use commonly recognizable icons suchă asă theă previouslyă mentionedă “play”ă andă “stop” buttons on gadgets represents a plausible solution. Another exampleăisătheă“Windows”ăfunctionăkeyăonăWindowsă Operating System compatible keyboards, which is now universally known to trigger the Windows menu. Considering the above remarks, the definition provided by LISA for internationalization found in the MultLingual Magazine encompasses much more accurately the current state of thisăphenomenon:ă“... the process of generalizing a product [or service] so that it can handle multiple languages and cultural [and legal] conventions (currency, number separators, dates) withoută theă needă foră redesign.”ă (LISA cited in Pym 2004a:29) Similarly to globalization, internationalization is rather a managerial decision involving financial and technical aspects. Pymă considersă internationalizationă ană “intermediary version”. The production of that intermediaryă versionă isă calledă “internationalization,”ă andă theă thingă producedă isă the “internationalized”ăversion.”ă(2014:121) In my opinion, this intermediary version should be based on human-computer interaction findings and any typical locale references be excluded. Pym appropriately names the process of producing an internationalized version delocalization or interlocalization, as all locale traces are to be removed. (2004a:30) Similarly, Cronin speaks of internationalization as a process of accomplishing maximum possible cultural neutrality. (cited in Jiménez-Crespo 2013: 32) From a TS perspective, Pym considers that equivalence between ST, on the one hand, and TTs, on the other, is achieved during the internationalizationăprocess:ă“equivalence [is] created not particularly by one-toone translation, but by the prior moment of internationalization.”(2004a: 31) Strictly referring to software internationalization, Microsoft uses for the same concept the term localizibility in its documentationăforăVisualăStudioă2005:ă“An intermediate step prior to localization is testing for localizability. In this step, you ensure that you have separated the application's resources that require translation from the restăofătheăapplicationțsăcode.ă”[14] 28 Apart from separating the resources that require translation, the main technical aspect involved in the internationalization process is to delimit and separate the localizable parts from the whole. In the case of a website, the developer has to establish variables - references or, more technically called, Uniform Resource Identifier (URI), both for menus, instruction and help files on the one hand, and for content on the other (text strings and media files). HTML5 standards use specific coding page elements that describe in a more appropriate manner what a certain page element includes - <article> for the main part of a page, <nav> for navigation menu, <section> for a more general type of content, <footer> for the footer of the webpage, and so on. For non-text content like images and animation, no text should be fixed, but dynamic. Texts should be displayed over pictures through dynamic code with the contentăofătheătextădependingăonătheăuser’sălanguage and locale, controlled from an XML file or, even more so, database driven so that it allows the usage of an administration interface through which non-programmer employees can operate changes. From a TS perspective on internationalization, the source text is produced in such a way that it can be straightforwardly transferred into the target languages. This is related to the concept of one-to-many: “This is a term for translation processes that go from an internationalized version to many target-language versions simultaneously.”ă(Pymă2014:120) 2.3.1 Internationalization and translation strategies In this subchapter I will examine several aspects that have been, more or less, overlooked by the academic community. While these issues are important for generally more efficient GILT processes, there are also numerous aspects that need to be considered from the perspective of various translation study theories. Thus, in order to maintain the same branding yet adapt the marketing strategy, text source content should be written in such a way as to be easily adapted for any target market. By undertaking this properly at this stage can save time and money, and the company will be ableă toă avoidă erroneousă translations.ă Foră example,ă byă usingă “plaină English”ă foră theă sourceă text, while simultaneously avoiding non-standard English, the translators may, with less effort, understand and then transpose the message into the target language. Texts that cannot be generalized, such as rhetorical figures, puns, alliteration, rhyme must be avoided. Content 29 which is difficult to understand is difficult to translate. Furthermore, Oracle.com, while referring to user interface labels (buttons, links, etc.), suggests that at the stage of internationalization it is crucial to avoid jargon, idioms, slang, colloquialisms, abbreviations, Latinate abbreviations (i.e., e.g.), contractions, possessives, capitalization (not all languages use capital letters). In the same article they suggest using the plural instead of the singular or addingăbothă theăsingularăandătheăplurală form,ă andăreplacingătheăslashăsymbolăwithă “or”ăandă “Ț”ă withă “and”ă [15].ă Cleară andă fullă phrases,ă simple,ă active-voice sentences, if the genre allows it, are the most efficient both from the perspective of localization and translation, on the one hand, and from the perspective of marketing on the other. If the message is understood by the translator, the translation as a product will have a higher degree of accuracy and the local market will be more receptive to the message. This is a move towards facilitating the comprehension of the source text, - that Chesterman (1997:92) incorporates under the “comprehensionă strategies” paradigm; they refer to analyzing and understanding theăsourceătextăasăcomparedătoătheă“productionăstrategies”ăwhichăfocusăonătheătranslationăasă an output2. Most companies require that the translations should be made by translators that are native in the target language. Consequently, in my opinion, more emphasis should be laid on the comprehension process of the source text, as the translator is often not a native speaker of the source language. Comprehension strategies include various sets of activities such as organizational, reading, text-analysis, terminological issues (which in the case of websites is determined by keyword usage), further information retrieval, consulting experts, etc. It is more beneficial to hire translators specialized in the field of activity required by the client, as mostăofătheăpreviouslyălistedăissuesăareăalsoăaămatterăofătheătranslator’săexperienceăinăaăcertaină field of human activity. Also, if the text is internationalized efficiently, machine translation can be useful as it works much more capably with texts written in a clear, unambiguous language (in a similar manner to technical texts). Throughout the internationalization stage the internationalization team/individual has to decide upon the global translation strategy (Bell 1998:188). This is the moment when features such as the style of the text, discourse, genre, and so on, that are to be applied to all localized texts in their respective target markets are decided. For instance, if an e-learning website providing math tutoring in the source language addresses users aged 10 to 12, the vocabulary 2 ăSeeăalsoăsubchapteră2.5ăTranslationăinăthisărespect. 30 used should be comprehensible for this age group - simple and clear sentences, active voice, friendly tone – even if it uses terms that are specific to math. Ifă weă adoptă Romană Jakobson’să perspectiveă ofă theă verbală signă (1987:ă 429)ă - rendering meaning in the same language, into a different language, or into a nonverbal code - during the internationalization stage of the source text, and more generally of the entire website, there are two processes that are used: - intralingual translation or rewording, to standardize the source text/content (See previous Oracle.com examples above [15]) and - rendering message by means of nonverbal code or iconization. Common examples on websites are the use of a house icon for the home or start page, several horizontal lines for opening the menu on a mobile device, an envelope for the contact page, and so on. Yet, another aspect that needs to be determined during the internationalization process, is the translation orientations, in other words whether it is necessary to adopt a domesticating (i.e. source-oriented) or a foreignizing (target-oriented) strategy (Venuti, 1998:240), and to what extent either of these approaches should be used. In the world of global e-commerce while theă veryă termă ofă localizationă suggestsă “domesticatingă strategies”,ă theă foreignizingă approachă would still be more suitable if a company decides that potential clients request information on the origin and the story behind the products they intend to buy. However, this may be dependent on the target market. The foreignization-domestication dichotomy, in the case of ecommerce websites, has more to do with the culture-specific terms, references to location and time, relationship between communicative partners, measuring conventions, formal conventions, text-type and genre conventions, conventional forms of address, salutation formulas, and structural differences in vocabulary, syntax, and supra-segmental features of the two languages. Often foreignization may be applied only to slogans or advertising campaigns. German cars are much appreciated all over the world as reliable (and on the Romanian market even more so, as there is a frequent distrust about national products). On the other hand, aăquickăsearchăonătheăInternetăshowsăthatăVolkswagen’săslogană“DasăAuto”ăisă used on most international markets as much as the other no less famous previous slogan, “Ausă Liebeă zumă Automobil”.ă This can be useful in the case of other German products as well, as potential buyers would infer that any other German technology related product would riseătoătheăsameăqualityăasăthatăofăGermanăcars.ăMcDonald’săandăPepsi,ăwillăoftenăuseăEnglishă words in their slogans or advertisements on international markets. Therefore, when the origin 31 of the product is important, the foreignizing approach is often used across the world especially in the case of global brands. This strategy may enjoy even a higher degree of success when the target market suffers from the “countryă ofă origin”ă effectă (Pucci et al. 2012:155). Mirroringătheăsameăidea,ăinăTSăterms,ăIăfindăappropriateăHumboldt’sădefinitionăofă translation:ă “Translation should indeed have a foreign flavour to it, but only to a certain degree; the line beyond which this clearly becomes an error can easily be drawn. As long as one does not feel the foreigness (Fremdheit) yet does feel the foreign (Fremde), a translation has reached its highest goal; but where foreigness appears as such, and more than likely even obscuresătheăforeign,ătheătranslatorăbetraysăhisăinadequacy.”ă(cited in Dimitriu 2002:21) Standardizing the source text does not imply that the localization and (intratextual) translation - for the same language but different cultural values, should not allow puns, rhyme, culturebound terms and other text specific operations that the localization process involves. If the standardized source text is English, and in most cases it is, localization will involve changes at text level for the US market, the UK or Australia. This means that localization and personalization depends on the marketing strategy. If the language of a product or service provider is not English, English will become the pivot/bridge language. A pivot language in the machine translation terminology is defined as an intermediary language to achieve translation from one language into another. Thus, in order to translate from Romanian into Chinese, one could opt for English as a pivot language – Romanian would be translated first into English and then the English text into Chinese. Case study Bitdefender.com is a Romanian antivirus software developer. Under the language option menu there are listed several languages. One could consider that the most delocalized language variant of the website is WorldWide English (form found on bitdefender.com). This English variant can function as the TS for translation or as a pivot language. All the other language options would be considered localized versions. However, a more in-depth look reveals that choosing from any of the following options: WorldWide English, United States English, or Canada English, displays the same content. By further analyzing the source code and text on the home page for each of the above mentioned options and the other English speaking countries, it is obvious that the differences are only URL related (.com.au, .co.uk). There is no localization involved, they use an all-around standard version. 32 The standardization of the text is somewhat similar to the process of pre-editing the source text, so that the pre-edited text is easier to be translated by using machine translation. Accordingătoă Hutchins,ă “pre-editing of the input, using a controlled language, or restricting the system to a specificăsubjectădomain”ă(2005:ă3)ăwillăensureătheăspeedăofătheătranslationăandă localization process and also ensure a higher quality of the output. Similarly to Oracle.com [15], Hutchins (2005: 8) lists several pre-editing procedures and controlled/standardized source text input best-practices which apply to MT and thus to the more efficient website localization process. Cardey et al., referring to MT, speak about controlled language of the input language, which through extrapolation applies to the source text of the website to be localized, during the internationalization process. (2004: 38) Similarly, Pym considers that “internationalization can make a text simpler, reducing surface-level variation through the use of controlled language.”ă(2014:122) A simplified language can be translated more easily and more efficiently through MT. In this way translators could return the translations more rapidly as presumably only the revision of the MT output would be necessary. Pym also considers during the internationalization step expanding the start text so that it allows a multitude of localization possibilities. (2014:121) However, the localized content would be then the result of an assimilation or gisting translation approach3 and the text may be perceived by the translators for each of the target texts in various ways. Also, branding at global level may show deficiency in unity. While MT is of use to the translation process, oversimplified language of the source text will require extra effort on the part of the translator if the text is to be perceived as appropriate and felt as created in the target language. What makes a language and culture specific are precisely the rhetorical figures, puns, alliterations, rhymes etc. Under these circumstances, what would be the most appropriate approach? An attempt to answer this question through a conceptual analysis is made in subchapter 2.5.1, Translation, adaptation, copywriting, while empirical evidence is provided in subchapter 4.4, Which Way Website Localization: Translation or Copywriting? One last important issue to be elucidated during the internationalization process is to study how the information on the webpages is accessed and read by each of the target nations. There have been numerous studies regarding user interaction with the online medium. One of 3 ăSeeă2.5.4,ăMachine translation for website localization 33 the most important research groups in this area is the nngroup.com, led by Jakob Nielsen. They have hundreds of studies regarding usability, user interface and user interaction. Some of the group findings must be considered during the internationalization process. They relate to reading patterns, content prioritization, types of e-commerce web-users, international usability (effectiveness of user interface used in a target market different than the one in which user interface was created) (Nielsen 1996), localized website testing, level of Internet maturity among various nations (rapidly levelling over the last years), and many others. Regarding internationalization, Nielsen found in one of his studies (Nielsen 2011) that, in general, the basics of the interaction with website interface are the same all over the world and founded on human-computer interaction findings, regardless of the country of origin or of their language and cultural background. Reading from right to left, in the case of Arabic or Hebrew, is only a matter of direction, as the users will still focus their attention more on the beginning of the message, hence, it is a mirrored F-pattern (Nielsen 2006)4. 2.3.2 Internationalization and text length The internationalization process stage of the source text in web localization must also consider the various character sets necessary for displaying the multitude of international languages: variants and extension of the Basic Latin alphabet, Cyrillic, Japanese, Chinese, Arabic, Greek, etc. While there are similarities with the keyboard localization issues previously mentioned, there is a necessity at this stage that all possible languages should be accounted for. In my opinion, very often, even pictures need to be changed depending on the target locale, but for such a case a URI (Uniform Resource Identifier) needs to be set from this stage. For instance, it would look inappropriate to use black people to advertise insurance services, buying electronics or employ them to be part of contextual pictures on a website on the EastEuropean market, where black populations are insignificant. It would not convey consistency with the beliefs and common knowledge of the potential clients; it would appear peculiar, or For further details, see also 2.4.7, Grice’s Maxims applied to localization, specifically the maxim of manner and subchapter 3.2 Text and E-text 4 34 US-like. Thus, knowing that a picture must be kept dynamic within a design is as important during the internationalization stage as separating text strings from code. Very often, local celebrities are used to convey to theă audienceă “Iă amă likeă youă andă Iă useă it. Follow my example!”.ă Consequently, the spot for an image or a video on a webpage should also be considered from the perspective of the internationalization process as it will be significantly easier to change and test various non-linguistic elements when reaching the localization process. Ină websiteă localizationă webă designă andă theă websites’ă layoută mustă beă consistentă acrossă theă various target languages. There have been numerous studies in Machine Translation and Comparative Corpora showing that the target text (TT) is often longer than the English source text (ST). The length of the text can be measured either by the number of words or by the number of characters. A straightforward free online tool to test the length of ST and TT is Side-by-Side SEO Comparison Tool [16]. For instance, one can compare the text length of the English and Romanian versions of two pages from the EU official website (europa.eu). If we consider languages other than English, then the TT can be longer or shorter. However, in the case of menus and other buttons, there is often a limited space for the text, and the target language may use certain corresponding terms, from several competing synonyms that are longer than the available space. Synonyms may be used but they may not in all cases be perceived as natural replacements. This is called Text expansion and is defined by Roger Chrissăasă“anăincreaseăinătheălengthăofătheătargetătextăasăcomparedătoătheăsourceătext.ă[…] but must be given due consideration by graphic artists and desktop publishers who want to use the same format or templates for both the source text and translated text. Similarly, text expansion must be taken into consideration when translating software. Since dialog boxes andă windowsă mayă haveă toă beă resizedă toă accommodateă theă translatedă text.”ă (2006:202)ă Ită isă also worthwhile mentioning that, conversely, there is also a possibility for the TT to be shorter for certain language pairs. High-context cultures require less text. It would be useful to conduct a research on verbosity with regards to the target market. If, for example, a lowcontext culture would require producing 500-word-long textual content, in the case of a highcontext culture 300 would suffice. Also, it would be useful to determine if the difference in text length is compensated for with images or other non-textual elements. Nielsen (2011) also reveals some important data relating to the “degreeăofăverbosity”ăbetweenă languages. Some languages are seemingly more verbose than others. Arabic and German 35 would need more space to display, and this is not a matter of translation but a cultural issue, which means that the translator should be a native speaker of the target language and use a lengthier and wordier text in a natural manner. This might imply copywriting skills as well. There is also a culture specific issue linked to textual length, as there are two main types of communication which also influence the length of text but from a different perspective (cf. Hall 1976: 105-116): high-context (HC) and low-context (LC) cultures. LC cultures use clear, direct and unambiguous communication as well as an argumentative style, whereas HC cultures require less wording but more context. German, for instance is an LC culture, so there is no doubt it requires more display space on webpages. More details regarding the cultural impact on communication, translation and localization in later chapters. With the ever increasing usage of mobile devices, apart from connection speeds, it is very important to provide content laid out properly according to the screen resolution and size of the device used. Even the buttons used for the menu must be resized to be pressed easily usingătheătipăofătheăindexăfingerăonăaă4.52”ăasăcomparedătoăpressingăbuttonsăofătheăsameămenuă using the mouse pointer on a 24" desktop monitor. Fortunately, this can be automated if so desired. Through a programming framework (such as Bootstrap - http://getbootstrap.com/) screen resolution, menu position and content layout are adjusted accordingly, maintaining a user-friendly experience across as many screen sizes as possible. This kind of option should be considered as one of the technical aspects of localization. In technical terms, it means applying a responsive design. Responsive design adjusts content layout according to the device on which the information is displayed. Thus, apart from dealing with source content that needs to be displayed correctly on a multitude of devices, textual length in the target language must also be considered during the stage of internationalization. Thorough planning at this level can save time and money. 36 2.4 Localizationă(L10n) As previously shown, localization (L10n) is a process that cannot be achieved without going through the globalization and internationalization processes first. Cadieux and Esselink defineă theă termă asă “...adaptingă aă thingă toă theă needsă ofă aă givenă locale.”ă (2004:2) The online MultiLingual Magazine, defines in its terminology section localization as “theă processă ofă adapting a product or software to a specific international language or culture so that it seems natural to that particular region. True localization considers language, culture, customs and theăcharacteristicsăofătheătargetălocale.”[17:50] Although the software industry began to be interested in localization as early as the 1980s, scholars showed no interest in it prior to Anthony Pym (1999) and David Brooks (2000). Later on, several researchers began tapping the subject, either from a linguistic perspective (David Crystal 2006), a technical one, namely web-user behavior (Justin Cutroni 2010) or from a multidisciplinary perspective which, nevertheless, involves translation studies to a high extent (Dianne Cyr 2003, Daniel Gouadec 2003, Anthony Pym 2004, 2005, 2010, Miguel A. Jimenez-Crespo 2013, etc.). Website localization is a more specialized type of localization, which is extending rapidly as website content proliferates at a fast rate. Most often, when speaking about localization, scholars in the humanities refer only to language and cultural issues. The language aspect is in general enclosed within the translation process and may be rightfully considered to be the most time consuming part within the localization process. Culture is acknowledged to play an important role as there can be significant cultural differences from one language community to the next, even when they use the same language. For example, American English and British English have both their language and culture particularities, due to different environmental, historic and economic conditions. The same applies to French. Canadian French is somewhat different from that used in France or Belgium. Customs can lead to even more particularization, as they can vary from region to region. Sprung views localization asă “theă adaptationă ofă anyă goodă oră serviceă toă aă targetă market”ă (2000: xviii). LISAă(2003)ă[9]ădefinesălocalizationăasă“theăprocessăofămodifyingăproductsăoră servicesă toă accountă foră differencesă ină distinctă markets”.ă Similară definitionsă areă provided by Esselinkă(2000)ăandăYunkeră(2003).ă“Localizationăisătheăprocessăofăadaptingăproducts that are 37 part of global distribution networks to the linguistic and cultural requirements of a given locale” (Mazură 2007).ă O’Hagană andă Ashworthă consideră ită asă “aă processă toă facilitateă globalizationăbyăaddressingălinguisticăandăculturalăbarriersăspecificăto the Receiver who does notăshareătheăsameălinguisticăandăculturalăbackgroundsăasătheăSender”ă(2002:ă66-67). Pym correctly observes that all definitions on localization refer to “products rather than texts,ăandădescribingătheăprocessăinătermsăofătheă“preparation,” “tailoring,”ăoră“adaptation”ăofă the product for a new situation. That shift is important.”ă (original emphasis, 2014:119). By acknowledging that the shift from texts to products is important, it means that the main focus of localization is on products or services. Then, it could be argued that localization is a new, major shift in Translation Studies, alongside similarly important turns such as Nida’să dynamic equivalence (1959: 19) as a basis for later modern TS theories such as the functionalist theories. (Dimitriu 2009: 23-41) However, this is not the case, as Pym does not include localization under the concept of translationă bută viceă versa:ă “Theă entireă processă isă thenă calledă “localization,”ă ofă whichă translation is a part.”ă (2014:118)ă Therefore, the focus of the definitions should be of a different nature. Thus, I propose this concept be defined as follows: localization is adapting theăcommunicationăandămarketingă strategyătoătheăcustomers’ăexpectations,ăbasedăonăaă locale mix. Unlike Mazur (2007) oră O’Hagană andă Ashworthă (2002),ă whenă discussingă localization I intentionally exclude references to global and globalization as I will show that going global is made in steps, and new markets are gradually, not simultaneously approached, most often because of financial issues or information deficiency. Covering two or three markets does not mean being global. All international brands initially emerged on the market of origin, gradually covering markets with similar cultural values. Thus, globalization tends to be achieved through contamination being determined by financial factors 5. Also, globalization is a separate, more general process that includes localization. I also propose focusing not on products but on adapting the communication and marketing strategy as the products or services are not always modified in terms of their functionality and purpose. Most often the products and services are the same so that production costs are reduced. Thus, translators and localizers do not adapt products and services but the 5 ăForăfurtherădetailsăseeăsubchapteră2.4.2ăEconomic factors involved in the localization process 38 interface of the products or services and how they are marketed/presented to the customers of the target market. As a counterexample, one might mention the modification of the product in the case of cars for instance, if we consider left side driving wheels versus right side driving wheels. As shown in subchapter 2.3.1, Internationalization, I consider this as part of the internationalization process. The very essence of a car still remains the same. Most often differences are determined by financial factors (for instance, buying power). On the other hand, within the same market, products can vary in terms of features. Cars are offered with a basic package of features and, depending on each of the customers, further features can be added. All modern products and services are modular. Adaptingătoăcustomers’ăexpectations is not a new principle. It is similar to Nida’să“focus on theăreceptors’ăresponse”ă (Dimitriu 2009:26) and to subsequent receptor-oriented theories. It also means closing the advertiser-client gap. (Lakó 2007a:244) However, by using the term customer, it is implied that they buy products or services. Theă customers’ă expectationsă areă always determined locally and globally6. The local mix is encompassed under the concept of locale. Sandriniă definesă ită asă “aă collectionă ofă peopleă whoă shareă aă language,ă writingă systemă andă anyă otheră propertiesă whichă mayă requireă aă separateă versionă ofă aă product.”ă (2005:1)ă Aă locale is determined by its geographic area often referred to as geo-targeting by the industry. While one major determinant of a locale is always the language factor, there are cases when language related localization processes are not required (i.e. Anglophone countries, or German communities from Germany, Austria and Switzerland, or Romania and Moldova). While linguists might claim that there are numerous linguistic differences, in practice, whenever possible, companies are quick to reduce costs and not consider language issues. In subchapter 4.3, Is it GILT all the way, when it comes to website localization?, I analyze the privacy page of several websites to see if the same content is used for countries that share the same language. In a more recent article, Sandrini defines a locale in more general terms “Localeă isă aă setă ofă parametersă usedă toă identifyă theă userțsă language,ă countryă andă otheră preferences”ă (2008:2).ă Ină theă sameă articleă heă admits that language changes are not always necessary for different markets. Similarly to multilingual websites, they may address the same locale. Institutional websites are the most articulate examples that confirm that websites in multiple languages do not imply localization. The EU official site is in multiple languages but it addresses all of the EU citizens, informing them about a common zonal locale. A more 6 ăForăfurtherădetailsăseeăsubchapteră2.5.10ăGlocalization or „looking in both directions” 39 specific example is that of bi- or multilingual city hall websites. Localization is not always implied in the case of websites, therefore a site may be multilingual but may not need to be localized as its audience is from the same locale. Case study http://www.tirgumures.ro is in three languages, Romanian, Hungarian, and English. Although there is a language difference, there is a common locale to a more or less extent, as the audienceăisăthatăofătheăcityăofăTg.ăMure (150,000 inhabitants). As for the usage of English on the same website, menus are in part translated but most of the content displayed is still in Romanian, the default language of the website. If the implementation of the English language had been done considering the potential target, i.e. foreign tourists, even content should have been different. The specific approach would have been to offer tourism related information. In this particular case English becomes useless. Sandriniă definesă websiteă localizationă asă „theă processă ofă modifyingă aă websiteă foră aă specific localeă accordingă toă theă goalsă outlinedă byă theă client.”ă (2005:5).ă Ină myă opinion,ă becauseă theă locale can be tracked and recorded, from a search engine optimization (SEO) perspective website localization can be further subdivided into prescriptive and descriptive localization. 2.4.1 Prescriptive versus descriptive website localization In the field of translation studies (TS) there are descriptive and prescriptive theoretical approaches to translation. As far as localization is concerned, it should benefit from both perspectives. TS is more interested in prescriptive localization, whereas the industry is more open to approach website localization from a descriptive point of view. While the prescriptiveness vs descriptiveness dichotomy in TS refers to methods of research, my thesis on website localization focuses on the customers and how they are retrieving content by using certain signifiers or keywords, thus resorting to a descriptive approach. The prescriptive approach might force onto web-users certain signifiers either by requirements imposed onto the translator by the commissioner or directly by the SEO unaware translator. 40 Case study In 2009, I was asked by a company to provide them with a list of the most used keywords on the Romanian online market regarding free games and free online games. (Lakó 2009b:762763) While free games in Romanian translates as jocuri gratuite or jocuri gratis the majority of Romanian users looked for free games using a combination of Romanian and English as a keyword: jocuri free. This means that a translator should have used jocuri free as the main keyword in the translation of an article from English into Romanian. The use of such a keyword would have been legitimate if young people had been the target audience. Redoing the same keyword research in 2013 and 2014 has shown that jocuri free decreased in Google statistics significantly. The average monthly search for jocuri gratis is over 140.000 and for jocuri gratuite is approximately 22.000, whereas jocuri free is less than 2.000. While web-users do not always adhere to linguistic norms, a company may lose potential clients if they do not make use of common signifiers. On the other hand, complying with linguistic norms is desirable; however, the reality is different and shifting7. Companies use the prescriptive approach because professional translators are educated people and use the standard institutionalized language. They also use the prescriptive approach because one of the roles of a company is to educate its customers. The descriptive approach is used by affiliate websites. They find the common signifiers and then direct users towards the official websites and thus earn money. Nowadays, numerous companies use both an official website and a blog. While official websites are translated, the localized blogs are either gist translations or content created through copywriting. Search engines may be said to use both approaches. For instance, when a keyword is misspelled Google will suggest the correct version as well. Figure 3: Google using prescriptive approach for search suggestions 7 ăSeeăalsoăsubchapteră2.6ăWebsite localization and search engine optimization 41 The prescriptive or descriptive approach can also depend on the type of the webpage of an official website. The more technical the content, the more appropriate the prescriptive approach. If the content is closer to the advertising style the descriptive approach is more appropriate. Very often prescriptive translation strategies are associated with paid translation services, whereas the descriptive approach is often correlated with volunteer translation.8 (Boéri and Maier cited in Pym, 2014:118). However, the tendency of companies is to use both descriptive and prescriptive approaches in order to more effectively cover the potential market. 2.4.2 Economic factors involved in the localization process While all the issues mentioned above are very important, there are other aspects to consider when proceeding with website localization towards various markets. Tackling various markets through localization is, yet again, a matter of management decision. Management decisions are made based on figures and cost efficiency. Companies decide on entering new markets by following certain patterns. According to Forrester Research [18:16], companies begin with their homeland. The U.S. is the most targeted market, with Canada, the UK and the other English speaking countries following naturally as there are a multitude of cultural affinities and there are no significant language barriers. In Europe, the next markets are France and Germany and the rest of the EU countries, and while there are various country specific localization issues within the EU market, the cultural affinities and the EU common market features may at times prevail over considering larger markets such as China, Japan, Russia, Brazil or India. However, in recent years, there has been an important change as to the ways in which various markets around the globe are considered. Chris Giles, economics editor at Financial Times pinpoints in his online article, China poised to pass US as world’s leading economic power this year [19], that the economies of various nations around the globe should be considered from a different perspective. The purchasing power parity or PPP concept [20] alters the managerial decision making into entering or not certain markets. Not only has China become the most important producer of goods but it is also the second largest consumer of goods and 8 ăSeeăvolunteerătranslationăinătheăMachine translation for website localizationăsubdivisionă(2.5.4) 42 services. This is also seen in the proliferation of online retail sales, placing China second after the US online market [18:12]. According to the same report the following aspects are also considered at managerial level prior to penetrating a new market [18:33]: Language Product range Currency Marketing Merchandising Analytics Payments Packaging Trustmark Delivery Fraud detection Returns Customer service All of these aspects need to be taken into account from the perspective of language, national laws, regional regulations such as the EU cookie law [21], logistics, and so on. Fraud detection, for instance, is a major factor. Countries that are deficient with regard to legislation for online criminal activities are not considered. Romania, for instance, because of numerous online frauds on eBay.com, had to wait quite a while before PayPal (paypal.com), a subsidiary of eBay, officially entered the Romanian market in 2007. eBay itself has not arrived yet. 2.4.3 The Media and the Internet in the localization mix Characteristics of the target locale include national, regional or local laws and regulations, media channels for communication (TV, Radio, Magazines, street advertising, etc.), population setup (mono-, bi- or multilingual), technological restraints, etc. For instance, mobile phones in the U.S. may malfunction on the EU market. Sockets used on mainland Europe will not comply with the requirements in the UK. The ultimate localization for content, on TV, on the Internet or in newspapers would be to offer news referring to local events but this is no longer translating but rather creating content. On TV, this is achieved by first broadcasting the national news and then by simultaneously broadcasting the local news from the local TV stations. On the Internet, the publisher of a website displays news based on 43 theă IPă (Internetă protocol)ă ofă theă user’să computer.ă Foră instance,ă ifă theă IPă addressă ofă myă computer is 82.137.13.198, the provider of the news knows that I am accessing the website fromăTârguăMure ,ăasăthisănumberăisăwithinăaăcertainărangeăofăpossible numbers dedicated to this region of Romania. Thus, it may automatically send me content and advertising targeted for this region. This same procedure is used for multilingual websites. When I access http://www.samsung.com the site will be redirected automatically to the Romanian page (http://www.samsung.com/ro/) based on my IP address compared against a range of IP addresses. Newspapers offer now hardcopy or online regional news to their users based on user choice: http://www.adevarul.ro/locale/targu-mures/, http://www.adevarul.ro/locale/iasi/ and others. These are examples of localization but clearly there is no translation involved; only customized content by means of copywriting. Furthermore, all the websites use cookies toămonitorăuser’săbehavior,ăallegedlyătoăimproveăuser’săexperience.ăAăcookie is defined by the MerriamăWebsterăDictionaryăasă“aăsmallăfileăorăpartăofăaăfileăstoredăonăaăWorldăWideăWebă user's computer, created and subsequently read by a website server, and containing personal information (e.g. a user identification code, customized preferences, or a record of pages visited)”ă[22].ăCorporateăwebsitesăprovideămoreădetailedăinformationăaboutăwebăcookies.9 2.4.4 Norms and conventions in the localization process Pym (2010:3-4) lists the technical issues involved in localization: date and time formats, calendar settings, currency formats, number formats. In the case of address formats, telephone number formats and even name formats, there should be no changes, as they might generate faulty communication. Although there are important differences, a multinational company would use either a local representative, hence a familiar format will be provided, or if, for example, there is only a single office for clients all around the world, dealing with public relations from the U.S., the U.S. address and telephone are left completely intact in order to avoid rendering the postal address unrecognizable by the U.S. postage services. Names can be considered in a similar manner. If, for instance, the name of the person who invents a product is important for marketing, the name will be kept intact, and the name of a representative will not be changed either. You could still change the name in a dubbed video advertisement and have it look quite natural in the target language if names are close in terms 9 ăForăfurtherădetailsăseeăsubchapteră2.4.5ăPersonalization 44 of sounds and length. So, the English John Johnson would translate into Romanian to Ion Ionescu but Spanish would commonly use three names, while in Hungarian the norm is surname first and first name second. As a general rule, whenever there is information about contact details, it should be kept as it is used in the country where the contact person/office resides, while when it relates to content this type of information should be localized. Converting units of measure is important as references are different especially between Great Britain and the U.S. on the one hand, and the rest of Europe on the other. However, paper sizes for printouts can now be dynamically converted depending on the IP address. For instance, an online drawing application for children can be set up in such a manner that it tracks if the user is from the European mainland, or from the U.K., the U.S. or Australia. In this case, it may automatically set the layout of the drawing to be printed to fit to A4 paper size or to letter size. As for color conventions, they may have cultural connotations but on the web the colors are usually related to content. Red may mean danger in European cultures but it could also be appear on sites that sell beauty products or on adult sites. Iconic conventions are becoming increasingly globalized, and dominant cultures will impose their iconic system onto other cultures especially by exporting goods that would be quite costly to be localized (See previous keyboard and gadgets localization examples). On the Internet a house icon would be understandable for any user as referring to the main page of a website, or an envelope would lead to a contact form or a @ symbol suggests emailing; these would obviously be understandable to any computer user around the world. Icons would be the ideal choice as a convergent localization method for navigation menus. Media files would need to be separated from sound, images and text, thus opening the way for their recombination by user preference or automated by the application that delivers the media files. As with films on DVD, online video content may allow changing audio language (if available) and text language; it should also provide more control over viewing or skipping certain scenes from an online movie file. Some website already use advanced media players that use cue points, which show a representative preview with a static image of the scene plus, at times, text information. 45 Research on norms has been quite dynamic in Translation Studies, the classifications Gideon Toury, Andrew Chesterman and Christiane Nord being among the most popular. Gideon Toury distinguishes three types of norms (in Dimitriu 2006:49-50): - initială normsă (theă translatoră decidesă toă adhereă toă theă rulesă ofă theă STă –ă calledă adequate translation,ă oră theă rulesă ofă TTă –ă calledă acceptable translation).ă Asă Dimitriuă notices,ă actuallyă translatorsă cannotă optă foră oneă ofă theă alternativesă only,ă theyă ultimatelyă adoptă aă mixedă approachă ină whichă oneă orientationă prevails.ă Duringă theă localizationă processă theă normsă shouldă followă toă aă greată extentă theă normsă ofă theă TTă andă TC.ă (Seeă Pymă 2010)ă However,ădependingăonătheătypeăofăSTăwithinăaăcertainăwebpage,ănormsăfromătheăSTăcană oftenătrespassăintoătheăTTătoăvariousăextents. - preliminaryănorms:ăTouryădiscussesăaboutătheăacceptanceăorărejectionăofăcertainăSTsăandă SCsă (Foră instance,ă afteră 1989ă thereă hasă beenă aă strongă generală anti-Russiană trendă acrossă EasternăEurope10)ăonătheăoneăhand,ăandătheăusageăofăaă“third-party”ălanguageătoătranslateă fromălanguageăAăintoălanguageăB,ăonătheăother.ăInăwebsiteălocalizationăandăspecificallyăină theăfieldăofăe-commerceătheăacceptanceăorărejectionăratherădependsăonăpolicies.ăChina,ăforă instance,ă hasă blockedă severală multinatională websites.11ăAsă foră theă usageă ofă intermediaryă languages,ătheălocalizationăindustryămayăoftenăresortătoăwhatăMTăterminologyărefersătoăasă pivotălanguage.ăSmallăcompaniesăwithălimitedăbudgetsămayăoptăforăsuchăanăapproach,ăevenă ifăforătheăTCăităisăacceptableăorănot,ăasăcompaniesăattemptătoălimitătheirăexpensesăasămuchă asă possible.ă GoogleăTranslate,ă evenă ifă ită isă statisticală MTă softwareă isă saidă toă optă foră thisă approach,ă especiallyă whenă thereă isă insufficientă parallelă corpusă betweenă twoă specifică languages. - operatională normsă referă toă theă actuală translationă process.ă Theseă areă furtheră subdividedă intoămatricialănormsăandătextual-linguisticănorms.ăMatricialănormsăinăwebsiteălocalizationă mayăbeăsetăduringătheăinternationalizationăstageăandăreferătoătheăstructureăofătheămenusăină theăSTăversusăTT,ăpredilectionăforălengthyăparagraphsăversusăparagraphsămadeăupăofăshortă clearăsentences,ăpageălayoută preferences,ăetc.ă Textual-linguisticănormsăshouldă adhereătoă theăTTănormsăandăthusătheăoutputăisăanăacceptable translation.ă(seeăaboveă–ăinitialănorms)ă Theă implementationă ofă theseă normsă dependsă onă theă knowledgeă andă experienceă ofă theă translator,ă asă theseă normsă operateă ată sentenceă level.ă Theyă cană beă appliedă ină aă similară mannerăbothăinătheăcaseăofătraditionalătextsăandăe-texts. ăHowever,ăthereăareăstudiesăthatăshowătheăcontrary;ăforăinstance,ătransalationsăofăDovstoievskiăintoăRomaniană increasedăăfromă1996ătoă2002ă–ăMagdaăJeanrenaudă(2006)ăUniversaliile traduceriiăIa i:ăPolirom,ăp.ă211 11 ăhttp://en.wikipedia.org/wiki/Websites_blocked_in_China 10 46 Another useful distinction Toury makes is that between textual and extratextual norms. Textual norms asă “regularities of behavior”ă (citedă ină Dimitriuă 2006:50) in the case of website localization can be determined statistically. One example, even if from the field of IT&C, is the appearance of HTML5 as a web standard. The generic <div> tags from HTML4 have been replaced by more specific, semantic tags, such as <article>, <aside>, <details>, <figure>, <footer>, <header>, <nav>, etc. The choice of these specific terms is based on collected data from millions of websites. Thus, W3C (The World Wide Web Consortium http://www.w3.org) suggested using simpler and more straightforward coding syntax. For example, <divă class=ă “footer”></div>ă was simplified to <footer></footer>. Extratextual norms in the case of website localization refer to the norms imposed on the outcome by the commissioner, project manager, team managers, QA testers, translators. One important human factor in determining these norms in the case of my approach to website localization (keyword oriented) is the end-user and their input in search engines. Chestermană furtheră elaboratesă onă Toury’să normsă andă distinguishesă betweenă product/expectancy norms and production/professional norms. (in Dimitriu 2006:52). Expectancy norms, as the term suggests, are determined by the (web content) readers. This type of norms is important for the user-oriented approaches in TS. In the case of websites the expectations of the web-users can be determined statistically by using specific tools. By intentionally using in their translations keywords used by web-users, SEO aware translators fault prescriptive norms of norm-authorities.12 Also, expectancy norms can be correlated with what I called the anticipation layer as a layer of text.13 Professional norms control the translation process. These norms are further subdivided into accountability norm, the communication norm and the relation norm. I consider that these norms apply to website localization as well and there is nothing specific that can be added. Nord examines norms from a functionalist perspective and differentiates between constitutive and regulatory conventions (in Dimitriu 2006:55). Constitutive conventions concentrate on the output and on the extent to which a text can be considered a translation by the target reader. However, for the purpose of my thesis I consider thatăNord’săfocusăonătheăreaders is of more interest: “what the users of translations expect from a text which is pragmatically 12 13 ăSeeătheăexamplesăină2.4.1 ăSeeă3.4.7ăAnticipation layer 47 markedă asă aă translation”ă (citedă ină Dimitriuă 2006:55).ă Thisă corroboratesă Chesterman’să expectancy norms, but the idea of a text marked as a translation is, to a certain extent, not in accordanceăwithăNord’săviewăonăinstrumentalătranslationă(1997:81);ătheăTT outcome should be independent of the ST. 2.4.5 Personalization Schäler (cited in Baker, 2009: 162) observes that a specialized type of localization is personalization. Applications allow users to choose from a range of options. For instance, Yahoo Mail allows you to change the lookout of the email service (colors, fonts). The Firefox Internet browser allows changing its background image. Some sites allow you to modify the font size usually between small, normal and large, making the text readable for senior citizens as well. Text size can be changed by users as well without a special set of buttons on the site. However, most of them are not aware that for most websites they can modify the font size themselves by pressing the CTRL key on the keyboard and rotating the mouse wheel. Other applications, like Adobe Flash, let the user modify the layout of the menus and toolbox windows (re-placing menus, minimizing or expending certain sub-menus); the user can opt for a designer or a programmer layout, or s/he can set his/her own layout. This type of localization is called by Schäler (ibid.) development localization. This does not necessarily involve translation, as the designer layout and programmer layout could be used by a designer and a programmer working in the same office and using the same native language. On the other hand, software localization which includes translation as well is a standard these days. In recent years, numerous software applications have been moved online. Many website building applications are used online to set up and maintain a website. One of the most popular such applications is WordPress. Upon installation WordPress detects the location of my server, based on its IP (Internet Protocol Address) and sets the language of the administration pages to Romanian. This is desirable as long as you are not already accustomed with similar systems that you used in English. It is similar to using Microsoft Office for several years with the menus and help files in English and then coming across a computer that uses the Romanian version of the same software. English terms, in such cases, are already used as specialized terms in the field of text editing or programming and changing 48 it to theă user’s native language would mean relearning the specialized terms. The usage of native language is beneficial only in the case of new users of such web software. On the other hand, if you need support and you look for information using the translated terminology you might not find the answers to your questions as most probably the help and support pages have not yet been translated into your language. This is most often the case with free and open source software. Furthermore, apart from using the core WordPress functionalities, a programmer may choose to install various modules to run on his/her website. While all modules use English by default, they may be developed by any developer from around the world. Thus, most likely there will be no translation and localization in theă user’s own language. Nowadays widgets are also very common. They offer various types of information (content). For instance, a user can download for his/her desktop from his/her Microsoft Windows operation system weather, calendar or currency fluctuation widgets and have them display useful information or s/he can remove them altogether. Some websites, like emag.ro use cookies (bits of code) to "memorize" the products a user looks at and then, when s/he returns to the website, a range of similar products are displayed at the top of the page. However, there are several types of cookies. If, for instance, we study the information on cookies provided on the Apple website, they categorize cookies into Strictly Necessary Cookies – essential for the core functionalities of the website or for ecommerce features, Performance Cookies – used for optimized usage, Functionality Cookies – especially set up by the user, like font size, background, etc. [23]. Other websites use a similar description for the usage of cookies. 2.4.6 Culturability Many of the above mentioned aspects related to localization, especially to website localization are referred to as culturability ină someăacademicăpapers:ă“<<culturability>> is the merging of culture and usability and represents a relationship between design elements andăculture” (Barber and Badre (2001) cited in Cyr 2003:4). 49 In a similar manner, Nielson (cited in Pym, 2010: 7) enumerated in 1994 the features to be considered when considering website usability (often referred to, by website or application developers, as user-friendly design and content):  Learnability seen as easiness of usage of the design the first time. Design should be simple and intuitive. Iconic menu buttons are such examples. Also, tooltips can be added so that the user can quickly discover the information about the tool. Figure 4:Showing tooltip In this particular application (developed by me in Adobe Flash), when holding the mouse motionless for two seconds over a button a very short description is displayed (http://scienceheroes.com)  Efficiency refers to how fast a task can be accomplished with minimum of time and effort. On a website or in an application, testers often test for the number of clicks required to achieve a certain task, for instance, accessing a certain piece of information on a site. Google recommends clicking no more than twice for reaching any page of a website. The combination of learnability and efficiency is called accessibility by some publishers  Memorability refers to the ability of a user to recall the usability of a design after a longer period of non-usage. I consider this is not relevant for the fast moving economy of computer technologies, as technologies are continuously evolving, always adding more and improved features for the users. For instance, the design of Microsoft Office Word 2007 was considered by many to have been quite radically changed, as compared to its predecessor, Word 2003. Even if in the 2007 version menus were organized in a significantly superior manner, for those used to the 2003 version it was rather difficult to familiarize themselves with the new interface. Consequently, it is important to avoid steep 50 learning curves. Intermediary versions would be useful but would add to the development costs, so, they would not be economically viable. Figure 5:Office Word 2007 vs. Office Word 2003 By errors, Nielson refers to the number of errors and their severity. This aspect has become somewhat unnecessary nowadays. Through careful planning and design most potential errors can actually be excluded. I suggest guiding the user by restricting his/her input and actions. For instance, in the case of sign up forms a step-by-step, subsequent order for input is often used. The images below are an illustration of such designs, showing a directed-user approach; it is an application I developed for http://www.web-study.org/. The user can input only what the application allows and only when it allows it. Step1 in a directed-user approach Step 2 in a directed user approach Step 3 in a directed-user approach Figure 6: Example of directed-user approach 51  According to Nielson, satisfaction refers to the degree of fulfilment when using a design. Based on the explanations and examples above I consider that satisfaction in usability actually depends on all the features previously mentioned. It is a sum up of learnability, efficiency, memorability, and error-free functionality of the design. All these usability features need to be considered in the process of localization in general and personalization in particular. Consequently, I would add personalization as well to Richard Sikes’ă graphică belowă thată showsă theă interconnectionă betweenă globalization,ă internationalization and localization (from the Multilingual.com guide: Localization – 2009 issue [24:4]) Figure 7: Personalization as a custom type of localization The two graphics show the interdependence of the four processes. Globalization is the building block, on which all the other processes are based. The closer to the tip of the pyramid, the more specialized and specific the processes and skills involved. I added personalization at the top as it is a specific type of localization. Personalization is one step closer in achieving website localization targeting the end-user. Businessdictionary.com defines (web) personalization asă “Tailoringă theă presentationă ofă a website's content to match a specific user's instructions or preferences. This custom tailoring is accomplished either by the user choosing from a menu of available alternatives or by tracking his or her behavior (such as which pages are accessed and how often)ăonătheăsite.”ă [25] Many are already offering this type of personalization either through navigational options or through the use of cookies, as seen in previous pages. 52 Figure 8: Stratifications of GILT processes The above figure represents the layered visualization of the GILT processes; I included Copywriting next to Translation as an alternative to the latter. They can be exclusive (one approach or the other – more details in the subchapter 2.5.1 Translation, adaptation, copywriting) or simultaneous, to various degrees. Personalization, at the very center, is very important within the current inbound marketing strategy trend - users look for services and products when they choose to do so. Many website owners consider that including the name of an authenticated user at the top of the page is personalization. It is, but at its most rudimentary and unprofessional level. The team at Oracle clearly delimits how personalization should be implemented. They define personalizationă asă “deliveringă theă rightă contentă toă theă rightă personă ată theă rightă time,ă acrossă interactions,ă throughoută theă fullă customeră lifecycle.”[26]ă Theă right person should not be taken as an individual but as a broader group of people sharing common interests, so the merchandiser should target a niche market. This means segmenting the market. Pym also acknowledgedăpersonalizationăasă“adding locale-specific content”.ă(2010:4)ă Localizing with added personalization is the most appropriate manner to reach global customers, in other words to focus on the end-users and their needs. 53 2.4.7 Grice’s Maxims applied to localization Search engines14 are the medium through which potential buyers on the one hand, and companies, on the other, are communicating. Over the years, Google has improved this communication medium, so that it offers the optimal communication outcome between the two sides. Thus, whenever a user looks for a certain product or service, Google displays results from pages created by various companies. Listed webpages contain the keywords entered by the search engine user. For optimal results, Google filters and orders the results based on some parameters. Some of these parameters, of a linguistic nature, are in accordance with PaulăGrice’sătheory.ăGoogleăhasăevenăofferedăseveralăguidesă[28][29]ăonăhowăcompaniesă shouldăcreateătheirăcontentăsoăthatăpotentialăbuyers’ăexpectationsăareămetăon the webpages of the companies, via the Google search engine. Grice states (1975) that during a communication process, those involved have certain expectations, which he calls conversational maxims. Also, any type of communication is based on mutual collaboration, which he calls the Cooperative Principle. The cooperative principleărefersătoămakingă“yourăconversationalăcontributionăwhatăisărequired,ăatătheăstageăată which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged.”ă(45) The Gricean maxims derived from the above statement are as follows (1975:45-46): The maxim of quantity: - Makeăyourăcontributionăasăinformativeăasăisărequired.ă - Doănotămakeăyourăcontributionămoreăinformativeăthanăisărequired.ă The maxim of quality: - Doănotăsayăwhatăyouăbelieveătoăbeăfalse.ă - Doănotăsayăthatăforăwhichăyouălackăadequateăevidence.ă The maxim of relevance: Be relevant. The maxim of Manner: - Avoidăobscurityăofăexpression. - Avoidăambiguity.ă ăIăwillăreferăespeciallyătoăGoogleăasăităisăusedăbyătheămajorityăofătheătotală Internetăusersă(65.2șăinăDecemberă 2012ăaccordingătoăcomScoreăcitedăbyăsearchengineland.comă[27] 14 54 - Beăbrief.ă - Beăorderly.ă If we assess bothăGriceanămaximsăandăGoogle’săWebmasterăGuidelinesă[28]ăthereăareămanyă similarities. Thus the maxim of quantity translates in Google terms into writing content of at least 300 words. This number is mentioned by certain SEO (search engine optimization) plug-in (software component), for instance the WordPress compatible Yoast plug-in. However, at least in the English speaking online world, this number has increased to over 1000 words, while some SEO companies claim that the most appropriate length is 2000 words. These figures are based on SERPs (search engine results pages) that position webpagesăthatăcontainăarticlesăofăaăcertainălengthăonătheătopăpositions.ăUltimately,ăGoogle’să response to lengthier articles is based on tracking web-users’ă behavior.ă Consequently, for instance the time spent on a page can be a matter of quantity (and quality) of information found on the page as related to what the user looked for on the webpage and where in the SERP a result leading to a webpage is located. Case study The purpose of this case study is to compare and determine the correlation between text length and positioning of webpages in SERPs. Therefore, I analyzed the top ten positions for the keyword green tea benefits for Romanian, British English, American English, German and French, on each of the corresponding localized version of Google. Min. Max. Avg. length length length* beneficii ceai verde 246 2318 993 Google.co.uk green tea benefits 339 3193 1065 Google.com green tea benefits 162 3193 1034 Language Search engine Keyword 1. Romanian Google.ro 2. British English 3. American English 4. German Google.de grüner tee nutzen 315 6365 1740 5. French Google.fr thé vert bienfaits 194 5514 990 Table 1: Analysis of article length in SERPs * Calculation based on the top ten results in the search engine results page The table above illustrates that the average length, is around 1000 words. For German the average is significantly higher. For Romanian, British and American English, the length of 55 the articles on the first position is around 2000 words. Google.fr and Google.de display on the top position articles of 309 and 6365 words, respectively15. At the same time, the top ten results include articles of fewer than 200 words (American English and French). Comparing the three numeric columns French shows the highest contrast as compared to the average article length. These discrepancies are determined by how the publisher decides to publish their content, either divided up in several smaller parts, or displayed in a compact manner. Other publishers display on the same webpage several related articles written at various times yet grouped under the same topic. With regard to British and American English, the results of the case study demonstrate that Google displays results based on user location. However, 7 out of 10 results are similar, yet positioned differently. Thus, Google displays the information in a localized manner. In summary, the case study demonstrates that currently the average article length is 1000 words. By extending the range of the websites analyzed for a certain industry, more relevant data can be obtained. Similarly, a case study to determine the current most efficient article length could be based on analyzing the articles displayed on the first place for several top internet industries. The trend is to provide users with lengthier and more elaborate articles. The purpose of this subchapter does not include discussing all the aspects regarding webpage ranking in SERPs as it is a complex issue requiring separate consideration. Also, the content creator can use tools such as Google Analytics to verify if the content created was of interest or not to web-readers, and offer sufficient useful information, but not overwhelming them with excessive data. Extra information should be always provided on request, that is, in terms of websites, allowing users to click on links, thus complying with the maxim of quantity. Considering this maxim from the perspective of localization, I have stated above that the length of the text can be specific to various target audiences, but tends to be dictated by the search engine operating on a certain national market and on the maturity of the Internet in terms of content in that particular language. Therefore, if the source language is a major language, like English, it is convenient to maintain the same length of texts. On the other hand, if the direction of localization is from a minor to a major language/culture, for instance Romanian to English, if the Romanian text is only 500 words long, the translator into English ăTheăresultsăcorroborate,ătoăsomeăextent,ăHall’săLowăcontextăversusăHighăContextăculturesătheoryă(1976:ă105116) 15 56 should be required to extend the translation to a 1000 words long text (.com domains as compared to .ro domains are highly competitive and keyword research and content optimization needs to be more precise). Doubling the length of TT implies from the part of the translator creative writing and copywriting skills. Also, it is important to mention that whereas search engines crawl and examines the entire page, the web-user will not read the full text but only scan it. According to Nielsen reports [30] only 20% of the content is read on a page. For instance, out of 300 words, only 60 will actually reach the user. Apparently, it is a waste of resources, but increasing content both for the search engine and the users means higher conversion potential. More important in this case is the maxim of manner as will be shown below. The device on which content is accessed is also important, as the screen size varies. Internet research companies [30] [31] state that 80% of the time on a page is dedicated to content above the fold, whereas only 20% of the time is spent for content under the fold.16 So, articles should not be longer than what the fold allows. While this is doable on desktop screens, mobile screens will involve scrolling. Anyway, the essence of the articles should be positioned above the fold and on the left side of the screen. The maxim of quality in website content terms means not deceiving users, making webpages primarily for them and not for search engines, avoiding pages with irrelevant keywords, adding sufficient value to webpages, avoiding making claims not supported with evidence, avoiding false claims and so on. In Google terms: - “Createă aă useful,ă information-richă site,ă andă writeă pagesă thată clearlyă andă accuratelyă describeăyourăcontent. - Thinkăaboutătheăwordsăusersăwouldătypeătoăfindăyourăpages,ăandămakeăsureăthatăyoură siteăactuallyăincludesăthoseăwordsăwithinăit.”ă[28] Any website should present honest data about its services or products. Referring to content marketing and localization a website should add new information or a different perspective or add value on a certain topic and stand out from the competition in a certain field while transferring meaning of the source content into the target language – culture pair. Not only ăTermătakenăfromătheăprintedămediaărefferingătoinformationăfoundăaboveătheălineăfoldăresultedăfromăfoldingăaă newspaper.ăOnline,ăitărefersătoăwhatăisăvisibleăwithoutăscrollingătheăpage . 16 57 can this maxim be applied to localization consistently, but through the localization process of some multinational websites the local business community can benefit from the value of leading practices from the markets of their origin. For instance, in the online area, Google itself, while currently not performing so well in terms of search engine in Romania as it does for the .com domains, it certainly influences in a positive manner the practices of the Romanian market. The maxim of relevance. While this is very much dealt with already at search engine level, there may be cases when through spamming and other malicious strategies, users might access certain sites that are irrelevant to their search. Content scraping (taken from other websites) is a technique often used but currently penalized by search engines. The usage of such techniques depends on the maturity of the market in which the search engine operates. For instance, the Romanian Internet content is still at an incipient level, and in numerous fields of activity there is a content deficiency.ăAsăaăresult,ămanyă“blackăhatăSEO”ătacticsăcană be employed to achieve high ranking in SERPs. Relevance issues can at times arise from language specific issues such as homonymy and polysemy. In the case of homonyms, for some languages, such as English, there is a further differentiation between homophones and homographs. With the latest searching technology by voice input, both can cause problems. If we look for an image and type in the search engineă underă theă imagesă tabă “bat”ă Googleă willă notă knowă ifă weă referă toă theă animală oră theă object.ăTheăresultăwillăbeădisplayedăbasedăonătheăusers’ătrackingăhistory,ăandăasămostăofăusersă had previously looked for the animal, the great majority of images will display images of these animals. The user can obtain more specific results if longer 2 or 3-word-long expressions are provided17. 17 See subchapter 2.6.4 Web content translation and search engine optimization (SEO) 58 In terms of localization, websites must create relevant information content, which isă based,ă however,ă onă theă users’ă queriesă ină the search engines. Irrelevant advertising is still abundant in the Romanian online media. This is due either to absence of knowledge or to misleading users on purpose. For instance, looking for a specific bicycle for kids, using the keyword bicicleta - 123 GLN the SERP displays both organic results (results based on the free submission and listing of websites) and paid resultsă(“sponsored”).ă Figure 9: Irrelevant sponsored results Whereas the webpages I was directed to when clicking on the results from the organic results displayed the bicycle I was looking for, the paid results took me to webpages that contained absolutely no information on children bicycles. Also, the description of the sponsored results did not contain my keyword. Furthermore, it was unexpected to notice that compari.ro showed up in both the organic and paid lists, yet only the organic result was appropriate for my search. This example is relevant for the process of localization in terms of the keywords a translator is supposed to use in order to get content to be displayed in SERPs18. The maxim of Manner and its four requirements (avoid obscurity of expression, avoid ambiguity, be brief, be orderly) is a matter of selling principles, making the potential buyer understand what the website has to offer in a manner which makes it clear that content is built for the target audience. These principles can be applied directly to the localized text but also to the process of internationalization as already shown in subchapter 2.3 Internationalization. For instance, when applied to the localization process, the localizer or translator needs to knowăthatăsomeălanguages/culturesădoănotăfollowăNielsen’săF-pattern (2006) but a reversed Fpattern. Languages such as Arabic and Hebrew are read from right to left, while Chinese is 18 More information on these issues in subchapter 2.6.4, Web content translation and search engine optimization (SEO). 59 read from top to bottom but starting with the right-hand column. Manner can also be reflected in the style of the text as discourse19. These maxims need to be followed exactly in content marketing, if a website is to be successful with its content marketing campaign. However they are often faulted in online advertising that is closer to traditional advertising such as banner advertising or video advertising. 2.4.8 Content marketing – delimitations in the sphere of web localization “Contentă isă King”ă areă presumablyă theă wordsă ofă Billă Gates,ă whoă backă ină 1996, correctly predicted that whoever provides content can make money. [32] While content can be any form of information or entertainment my thesis focuses on text content. Why focus on content marketing? There are several objective reasons for analyzing content marketing and how it relates to localization and translation. Contentă Marketingă Instituteă [33]ă definesă contentă marketingă asă “theă marketingă techniqueă ofă creating and distributing relevant and valuable content to attract, acquire, and engage a clearly defined and understood target audience—with the objective of driving profitable customeră action”. Nielsen and Tahir also acknowledge text and text production as determining factors of a website. (in Jiménez-Crespo 2013: 37) Several recent studies from the marketing industry show that potential clients no longer tolerate interruptions from advertisers, or if companies insist on disturbing (intrusive marketing) they will be ignored. For instance, many reported that while advertisements were displayed on TV (15 minutes on average – now regulated by the EU law to up to 12 minutes within the span of an hour) viewers chose to attend to various types of housework: like watering flowers, making a sandwich and similar other activities. The equivalent of this on the Internet is web advertising affected by banner blindness – web-users ignore completely advertisements displayed (Nielsen 2007). This phenomenon may be the result of theă“phobia of advertisements among web-surfers”ă phenomenon. (Lakó 2007a:242) Marketo.com, MarketingSherpa and other companies have shown that the manner by which marketing 19 ăMoreăinformationăinăsubchapteră3.2ăText and E-text. 60 functions nowadays has changed dramatically, from finding customers to being found, that is moving from intrusive to unobtrusive marketing (called permission marketing). Through all the Internet channels available nowadays, there is a massive abundance of information, as compared to TV and printed media a few decades back. Unobtrusive marketing is also called by the industry inbound marketing (or non-intrusive advertising, Lakó 2007a) and ensures potential buyers discover services and products they look for by using various channels. Fishkin and Høgenhaven speak about inbound marketing as “organic, earned marketing, for which you do not pay, and is the sum up of all the channels that direct web-users to a website. It is to <<earn traffic and attention>> without directly spending money” (2013:3). There is not a finite list of what inbound marketing means but most people from the industry agree on listing the following: Branding, Content Marketing, Search Engine Optimization, Social Media Marketing, Video Marketing, Email marketing. All of these do require some kind of content creation. However, branding is no longer so important. According to marketingcharts.com the perception of quality based on brand has diminished from 43% in 2010 to 29% in 2013 [34]. This change has been determined by the results on SERPs. According to the same website a study conducted by Google and Ipsos MediaCT companies that were positioned on top positions were more highly appreciated by web-searchers [35]. Brand success is linked by web-users with positioning in SERPs. Top positions are considered to be the highest performing brands in a particular field. From the listed channels of inbound marketing, branding is not an option as it is costly and it is the result of years of traditional intrusive advertising. It is appropriate for multinationals with large budgets for branding marketing. For the small and medium companies Content Marketing, Search Engine Optimization, Social Media Marketing, Video Marketing, Email marketing are the most lucrative. However, the list can be further reduced. Video marketing requires technology and specific skills and most often cannot be achieved by the companies themselves. It requires outsourcing, hence further costs. As for email marketing, you need to have a user pool to which to send out emails. Obtaining lists of email addresses can be achieved either by convincing people to subscribe or through acquisition. The latter is often considered inefficient as recipients did not sign up themselves, so, it is indirectly intrusive. Social Media Marketing is useful but studies show that Likes do not always imply clicking and reading the articles posted. My personal experience with social media shows that friends and acquaintances of authors, like articles without even bothering to read them. Advertising 61 on Facebook is seemingly not always intrusive as advertisements are intercalated among friends’ăposts. To trigger higher user response, at the very top of the advertisements Facebook also displays a list of friends that had already liked a certain ad. However, if users are not specifically looking for that product or service, they become as immune to the inline advertising as to the advertisements on the right. Figure 10: Facebook inline advertising vs. right hand intrusive rotating banner This demonstrates that creating content (primarily visual) for social media marketing purposes is similar to advertising on TV; it is not targeted and contextualized to what the user is looking for on a socializing platform. MarketingSherpa [36:2] found that Social media marketing is only 20% in terms of marketing effectiveness as compared to 60% in the case of website content marketing. There are also several other reports that show which inbound channels are the most successful ones. Vocus [37:2] in its research report The State of The Marketing 2014 shows which are the most lucrative distribution channels. 62 Figure 11: The most popular channels The website approach is the most lucrative, according to this graphic. Advocate of the same findings is MarketingSherpa, as previously mentioned, which has also released several research reports on this topic. Figure 12: Channels of marketing. Source MarketingSherpa [36] 63 Considering the above graphic which represents the three assessment factors selected by MarketingSherpa, i.e. level of effectiveness, degree of difficulty and level of usage, webpages (websites) are second-best in terms of effectiveness, on the lower end of the difficulty axis (40% to 55% - less effort required to achieve it), and it is the most used method in terms of usage (largest sphere size). While multinationals may opt for all the inbound channels when localizing their services, smaller companies will most likely commence with only website content localization. Consequently, I will focus on websites content marketing. On the other hand, content marketing does not refer only to text, it refers to any type of content found on a website: text, images, graphs, video, sound and even style and layout. However, the main focus of the thesis is text content for several reasons: - “Linguistică communicationă isă theă strongestă possibleă formă ofă communication:ă ită introducesă ană elementă ofă explicitnessă whereă nonverbală communicationă cană neveră beă moreăthanăimplicit”ă(SperberăandăWilsonăcitedăinăDimitriuă2002:44) - textă contentă offersă contextă soă ită isă easieră toă beă reachedă byă usersă onă searchă enginesă irrespectiveăofătheăwebsite,ălocalizedăorăstandard - accordingă toăAlexa.comă (provideră ofă commercială webă traffică data)ă oută ofă theă topă 25ă mostăusedăwebsitesăatăglobalălevelă-ăwithăGoogleăatătheăveryătop,ă13ăareăInternetăsearchă related.ăThisăshowsăandăconfirmsăMarketingSherpa’săoutcomeăthatătheămostăeffectiveă methodă ofă inboundă marketingă isă throughă webă textă content,ă searchableă byă searchă engines. - anotherăreasonăisăthatăwhereasăonlineăadvertisingăhasăaălimitedălifeăcycle,ăoriginalăandă valuableăcontentă remainsărelevantă forălongerăperiodsăifăoptimizedăappropriatelyăwithă theăsuitableăkeywordsăforăeachălocalizedăwebsite.ă - consideringăaăcustomer’sălifecycleă(reach,ăacquisition,ăconversion,ăretention,ăloyalty)ă advertisingă cană possiblyă achieveă onlyă theă firstă twoă stages,ă whereasă aă well-writtenă informativeăandăeducativeăcontent,ăasălongăasăităisădisplayedăonăSERPs,ăcanătakeătheă userăthroughăallătheăstageăofătheăcycle.ă - bothăB2CăandăB2Bătypesăofăcommerceăexchangeăuseăwebsitesăasăprimaryăinformationă source,ă regardlessă ofă theă firstă contactă channelă (reachă stage):ă socială media,ă searchă engines,ăvideoăchannels,ăfriendăsuggestion,ăetc. More information on website content marketing and content production from the perspective of translation and localization can be found in the subchapters devoted to Translation (2.5) and Text and E-text (3.2). 64 2.4.9 Classification of localized websites Over the last decade there have been several attempts to classify the localization process according to various criteria. Some of the criteria are more general and are based on linguistic factors, but a classification can be built from a marketing perspective as well or from the point of view of the technical factors involved. 2.4.9.1 Classification by language Theămostăobviousăclassificationăcriterionăisăaccordingătoătheănumberăofălanguagesăused.ăFromă thisăperspective,ălocalizedăwebsitesămayăbe: - monolingual,ăwhenăonlyăoneălanguageăisăused;ăthisătypeăofălocalizedăwebsiteăisăveryă rareă nowadays,ă beingă encounteredă withă websitesă withă globală marketingă tendencies;ă theirălanguageăofăchoiceăisăusuallyăEnglish; - bilingual:ă foră theseă websites,ă theă useră cană chooseă betweenă twoă languages,ă usuallyă Englishăandătheălanguageăofăanotherăaffluentănation; - multilingual:ă threeă oră moreă languagesă areă usedă onă theă sameă website;ă Englishă asă theă firstălanguageăplusăseveralăotherălanguages. The constant element in this classification is the English language due to its international status (Crystal, 2006). However, English is not always the first choice, especially when a business originates from a non-English speaking country and initially targets only the local community. However, when going global the first choice is English. The choice of subsequent languages is most often "imposed on" by the number of users from a certain language community and the GDP (Gross Domestic Product) of the target area; more recently it has been based on the PPP (Purchasing power parity) [20]. For instance, one can review the evolution of a website by using http://wayback.archive.org, a service that records on a regular basis snapshot of public websites (it can be used as a tool for inspecting the localization history of any website). By investigating the history of payoneer.com one can see that initially, back in June 2004, the language of the site was Vietnamese and even the content offered was of a different type. In September, the same year, it was used as a link farm20 and the language used was already English. In June 2005 the site began to offer the services that it 20 ăAălinkăfarmăisăaălistăofălinksătoăotherăpages. 65 still offers today – payment services. The second language used for this banking service was Chinese. Later, Hungarian, Russian and Spanish were introduced, probably according to the countries to which most U.S. employers outsourced at that time, also forecasting the role China would have in the future of the global economy. Currently, even more languages are available. Figure 13: Language options 2.4.9.2 Classification by region Technicallyăreferredătoăasăgeo-targeting,ălocalizationăcanăbeăfurtherăclassifiedăbyăcountry,ăbyă statesă withină aă country,ă oră countiesă oră evenă byă majoră cities.ă Thisă typeă ofă localizationă isă especiallyăusedăinăonlineăadvertising.ă Ină addition,ă stillă somewhată relatedă toă regions,ă Iă suggestă furtheră classifyingă localizationă byă culturală setsă ofă values:ă West-European,ă East-European,ă Chinese,ă Arab,ă etc.;ă byă socioeconomicălevel:ăaffluentăandăemergingăcountries;ăbyăeducationă(contentăwouldăbeădifferentăforă theăgeneralăpublicăasăcomparedătoăexperts);ăbyăuserăinterestsă(traceableăbyătheăwebsitesăbeingă visitedă andă byă searchă engines);ă byă levelă ofă expertiseă betweenă theă agentsă involved:ă B2Că (businessă toă customer),ă B2Bă (businessă toă business),ă C2Că (customeră toă customeră suchă asă ebay.comă oră olx.ro)ă asă theă communicationă style,ă foră instanceă wouldă beă differentă ifă eachă ofă theseă variablesă isă takenă intoă consideration;ă byă ageă (e.g.ă contentă foră childrenă wouldă beă veryă differentă fromă contentăforăteenagers).ă Consequently,ăsimilarlyătoămovies,ăină theăfutureăthereă willă beă aă recommendationă fromă theă publisheră asă toă theă ageă groupă theă websiteă contentă addressesă (Googleă isă alreadyă filteringă oută adultă contentă byă default).ă Thisă typeă ofă categorizationă foră localizationă isă usuallyă connectedă withă marketingă researchă andă marketingă decisions.ă 66 2.4.9.3 Classification by recipient profile Whereasăinătheoryătheăcategoriesăsuggestedăunderă2.4.9.2ăareăpossible,ăinăpracticeătheyămakeă upătheărecipient’săprofile.ăWhenăităcomesătoăwebsites,ătheărecipient’săprofileăisădeterminedăbyă previouslyă accumulatedă mentală baggageă (Nielsenă andă Tahiră citedă ină Jiménez-Crespoă 2013:ă 38),ă oftenăreferredătoă asă backgroundăknowledgeă contextă ină linguisticăpragmatics.ă InăTS,ă theă firstătheoristătoămoveătheătranslationăprocessătowardsătheăreaderăwasăEugeneăNida.ăHeăcalledă thisătypeăofătranslationădynamicăequivalenceăandălaterăfunctionalăequivalence.ăNidaă(2000:ă 128)ă analyzesă theă recipients’ă profileă fromă twoă perspectives:ă theiră decodingă abilityă andă theiră potentialăinterest.ăAccordingătoătheădecodingăabilityăheădeterminesăfourămainălevels:ă - theăcapacityăofăchildren; - theădouble-standardăcapacityăofănewăliterates; - theăcapacityăofătheăaverageăliterateăadult; - theăunusuallyăhighăcapacityăofăspecialists. InătheăproposedăSEOăawareăwebsiteălocalizationătheădecodingăabilityăofătheărecipientăcanăbeă determinedăbyătrackingăuserăbehavioră(usingăspecificăsoftware)ăandăfeedbackă(userăcomments,ă socialăsharing). Regardingătheăusers’ăpotentialăinterests,ătheyăcanăbeădeterminedăbyăexaminingăusers’ăsearchă queriesăinăsearchăengines.ăDeterminingărecipients’ăprofileăfromăthisăperspectiveăisăveryăsimilară toăestablishingătheăpotentialăbuyers’ăprofileăinămarketing. Ină aă similară manner,ă Newmarkă distinguishesă threeă readeră types:ă theă expert;ă theă educatedă layman;ătheăignoramusă(1988).ăWhileătheseădistinctionsăapplyătoătranslatorsăasăreaders,ăduringă theă analysisă ofă theă ST,ă thisă thesisă especiallyă focusesă onă howă theseă categoriesă applyă toă theă recipientăofătheătranslation.ăInămyăopinionătheăaverageăweb-user’săprofileăisănotă“caught”ăină anyăofătheseăcategories.ăHowever,ăinăwebsiteălocalization,ăeachăofăNewmark’săcategoriesăcană beă consideredă asă broaderă typesă ofă nicheă targetsă thată cană beă furtheră subdividedă accordingă toă web-users’ăinterests. Pymă considersă readers/usersă fromă aă socio-culturală context:ă participative,ă observatională andă excludedăreceivers.ă(2009:ă317-332)ăApplyingătheseădistinctionsătoăwebsiteălocalization: 67  participativeăusersăareăthoseăthatăinteractăwithătheăcontentăbyămodifyingăită(reduction,ă replacementă oră addition),ă althoughă ită mayă beă rightfullyă claimedă thată participatoryă   actionsăareăanyăinteractionsăwithătheăe-text; observationalăusersăbenefităfromătheăreferentialăfunctionăofăaătext,ăwhereas ă(self-)excludedăreceiversăareăsimilarătoăNord’săaddressees.ă(seeăbelow) Nordă examinesă audienceă fromă aă different,ă bipolară perspective.ă Sheă distinguishesă betweenă addresseeă andă receiver:ă “Theă addresseeă isă theă prospectiveă receiveră seenă fromă theă textă producer’să standpoint;ă theă receiveră isă theă person,ă groupă oră institutionă thată actuallyă readsă oră listensătoătheătextăafterăităhasăbeenăproduced.”ă(1997:22)ăThisădistinction,ăappliedătoămyăthesis,ă differentiatesăbetweenăsearchăengineăusersă(theăaddressee)ăwhoăseeătheăresultsăinăSERPsăandă thoseăwhoăactuallyăclickăonăanyăofăthoseăresultsătoăreadătheăactualăcontent. 2.4.9.4 Classification by direction Localizationăcanăalsoăbeăcategorizedăbyăitsădirection.ăConsequently,ămentionăshouldăbeămadeă ofătheăfollowingăsituations:ă  localizationăfromă aămajorălanguage/cultureătoă peripherală language(s)/culturesă(Spanishă andăCatalan);  localizationă fromă aă peripherală languageă toă majoră language(s)/culture(s)ă calledă byă Schäleră (2002)ă reverse localization (Romaniană andă English,ă ană approachă specifică toă RomanianăITăcompanies).ăFromăaăTSăperspectiveăităcanăbeăcomparedătoăUlrych’sămodeă ofă mediation,ă specificallyă alteringă andă producingă textă ină non-nativeă languageă (2009:ă 219-234) I also suggest the following type:  localizationă fromă aă majoră language/cultureă toă anotheră majoră language/cultureă (foră instanceăwebsitesăinăCanada);  localizationă fromă aă peripherală languageă toă anotheră peripherală languageă (aă Romaniană websiteă localizedă toă Hungariană oră viceă versa,ă achievableă byă companiesă thată wishă toă coverăbothătheăRomanianăandăHungarianămarkets. Also, from a directionality perspective, website localization can be achieved from a centralized or decentralized perspective. (O’Hagană andă Ashworthă 2002: 74) Centralized 68 approaches are rather brand focused and would use as text production strategy translation. A decentralized approach allows a higher degree of locale features. The text production strategy would be copywriting. 2.4.9.5 Classification by level (degree) Accordingă toă theă levelă ofă localization,ă Singhă andă Pereiraă (2005)ă distinguishă theă followingă typesăofăwebsites:  Standardizedă (sameă websiteă andă sameă languageă employedă foră allă users):ă foră instanceă vworker.com,ă aă freelanceă U.S.ă basedă companyă offeredă theiră workă mediationă servicesă onlyăinăEnglish,ăalthoughăemployersăandăemployeesăalikeăwereăfromăallăoverătheăworld.ă AnăexplanationămightăbeăthatătheăcompanyăthatăownedătheăwebsiteăhadătoăadhereătoăU.S.ă lawsă andă regulationsă ină whată workă mediation,ă moneyă transferă andă taxesă wereă concerned.ăHowever,ăoneăofăitsă competitors,ăelance.com,ă offersăitsăservicesăină 8ămoreă languages,ă andă termsă specifică toă theă U.S.ă areă translatedă intoă thoseă languages,ă throughă explicitation;ă(vworkerăwasătakenăoverăbyăfreelancer.comăină2012)ă  Semi-localizedă (oneă websiteă foră allă targetă markets,ă bută differentă languages;ă contentă isă notăentirelyătranslatedăintoăallătheătargetălanguages):ăsuchăanăexampleăisăpayoneer.comă whichă hasă translatedă onlyă theă postă logină pages.ă Pymă namesă thisă typeă ofă localizationă partial localization.ă(Pymă2014:120)  ăLocalizedă (contentă ofă websitesă isă entirelyă translatedă foră allă targetă markets)ă Toă distinguishă betweenătheă differentă localizedăversions,ătheăwebsiteădevelopersăuseăeitheră subdomainsă(de.elance.com,ăes.elance.com,ăetc.ăversusăwww.elance.com)ăorăaăsubfolderă systemă(www.samsung.com/ro/,ăwww.samsung.com/hu/,ăetc.);  ăHighly-localizedă (theă contentă ofă websitesă isă entirelyă localizedă foră allă targetă markets;ă bothă textă translationă andă technicală localizationă issuesă areă complete).ă Foră instance,ă theă Samsungă siteă foră Germanyă isă locatedă ată www.samsung.deă andă evenă theă designă isă different.ă Asă foră theă productă range,ă thereă areă bothă productsă availableă globallyă andă productsă customizedă foră theă Germană speakingă countriesă (Ledă TVă 6500ă seriesă versusă LedăTVă6200ăseries).  Culturally customizedă (theă websiteă isă perceivedă byă usersă asă completelyă belongingă toă theirălanguageăcommunity)ăInătheăcaseăofăwebsites,ăthisăisăclearlyămarkedăbyătheădomaină extension:ă .co.uk,ă .ro,ă .us,ă .de,ă etc.ă However,ă ină theă caseă ofă brandingă websitesă thisă cannotăbeăentirelyă achievedăbecause,ăforătheăsakeăofăbranding,ăsomeăcommonăfeaturesă 69 mustă beă maintainedă onă allă theă localizedă versionsă ofă theă website.ă Foră theă timeă being,ă thereăareănoăobviousăwebsitesăthatăareăculturallyăcustomized.ăSuchăanăexampleăwouldăbeă settingăupătwoăwebsites,ăownedăbyătheăsameăcompanyăandăofferingăsimilarăweb-hostingă services,ăinătwoădifferentăcountriesăandăwhereăităisănotătheăbrandingăthatăisăimportantăbută theătypeăofătheăservice.ă Theă followingă examples,ă althoughă possiblyă real,ă areă justă usedă foră demonstrationă purposes,ă andă theă websiteă pairsă shouldă belongă toă theă sameă owner.ă Foră instance,ă www.ukwebhosting.co.ukă foră theă U.Kă andă aă Romaniană equivalent,ă www.gazduire-web.ro.ă Similarlyă newvehicles.co.ukă andă autovehiculenoi.ro.ăTheă moreă generalizedă theă nameă ofă theă service,ătheăhigherătheăpotentialătoăculturallyăcustomizeăaăwebsite.ă Aă lessă generalizedă nameă ofă service,ă especiallyă ifă theă origină ofă theă productă isă necessaryă foră marketingăpurposes,ăwouldănaturallyănoălongerăbeăperceivedăasăbelongingătoătheăcommunity,ă asătheăpurposeăisătoăhighlightăthatăităisăană“alien”ăproductăorăservice.ăSo,ăromanianwines.co.ukă andăvinuriromanesti.roăwillănotăbeăfeltăbyăU.K.ăcustomersăasăaăculturallyăcustomizedăwebsiteă becauseătheămarketersăwouldăratherăemphasizeă“theăadvantages”ăofădrinkingăRomanianăwines.ă However,ăifătheăRomanianăwinesăareărecommendedăinăaănaturală BritishăEnglishădialect,ă andă byăaăBritishăcitizen,ăhigherăsalesăareămoreăprobable21.ăAsăPrinceăCharles’ăvisitătoăTransylvaniaă hasă beenă beneficială toă tourismă andă toă theă locală market,ă soă anyă productă withă origină basedă marketingă wouldă flourishă ifă aă famousă personă hasă “tested”ă ită previously.ă Theă “countryă ofă origin”ăeffectă(Pucciăetăal.ă2012:155)ăcouldăalsoăhelpăthroughăcontaminationăfromăoneăserviceă orăproductătoăanother. Anotherăissueăworthămentioningăisăthatătheătranslationăcomponentăofătheălocalizationăprocessă would,ă mostă probably,ă beă replacedă byă copywriting ină theă caseă ofă culturallyă customizedă websites. Also, Pym (cited in Dimitriu 2002), suggests moving from a purely linguistic perspective to a sociological and economic one, as in the case of websites, more often than not, the driving engine is generating sales. Considering that globalization is already sociological and economic, the direction of translation within localization is already a sociological and 21 ăForăfurtherădetailsăseeăsubchapteră4.6,ăReverse localization: Romanian companies targeting foreign markets. 70 economic phenomenon; this is also corroborated by the findings of Forester Research [18]. In the case of e-commerce websites the sociological and economic factors are rather overstated, as their very specific purpose is to generate sales. All in all the classification of localization processes can be useful in determining the appropriate strategy for gaining access to particular markets based on the type of products or services that are to be made available in the target culture. 2.4.10 Localization and diacritics usage: a case study It is at localization level that it has to be decided if for the translation, as a product, the text will be using diacritics or not. The translator will simply provide the text according to this requirement. The usage or non-usage of diacritics is especially true if the localization direction is from a major language (English, German, French, Spanish, etc.) to a minor language (Romanian, Hungarian, etc.). Search engines represent an intermediary channel in the communication between web-users on the one hand, and website content on the other. Therefore, it is important to observe how search engines manage the issue of diacritics. Obviously, not using diacritics is not user-friendly, as some words may become homographic andăreadingăbecomesăsomewhatămoreădifficultă(notăinăaccordanceăwithăGoogle’săguideă[29]).ă The web-user must rely on co-text.22 In what follows, I will consider some aspects relating to the usage of diacritics: - Diacriticsăandăweb-users:ămostăkeyboardsăaroundătheăworldăuseătheăstandardăEnglishă keyăsetting,ăespeciallyăinătheăemergingăandăpoorăcountries.ăAlso,ăvirtualătouchăbasedă devicesăuseăaălimitedănumberăofăcharacters,ăandăifăyouăneedăaăcharacterăwithăaădiacritică youă cană optă foră variantsă (pressingă foră longeră theă keyă „a”ă onă aă Romaniană keyboard,ă displaysă theă options:ă ,ă â).ă Theseă factsă makeă theă usersă inpută theă keywordsă withoută diacritics.ăFortunately,ăsmartăphonesăautocorrectătyping,ăthusătheătendencyăisătoăinpută keywordsăwithădiacriticsă(aăprescriptiveăapproach). - Diacriticsă andă searchă engines:ă usersă aroundă theă worldă mayă useă eitheră .comă searchă enginesă oră localizedă versions.ă Ană analysisă ofă theă statisticsă byă countryă providedă byă Alexa.comă[38]ăcanărevealăifătheăusersăpreferătheălocalizedăGoogleăsearchăengineăoveră 22 See subdivision on Co-text (3.4.2.1) for further details. 71 theă Google.comă generală searchă engine.ăAfteră Iă analyzedă theă positionă ofă google.comă versusălocalizedăversionăofăGoogleăforăeachăofătheăcountries,ăoutăofătheă117ăcountriesă listedăbyăAlexa.comăină83ăcountries,ăweb-usersă“prefer”ătheălocalizedăversion,ăthatăisăaă 70,94șă ofă theă total.ă However,ă google.comă isă almostă alwaysă positionedă eitheră immediatelyă underă theă localizedă Googleă version,ă oră twoă oră threeă positionsă lower,ă whichă isă consistentă withă theă firstă positionă ofă google.com.ă Thisă isă consistentă withă Alexa’sălistingsăforăwebsitesăatăglobalălevelă[39].ăEarlierăIăusedă“prefer”ăbecauseăusersă areănotăactuallyăofferedăaărealăchoiceăinăusingătheă.comăorătheălocalizedăversion,ăasătheă searchă engineă automaticallyă directsă youă toă theă localizedă version.ăYahoo.comă isă alsoă positionedăatătheătopăofătheămostăusedăsites,ăbutătheyădoănotăuseăaălocalizedăversionăofă theă URLă byă changingă theăTLDă (topă levelă domaină foră eachă countryă asă Googleă does,ă e.g.ă.ro)ăbutăaăsubdomainăapproachă–ăro.yahoo.com.ă - DiacriticsăandăSERPsă(searchăengineăresultsăpages):ăReturningătoăGoogle,ăifăthereăareă severală versionsă ofă theă searchă engine,ă doesă ită meană thată weă obtaină differentă resultsă dependingăonătheăGoogleăversionăused?ăHowădoesătheăusageăorănon-usageăofădiacriticsă affectătheăresults? Apparently the academic community has not considered the usage or non-usage of diacritics to be a relevant issue,ăasătranslationătrainers’ăapproachăisămostăoften a prescriptive one. The usage or non-usage of diacritics is connected with search engine based localization. As there is no academic research regarding website localization based on web-user keyword input, there are no discussions on diacritics either. On the other hand, the localization industry and the search engine optimizers have been considering the issues of search engines and the issue of using diacritics as a means of success or failure of a localized website23. With regard to the usage of diacritics, the industry is rather divided, but I would assert that this issue is related to the maturity of the market in a specific country. In theory, German, French, Spanish should work using diacritics because there is already significantly more content written with diacritics in those languages indexed by Google. Thus, the search engine adapts to existing content and even if the German web searchers did not use diacritics, the search engine would still “guess”ă theiră intention.ă However,ă theă greată majority of German keyboards use the German key layout. 23 More on this subject in subdivision 2.6.4 Web content translation and search engine optimization (SEO) 72 The SEO professionals in Romania are also divided, whether to use diacritics or not. If we wereă toă consideră Google’să maină directive,ă toă writeă firstă ofă allă foră theă web-users, then we should use diacritics in Romaniană texts;ă however,ă onă theă otheră hand,ă Google’să suggestionă tool does not use diacritics, because it displays suggestions based on the highest number of queries entered by web-users (who mostly use standard U.S. keyboards). Figure 14: Absence of diacritics in Google suggestion for Romanian Nevertheless, Google is ready for users that opt for typing their input using diacritics as well. If I type ceai de pă the sugestions will be using diacritics: ceai de păpădie, ceai de păducel, ceai de pătlagină, ceai de pătrunjel. Also,ă Google’să Keywordă Planneră showsă thată theă Romaniansă doă notă useă diacriticsă whenă making a query. For instance, if we look for suggestions for keywords like ceai de papadie and ceai de păpădie (both meaning dandelion tea, but only the second is typed in standard Romanian, i.e. with diacritics) we can notice that users have actually looked for ceai de papadie - without diacritics - (and other related keywords such as ceai papadie, papadie ceai, ceai de papadie contraindicatii, ceai de papadie proprietati, ceai de papadie indicatii, ceai din radacina de papadie) on average 540 times a month. If we compare this to the variant with diacritics, ceai de păpădie, the monthly average search is only 10. Doing the same experiment on google.de the outcome is different. Figure 15: Google suggesting words with diacritics for German 73 If I intentionally write the German term for dandelion in German without diacritics, Google willăstillăsuggestătheăstandardădiacriticalăversion.ăLookingăforăkeywordăsuggestionsăGoogle’să Keyword Planner will reveal for dandelion tee that regardless of the variant (löwenzahntee or lowenzahntee) the suggestions from Google will be reversed as compared to Romanian, löwenzahntee with 1600 queries (1930 with extended keywords) whereas lowenzahntee and loewenzahntee (spelt with oe) showing only 10 and 20 queries, respectively. These findings further stress that most keyboards in Germany are localized for the German users (See Figure 1 in subchapter 2.2 Globalization). They also mean that German Internet content is written with diacritics. For Hungarian, there are mixed findings. I looked for the Hungarian term for dandelion first on Google.ro, as it is the webpage to which it directs me automatically, and the spelling suggestion was both with and without diacritics, i.e. gyermekláncfű and gyermeklancfu. However, changing Google.ro to be used in Hungarian, or using Google.hu results in suggestions of keywords spelt with diacritics. More important than the previous tests is to observe for each localized target language, how certain keywords perform on results pages; specifically, to analyze if results depend on whether the queries use diacritics or not. While for German and Hungarian the obvious choice is to use diacritics, for Google.ro the localization individual or team must perform further tests. First of all, although Google.com redirects by default to Google.ro, many Romanians still prefer using Google.com. According to Alexa.com, Google.ro is the most used website in Romania, but Google.com is the third most used. This means that tests should be conducted on both the general and localized variants of the search engine, especially if we consider that in the case of Hungarian, the issue of using diacritics depends on the search engine variant. In order to collect results unbiased by user location, it is recommended to use the browsers in privacy mode. If we first use Google.ro and search for ceai de păpădie and ceai de papadie (same keyword without diacritics) we obtain the following results: 74 These results show in the top ten results mixed spelling for the word păpădie. Note that the standard spelling lists 113,000 results, whereas the non-diacritical version yields approximately 66,100 results - half of the initial figure. Therefore, by comparing the results, we may conclude that website content using standard Romanian spelling is more beneficial to users; yet, unfortunately, Google does not suggest the use of diacritics, as shown previously. Also, we can see that the first 9 results are the same for both spellings. The tenth result is different, but strangely, ceai de păpădie displays in the tenth position a result without diacritics while ceai de papadie shows a result with diacritics. This means that, in comparison with previous years, when Google.ro favored non-diacritical characters (verifiable by examining the date when content was published on various websites and Romanian SEO forums), there has been a shift towards listing in the SERPs results both with and without diacritics. For this particular keyword, 2 out of 10 results use diacritics. 75 Google’să blendedă solutionă couldă beă aă reflectionă ofă theă content crawled on Romanian websites. Clicking on the first result shows that the article uses both standard spelling and non-diacritical spelling in titles, headings and the main body of the text. The 7th result also uses diacritics but only in the meta description (short description used to be displayed on the SERPs) while clicking on the link displays a page that does not use diacritics and the content is not entirely related to the query. Examining the 10th result for ceai de papadie in the SERP shows that the text uses characters with diacritics. However, the content is related to the query only about 20% of the whole when accessing the page. By extending the results taken into consideration to 100, we can see that only 17 out of 100 pages use diacritics when searching without diacritics. By using diacritics in the search query the results improve to 19% but no relevance factor has been considered. Also, there are differences in webpage positioning on the result page, which should not be determined by the usage or non-usage of diacritics. Using Google.com for the same keyword shows the same results when considering the top 10 results, depending on the standard or non-standard input. The reason is that Google.com recognizes it as a Romanian expression. While using private browsing mode, if we use a Romanian expression without diacritics, which is applicable to other languages as well, then the search suggestions for both .ro and .com are from several languages, based on input queries at global level. Typing canal de will suggest Spanish (canal de [los] …) and English (canal definition) terms. However, if we are loggedăin,ăGoogle’săsuggestionătoolăwillăsuggestă keywords specific for the location of the device. In my case it will even suggest different spellings (Kanal D – homophone of canal de). As the tendency is to care all the time more about privacy issues (see EU directive and its implementation [40]), it is important to lay emphasis on SEO localization for keywords made up of more than two words24. 24 ăSeeămoreăinformationăinăsubdivisionă2.6.4ăonăWeb content translation and search engine optimization (SEO) 76 Considering the findings in this subsection I recommend localizers to reflect on the issue of diacritics usage, separately for each target market, first of all, because a translator would by default use the standard spelling, as s/he was taught in school. For some of the markets, standard spelling is the present method of delivering content. On the other hand, there are languages for which content is either scarce or input in previous years was without diacritics, because the need to be listed on the first page in search engines determined it that way, as with Google.ro. However, the tendency is to move towards standard spelling. The current state of Google.ro (2014) requires a mixed approach (both keywords with and without diacritics). The future is likely to favor those who use diacritics if Google is to conform itself to its own principle, that of writing, first of all, for its users. Of course, on the other hand, Google is considering how the vast majority of users input queries, and that is without diacritics, but the role of such corporations is also to educate and stick to language standards. If in the case of German and Hungarian they are already implemented, it would be natural for Romanian as well. The ever-increasing usage of smart phones to look for information on search engines will depend all the time more on word input with diacritics because the smart phone system will automatically correct input (as already stated above). It would be valuable to see other studies on how diacritics are used for website content in other European languages but also in Vietnamese or Turkish, and if these findings will make a difference in SERPs. 2.4.11 (Quasi-)Automatic webpage localization software A natural step of MT developers is to develop localization software. While the language input is already in place from the MT itself, the elements specific to localization must be entered manually (currency conversion rates, VAT, duty taxes, etc.). Furthermore, cultural references cannot be automated. Most often such website localization software cannot outperform software such as Google Translate[41] or Microsoft Translator[42] (more details on these tools in 2.5 Translation). However, the approach is different. While the two search engine MTs are based on statistical translation, IBM WebSphere, software like Passolo, InterTran Website Translation Server are based on aligned TMs. Also, while search engine based MT is often quality deficient, by default, TM based software is aligned by humans. Irrespective of the system type used, most effort is required at language level, i.e. translation. For more technical, logistic and law related localization issues I recommend using content 77 management systems (CMS) and e-commerce platforms. Such systems have already various built-in localization specific features. From an administration page one can set the target market, language and locale specific elements. Usageă ofă CMSă “interrupts the syntagmatic flow. The technology imposes the paradigmatic on the syntagmatic. All translation technology does this to some extent.”ă(original emphasis, Pym 2014:124) This paradigmatic shift may have a negative impact on the translation, as the translator does not perceive the text as whole but as chunks of texts of various lengths. (Pym 2014:124) However, this deficiency can be overcome by using an integrated module similar to the application I suggested Figure 18, in 2.5. Using such a module re-contextualizes the text chunks to be translated. The major drawback is that you have to input the localized text yourself, although I assume that such CMS will use Google Translate [41] or Microsoft Translator [42] in the near future for more efficient workflow. Their current state requires extra effort, namely post-editing. Such an e-commerce platform is PrestaShop [43]. PrestaShop has a localization module and you can target 60 markets from all around the world [44]. The following image shows the level of localization that can be achieved. Figure 16: Prestashop's advanced localization package As one can see, everything that is essential from a commercial perspective can be included in the PrestaShop powered website. This platform is free and translation and localization information is supplied by the people using it. This is called by the industry crowd translation or massive online collaboration (MOC) and is a viable alternative to machine 78 translation as translation is addressed by humans25. Pym calls this type of translation “voluntaryă translation” (2014:128) Prestashop even extended massive online collaboration to localization. On http://www.prestashop.com/en/contribute-prestashop-localization, as the URL of the page suggests a PrestaShop powered website owner, or anyone else for that matter, can contribute with information on a target locale. One can submit information on logistics (payment solutions, shipping providers, and major industry players), local laws and compliance, e-commerce specific laws or even share general tips and make suggestions for the target market. Thus, whatăresearchersăinălocalizationă(O’HaganăandăAshworthă2002:ă71,ă [9: 16], Pym 2004b: 2) highlighted as localization issues are now accomplished and updated automatically. Interface, both front- and back-office are also translated by Prestashop driven website owners and is open for improvements. Product information requires translation as with any other website. For the translator, localizer or company similar platforms can be used to speed up the localization process or they can be simply used in order to find useful localization related information. 25 ăMoreăinformationăonăMOCătranslationăinătheăMachine translation for website localizationăsubdivisionă(2.5.4) 79 2.5 Translation Soăfar,ătheăthesisăhasămostlyădiscussedătheătechnicalăaspectsăofălocalization.ăForătheăfieldăofă Translationă Studies,ă however,ă translationă isă theă mostă importantă aspectă ofă theă localizationă process.ă Ină termsă ofă separatingă textă fromă theă otheră aspectsă ofă localization,ă Massion’să textă translationăschemeă(2011:40)ăcanăbeăappliedăsuccessfully. isă designed.ă First,ă aă contentă managementă systemă mustă beă developedă (usingă Drupal,ă Magentoă oră otheră availableă contentă managementă systemă -ă CMS)ă whereă theă translatoră cană seeă andă edită onlyă theă text,ă whileă textă stylingă wouldă remaină consistentă andăchangeableăonlyăbyăaăprogrammer,ăifăsoă required.ă Foră instance,ă allă titlesă wouldă beă darkăblue,ăfontăsizeă20,ăsubtitlesădarkăgray,ă fontăsizeă14,ăwhileămainăbodyătextăwouldăbeă setăblackăandă12ăforăfontăsize.ăInătheăcaseăofă Chinese,ăArabicăorăRussianătheăsetupăofătheă textă stylingă mustă beă differentă becauseă theă alphabetăcharactersăareădifferentăandăatăleastă fontă sizeă mustă beă different.ă Withă Asiană characteră setsă evenă theă verticală lineă heightă Figure 17: Massion’s diagram isă differentă andă thereă isă noă spaceă characteră betweenă words.ă However,ă ifă duringă theă Mostă ofă theă concernsă thată Massionă internationalizationă stage,ă theseă issuesă areă expressesăinăhisăarticleăcanăbeăavoidedăifătheă accountedăfor,ăproblemsăareăhighlyăunlikelyă appropriateă systemă foră theă translationă taskă toăappear. Next,ă ită isă importantă foră theă translatoră toă haveă accessă toă theă siteă ină theă intermediaryă internationalizedălanguage26,ăalsoăknownăasăaăpivotăorăbridgeălanguageăinămachineătranslationă terminology.ăWhetheră consideringămachineătranslationăorălocalization,ăEnglishăisă theăchoiceă ă Seeă subchapteră 2.3ă Internationalization (I18n)ă andă moreă specificallyă theă subpartă dealingă withă theă internationalizationăofătheătextă(2.3.1) 26 80 foră bridging.ă Theă Internetă isă theă largestă sourceă foră parallelă corporaă andă Englishă isă stillă theă languageă mostă widelyă usedă aroundă theă globeă oră theă languageă ofă choiceă foră onlineă contentă (Crystală 2006,ă [45]),ă withă Chineseă showingă ană upwardă trendă ină recentă years.ă Theă statusă ofă EnglishăasălinguaăfrancaăisăaădecisiveăfactorăinăfavoringăEnglishăasăaăpivotălanguageăinătheă caseăofăwebsiteălocalizationăasăwell. ă Finally,ăităisăimportantăforătheătranslatorăhim-/herselfătoătestăoutătheătranslationăinărealătime,ăonă aă non-publică pageă bută onă whichă s/heă cană seeă whată theă translatedă textă looksă likeă withină theă localizedă designă ofă theă web-page.ă Byă usingă thisă strategy,ă companiesă cană saveă substantială amountsăofătimeăandămoney,ăasătheămanagementăsystemăwillăallowăimmediateăfeedbackătoătheă translator,ătestingătimeăwillădecreaseăsignificantlyăandălocalizationăcontentămanagementă willă beăerrorăfree.ăTherefore,ătheătranslatorăwillănoălongerăexperienceătheăpressureăofădealingăwithă technicalăissues.ăS/heăwillăhaveătoăfocusăonlyăonătheătranslationăprocess.ă Below,ăIăsuggestăaă simpleădesignăforăsuchăaăsystem: Figure 18: Code free translation tool with preview feature . 81 2.5.1 Translation, adaptation, copywriting Thisă subchapteră discussesă strictlyă translationă relatedă issuesă fromă theă perspectiveă ofă websiteă localization.ă Eugene A. Nida defines dynamic equivalence in translation as “producing in the target language the closest natural equivalent of the source language message, firstly with respect to meaningăandăsecondlyăwithărespectătoăstyle.”ă(Nidaă1959:ă19)ăOn the other hand, meaning can be conveyed only when there is a common reference system (Lakó 2009a:221-227). As language to language and from culture to culture references are never entirely the same, extra operations are required, namely adaptation. Adaptation is used in addition to translation to make necessary modifications in the target language "whenever the context referred to in the original text does not exist in the culture of the target text, thereby necessitating some form of recreation. This widely accepted definition views adaptation as a local rather than global strategy, employed to achieve an equivalence of situations wherever cultural mismatches are encountered". (Vinay and Darbelnet cited in Bastin, 2009: 3) This view on adaptation as a local strategy is very much like the localization process. As previouslyăseen,ătheădefinitionăofălocalizationăisărelatedătoă“adapting".ăTheămainădifferenceăisă that adaptation refers only to text (as a method of translation), while localization also involves numerous technical aspects. In what follows, I shall analyze the extent to which Bastin's adaptation procedures (2009: 4) applies to web localization:  transcription of the original - for example adding Google's website translator (Machine Translation) to webpages for instant access to automatic translation of the pages. This applies rather as a word-for-word translation. The more technical a text, the higher the quality of the translation, but translation quality depends on the source and target languages as well, on the existing corpora in the two languages, crawled by Google. One disappointing example, is http://www.masterstudies.ro/Master-%C3%AEn- %C8%99tiin%C8%9Bele-Lingvistice/Finlanda/UEF-Philosophical-Faculty/, which is 82 just machine translated, and yet is in top ten in Google SERP for the masterat traducere Romanian keyword.  omission: texts on the web, especially on commercial websites, are already short and optimized for screen reading, so further omissions are unlikely to occur. This is especially true on homepages of e-commerce websites. Inferior level webpages (level 3 or 4) may use omission as the longer the source text is the higher the probability of using this strategy especially if the target language (TL) is by default more verbose than the source language (SL) and the layout is not flexible. Omission in text can also occur when ST is LC culture and TT is HC culture (Hall 1976: 105-116). Also, parts of texts can be replaced by other semiotic systems – for instance using pictures or icons.  expansion: with the exception of online advertisements, which require complete recreation, web-texts are usually very clear and require no further explicitation (for instance, product or service description). However, as will be seen in Chapter 4, the requirement to use certain keywords may entail lengthy additions - even full sentences to the initial text. Extra information can occur in various forms:ă “Theă additională information a translator may have to add to his version is normally cultural (accounting for difference between SL and TL culture), technical (relating to the topic) or linguistic (explaining wayward use of words), and is dependent on the requirement of his, as opposedătoătheăoriginal,ăreadership.”(Newmark cited in Dimitriu 2002:23)  exoticism is used especially for the translation of online advertisement banners. In the case of website localization, during the internationalization process such elements are removed, but they can be re-added through copywriting during the actual localization process.  updating: in the case of online texts in theory there is no extra need for this procedure as the source text itself is updated regularly; in practice TT may require updating if keyword usage in search engines changes (likelihood is however low as information in ST often becomes obsolete )  situational or cultural adequacy: this procedure is already inherent to the localization process;  creation: this is a common procedure on websites, especially on those which are localized or culturally customized (Singh and Pereira 2005). Also, the decision to create, as an adaptation procedure, is made based on keywords used in search engines. I will develop on how search engines should influence the translation process later on. 83 Further on, Bastin (2009: 5) speaks about factors which lead to using adaptation procedures:  cross code breakdown: when at the lexical level there is no equivalence in the target language, in comparisons and metaphors, when the concept is nonexistent or when the level of gradation for a concept or object is not the same (for instance the considerable number of words for snow in the Inuit language [46]  situational or cultural inadequacy: this is similar to the absence of a common reference system;  genre switching: it is very rare in case of website text translations, because if the source text is intended for adults it will still be for adults in the target website content;  disruption of the communication process is irrelevant in the case of commercial websites as information in the source text is always the latest available. For the purposes of this thesis, i.e. from the perspective of localization, I am more interested in the restrictions that Bastin lists, still adapting them to translating web content:  the knowledge and expectations of the target reader: in the case of websites the adapter has no longer to decide if the source text is new or not, but rather verify, using search engine statistics (Google KeywordPlanner) and thus responds to the target reader's expectations27. Also, the target reader expects to benefit from the latest information so there is a continuous process of maintenance (Pym 2010);  the target language: the type of discourse is kept the same in the target text; style and addressing (formal or informal) may vary depending on the profile of the TL (for instance the discourse in Romanian may be using second person singular pronouns and verb forms – informal, or polite formal forms.);  the meaning and purpose of the source and target texts: in web localization they are the same, to inform and generate sales. There are two main types of adaptation: local and global. While they coexist in practice, local adaptation models are used for semi-localized websites, whereas global adaptation is more appropriate for culturally customized ones. Ină myă opinion,ă weă cană correlateă Bastin’să adaptation strategies with Singh and Pereira classification of websites according to their level of localization (2005) 27 ăMoreăinformationăinăsubchapteră2.6ă-ăWebsite localization and search engine optimization. 84 local adaptation <--------------------------------------->global adaptation standard -> semi-localized -> localized -> highly-localized -> culturally customized Figure 19: Adaptation methods related to localization The closer to the culturally customized state a site is, the more likely it is for global adaptation techniques to be used as opposed to local adaptation. This is also true when referringătoăauthorship.ăMoreover,ăonăSinghăandăPereira’săaxisăofăwebălocalization,ătheămoreă creative the translator, the more interventions are operated at lexical level. Relevance and purposeă areă moreă importantă thană “truth”,ă whichă explainsă whyă theă resultingă adapted text is closer to copywriting. This is also in agreement with the statement that an efficient translation should be perceived as anăoriginal;ăfromăNord’săfunctionalistăperspectiveăsuchătranslationsăareă called instrumental (Nord 2005: 81). Copywriting.com defines copywriting asă“theăartăandăscienceăofăwritingăwordsătoăpromoteăaă product, a business, a person or an idea; and carefully selecting, editing, weaving and constructingăthoseăwordsăinăaăwayăthatăthey’llăpersuadeătheăreaderăintoătaking a specific and measurableă action.”[47]ăThisă definitionă isă closeă toă theă definitionă ofă textă adaptationă foră webă localization, the only (significant) difference being that copywriting happens in the target language and it is already under the influence of the specific locale. Also, this does not include the concept of keywords, which are very important in web content marketing. In the economics of website localization the translator will often have to become a copywriter in the target language and show proficiency both in language and culture specific issues. Deficiency in copywriting skills may lead to contract loss once the employer realizes that it is more cost-effective to hire professional native-speaker copywriters for the target language and culture. In website localization the employer would set the task quite similarly and the target text outcome would be expected to yield the same response from the users in the target language/culture. To determine which approach is more effective, in Chapter 4, devoted to case studies, subchapter Which Way Website Localization: Translation or Copywriting? (4.4) I conducted two experiments. In terms of return on investment (ROI), the study determines the efficiency of translators versus native copywriters. The table below synthesizes the conditions in which copywriting can be used as an alternative strategy to translation. The factors are listed and grouped from general to specific elements. 85 factor translation copywriting Communication strategy B2C C2B semi-localized / localized highly-localized / culturally Website localization level customized product page home page centralized/brand focused decentralized prescriptive descriptive foreignizing domesticating required NA Text standardization required NA Localization strategy Text comprehension strategy Text production strategy Orientation ST user and market, by default Initial Source ST various texts from TT Text type referential expressive/appellative Creative writing skills optional required Keyword implementation some degree of difficulty natural keywords keywords + CTA semasiological onomasiological Naturalness original content traceable Production speed topic dependent natural higher speed Mediator Text producer profile layman expert Authorship irrelevant relevant User profile observational participative Table 2: Translation vs. copywriting 86 In short, text adaptation, in the context of commercial website localization, is the translation strategy with the highest yielding results. Adaptation is one of most important translation methods and translation is part of the localization process (as shown previously). However, in practice, translation is often replaced by copywriting. 2.5.2 Translation and business discourse Business discourse refers to communication through oral or written channels (or simultaneously both – on chatting platforms you can concomitantly use verbal communication and send documents or links). It occurs in a commercial organization or in a commercial transaction setting, during the stage of reaching towards a potential customer or business partner and establishing a business relation. Establishing such a relationship implies a specific social context, thus, business discourse is social action in a business context. From a TS perspective, Nida was the first to acknowledge the importance of social factors: “Aă communication is not intelligible if it is treated as an event abstracted from the social context ofăwhichăităisăaăpart”. (cited in Dimitriu, 2009:29) On websites, we can see the ever increasing importance given to social media. When speaking about business discourse there are two main directions in terms of discourse, depending on the relationship established between the partners of the commercial transaction: - businessătoăbusinessă(B2B) - businessătoăconsumeră(B2C) While in the traditional way of B2B versus B2C communication there are significant communication pattern differences, when referring to content on webpages and online communication in general, B2B and B2C communication blur into person to person (P2P) communication, as in both B2B and B2C, there are people behind the decision making process, and they adhere to the same general patterns of communication and negotiation but with an added layer of specific culture28. P2P is even more noticeable in C2C (customer-tocustomer) communication. C2C communication is very popular on sites where second-hand products and not only are on sale (ebay.com, tocmai.ro, olx.ro). Users on such sites are both customers and buyers and the style of the communication is often informal, limited in terms 28 Forăfurtherădetailsăseeăsubchapteră3.4.3ăThe cultural layer. 87 of text verbosity but compensated by the usage of images of the product on sale. While B2B and B2C communication often makes use of keywords and SEO optimization, C2C communication is employed only within the same website; thus it does not require special planning. It is rather characterized by spontaneity. If we further compare the traditional ways of advertising to websites, the latter form is characterized by multimodality. Multimodality isă definedă asă “theă useă ofă severală semiotic modes in the design of a semiotic product or event, together with the particular way in which theămodesăareăcombined”ă(KressăandăLeeuwenă2001:20).ăMultimodalityăisăcharacterizedăfromă the perspective of content and expression. Content can be further subdivided into discourse and design while expression divided into production and distribution. Kress and van Leeuwen (ibid.) perceive discourse asă aă “resource”ă oră aă “semiotică mode”,ă emerged and developed due to human activity in a certain field, within a certain sphere of common knowledge, in a specific context (for instance, experts in communication) and can be achieved orally or in writing or both, through various media channels (newspapers, TV, websites). Design,ăisădefinedăasă“semioticăresources,ăinăallăsemiotic modes and combinations ofăsemioticămodes”ă(2001:5),ătoăproduceădiscourseăinătheăcontextăofăaăspecificăcommunicationă situation. On the expression panel production refers to the material forms through which semiotic products or events are produced. Production creates additional meanings in discourse. Using lists on a webpage or a certain color scheme, or a certain font-size will indicate to the user if it is an important piece of information or a secondary one. Media are semiotic resources or channels of communication through which simultaneous discourses can be created. The other subdivision of expression, distribution, refers to the technology used in the recoding of semiotic signs. On the Internet, production and distribution are merged together. The semiotic modes on commercial websites transform from textual semiotics into various different modes; text is in part replaced by images, layout, design features, especially in the case of homepages. One important website specific feature is hypertextuality, which obviously influences ecommerce websites as well. Hypertextuality is characterized by multi-linearity, multi88 directionality, and interconnection. Hypertext is non-linear and offers to web-users options regarding the reading path, through hyperlinks.ă Nielsonă definedăhypertextualityăsimplyă asă aă “nonsequential”ămanner ofăaccessingăinformationăwhichă“presentsăseveralădifferentăoptionsătoă the readers and the individual reader determines which of them to follow at the time of readingă theă text”ă (Nielsenă 1990:1).ă Websitesă areă theă mostă flexibleă typeă ofă mediaă thată cană encompass all the other media types and consists of several layers of reading that, ideally, convey the same meaning, i.e. a whole of convergent semiotic elements. Also, hypertext allows accessing, amending and exiting the text at any moment. Lemke combined multimodality and hypertextuality in hypermodality. (Lemke, 2002) He suggests that the different modalities (that is to combine all the elements of a website) must share a common source of meaning-making. Lemke suggests that there are three sets of interdependentă meanings:ă presentational,ăorientatională andăorganizational.ă Lemke’sătypesă ofă meaning are quite similar to text types categories determined by Reiss (cited in Munday, 2008: 72): presentational meaning is similar to the informative function of text, orientational meaning to the expressive function, while organizational meaning refers to the coherence and cohesion characteristics of the text (Lemke 2002: 304) On commercial websites, business discourse is often intermingled with the technical discourse, especially in the field of marketing text content, where the discourse is based on clear, simple and informational data. Also, because English is still the dominant language on the Internet, [45], it carries along cultural traits especially from the North American culture (they are also the creators of modern communication technologies). According to Hall (1976: 105-116), North Americans alongside Germans are at the extreme end of Low-Context cultures, which means that more or less, the low-context communication type has been spreading rapidly, together with the spreading of communication technology. Analyzing e-commerce discourse, one can see it is in accordance with the recommendations of oracle.com [15], the features of low-context cultures (Hall 1976: 105-116),ăandăNielsen’să reading pattern findings (2006): simple sentences, clear and concise messages, lists, a few paragraphs. However, there is a difference in discourse construction between top level pages such as the main menu related pages and more specific pages such as product full description pages or product review pages. 89 If we analyze the percentage of textual content versus contextual content (images, graphics, videos, etc.) on webpages from a multimodality perspective on various levels (from main menuă relatedă pagesă toă productă descriptionă pages)ă andă compareă themă toă Hall’să high-context low-context axis (1976: 105-116) we can see some similarities: HIGH-CONTEXT CULTURES 1st level webpages (homepage) Japan China 2nd level webpages(category page) Arab Countries Greece Spain 3rd level webpages Italy (subcategory page) England France 4th level webpages North America (full description product page) Scandinavian Countries German-speaking Countries LOW-CONTEXT CULTURES The more specific a webpage, the more accurate, verbose and technical it is. Thus, it can be said that websites use a mixture of LC and HC communication strategies, depending on the purpose and function of the webpage. However, the general tendency in e-commerce discourse is to have the same discourse for both B2B and B2C (as the role of webpages is to inform the potential customer, so style and register are no longer differentiated on websites either). Thus, there is no impact on localization and translation as communication patterns have become simpler and more unitary. However, subchapter 4.5 on Website localization for the Romanian market – a cultural approach will show that there are some aspects that need to be considered. 90 2.5.3 Translation theories applicable to website localization HansăVermeer’săSkopostheorieă(2000)ăsetăaănewătrendăinătranslationăstudiesăinătheă1980's. The translational action aim of the translatum is imposed on by or negotiated with the commissioner (the person requiring content to be translated). The purpose in commercial web localization is the same both in the source text and in the target texts, which is to generate traffic or sales. On the other hand, the main difference lies in the fact that web-texts impose restrictions on translators, such as maximum text length - due to web-page design. Also explicitation may not fit in, instead one can use hypernyms, replace subordinates with gerund constructions or even with terms that bear a more concentrated meaning. When considering the factors that influence the composition of the original text, the main difference between a general text to be translated and web content translation is that the author of the former does not (generally) plan to be translated into several languages; the translation as part of the localization plan on a commercial website, will take into account from early stages, that is, from the globalization and internationalization stages, that the text will be localized to several locales. For example, the producer of the original text, instead of using American English, could use International English or another standard, a language free of slang, of vocabulary specific only to certain regions, or particular grammar structure patterns. Thus, further to the internationalization of an application or website, I discussed in the Internationalization and translation strategies (2.3.1) subchapter about the internationalization of the text. It can be achieved both by using a more general language and by finding/delimiting text segments. In practice, it is difficult to clearly establish text structure and styling. Yet, on websites one can distinguish between button texts, menu texts, short informative texts, full-length texts, texts used with media elements, etc. Another important functionalist theory that leaves ample room for application to web localization is Christiane Nord's theory (1997). Of special interest is the distinction she makes between documentary and instrumental translations (Similar to House’să overtă /ă covertă model – in Dimitriu 2002:65). Instrumental translation “serves as an independent message transmitting instrument in a new communicative action in the target culture, and it is intended to fulfil its communicative purpose without the recipient being conscious of reading or hearing a text which, in a different form, was used before in a different communicative situation.” (2005: 81) In the case of website localization, this is only applicable to Culturally 91 customized websites (Singh and Pereira 2005).29 A Romanian web-user visiting Samsung.com/ro/ will be immediately aware that the text s/he is reading might be based on a ST. It is not because the translation is not accurate or culturally adapted, but because the Romanian content is on the same server with the original content. As previously mentioned, only the German version of the Samsung site is completely customized as it is hosted on its own .de domain. Instrumental translation accomplishes the same functions as the source text (convincing users to take action, buy a product or service) and is called by Nord “function-preserving translation”. Additionally, Nord distinguishes three forms of this translation type: equifunctional (e.g. instructions for use), heterofunctional (e.g. Gulliver's Travels for children) and homologous translations (e.g. poetry translated by a poet) (see Nord, 1997: 4752). Translations of commercial webpages maintain the same function as the source texts across the various localized webpages, thus, in this case the strategy applied is similar to equifunctional translations, ensuring business to consumer or business to business communication. Nordă defines documentary translation asă aă “typeă ofă translationă processă whichă aimsă ată producing in the TL a kind of document of (certain aspects of) a communicative interaction in which a source-culture sender communicates with a source-culture audience via the ST under source-culture conditions. (Nord, 1997: 138) In web localization documentary translation is used for product features and specifications, “aboutăus”ăandă“contactăus”ăandăotherăpagesăalike.ăIfătheăbranchăinătheătargetăeconomicăareaăisă separate from the headquarters the instrumental approach would be more appropriate as the information contained will be specific. (different address, different contact details, etc.) Instrumental translation isă definedă asă aă “typeă ofă translationă processă whichă aimsă ată producing in the TL an instrument for a new communicative interaction between the sourceculture sender and a target-cultureă audience,ă usingă (certaină aspectsă of)ă theă STă asă aă model.”ă (Nord, 1997: 139) 29 ăSeeă2.3.1.1,ăClassificationăbyălevelă(degree) 92 In my opinion, in the case of website localization with a focus on the end-user the translator has to produce an output that should be perceived as an original in the TT and furthermore uses as a starting point keywords and references used by the web-users in search engines30. This view is in accordance with Mona Baker’săviewăonătheăroleăofătheătranslator:ă“Like any writer, a translator has to take account of the range of knowledge available to his/her target readers and of the expectations they are likely to have about such things as the organization of the world, the organization of the language in general, the organization and conventions of particular text types, the structure of social relations, and the appropriateness or inappropriateness of certain kinds of linguistic and non-linguistic behaviour.”ă (Citedă in Dimitriu 2002:13) This approach is essential for online businesses in the era of inbound marketing. (Fishkin and Høgenhaven 2013) Equifunctional translation is used on websites in cases such as product manuals. Heterofunctional translation in commercial website localization may primarily be used if the target market based on age is different from one market to the other. This is not the case as a website will already have adapted its source content to various age groups; hence the localizer will localize content already adapted to an age group. For example, Physical products may be built modularly and have simpler modules for children and more advanced modules and features for adults. An even more illustrative example of a modular product, aimed at various age groups, may be that of online image editing software. For children, there may be fewer buttons and controls. Most webpages, however, will require a homologuos translation approach as the content must be localized entirely to the target locale and include popular search terms, all to convince buyers from the target market to buy. Numerous webpages will be composed as advertisement pages (content marketing), and advertisements are like poetry translation completed by a poet, that is, translating ads by an advertiser. The role of instrumental translation is to function as a communicative act in its own right, independently of the source text, and the outcome, the translation as a product, is assessed based on how the communicated message performs. On websites, Google Analytics [48] is one tool that can be used to analyze how content performs in terms of ROI. 30 ăForămoreăinformationăreadăsubchapteră2.6,ăWebsite localization and search engine optimization 93 InămyăopinionăNord’săhomologuosăinstrumentalătranslationăisătoăaăgreatăextentătheăequivalentă ofă copywriting.ă Bothă processesă areă basedă onă theă sameă instructionsă fromă theă commissioneră (sameăsubjectămatter,ăwhichămayăbeărequestedăasăaăspecificăparagraphăorăpopularăsearchătermsă toă beă included),ă andă theă stressă isă onă theă receiveră ofă theă message,ă whoseă responseă cană beă assessed.ăTheădifferenceăbetweenăaăcopywriterăandăaătranslatorăisăthatăwhileătheătranslatorăhasă toă beă proficientă ină bothă theă SLă andă theă TL,ă theă copywriteră needsă excellentă languageă andă culturalăskillsăonlyăinătheăTLă(usuallyăhis/herămotherătongue).ăAlso,ătheăoutputăinătheăcaseăofă translationăisă STădependentă whereasăină theăcaseă ofăcopywritingăthereă areănoăSTădeterminedă restrictions.ă Figure 20: Services required by small companies. Source: www.elance.com 94 Aă usefulă studyă wouldă beă toă determineă theă numberă ofă employersă thată requireă translationă services,ă asă comparedă toă thoseă requiringă copywritingă (ină theă targetă language)ă services.ă Awarenessăofătheăadvantagesăandădisadvantagesăofătheătwoăapproachesăcouldăalsoăbeăincludedă inătheăanalysis. 2.5.4 Machine translation for website localization In a global community, mediated through the Internet, everybody needs to understand or produce information efficiently. From the perspective of web-users navigating the web of international information it is essential to understand various texts in unfamiliar languages that s/he may discover accidentally. E-businesses tend to come forward and provide information in the language of the visitors. Translators, with an increasing demand of their services, need tools for delivering translations ever faster. Translation technology has evolved as an answer to the need of the various types of users and is often referred to as machine translation (MT). MTă isă definedă byă Theă Europeană Associationă ofă Machineă Translationă asă “theă applicationă ofă computersătoătheătaskăofătranslatingătextsăfromăoneănaturalălanguageătoăanother”ă[49].ăArnoldă etăal.ădefineăMTăasă“theăattemptătoăautomateăall,ăorăpartăofătheăprocessăofătranslatingăfromăoneă humanălanguageătoăanother”ă(1994:ă1).ăDependingăonătheăgoalăofătheăusersăasăseenăabove,ăMTă can be used for several tasks. According to Austermühl (2011:5), MT involves:      Assimilationă(Gisting) Dissemination Post-editingă(Revising)ă Pre-editingă(Controlledălanguage)ăă Trainingă(Terminology)ă Of interest for the website localization process are the first four. Indeed, one can use for instance Google Translate [41] or Microsoft Translator[42] by pasting in text from various sources for assimilating information. If on a website, depending on the browser used, Chrome or Internet Explorer, the web-user can access the page to be translated through the contextual menu, by right-clicking anywhere on the page. 95 Figure 21: Website contextual menu for translating pages in Chrome and Internet Explorer The picture above shows a simple mode of accessing the translation tools of Google and Microsoft. Internet Explorer, on the right, offers more contextual options, like allowing from this menu the choice of languages. Assimilation is a type of translation based on unassisted machine translation as the user will rather benefit from lexical translation only, especially if there are no sufficient corpora in the pair languages and is useful only for gisting.ăFromăSager’săperspectiveăthisăisătranslatingăwithă a new function in relation to the original (1983: 124), but accomplished automatically. MT is of much more use for websites that wish to automate to some extent the translation of their content into an acceptable publishable format. Dissemination is the reverse approach to assimilation as companies through translating content disseminate the information in the language of the potential users. Due to the fact that MT alone cannot produce an acceptable translation human intervention is required and is called assisted machine translation. Hutchins and Somers (cited in Austermühl 2011:4) distinguish between human assisted machine translation and computer assisted human translation. Both fall in the more general category of CAT (Computer assisted translation). Pym considers that nowadays all translations are to some extent computer assisted. (2014:120) However, in my opinion, delimitation between the two approaches is necessary. The main difference between the two approaches is the percentage of human effort input or as the terminology suggests, which of 96 the parties involved, the technology used or the translator, bears the greater importance in the productionă ofă theă translation.ă However,ă qualityă ofă theă finală outpută isă theă translator’să responsibility. Quality is achieved by the translator or translator supervisor through Postediting (Revising). The scale of post-editing is determined by the corpora used by the MT. In my opinion MT is only relevant in terms of the speed at which the translation can be completed. The speed of achieving a final deliverable translation depends on the language pair. Pym and Garcia, cited in Pym (2014:127), demonstrate through empirical research that for some language pairs, machine translation technology has been improving to the phase when it can produce satisfactory outcomes. I consider that efficiency depends on the usage of languages for websites content [45]. The higher the percentage of content in a certain language, the higher the efficiency of the MT, as both Google and Bing crawl and add to their databases website contentă ină thoseă languages.ă Austermühl’să experimentă (2011:10)ă bothă confirms and contradicts this statement. In the case of German and Spanish the speed increased significantly, over 30%; however, Chinese, the second most used content language after English showed that the translation produced by means of MT cannot be used. Pre-editing (using controlled language) as already seen in subchapter 2.3 Internationalization (I18n) is a concept that can be applied successfully during the website localization processes. MT applies successfully to website localization all the more so as tools such as Google Translate [41] and Microsoft Translator [42] are freely available and use the largest existing corpora,ă theă Internet.ă Qualityă dependsă onă theă translators’ă experienceă andă theiră skillsă ină localization but also on the pre-editing of the source text. According to Lionbridge [50] there are three types of approaches to MT: rule-based, statistically-based, and example-based. While rule-based MT requires mapping grammar rules from one language to others, statistical machine translation and example-based machine translation require large volumes of electronic text of similar content, which are parallel corpora. I personally opt for the free online machine translation tools as they use a descriptive approach to translation by collecting corpora from the existing websites crawled by the search engines. Some language pairs can be translated more efficiently because crawlers 97 added to the databases parallel corpora from the multilingual EU website. Most often this applies successfully to law related content. As Austermühl suggests in the same article (2011:15) Massive Online Collaboration (MOC) content translation or crowd translation is also a type of translation showing an upward trend, not produced by professional translators but by professionals and enthusiasts in a certain field. For this reason, such translations often are of highest quality. Similarly, Peter Newmark in an interview mentions that professionals mayă produceă aă betteră outcomeă thană aă professională translator:ă “The person who is a doctor would produce a better result. She would make sure the text makes sense, then a nonspecialist could touch up the work.”ă(David Shea 2005:394) This type of text production is also crawled and added to the parallel corpora of Internet and that makes other non-law content readily available for MT. The interface of Facebook is an example of MOC translation. Considering e-commerce, there are hundreds of renowned multilingual sites that can also be used as a source for parallel corpora. Amazon, eBay, Apple Store, Google Play and all the websites discussed in subchapter 4.2, Benchmarking website localization, an SEO perspective, are crawled by search engines and the quality of the content is human generated or mediated, either through translation or through copywriting. All the content of these multilingual websites add up to the database available for machine translation and thus the efficiency of online machine translation tools is improving noticeably from year to year. 2.5.5 Translation memory and machine translation Translation Memory can be an alternative to machine translation. TM is basically a software product that aids translators to automate to a certain degree the translation and localization processes.ăLionbridgeădefinesăităasă“aădatabaseăthatăstoresăsegmentsă(usuallyăsentences)ăthată your translators have already translated. Translators use your TM to automatically populate sections of text with previously approved translations. Previously translated material that is similar, but not identical (a <<fuzzy match>>) is plugged in and reviewed by the translator – butăitădoesn’tăhaveătoăbeătranslatedăfromăscratch.”ă[51] If we compare TM to MT one important aspect is that TM is based on aligned translation units (words, phrases, idiomatic expressions), meaning that one source text unit has a corresponding approved target text unit. A TM feature that is beneficial for the translator is 98 the repeatability of text. Of importance to the website localization process is the measure of the translation unit (TU). According to Hatim and Munday, a TU “mayă beătheăindividuală word,ă group,ă clause,ă sentenceă oră evenă theă wholeă text”ă (2004:17).ă Newmarkă (1981:66-68) considers translation units to be hierarchical: full text, paragraph, sentence, clause, group, word, and morpheme. While all these units are useful in terms of human translators from the perspective of TM the smaller units are preferable. Vinay and Darbelnet (cited in Hatim and Mundayă2004:17)ădefineăTUăasă“theăsmallestăsegmentăofătheăutteranceăwhoseăsignsăareălinkedă ină suchă aă wayă thată theyă shouldă notă beă translatedă individually”.ă Fromă aă TMă perspectiveă Malmkjær (2001:286) considers the unit of translation the equivalent of “theăTTăunităthatăcană beă mappedă ontoă aă STă unit”.ă Thisă shouldă beă alsoă theă unită levelă foră MT.ă Ină keepingă withă Malmkjær’să definitionă ofă TU,ă ină aă controlledă languageă inpută environment,ă whereă qualityă isă sought for, larger TUs are not viable. In Translation Memory, TUs are aligned at expression level or even sentence level depending on the recurrence of various units, for example, translating products and product characteristics. In website marketing search engines value uniqueness and added value. Repetition of content (be that at expression, sentence or paragraph level) is penalized. The larger the TU, the higher the possibility of being categorizedăasă“self-plagiarist”.ă Inăwebătextă termsă ităwillăbeăconsideredăasăduplicateăcontentă and thus not score high in SERPs. Legal texts can easily be faulted as there is little room for originality. Even if statistical MT is used, the results show that TU cannot be clearly determined. Kenny concludesă thată TUă areă “repeatedă stretchesă ofă sourceă textă consistently receive[ing] the same translation in a target language, then this can be taken as quantitative evidence that there are units of source and target texts between which a relatively stable translation relationship exists”ă(2009:304). 99 Figure 22: Statistical machine translation determined translation unit (TU) In the case of Google Translate [41] the TU can range from one on one word mapping to several to several words. Please note that while the translation needs revision, the MT is aware of intercalated TUs. In the ST quantitative is intercalated, while in the TT cantitative is at the end of the TU (syntactically correct, morphologically incorrect: the adjective form should have been in the feminine singular – cantitativă). Unlike MT, in the case of human translation the TU in ST and TT can differ (a one-word TU in a ST may need explicitation and transformed into a sentence unit in a TT). This corroboratesă Catford’să opinionă thată translationă unitsă shift.ă (ină Dimitriuă 2002:96)ă Ită isă evenă more so in the case of SEO aware website localization. From the perspective of my approach to website localization, that of shifting focus onto the end-users, translation units are keywords found in search engine statistics. Mapping of TU from ST onto the TU from TT is determined by the web-user from the target culture.31 One drawback of the TMs is that, unlike MT, they do not provide instantaneous translation. They do automate to a certain degree the translation process, but while MT can use several TMs, TM cannot include MT. MT is a translation engine, while TMs are databases. TM and MT are not exclusive but complementary tools to speed the turnaround of translations and localization. Besides TMs and MT any translator should use other general and specific sources as well, depending on the domain in which the source website operates. Apart from using Google Translate or Microsoft Translator, translators can opt for other free or proprietary solutions such as itranslate4.eu [53], Tapta4UN [54] XMT Cloud, CLWE, use 31 For more information see subchapter 2.6, Website localization and search engine optimization 100 general reference websites like Wikipedia, terminological resources such as TermWiki, and OmegaWiki, and corpora and translation memories like My Memogg, and Wordfast's VLTM. 2.5.6 Translation through parallel corpora and the problem of originality A parallel corpus is composed of two or more texts, in the same language (a text written in an archaic form and in its modern version) or in two or more different languages. Like TMs, texts in parallel corpora can be aligned at various levels. Such examples of parallel corpora are the bilingual editions of classic authors for the purpose of second language learning. In suchăcasesă evenătheăsourceătextămayăbeăalteredăsoăthatăităcorroboratesătheăstudent’sălevelăofă knowledge in the second language. In my opinion in such a case one could further speculate onăNord’săheterofunctional translation (1997: 47-52) and speak of heterofunctional intratranslation (changingăSwift’săoriginalătextăintoăaăsimplerăandămoreăaccessibleăEnglishătext)ă and heterofunctional inter-translation (Swift in Romanian adapted to children). The size of the translation unit (TU) can be variable in the case of a parallel corpus, but most often it is at sentence level - Déjà Vu (http://www.atril.com) and other similar tools align a text and its translation at sentence level; algorithms are written for sentence alignment: Tufi ă et al. (2006), and Gale and Church (1993). Kenny cited in Munday (2008:181) also acknowledges that parallel corpora is aligned at sentence or paragraph level. While alignment of corpus at sentence level can help at reusing translation segments and ensure the high quality of the output, the translator may be confronted with originality issues with regard to the final output. Translation Studies by default do not include the issue of originality in translation (with the notable exception of literary text translations). The extensive usage of parallel and aligned corpora and TM is a proof in support of my point of view. However, as previously mentioned, in website localization the originality of content is important. One might claim that if the source text is original then the TT will also be original. With the usage of TM, parallel corpora and MT - automated translation in general, texts ultimately converge into being translated in a similar manner. It results in either duplicate content within the same website or duplicate content shared with other websites. The translator unaware of SEO will faithfully replicate the ST in the target language maintaining high quality in terms of grammar, spelling, terminology and context but through the repetitive usage of the same 101 terminological databases, within the same field of human activity, translated content will be more or less the same, the translation going only as far as a factual replication, both from human and search engine perspective. ThisăstatementăcorroboratesăHatimăandăMunday’săviewă onă Toury’să law of growing standardization – “TTs generally display less linguistic variation than STs”ă (2004:7). This will lead to penalizations in SERPs. Search engines measure texts from a statistical point of view. They compare strings of texts found in texts on the same website and different websites. Consequently, a more efficient approach is transcreation. 2.5.7 Transcreation in localization. Transcreation is seen as a specific type of translation that entails more than the simple translation of ST. It involves creativity, particularization, and emotional transfer, where parts of the ST are translated while others are adapted to the local market. The closer in terms of cultural values the ST and the TT, the more straightforward the output of the TT. Apart from being proficient in translating, the transcreator must be skilled in creating content for the local market while maintaining brand unity, through the use of specific terminology and style guides. Transcreation is rather an industry coined term. Common Sense Advisory defines transcreationăasătranslationă“appliedătoămarketingăandăadvertisingăcontentăthatămustăresonateă in local markets in order to deliver the same impact as the original. The term may be applied when either a direct translation is adapted, or when content is completely rewritten in the locală languageă toă reflectă theă originală message.”ă (Rayă andă Kellyă 2010:ă 2).ă Considering this definition, one can see that transcreation also shares features with the localization process itself. However, while transcreation acts more at content level, localization acts at a more general level. Transcreation is seen in the above definition as an encompassing concept for both adaptation (Bastin 2009) and instrumental translation (Nord, 1997) or copywriting [47]. O'Hagan and Mangiron use extensively the term transcreation in the context of game localization in order toă embraceă “[…]ă theă highly structured use of multimedia and multimodality, involving the verbal and the non - verbal”ă (2013:198).ă Ină myă opinionă theiră perspective is a rather related to discourse and semiotic issues. 102 From a more traditional Translation Studies perspective, transcreation is in consonance with Nord’să functionalistă approachesă (1997:ă 47-52).The final output must be created as if it had been produced within the target culture. As shown in subchapter 2.5.1 Translation, adaptation, copywriting, copywriting mostly involves the same processes as transcreation. Copywriting on the one hand, and transcreation or functionalist approaches on the other, are not different in terms of the outcome. However, in the case of copywriting the influence of the ST is non-existing, and a copywriter does not need to be a translator as well. In the case of transcreation and functionalist approaches, even if the orientation and focus are on the TT and on the context in which it will function, there is a significantly higher possibility of contaminating the TT with cultural elements from the ST. The copywriter is by default rooted in the target culture32. Transcreation should be used in website localization depending on the level of webpage. The more general the page, the more transcreation can be used (homepage versus product description page). The closer to advertising-like style, the more open the translation approach. While transcreation is more appropriate for the advertising style, translation is appropriate for instructional and legal content, where accuracy is primordial. In the case of websites, due to multimodality and the semiotic complex, transcreation may be a more appropriate term to describe the modifications that must be operated on text content. As seen when discussing GILT processes, transcreation is actually localization applied to content, that is localization applied to everything but the technical and legal aspects of the website. 2.5.8 Reader/user-oriented translations and search engines In website localization, the focus of all the processes should be on the web-user; therefore, the translation must also focus on how users might be looking for a certain piece of information. According to Newmark, depending on the text-type, an expressive text concentrates on author and style, an informative text on content, a vocative text on the reader (1981: 12-15; 1993: 39-42). These text-types can all be found on e-commerce websites: the expressiveness of the original translates in preserving branding features, both in the ST and 32 For further information on copywriting and translation see Which Way Website Localization: Translation or Copywriting? (4.4) 103 TT, informative texts are on any service or product page, while vocative texts are common to homepages and other pages that use call to action words. Newmark defines communicative translation asă “attemptsă toă produceă on its readers an effectă asă closeă asă possibleă toă thată obtainedă onă theă readersă ofă theă original”ă (1981:ă 39);ă thisă definitionăisăveryăcloseătoăNida’sădynamic equivalence (1959:ă19).ăVermeer’săSkopostheorieă (1980),ăandăNord’săfunctionalistăapproachesă(1997)ăare more recent theories that shift focus onto the target text and hence on how it can be optimally received by the target audience. The key benefit of nowadays technology, namely tracking users, is that it assists translators and localizers to record, modify and improve continuously the translation from the webuser’s perspective33. Also, reader/user-oriented translations should be seen from the perspective of how the translated text is crawled and indexed by search engines. Search engines are the main link between the translated text and the potential reader/customer. If the translated text is not appropriately optimized for search engines the translation itself is worthless as no one will possibly find it. The translatum needs to stand out through the useful information it provides otherwise it will disappear inătheă“noise”ăofătheăInternet34. As can be seen, in website localization, user-oriented translation is desirable, achievable, and improvableăthroughătheătechnologyăthatăisăatăeveryone’sădisposal.ăTranslationăandălocalizationă should be web-user and search engine oriented. 2.5.9 Semasiology and onomasiology Semasiology and onomasiology are two branches of lexicology which represent two different approaches to the term-concept relation. Semasiology is defined by Oxforddictionaries.com asă “The branch of linguistics that deals with words and phrases and the concepts that they represent”ă[55].ăTheăsameăonlineăsourceădefinesă onomasiologyă asă“the branch of linguistics that deals with concepts and the terms that represent them, in particular contrasting terms for similar concepts, as in a thesaurus”ă[56]. 33 34 ăForămoreădetailsăseeăsubdivisionă2.5.9,ăSemasiology and onomasiology ăMoreădetailsăinăsubdivisionă2.6.4,ăWeb content translation and search engine optimization (SEO) 104 Grzega’să andă Schöner’să viewă onă theă topică isă thată “Theă goală ină onomasiologyă isă toă findă theă linguistică forms,ă oră theă words,ă thată cană standă foră aă givenă concept/idea/object”ă (2007:7).ă Onomasiologyăisă“theăstudyăofădesignations,ăwhereătheălinguist starts with an extralinguistic conceptăandălooksăforăitsăformalăverbalizations.”ă(Grzegaă2012:271) Semasiology, on the other hand, examines terms first and then the concept or concepts designated. The semasiological direction in Saussurean terminology (2011) is from the signifier (term) to the signified (concept), that is to observe which concepts are associated with a specific word. Onomasiology analyzes the term-concept relation from an opposing perspective. The direction is from the signified to the signifier, from a concept to a word that may be used to represent or name the concept. Semasiology, applied to localization and translation, contribute to increasing the efficiency of these processes. The use of synonyms, hyponyms or heteronyms and the choices made according to the ways in which the users input their queries in search engines help website content producers to build the message in a more accessible manner making it reachable by users and search engines alike. Onomasiology is of particular interest to my research and, more specifically, synchronic onomasiology because it is during the localization process that the localizers of websites need to consider how users designate the concepts they are looking for. In Lakó (2010a), I proposed a similar approach. In these discussions on an onomasiological strategy to text production I will refer strictly to one-word units (simple or compound). However, the practice of using search engines shows that the naming process should be extended to expressions of two or several words. Grzega mentions in his English and General Historical Lexicology that some linguists viewăonomasiologyăasă“grammaticalăformsăthatăcanăstandăforăaă givenăfunction”ă -<<How can I express future time?>> or <<How can I greet somebody?>> “foră conversatională patternsă thată cană beă usedă ină aă givenă communicativeă task”ă (2007:7).ă Ină terms of localizing web content, the localizer should be thinking from the perspective of the web-useră “Howă cană Iă findă thisă product/service?”ă oră “Whată wordsă shouldă Iă useă toă findă thată information?” 105 The importance of the onomasiological approach to website content localization can be expressed in simple terms with examples such as naming the “theăthirdăseasonăofătheă year,ă when crops and fruits are gathered and leaves fall, in the northern hemisphere from SeptemberătoăNovemberăandăinătheăsouthernăhemisphereăfromăMarchătoăMay”ă[57] as autumn in the UK versus naming it fall in the US. Naming things or concepts differently can occur within the borders of the same country, with different naming of the same concept from region to region, called regionalisms. As shown in subchapter 4.3, Error! Reference source ot found., localization often will not even require translation, as localization can happen in the same language area. Reverting to my previous example, the localizer would use Autumn Sale for certain markets while Fall Sales for others. Figure 23:Term usage adapted to target Figure 24: Term usage adapted to target market - autumn35 market - fall36 As already shown in the introductory chapter, inbound marketing, through website content is the most efficient technique to deliver pertinent information that is straightforwardly accessible with the aid of search engines. Consequently, it is important to research the keywords by which web-users look for information that relates to the offer of products or services. Regardless of the type of localization (involving translation or not), keywords must be contained in web content and those keywords must match the signifiers used by the webusers to refer to common concepts (products or services). While this approach is not new, academics in Translation Studies most often touch upon the issue of onomasiology in passing and tend to overlook its importance in online communication. 35 http://www.malaysiadealsales.com/babyjaya-autumn-sale-damansara-utama-17-aug-2013-16-sep-2013/ http://coolcanucks.ca/motherhoodmaternity-ca-canada-canadian-deals-fall-sale-up-to-30-off-maternityclothing/ 36 106 A similarity with the onomasiological approach can be seen in Translation Studies in the case of terminology building: It is concept-oriented, when the person building the terminology provides one concept per entry. If localization is within the same language area (US, UK, Australia, etc.), during the Internationalization process a team can establish a terminological database where the internationalized or pivot source web content is paired with the various regional target web content. If translation is involved in the localization process, the keywords used by the search engine users should be reflected in the terminological database. Furthermore, inside the same target language, there should be several regional onomasiological entries, for each of the regions, if this may be the case, to mirror the naming of that particular region (a country, a state, a county, or a city). The process would still be keyword based. Case study The borrowing of a term or loan translationăthroughătheătranslatoră(Venuti’săforeignization),ă or by web-users that work in a specialized field may also be considered as relating to the onomasiological approach. In e-commerce sellers often target a group of specialized people, from a certain profession or with common hobbies. For instance, if research is conducted on keyword usage withă Google’să KeywordPlanner, sticker(e) is suggested, a non-standard yet high traffic term in Romanian (there is no such entry in the Romanian dictionaries – dexonline.ro). Such a term can be included into the Romanian target text, instead of etichetă, or adeziv or can be alternated with autocolant. If I compare baby-sitter with the Romanian bonă/bone using the same tool, the obvious main keyword is the autochthonous bonă or its plural form bone. Voucher ină Englishă isă definedă asă “Aă smallă printedă pieceă ofă paperă thată entitlesă theă holderă toă aă discountă oră thată mayă beă exchangedă foră goodsă oră services.”ă [58]ă Voucher was initially used by Romanian travel agencies as a certificate that entitles the owner to benefit from services bought at the agency. Its usage as a discount certificate began only later. There are several synonyms in Romanian borrowed previously that refer to the same concept, cupon (fr. coupon.), tichet (fr., engl. ticket), which are also used in marketing cupon reducere (discount coupon), or in an institutionalized governmental context tichet de masa (food voucher) (Monitorul Oficial [59]), tichet rabla(car voucher) [60]. However, according to Google Trends [61] and Google KeywordPlanner [62] voucher is the preferred term, even if it is a luxury loan. This preference also overlaps the institutionalized tichet so keywords such as voucher rabla, and its plural forms vouchere rabla are more frequently used than 107 tichet(e) rabla,ăevenăifăthereăhasăbeenăaăcontinuousăinstitutionalizedă“educational”ăinitiativeă with regard to the car renewal program ever since 2005. Figure 25: The term Voucher vs. cupon and tichet in Google Trends Cupon and tichet,ăisăaboută25șăofăvoucher, representedăinăblue.ăConsideringătheăgraphăfromăaă diachronică perspectiveă ită showsă that,ă untilă Marchă 2010,ă theă usageă ofă theă termă tichet outperformedăvoucher and cupon inătermsăofăqueriesăusedăbyăsearchăengineăusers.ă Fromătheăexamplesăabove,ăoneăcanăobserveăthatăcompetingăsynonyms,ăusedăasăsignifiersăforă theă sameă concept,ă thing,ă oră idea,ă cană beă eitheră recognizedă asă theă normă oră not.ăAă synonymă fromă aă groupă ofă severală synonymică wordsă andă expressionsă mayă becomeă theă dominantă one,ă whileătheăothersăfadeăaway.ăWeb-usersăhaveătoăbeăspokenătoăinătheălanguageătheyăcommonlyă useă evenă ifă thată isă notă aă standardă acceptedă usageă oră theă norm.ăThus,ă web-usersă imposeă theă signifiersă theyă useă ontoă theă onlineă marketingă professionals.ă Whileă academicsă favoră prescriptiveăapproachesăandătryătoăimposeăonăwebăcontentăuserălanguageănorms,ătheăindustryă appliesă bothă approaches,ă prescriptiveă ină theă caseă ofă brandă websites,ă andă aă descriptiveă approach,ăespeciallyăinătheăcaseăofăaffiliateăwebsites. While considerations at word level are useful from a diachronic and synchronic point of view for researchers in linguistics, to the industry of localization keywords made up of two or more words (for SEO purposes) are more relevant. Overall, both approaches, semasiological and onomasiological, should be considered and combined for a successfully personalized web localization outcome. The semasiological approach is a prescriptive method (its purpose is to unify, institutionalize and standardize language). By using as a precautionary measure various signifiers (auto, automobile, car, 108 compact, limousine, limo, machine, motor, motorcar, ride, sedan, station wagon, touring car vehicle, wagon, wheels) for the same signified the localizer could communicate to as many potential web-users as possible. The semasiological approach is used primordially by wellknown brands when they educate or inform potential clients. The onomasiological approach is a descriptive method as it perceives and examines the appearance of terms as inputted by web-users in search engines. 2.5.10 Glocalization or "looking in both directions" Language, more than ever is subject to changes, as it shapes to the needs of those who use it. Source texts, usually in the language of a dominant culture tend to influence to a greater extentă theă “minor”ă targetă languages.ă However,ă theă processă ofă translatingă hasă focused even more on the receiver of the translated text, culture and specific context. The beneficiary of the translatedă informationă “dictates”ă whată signifiersă andă whată contextsă shouldă beă usedă ină theă process of conveying the message into meaningful bits, even if misspelt or grammatically incorrect. The same applies to all the elements of a certain website. For instance, its design and usability can influence to a great extent the communication process. So, the translator needs to cover a gap between the source message and the receiver of the message who is dependent on certain pragmatic contexts. These contexts are to some extent known to the translator with the assistance of search engine tools. The translator should no longer have a prescriptive role in translating, in terms of correctness at word level, but should employ terms used by searchers. Still, at all the other levels the message should probably be as close to the target culture as possible, especially when the translator becomes a localizer. The localizer must find the balance between global influences and local features Although English is nowadays the lingua franca (the global element and main carrier of global values), the vast majority of Internet users opt for reading information on products in their own language (locales). This user preference translates into the decreasing of the percentage of webpages written in English in favor of national and regional languages. In recent years there has been a steady decrease of the percentage of webpages written in English, while other major languages gained exposure. According to http://www.internetworldstats.com/ English is still the most widely used but Chinese and Spanish follow closely. In 2010 there were 536.6 million pages in English, 444.9 million in Chinese and 153.3 million in Spanish. Japanese, Portuguese and German follow with 99.1 109 million, 82.5 and 75.2 million, respectively. Arabic comes on the 7th position with 65.4 million, followed by French and Russian ranking very close by the number of webpages with 59.8 million and 59.7. The last in the top ten is Korean with 39.4 million. All the other world languages make up the rest of 350.6 million webpages. This classification is a result of a more complex mix: - numberăofătotalănumberăofăspeakersăofăaăcertainălanguage - levelăofăInternetăaccess - broadbandăpenetrationărate - Internetătechnologyădevelopment - affluence - investmentăinătechnology - ICȚTăgovernmentalăprogramsă - pricesăofătelecommunicationăservices - percentageăofăpersonsăemployedăwithăICȚTăuserăskills - teleworkingăandăorganizationăculture - onlineăbuyingăactivitiesăandătheirăpercentageăfromătheătotalăsales Theăcombinationăofăallătheseăuniqueă(local)ăelementsăconstitutesătheăglobalăInternetămedium.ă Theăinteractionăbetweenătheălocalăandătheăglobalăelementsăisăcalledăglocalization. Theă termă glocalizationă wasă coinedă byă Sony’să chairmană Akioă Morita.ă Ită isă theă resultă ofă blendingăglobal andălocalization,ăandărefersătoăglobal localization.ăGlocalizationăisăaăspecială typeăofălocalizationăasătheoristsăinăsocialăandăculturalăstudiesăseeătheălocalizationăofăproductsă andăservicesăasă“insiderization”ăandă“lookingăinăbothădirections.”ă(OhmaeăcitedăinăNederveenă Pieterseă2009:52)ăInăwhatăfollows,ăIăwillăanalyzeăseveralădefinitionsăforătheătermsăglocalăandă glocalization. Oxforddictionaries.com defines glocal asă “Reflecting or characterized by both local and global considerations “[63]ă whileă glocalization isă definedă byă theă sameă sourceă asă “Theă practiceăofăconductingăbusinessăaccordingătoăbothălocalăandăglobalăconsiderations.” 110 Whileă Oxforddictionaries.comă definitionsă areă fromă aă business-likeă perspective,ă thereă areă severală theoristsă ină socială andă culturală studiesă thată defineă glocalizationă fromă aă differentă perspective.ă Friedmană (2000:295)ă definesă ită asă aă coherent,ă non-tensioned,ă naturală process:ă “theă abilityă ofă aă culture,ă whenă ită encountersă otheră strongă cultures,ă toă absorbă influencesă thată naturallyăfităintoăandăcanăenrichăthatăculture,ătoăresistăthoseăthingsăthatăareătrulyăalienăandătoă compartmentaliseă thoseă thingsă that,ă whileă different,ă cană neverthelessă beă enjoyedă andă celebratedăasădifferent.”ăWhereasăFriedman’sădefinitionăfocusesăonătheăactorsăthatăinfluenceă theă glocalizationă process,ă Theă SAGEă Dictionaryă ofă Culturală Studiesă refersă toă productsă (culturală productsă –ăaămovie,ăaăservice,ăoră goods)ăinvolvedăină theăprocess:ă “Theă conceptă ofă glocalization,ăinăoriginăaămarketingăterm,ăhasăbeenădeployedătoăexpressătheăglobalăproductionă ofă theă locală andă theă localizationă ofă theă global.ă Theă globală andă theă locală areă mutuallyă constituting,ăindeed,ămuchăthatăisăconsideredătoăbeălocal,ăandăcounterpoisedătoătheăglobal,ăisă theăoutcomeăofătranslocalăprocesses.”ă(Barkeră2004:77) Ritzer,ăonătheăotherăhand,ărefersătoăglocalizationăasăanăoutcome.ă“Glocalizationăcanăbeădefinedă asă theă interpenetrationă ofă theă globală andă theă locală resultingă ină uniqueă outcomesă ină differentă geographicăareas”ă(2011:168).ăForătheăsameăconcept,ăRobertsonăusesătheătermăglobalizationă-ă “theănotionăofăglocalizationăactuallyăconveysămuchăofăwhatăIămyselfăhaveăpreviouslyăwrittenă aboută globalization”ă -ă toă reflectă theă sameă phenomenon:ă “globalizationă hasă involvedă theă simultaneityă andă theă interpenetrationă ofă whată areă conventionallyă calledă theă globală andă theă local,ăoră -ăinămoreăabstractăveină -ătheăuniversală andătheăparticular.”ă(2010:336)ăHowever,ăină thisăthesisă Iăusedătheătermă globalizationăasăanăintegratedăpartăofăGILT,ăasăaăpre-localizationă step.ă Ină myă opinionă globalizationă ratheră refersă toă theă convergenceă ofă allă theă elementsă thată constituteătheăglobalăamalgam. Aă definitionă encompassingă allă theă activitiesă fromă theă humană sphereă isă providedă byă Mendisă “Theăwayăweă—ănationsăandăcommunitiesă—ărespondătoăanăever-changingăinterplayăofăglobală political,ă economic,ă social,ă religious,ă andă culturală ethosă ată differentă localitiesă isă theă uniqueă processă ofă <<glocalization.>>ă Glocalizationă isă essentiallyă aă hybridă ofă globalizationă andă localization.ăGlocalizationăisălikelyătoă empowerălocalăcommunitiesăthroughăstrategicălinkingă ofăglobalăresourcesătoăaddressălocalăissuesăforăpositiveăsocialăchangeăandătoăbalanceăchangingă culturalăinterestsăandăcommunityăneeds.”ă(2007:2)ă 111 Theămainăpointăinăallătheseădefinitionsăisăthatăglocalizationăisăperceivedăasăaăcontinuousăandă simultaneousă processă ofă mutuală influenceă betweenă locală features,ă onă theă oneă hand,ă andă characteristicsăthatăhaveăbecomeăglobalăonătheăother.ăLocalăfeaturesăcanăbecomeăglobal,ăwhileă globalăonesă eitherăsufferăchangesăunderătheăinfluenceăofătheălocaleă (s),ădisappearăorăsurviveă onlyă ină theă moreă conservativeă communities.ă Locală andă globală shouldă notă beă consideredă asă antithetical.ă Locală andă globală shouldă beă seenă asă aă symbiosis.ă Theyă shareă aă symbiotică relationshipă andă oneă cannotă ignoreă theă other.ă Locală cană beă globală andă globală cană beă particularizedătoăfunctionăasălocal.ă Thereă areă hundredsă ofă examplesă toă supportă thisă idea,ă evenă ifă weă onlyă adoptă aă linguistică perspective,ă andă moreă specificallyă referă toă inter-borrowings.ă Visitingă http://www.vocabulary.com/ă toă lookă foră borrowingsă fromă Spanishă intoă Englishă willă showă aă listăofă164ăwords.ăTheăSpanishăwordsăenteredătheăEnglishăvocabularyăată variousătimesăfromă variousălocales,ăprimarilyăthroughăcontactăofăAmericansăwithăMexicanăpopulations,ăbutăalsoă fromă earlieră Europeană English-Spanishă encounters.ă Withă theă “expansion”ă ofă Americană Englishăintoătheăworld,ămanyăofătheăSpanishăloanăwordsăhaveăbeenăexportedătoăotherăculturesă asă well.ă Wordsă likeă alligator, armada, avocado, banana, barracuda, barbecue, cannibal, canoe, guitar, guerilla, lasso, toreador, tornado, vanilla,ă haveă beenă universalizedă throughă Englishă andă adaptedă toă theă pronunciationă ofă eachă particulară language.ă Theă sameă interinfluenceăcanăbeăseenăatăanyălevelăofăhumanăactivity:ăcultural,ătechnological,ăetc.ăToădescribeă theă sameă phenomenonă thereă areă otheră termsă ină useă asă well:ă mélange,ă hybridity,ă syncretismă (Nederveenă Pieterseă 2004)ă oră heterogenization,ă creolizationă (Ritzeră 2011),ă bută theseă areă directedăratherătowardsăculturalăissues. Consideringăhowăglocalizationăisăinfluencedăandăinfluencesăe-commerceăwebsites,ăIăcanăassertă thată theă termă glocalization,ă asă ană interactionă ofă bothă globală andă locală factors,ă cană beă seenă almostăonăanyăwebsite.ăMostăoftenăweb-usersămayăbeăimposedătheătaskăofălearningăandăusingă English.ă Numerousă websitesă areă localizedă justă foră theă mostă widelyă usedă languagesă onă theă Internet;ăotherăsitesăareă ată differentă stagesăofălocalizingătheirăpageăcontentă (forăinstance,ătheă descriptionăofăaănewătoolăonăGoogleăwillăbeăfirstăavailableăonlyăinăEnglishăandălateră“rollăout”ă intoătheăworldăwithăitsălocalizedăcontent)ăforăallăofătheălanguages.ăTheănameăofătheăproductsăoră servicesă isă ină English,ă andă foră theă sakeă ofă marketingă consistencyă theyă remaină ină Englishă (Thereă wereă casesă whenă aă productă nameă hadă toă beă changeă becauseă ofă languageă oră culturală issues).ăWhileă foră ană Englishă speakeră ită isă facileă toă predictă whată Google’să KeywordPlanneră 112 [62]ăisăintendedăfor,ăinătheăcaseăofăaănon-Englishăspeakerăităisăaăcompletelyănewăwordăthatăhasă toă beă learntă andă assimilated.ă Ofă course,ă oneă couldă argueă thată ită isă aă termă thată entersă theă vocabularyăofăaăcertainăsocio-economicăgroup,ăthatăofăSEO;ăsimilarly,ăwithăregardătoăadobe,ă mostăpeopleăaroundătheăglobeăwillăthinkăonlyăofătheăcompanyăandănotăofătheătermăoriginatingă fromăSpanishăandămeaningăsun-dried brick. However, glocalization on websites refers to providing global information to users from local social contexts around the world and allowing them to understand, retrieve, organize, share and produce content following the characteristics of their given specific locale, while retaining its globally accessible feature. The usage of icons for the main page, envelope for contact, map icon for location is globally accepted by any web-user, similarly to the square form for stopping or the double horizontal bars for pausing the music from playing on a music device has been accepted and used ever since its invention. In the case of e-commerce websites, I believe that the term glocalization reflects genuinely the mode inăwhichătheyăareăbuiltăandăused.ăWhile,ăinăgenerală IăsupportăNord’săinstrumentală translation (1997), it cannot be always successfully applied in the case of e-commerce websites,ă asă theă products’ă semiotică representationă (text,ă pictures,ă video)ă cannotă beă entirely written in such a way as to be perceivedăasă“an independent message transmitting instrument inăaănewăcommunicativeăactionăinătheătargetăculture,ă[…]ăintendedătoăfulfilăitsăcommunicativeă purpose without the recipient being conscious of reading or hearing a text which, in a different form, was used before in a different communicative situation.”ă (2005: 81) For instance, I cannot possibly put on sale a device and use its picture in the source content and change it for target culture, unless the product’să specificationsă areă differentă foră theă twoă markets. I could instead replace the background image; for instance Dacia Duster on Transalpina for the Romanian market and the same car make next to Stonehenge if targeting the UK market. Also, in the case of brands, most often they already contain the traits of the culture from which they originate. Documentary translation is closer to glocalization means. Venuti’săforeignization-domestication dichotomy applied to the resemioticization of website communication is closer to glocalization theory, with the mentioning that the foreignizationdomestication paradigm, like glocalization on websites, is a moving semiotic system. The boundaries between what is perceived as foreign or global versus domestic or local elements are continually shifting and this is reflected in website content maintenance. 113 Asă alreadyă mentionedă ată theă beginningă ofă thisă subchapter,ă Iă consideră thată glocalizationă isă aă specială typeă ofă localization,ă i.e.ă globală localization.ă Unlikeă generală websiteă localization,ă whereă instrumentală translationă appliesă (usefulă foră affiliateă websitesă thată usuallyă operateă ată nationalălevelăonly),ăglocalizationătheoryăacknowledgesăthatăthereăisăsimultaneityăofăbothătheă globalăandălocalăfactorsăinvolvedăinătheăprocessăofăwebsiteăadaptationăfromăaăsourceăwebsiteă toătheătargetăwebsite.ă Fromă aă moreă technicală perspective,ă glocalizationă appearsă toă beă aă bi-dimensională processă involvingă concurrentă globalizationă andă localizationă processes,ă andă bidirectională asă globală traitsăinfluenceătheălocalătraitsăandăviceăversa.ăOnătheăotherăhand,ăwithinătheăGILTăapproach,ă theă processesă deriveă fromă eachă otheră andă theă directionă isă fromă theă moreă general,ă globalization,ă toă theă moreă specific,ă localization.ă Theă direction,ă ifă eachă stepă isă carefullyă planned,ăisăunidirectional,ăandăchangesăandăupdatesăareăinitiatedăfromătheăglobalizationăstage,ă andămovingăthroughăinternationalization,ăthenălocalization,ăandăfinallyătranslation.ăThisădoesă notăimplyăthatăsignalsăfromălocalămarketsăcannotăbeăacceptedăandăappliedăatăglobalălevelăandă thenă reappliedă toă eachă ofă theă locală markets.ă Aă possibleă scenarioă couldă beă thată aă localizedă banneră provesă successfulă onă aă locală marketă andă thenă ită cană beă testedă onă severală otheră locală markets.ă Figure 26: Glocalization versus GILT 114 InăFigureă26,ăIăillustrateătheădifferencesăbetweenăglocalizationăandăGILT.ăWhereasăGILTăisăaă systematic,ă step-by-stepă approachă (ină practiceă notă withă clearlyă delimitedă borders),ă unidirectională (G11nă >ă I18nă >ă L10nă >ă Translationă /ă Copywritingă >ă Personalization)ă andă reflectsă theă necessaryă stepsă foră localizationă ratheră asă processes,ă Glocalizationă involvesă synchronousă andă bidirectională processesă oră states,ă acknowledgingă reciprocală influenceă ofă bothă locală andă globală factors.ă Theă tableă belowă synthesizesă myă findingsă regardingă theă predominantăfeaturesăofăGILTăprocesses,ăonătheăoneăhand,ăandăglocalization,ăonătheăother.ă GILT Glocalization systematic somewhat hectic planned unplanned unidirectional bidirectional somewhat unresponsive to changes highly responsive to changes institutionalized decentralized translation based copywriting centered asynchronous state synchronous state predominantly prescriptive predominantly descriptive predominantly domesticating foreignization and domestication cohabitation locale oriented (locale) global + locales intermingled (glocal) 4 distinct processes combined bipolar processes communication direction: global to local local to global enabled methodical production strategy observational and analytical framework rather viewed as a set of processes rather perceived as a product Table 3: GILT vs. glocalization Allă ină all,ă whileă glocalizationă fromă aă technicală pointă ofă viewă isă notă aă validă approachă asă ită cannotă beă implementedă systematically,ă likeă GILTă relatedă processes,ă glocalizationă theoryă reflectsă realityă ofă theă modernă interrelatedă andă interconnectedă world,ă theă “globală village”ă shapedă byă twoă forces:ă homogeneityă andă heterogeneity.ă Also,ă glocalizationă doesă notă specificallyăincludeătranslation,ăwhichămeansăthatăactuallyăGILTăisăoftenăonlyăGIL. 115 2.5.11 General considerations on website translation Fromă theă perspectiveă ofă webă localization,ă translationă hasă stillă muchă toă offer,ă becauseă aă distinctiveăapproachăisărequiredătoădemonstrateăwhereăadaptationăendsăandăwhereăcopywritingă begins,ă oră howă search-engineă basedă webă localizationă comparesă toă plaină webă localizationă ină termsăofăinformationăusabilityăandăreturnăonăinvestment.ă Theă translationă processă isă primordiallyă aă communicationă processă throughă aă mediator.ă Theă mediatoră ină localizationă isă notă theă old-schoolă translator,ă becauseă s/heă alsoă needsă skillsă ină localization. Theă localizationă processă isă notă merelyă aă linguistică exerciseă ofă translatingă contentă fromă aă sourceă languageă intoă severală targetă languages,ă bută aă managerial,ă marketingă andă technicală endeavorăasăwell.ăHowever,ăalthoughătechnicalăskillsăareăwelcome,ăIăhaveădemonstratedăthată byăusingăintelligentăapplicationsă(Figure 6: Example of directed-user approach),ătheătranslatoră cană avoidă facingă stringsă ofă “unintelligible”ă codeă andă focusă onă translatingă andă testingă theă translatum.ă Whenăconsideringătranslationăproper,ătheătranslatorăwillăhaveătoăcomplyăwithătheărequirementsă imposedăonăbyătheăassignoră(employer,ămanager,ăandăteamăleader),ăchangesăinătheăsourceătext,ă keywordsă fromă searchă engineă statisticsă (seeă upcomingă subchapter),ă modificationsă ină theă searchă resultsă andă variationsă ină theă marketingă strategyă foră aă specifică locale.ă Designă restrictionsăalsoăaffectătheătranslationăprocessăandăseveralărevisionsămayăbeănecessaryăevenăforă public. Whileălocalizationărequiresăaăwiderăsetăofăskills,ătranslationăisăaămoreăspecificăprocessădealingă strictlyăwithătext.ăTranslationăisăanăintegratedăpartăofătheălocalizationăprocess,ăwhichăinăturnăisă in-builtă intoă internationalization,ă whileă theă entireă processă isă commencedă byă theă assignoră throughătheăinitiationăofătheăglobalizationăprocess. 116 2.6 Websiteălocalizationăandăsearchăengineăoptimization Inămyăopinion,ăoneăofătheămajorăthemesăinăcurrentăreader-orientedătranslationătheoriesăshouldă beă theă issueă ofă commonă reference,ă thată isă sharingă commonă signifiersă (issuesă coveredă ină subchapteră2.5.9,ăSemasiology and onomasiology),ăandătheămediumăthroughăwhichătheăreaderă canăaccessătheăinformation.ăTheămainămediumăisămadeăupăofăseveralăsearchăengines,ăofăwhichă theămostănotableăisăGoogle.com.ăWhileătheăcommonăgroundăbetweenătheăinformationăprovideră (translatorăorăcopywriterăinătheăcontextăofăwebsiteălocalizationăasăagentsă ofăaăcompany)ăandă web-usersă isă madeă upă ofă searchă engines,ă Iă willă alsoă consideră searchă engineă optimizationă issuesăasăweb-contentămustăcomplyăwithăcertainăsearchăengineărules. 2.6.1 Search engines Looking for definitions for search engine, it appears that it is perceived as a self-explanatory term that needs no further clarifications. Few books on search engine technologies provide a definition for search engine. Kevin Curran defines search engine asă “aă programă thată willă searchă foră keywordsă ină documentsă andă thenă returnă aă listă ofă theă documentsă thată containedă thoseă keywords”.ă (2009:275). Oxforddictionaries.com define search engine asă“Aăprogramăthatăsearchesăforăand identifies items in a database that correspond to keywords or characters specified by the user, used especiallyăforăfindingăparticularăsitesă onătheăWorldă WideăWeb.”[64]ăIădoănotă agreeăentirelyă with this definition as users most often do not look for a specific website but for a certain piece of information. However, one can use a search engine to look for information on a specific website, either by using the search engine of the website being visited, using a wellestablished but customized search engine or by using site: on Google. For instance, site:website.com keyword1 keyword2 will search for results only on website.com looking for keyword1 and keyword2. A more accurate and all-embracing definition is provided by the online Miriam Webster dictionary. It defines search engine asă“computerăsoftwareăusedătoăsearchădataă(asătextăorăaă database) for specified information; also: a site on the World Wide Web that uses such 117 softwareă toă locateă keyă wordsă ină otheră sites”.ă Aă moreă elaborateă andă completeă definition is foundă onă theă sameă webpage:ă “Toolă foră findingă information,ă especiallyă onă theă Internetă oră World Wide Web. Search engines are essentially massive databases that cover wide swaths of the Internet. Most consist of three parts: at least one program, called a spider, crawler, or bot,ăwhichă“crawls”ăthroughătheăInternetăgatheringăinformation;ăa database, which stores the gathered information; and a search tool, with which users search through the database by typing in keywords describing the information desired (usually at a Website dedicated to the search engine). Increasingly, metasearch engines, which search a subset (usually 10 or so) of the huge number of search engines and then compile and index the results, are being used.”[65]. Search engines are perceived as a common ground for communicating purposes between companies, on the one hand, and potential customers, on the other. (Lakó 2007a:357) Theyă areă essentială “to establish a common context necessary for successful communication between advertisers and clients”.ă (ibid.) This corroborates the secondary objective of the thesis, that of analyzing search engines, as the main channel for fueling keywords.37 While these definitions may be sufficient to most people, of major impact on how web content is listed in SERPs, be that original or translated and localized, are the algorithms by which the webpages are considered relevant or not to the queries entered by web-users. AccordingătoăGoogle,ă“[a]lgorithmsăareătheăcomputerăprocessesăandăformulasăthatătakeăyoură questionsă andă turnă themă intoă answers.ă Todayă Google’să algorithmsă relyă onă moreă thană 200ă uniqueăsignalsăoră“clues”ăthatămakeăităpossibleătoăguessăwhatăyouămight really be looking for. These signals include things like the terms on websites, the freshness of content, your region andă PageRank.”[66]ă Whileă theă definitionă isă ină layman’să termsă ită confirms that Google considers to a great extent the content of any webpage, by using over 200 indicators, but also verifies information such as background information on website owner and terms of use, creation of new content and updating of existing content, the region which is targeted by each localized version of a website and PageRank.38 ăItăisăworthămentioningăthatăGoogleăappliesătheăinboundămarketingăprinciplesăandăitsăsearchăengineăalgorithmsă toăitsăfreeăemailăservice,ăGmail.ăAdvertisingăisăcontextualized,ăandăbasedăonăkeywordsăusedăinătheăemailăthatăisă beingăread.ă(Lakóă2008:237-238) 38 ăGoogle’săassessmentăaboutătheăimportanceăofăaăwebpage 37 118 Asă oneă cană seeă theă maină factorsă ină determiningă aă page’să positioningă ină theă SERPă areă bothă human and machine determined. These are covered by the industry under the concept of search engine optimization or SEO. 2.6.2 Search engine optimization (SEO) The academic community, while admitting the existence of search engine marketing (SEM) including SEO, fails to recognize its significance for website localization. For example, in Jiménez-Crespo’să otherwise elaborate and well-written Translation and Web Localization, while specifically addressing website localization, SEM is only summarily mentioned among the website localization processes. (2013: 31) As one would expect, most information on SEO can be found on the Internet, from professional and experienced SEO practitioners. First I will reflect on several definitions, but interestingly enough, many consider SEO to be an art, or science, or a combination of the two. However, I disagree with definitions that consider SEO an art as most often the humankind encompasses under the concept of art everything that cannot be explained. SEO is based on exact, measurable factors: moz.com statesăthată“SEOăisătheăpracticeăof improving and promoting a website in order to increase the number of visitors the site receives from search engines. There are various aspects to SEO, from the words on your page to the way other sites link to you on the web. Sometimes SEO is simply a matter of making sure your site is structured in a way that search enginesăunderstand.”[67] Jerriă L.ă Ledfordă definesă SEOă asă “aă living,ă breathingă conceptă ofă maximizingă theă traffică thată yourăwebsiteăgenerates,ăandăbecauseăităis,ăthatămeansăthatăit’săaăconstantlyămovingătarget.ă[…]ă Search engines are constantly changing, so the methods and strategies used to achieve high search engine rankings must also change. (2009:19).ăSEOăisăalsoădefinedăasă“aăsetăofămethodsă aimed at improving the ranking of a website in search engine listings. The term also refers to anăindustryăofăconsultantsăwhoăcarryăoutăoptimizationăprojectsăonăbehalfăofăclientăsites.”[68] As a sum up of all the definitions above I would define SEO as a subcategory of search engine marketing consisting of laborious processes that facilitate webpages to rank high in search engines result pages. An SEO engineer predicts what human visitors search for and 119 understands how search engine algorithms work. Ranking high translates as higher volume and quality of traffic by means of search engines (inbound marketing). Next, I will list a series of factors that influence the positioning of a page in the SERPs. A short and descriptive list is provided by searchengineland.com. Their Periodic Table Of SEO Success Factors contains both the factors involved and the impact of each factor on page ranking. [69]. SEO methods can be divided into on-page optimization and off-page (offsite) optimization. On-page (onsite) optimization refers to processes that applies directly to each individual webpage (content quality, keywords, etc.) whereas off-page optimization primarily refers to link-building. Of interest to the translator and localizer is on-site optimization. According to searchengineland [70] and in accordance with Google SEO guide [29], the most important factor is content quality and uniqueness. Combined with keyword research and keyword inclusion in content plus using relevant titles and subtitles is what any translator and localizer should acknowledge and apply. While in theory these are activities that can be accomplished with minimum training by any translator, in practice SEO specialists provide the keywords for both the source content and for the localized content. The SEO professionals should be native in the language of the target market for outstanding results. Searchengineland [70] also marks as an important factor authorship. While it is listed under the off-page optimization column it depends indirectly by the author of the original content and by the translator or copywriter of the localized content. For more information on authorship please refer to the following subchapter. 2.6.3 Invisibility of the translator: when authorship matters I have shown in the previous subchapter that there are hundreds of factors that determine the position of a certain webpage on SERPs, regardless of the fact that content is original or localized. One recent factor that has gained importance in SEO is authorship. Through authorship Google’s intention is to further filter content written by authorities in a certain domain of activity or authors acknowledged by others through social media tracking. Therefore, the more visits to articles and the more comments generated, the more acknowledged the author by Google. Thus, employing a certain author to write a review aboutăaăserviceăorăproductămayăbeăveryăimportantăinăpushingăaăcompany’săoffersătoătheătop of 120 SERPs. Furthermore, from the perspective of website localization, if the original content is written by a highly qualified copywriter (author), there will be several issues when considering the localization of the original content. First, the original copy is not standardized or internationalized. I have demonstrated it is a necessary step for efficient localization. Second, a well-established copywriter is recognized as such only by a certain market. Translating and localizing the copy of a copywriter in terms of authorship would not yield any importance in a target market as most likely the author is not known in that locale. Third, most probably, if by abstraction we consider only authorship as a ranking factor, content written in English will rank differently on Google.com than the translation of the same content in Romanian on google.ro because an American author is not acknowledged by Romanians; it is not part of the Romanian locale. Therefore, the invisibility of the translator is no longer required. On the contrary, a skilled translator would in time be acknowledged by the community for which s/he translates and, as a result, by Google as well. In practice, I have not met translators that acknowledge the original authors as Google is far from being able to recognize translations as plagiarized content. This is a statistical measurable proof that translations are unique. Much non-English content appears over the web as actually unacknowledged localized content of original English content. (Even more so, apparently, there have been reports in Romania that research papers were actually translations of research work originally written in English or French.) While in terms of localization such content is close to perfect, in legal terms is plagiarism which cannot be tracked, hence is seen as original content. However, those who build this type of content are potential candidates for companies who are looking for localizers in a certain locale. The most appropriate technique to find them is to search in Google by the topic you are interested in but using keywords in the target language. Authors that rank high inăsearchăenginesăareătheămostălikelyătoăsucceedăwithăaăcompany’sălocalăwebsiteămarketing. I have demonstrated previously that localization does not necessarily involve translation. If the localization is directed towards an Anglophone market probably the same content may rank high in all local SERPs, for instance it would be similar for both the US and the UK market. 121 Considering the above mentioned issues, seemingly the most appropriate approach to website localization is to choose copywriting over translating. The author can be either a person, as seen above, or a publisher, representing a company. In the case of multinational companies and well-known brands most likely the option is to have the content displayed as belonging to that certain brand. In this case, the invisibility of the translator is desirable. In practice, blogs and affiliate websites use persons’ăname for the author approach, whereas company websites, be that small or big, will use publisher name/brad as the author. Also, copywriting is more appropriate for blogs and affiliates while translation is more suitable for company websites. 2.6.4 Web content translation and search engine optimization (SEO) Ină thisă partă Iă wishă toă emphasizeă theă importanceă SEOă mustă playă ină webă contentă translation.ă Theă noveltyă ofă theă approachă isă thată theă localizationă processă isă aă continuousă process,ă ină theă senseă thată theă translatoră hasă toă adjustă theă webă contentă toă theă requirementsă ofă theă searchă enginesăandătoătheăsearchăqueriesăinputăofătheăsearchăengineăuserăinăorderătoăadaptătheăofferătoă theă requestă (negotiatingă theă signifiersă andă contexts).ă Localizingă websitesă mustă beă aă continuousăprocess.ăAsătheăsourceăwebsiteăchangesăinăorderătoăadjustătoătheărequirementsăofă webăsearchers,ătheălocalizedăversionsăhaveătoăcomplyăwithătheărequestsăofătheăsearchersăfromă theălanguage/cultureăofătheătargetăwebsite.ăContentămaintenanceăhasătoăbeăachievedăbothăwhenă theă sourceă textă isă modifiedă (Pymă 2010:3)ă andă whenă theă statisticsă foră theă localizedă searchă engineădisplaysăaădifferentăsetăofăuserăkeywords.ă(Lakóă2009b) 122 Figure 27: Click-through rate based on position in SERP Theăpictureăaboveă[71]ărevealsătheăimportanceăofăpositioningăinăSERPs.ăTheăcloserătoătheăfirstă positionă(theăhorizontalăaxis)ătheăhigherătheălikelihoodătoăbeăreachedă(theăverticalăaxis).ăThisăisă calledăClick-throughărateă(CTR)ăbyătheăindustryăandăvariesăfromăroughlyă20șăforătheăfirstă positionătoă1șăforătheătenthăposition.ăTheăclick-throughărateăisătheănumberăofătimesăaăresultăisă clickedă onă dividedă byă theă totală numberă ofă timesă ită wasă displayed,ă calledă byă theă industryă impressions.ăClick-throughărateă=ăClicksă/ăImpressions How does all that relate to translation and localization? Ranking high in SERPs depends on keywordă usage.ă Consideringă Romană Jakobson’să interpretationă ofă signsă fromă aă translationă process perspective (1992:145) translation procedure can vary. Undoubtedly the most widely used type of translation is interlingual translation or translation proper, which is translating from one language into another. However, in the case of website localization, if initial interlingual translation of the source text does not include the desired keywords (SEO requirement) then an intralingual translation or rewording is required. As shown above, the approach employed is intralingual translation when maintenance of content is necessary, by adding or readjusting keywords by which users look for information in search engines. Another type of translation procedure is the intersemiotic translation or transmutation. This is more difficult to be achieved by the translator alone, even if through a user interface. If the 123 source content is from a low-context culture and the target content is for a high-context culture intersemiotic translation is necessary, because the percentage in usage of verbal and non-verbal signs is different. High-context culture individuals are more receptive to nuances and non-verbal signs. Like advertising for a high-context culture, web content marketing must use mainly local and cultural images while decreasing verbosity. High-context cultures favor fewer words, a more indirect style and a balanced tone of voice. Verbosity from lowcontext cultures can be successfully replaced by body language, mimicry and gestures displayed in images and videos.39 Apartă fromă beingă equifunctională withă theă sourceă text,ă theă translatedă textă mustă beă builtă ină correlationă withă theă queriesă inpută byă theă usersă ină theă searchă engines.ă Findingă theă mostă appropriateăwordătoăbeăusedăinătheătranslatingăprocessăbecomesăevenămoreăchallengingăwhenă SEOăengineeringăisăaăconditionăimposedăbyătheăcommissioner.ă ThisăisăanăadditionalăandănovelădimensionăofăTranslationăStudies,ăasălocalizationăbecomesăană interactiveă andă on-goingă process.ă Theă localizeră hasă toă focusă onă whată isă ină demandă byă enclosingătheătermsăthatăareăusedăbyătheăpotentialăclientsăthemselves,ăinăorderătoăcomeăforwardă andămeetătheirăexpectations,ăandăinătechnicalătermsătoărankăhighăinăsearchăengineăresultsăpagesă (SERPs).ăThisăapproachăcouldăbeăcalledăkeyword equivalence,ăandăităisăaăcombinationăofătheă Skopos theory,ătheăpragmaticăorientationsăandăNord’săinstrumentalătranslationă(1997).ă Searchă enginesă furtheră assistă theă translatoră toă mediateă betweenă STă andă TT.ă Theă translatoră mustăcoverătheăgapăbetweenătheăsourceămessageăandătheă targetătextăreceiverăofătheămessageă usingătheăinformationăavailableăfromăsearchăengines.ăTheătranslatorăshouldănoălongerăcomplyă onlyăwithăprescriptiveănormsăinătranslatingăbutăshouldăalsoăemployătermsăusedăbyăusersăandă thus,ăforăweb-pageătranslation,ăsearch-engineăorientedălocalizationăshouldăbeăapplied.ăThisă alsoăconformsătoătheăonomasiologicalătheories,ănamely,ăfindingătheăwordsă(signifiers)ăusedăbyă searchăengineăusersăforăanăidea,ăaăconceptăorăaăreferent.40 Byăusingăsearchăengineăinformation,ătranslatorsăcanăbecomeăevenămoreăawareăofătheălinguistică contextăusedăbyănativeăsearchăengineăusersăandăthusătheyămayă produceăevenămoreă“natural”ă contentăinătheătargetătext.ă 39 40 ăAlso,ăreferăbackătoă2.3.2,ăInternationalization and translation strategies ăSeeăsubchapteră2.5.9,ăSemasiology and onomasiology 124 Unfortunately,ăveryălittleăhasăbeenăsaidăinăTranslationăStudiesăsoăfarăaboutătheăimpactăofăSEOă onăwebsiteălocalization.ăInăLakóă(2009b),ăIăproposedăaăkeywordăorientedătranslationăstrategyă andă drewăupăaăschemeăthată illustratesătheăinterplayăofătheăfactorsăinvolvedăină thisăparticulară typeă ofă translationă strategy.ă Basedă onă thoseă findingsă andă ină theă lightă ofă theă localizationă processesăinvolvedăIăproposeăaănewăoutlineăonăwebsiteăowneră–ăwebsiteăuserăcommunication: Figure 28: Communication in the keyword based transation process Figureă28ăillustratesătheăcommunicationăprocessesăinvolvedăinăkeywordăbasedătranslation.ăTheă departureăpointăisătheăsourceătext,ăideallyăinternationalized.ăTTăoutcomeăisădeterminedăbyătheă translationăpathsă(howăisătheăcommunicationăachieved)ăandăfactorsăinvolvedă(whoăandăwhată influencesătheăprocesses).ă Translationăpaths: 1. ST>MEDIATOR>SEARCHăENGINES>TT>RECEIVERSă Thisă isă theă maină (levelă 0)ă path,ă fromă STă toă TT.ă Thisă typeă ofă translationă is,ă ină mostă case,ă interlinguală andă unidirectional,ă fromă STă toă TTă andă ită isă aă continuousă maintenanceă processă (Pymă 2010:3).ă Modifications,ă additionsă oră deletionsă ină STă determineă updatesă ină TT.ă Thisă translationă pathă impliesă textă pre-productionă keywordă inclusionă(levelă1) 125 2. TT>RECEIVER>SEARCHăENGINES>TT Levelă1ăprocessesăcanăbeăconcurrentăwithălevelă0ătranslationăpathăasăspecifiedăabove.ăItă cană alsoă representă ană intralinguală translation,ă i.e.ă modifyingă TTă basedă onă keywordă changeă ină users’ă preference.ă However,ă thisă typeă ofă translationă cannotă beă achievedă withoută theă interventionă ofă theă MEDIATORă (levelă 2ă processes).ă Thus,ă thisă typeă ofă processesăgeneratesăaădifferentătypeăofătextămaintenance.ăThisăisăaăcyclicalăprocessăandă involvesăpost-productionăinterventionsăonătheăTT.ăLevelă1ăandă2ăcanăbeăperceivedăasăaă continuousăloopăthatăcouldăbeărepresentedăgraphicallyăbyăanăinfinityăsymbolă(∞) 3. TT>MEDIATOR>SEARCHăENGINES>TT Levelă 2ă processesă areă determinedă eitheră byă levelă 1ă processesă (changesă ină useră keywordă usage),ă oră cană beă self-imposedă updates,ă usuallyă involvingă editorială processă suchăasăspelling,ălexicalăorăgrammaticalăerrors.ăTheseăprocessesăareăcyclicalăandăimplyă post-production. Factorsăinvolved: 1. STăcontentăandăsourceăcultureăcontext 2. MEDIATOR/TEXTăPRODUCER’sămixedăabilitiesăsuchăasămessageădecodingăandăreencodingăabilities,ătheăabilityătoăre-encodeămessageăbasedăonăkeywordsăusedăbyăwebusers,ăusageăofăstatisticsătools,ălanguageăskills,ăgeneralăcommunicationăskills,ăetc. 3. Variousă conditionsă imposedă onă theă MEDIATORă suchă asă commissioneră terms,ă technicalărelatedăissues,ăfinancialăandătimeăconstraints,ăetc. 4. SEARCHă ENGINESă influenceă howă contentă isă foundă byă web-users,ă andă ită isă determinedă fromă aă linguistică pointă ofă viewă byă keywordă usage.ă Changesă ină useră keywordăusageăandăvariationsăinăalgorithmsămayărequireăTTămodifications 5. TTă isă ofă outmostă importanceă ină theă businessă toă clientă communication.ă Theă websiteă userăhasăcertainăexpectationsăfromătheătextăandătheătextăneedsătoăreflectătarget-cultureă specificălocaleăelements.ăTextualăcontentăhasătoăbeăexpressedăusingătheăsignifiersăusedă byăweb-users. 6. RECEIVERSă areă theă targetă ofă theă communicationă process.ă Understandingă theiră profile,ănamelyătheirăcomprehensionăability,ăisăvitalăinătheăcommunicationăprocess,ăi.e.ă onă theă TTă outcome.ă Users’ă needsă andă interestsă areă communicatedă backă toă theă MEDIATORăthroughăkeywordăusage.ă 126 To sum up, SEO translation is part of a continuous asynchronous online communication process that involves keyword usage as the main feedback channel directed from users/clients to mediators and website owners. The ever increasing importance of online marketing will determine a change in the way SEO is perceived by researchers in the field of Translation Studies. On the other hand, the localization industry has already acknowledged the importance of SEO in the web localization process. For instance, Lionbridge.com in its guide Website Translation for Today’săDigitalăWorldă[72]ătalksăaboutăSEO Translation and the steps involved in the SEO oriented website localization, including keyword research. 2.6.5 SEO website localization SEO website localization means shifting focus onto the end-user. It is website localization based on website analytics and statistics with a focus on making web content reachable in the targeted local search engines. It is an interactive, maintenance and update set of processes that involve both linguistic and technical skills. The linguistic aspects can be covered by translators with minimal training. As shown previously the two basic components that are of highest importance for SEO purposes are content and keyword research. Producing content involves creativity. Regardless if content is a translation or an original text, the translator or copywriter uses their professional skills, skills acquired during translation and/or creative writing classes or personal experience in certain fields of activity. While the translator may stay closer to the source text, the copywriter is unconstrained, so s/he can build the content based only on some directions. Also, no matter what the approach used to generate content is, text structure and layout need to comply with certain e-text related issues.41 Keyword research is the other process that can be accomplished by the translator. While it may appear overwhelming, there are numerous online tools that can be used to achieve this task. In practice, keyword research is most often conducted by an SEO engineer. In the future, the more advanced the content management system employed by a certain website, 41 ăSeeăsubchapteră3.2,ăText and E-text 127 the more tasks can be overtaken by the translator or copywriter and thus leave the more technical issues to the SEO professional. SEO localization most often will imply at least two different languages. However, I have shown previously that localization and especially SEO localization may sometimes require intralingual translation or rewording, based on keywords. In a multilingual approach the translator could first translate the keywords from the source content and then a local SEO professional would review the target words and even suggest similar keywords (synonyms and synonymous expressions, hyponyms, hypernyms, etc.). Once agreement is reached in terms of the keywords to be used keywords form ST and TT can be paired in glossaries or translation memories. When changes occur in keyword usage in search engines, in either one local search engine or the other, the glossary or TM can be adjusted. By having set up such an aligned TM, keyword modifications within the TT and ST can be easier adjusted. Also, if the translation is attended to by several translators the TM will maintain translations consistent on all the pages of a website and updated with the latest keyword usage data. SEO website localization has become increasingly more important with the targeting of local customers. Search engines are placing significantly more importance on the IP of the device used to access the search engine. In the case of some types of searches, especially those related to services such as hotels or restaurants, apart from the algorithms used regularly in ranking results in SERPs, local related searches are displayed at the very top as a list and as a map. This is very similar to searching for POI (point of interest) in GPS software. Such results are very useful especially when travelling and using mobile devices. Case study Figure 28 below displays such results based on the hotel keyword. Similar results are generated for service auto (car servicing), spalatorie auto (car wash). Searching for more specific businesses provides top ranking in both organic results and on maps (kaufland, banca transilvania). 128 Figure 29: Google result page based on the IP of the device In order to benefit from this search engine solution a business must function as a brick-andmortar business. A website is not even necessary. Looking for dezmembrari dogaru (secondhandă cară partsă byă Dogaru’s)ă Iă foundă theiră locationă andă aă phoneă numberă within the services related results, as there is no website of the company (last visited in March 2014). Yahoo! and Bing operate in a similar manner. Also, an important step prior to content production is to analyze and determine the keywords used by competition, namely national and multinational companies that operate on the same local market. The local market can further be divided into areas of the country or cities, or even neighborhoods (considering cities with millions of inhabitants like London).42 There are toolsă (similară toă Googleă Analytics)ă thată provideă intelligenceă onă users’ă locationă (citiesă oră regions). Link translation is also an important factor that influences how webpages rank in SERPs. Case study Links and text anchors (the text which is clickable) indicate to both web-users and search engines about what a certain page is. There should be relevance between text anchor and the URL to which the user is taken. Using the same wording in anchor text, URL and in the 42 ăSeeăsubchapteră2.4.5ă(Personalization) 129 content of the targeted webpage is essential. Herbal tea may be translated into Romanian either by ceaiuri medicinale or ceaiuri curative. If there is no consensus on which solution should be adopted, one translator may use ceaiuri medicinale while another translator opt for ceaiuri curative. While Google encourages the use of synonyms and from the perspective of SEO you can cover both keywords, Google does not list the same results in the SERPs. So, in Google terms (alte) ceaiuri medicinale as an anchor text from a webpage like www.website.ro/ceai-de-tei/ should not take you to www.website.ro/ceaiuri-curative/ but to www.website.ro/ceaiuri-medicinale/. Even if one uses a TM which accepts both Romanian terms, linking translations may create problems. The worst-case scenario is not translating the URLs at all. For instance (alte) ceaiuri medicinale linked to www.website.ro/herbal-teas. Ultimately, the most important step prior to producing content, either through translation or by copywriting, is to conduct the keyword research properly. 2.6.6 Keyword research and website localization Keyword research is an ongoing process that is essential to website localization. Pym acknowledges that keywords are essential to website localization and mentions them under the peripheral texts category. (2010:2) Keyword research can be applied prior to translating or copywriting or it can be a post-localization process. Post-localization activities are what make website localization a continuous process. This makes it also a maintenance process. Pym’să viewă onă websiteă maintenanceă isă ratheră limited:ă “Translationă teamsă willă ideallyă beă working in parallel, often on the basis of ongoing contracts to handle the maintenance of a site over a period of time. In such a frame, translation operates not on a text, nor at the level ofăaăfiniteăproject,ăbutăasăpartăofăanăindefiniteămaintenanceăprogram”. (2010:3) Similarly, he considersă maintenanceă “aă long-termă localizationă “program” but only to the extent that the changes in the source text determine changes in the target texts as well. (Pym 2014:126) I do agree that, while initially a company will most often have a certain number of webpages, a project can become an indefinite maintenance process as new products may be added to the website on a regular basis. Maintenance may be required by providing product or service upgrade as well. If the features of a product are modified the source text is also changed and thus the target text also needs changing. This is obvious and appropriate. However, it is important to underline that, especially for the case of website localization, maintenance processes involve readjusting existing content as well through intralingual translation by 130 revisiting the latest keywords used by web-users on search engine. Thus, intralingual translation is part of the maintenance process. Case study In analyzing traffic to a website, the SEO engineer often discovers that a particular webpage may have been reached through a keyword that no one accounted for. Various synonymous keywords function as alternatives when the translators have to choose “theăright”ăterm.43 On the other hand, unlike with traditional translation where the decision factor is the translator, grounding their decision on experience and context, in website localization judgment about the preponderant synonym is made statistically. Moreover, maintenance can imply keyword and content testing as well. SEO implies experimenting, continuous testing, and improvement. Testing has become even more important since Google encrypted searches. Thus, keywords by which certain webpages were accessedăcanăonlyăbeă“guessed”. Choosing the exact synonym or synonymic expression from a list of several possible keywords is based on statistics, hence, somewhat automated. An important consideration regarding keyword selection is that the localizer should select the keywords that appropriately match the preferred type of visitors, in e-commerce namely buyers. Using general keywords, usually made up of one, two or three words (called by the industry broad keywords) may generate substantial traffic but it may not translate into customers. The more general keywords are used, the more a user is interested in information only. Therefore, marketers and SEO professionals advise on usage of longer keywords, made up of four or more words, called by the industry long-tail keywords. Some researchers in localization may claim that if the source content is using long-tail keywords the target source will also use long-tail keywords. In my opinion, in practice, this is not always the case as the SEO unaware translator may choose to use implicitness as a translation strategy for a particular keyword. Priorătoădiscussingăissuesăonălong-tailăkeywordsăandătheirătranslationăIăwouldălikeătoămentionă thată untilă recently,ă Googleă favoredă theă rankingă ofă sitesă thată hadă generală keywordsă ină theiră domainăname.ăForăinstance,ălookingăforăsfaturi practice (practicalăadvice)ăwouldăhaveărankedă sfaturipractice.ro,ăsfaturipractice.netăandăotherăsimilarăwebsitesăatătheăveryătopăinătheăSERPsă regardlessă ofă content.ă Thisă generatedă aă hugeă marketă andă revenuesă foră thoseă whoă boughtă 43 ăSeeă2.3.2,ăInternationalization and translation strategies,ăandă2.5.9,ăSemasiology and onomasiology 131 keywordă consistentă domaină namesă justă toă sellă themă toă otheră interestedă parties.ă However,ă Googleăadjustedăitsăalgorithmsăandănowăwebsitesăcanănoălongerărankăhighăjustăonătheăgroundsă ofătheădomainăname.ăTheăbenefităisăthatăbrandingăisăachievableăwithoutănecessarilyăincludingă keywordsăinătheădomainăname. 2.6.7 Long-tail keywords As already mentioned above long-tail keywords are expressions of 4 or more terms entered as search queries in search engines. According to moz.com [73] 70% of the searches are longtail queries. The more terms entered, the more specific a search. The specifity of a search process is determined by the number of words in a keyword. The same source states that the more specific a search query, the more prepared the user to buy. The graphic below illustrates various keyword lengths and related search volume (in the second column). The third column reveals the CTR (click-through rate) value. Source from www.calculatemarketing.com [74]. The most significant volume of searches is for 3, 4 and 5 words. However, recorded CTR reveals highest efficiency for 5 words or more. As 7 and more words are not very practical to implement, I would recommend implementing 4, 5 and 6 words keywords. Also, it is unexpected that CTR for 2 words keywords is higher compared to 3 words keywords, but they are broad keywords and rather useful for branding and informative websites. Tracking 4, 5 and 6 words in the source content and translating them in the target content, by finding the corresponding keywords used in the target language can be tricky. Figure 30: Keyword length and how it determines CTR In the applied parts of my research (Chapter 4) I discuss issue regarding keyword research and translation strategies. 132 2.6.8 Tools for keyword research Asă Iă mentionedă previously,ă keywordă researchă requiresă noă specială skillsă asă accessă toă dataă isă effortlessăandămostăoftenăunrestricted.ăTheădifficultyăconsistsăinăknowingăhowătoăuseătheădataă collected.ăThereă areă tensă ofă toolsă toă beă used,ă someă directlyă fromă theă source,ă whichă isă fromă searchăengines,ăwhileăothersăareăfromăprofessionalăsearchăengineămarketingăcompanies. TheăfirstăandămostăaccessibleătoolăforăkeywordăresearchăisăGoogle’săsuggestionătoolăfoundăonă theăsearchăengine.ăSuggestionsăcanăbeăseenăandăcollectedăasăsoonăasătypingăinătheăsearchăfield.ă Theăpictureăbelow,ăalsoăillustratesăhowăusersămayăbeăinfluencedăbyătheăsearchăengineăitself: Googleălistsăkeywordsăbyăsearchăvolumeăandăitălistsătheătopă10ăkeywords.ăThisăinformationăisă usefulă foră broadă keywordsă bută byă addingă oneă wordă ată aă timeă toă youră searchă queryă youă cană obtainălong-tailăkeywords.ă Aă similară tool,ă whichă isă actuallyă usingă Google’să suggestionă tool,ă isă calledă Ubersuggestă (http://ubersuggest.org/)ă andă usesă alphabeticallyă foră loopsă recordingă andă displayingă allă theă Googleăsuggestions.ăForăinstance,ăforă ceaiuri (teas)ăitădisplaysătheăsameăresultsăasăGoogle’să suggestionătoolăplusătheăfirstă10ăsuggestionsăforăceaiuri a…,ăceaiuriăb…,ăceaiuri c…ăandăsoăonă toăceaiuri z…,ătotalingăoveră300ăkeywords.ăUbersuggestăprovidesămuchămoreăvaluableăresultsă ifă Iă addă oneă moreă word.ă ceaiuri energizanteă yieldsă 33ă resultsă withă long-tailă keywordă suggestions.ăThenăyouăselectătheămostăappropriateăonesăforătheătargetăcontent.ă GoogleăTrendsăisăanotherătoolăthatăisăratherăusefulăforălargerămarkets.ăYouăcanăcompareăupătoă 5ăsearchătrendsădiachronicallyăandăsynchronicallyăand,ăasăitsănameăsuggests,ăităcanăbeăusedătoă 133 predictăfutureătrendsăinăsearch.ăWhileăităperformsăefficientlyăforălargeămarkets,ăforăRomaniaăită onlyăscratchesătheăsurface. Anotheră Googleă toolă isă theă Googleă KeywordPlanneră [62].ă Ită requiresă creatingă ană accountă withă themă bută ită cană beă usedă freelyă toă gatheră informationă foră organică searchă ranking.ă Ită isă preponderantlyămeantăforăpay-per-clickăadsă(PPC),ăandăsomeăcompaniesămayăevenăuseăităforă paidă searchă ină orderă toă gatheră insightsă intoă usingă certaină keywords.ă Theă toolă providesă suggestions,ămonthlyăsearchăvolumesăandă locationărelatedădataăforăcitiesăandăneighborhoodsă (irrelevantăforăRomaniaăandăotherăsmallerăe-markets). Bingă Keywordă Researchă fromă Microsoftă [75]ă isă anotheră usefulă toolă thată cană beă usedă foră suggestionsă andă foră reportingă keywordsă byă whichă aă websiteă wasă accessed.ă Thisă isă aă verificationătechniqueătoăknowăifăcontentăranksă highăforătheăintendedăkeywordsăorăforăotheră non-relatedă oră broadă keywords.ă Unfortunately,ă Googleă eliminatedă thisă featureă fromă Googleă Analyticsă byă securingă searchesă ofă loggedă ină usersă and,ă thus,ă aă significantă amountă ofă informationă isă displayedă asă “notă provided”.ă Thereă areă someă workaroundsă bută requireă programmingăskills. Otheră toolsă whichă coveră severală searchă enginesă areă http://soovle.com/,ă http://www.wordstream.com/,ă http://tools.seobook.com/keyword-tools/seobook/.ă Theyă workă forăRomanianăasăwell.ă Thereăareăvariousăotherăusefulătoolsăforăkeywordăresearchăbutătheăenhancedăonesăhaveăbecomeă paidă tools.ă Priorătoă committingă toă aă subscriptionă withă anyă ofă theă payă toolsă Iă recommendă toă signă upă foră trială periodsă firstă toă verifyă ifă theă toolsă supportă theă languagesă youă requireă foră localization.ă Theă smalleră theă marketă theă higheră theă likelihoodă thată theă targetă marketă isă notă covered.ă Usingă Googleă foră furtheră keywordă researchă toolsă mightă produceă otheră usefulă results.ă However,ăIărecommendăbeginningăwithătheăGoogleăandăBingătoolsăandăuseătheăotherătoolsăforă refiningăinitialăfindings. Also,ă whenă doingă keywordă researchă isă oftenă usefulă toă studyă theă competitionă ină theă locală market:ăkeywordsăforăwhichătheyărankăhigh,ăcontentăstructure,ăkeywordădensity,ăetc.ăSuchăaă 134 toolă isă http://www.internetmarketingninjas.com/seo-tools/seo-compare/.ă Foră otheră suchă toolsă lookăforăSEO compareăonăGoogle.ăWhileătheăsameăkeywordsăasăthoseăofătheăcompetitionăcană beă used,ă theă contentă producedă mustă beă completelyă original.ă Otherwiseă theă client’să websiteă risksăbeingăpenalizedăorăsuingăforăduplicateăcontent.ăDuplicateăcontentăcanăbeăeasilyătracedăbyă usingătoolsăsuchăasăCopyscapeă[77]. Onceă theă appropriateă keywordsă areă foundă theyă mustă beă revisedă withă theă client.ăAfteră clientă acceptanceă theyă mustă beă incorporatedă intoă aă glossaryă oră aă translationă memory,ă soă thată bothă consistencyăandătranslationăspeedăisăimproved. Asă mentionedă alreadyă severală times,ă keywordă researchă isă aă majoră stepă ină producingă theă localizedă versionă ofă aă sourceă website.ă Ifă sufficientă timeă isă investedă ină keywordă research,ă ină contentă production,ă andă ină on-siteă optimizationă theă localizedă websiteă generatesă qualifiedă traffic.ă 2.7 Conclusions InăthisăchapterăIăanalyzedătheămainăconceptsăinvolvedăinătheălocalizationăprocessăandăatăallătheă relatedăprocessesăappliedătoătheăspecificăfieldăofăwebsiteălocalization.ăIă beganăwithătheămoreă generalăconcepts,ăglobalizationăandăinternationalization,ăanalyzingăspecificăissuesăsometimesă overlookedăbyătheăacademicăcommunity,ăsuchăasătheăinternationalizationăorăstandardizationăofă theăsourceătextăsoăthatătheăefficiencyăandăspeedăofătheătranslationăintoătheătargetătextsămayăbeă increased.ă Inăsubchapteră2.5ăonăTranslation,ăIăexaminedăseverală conceptsă fromătheăfieldăofăTranslationă Studies.ă Myă maină contributionă hereă isă myă attemptă toă defineă andă distinguishă amongă translation,ăadaptationăandăcopywritingăwhileăprovingăthatătheălocalizationăprocessă doesănotă alwaysă includeă translation.ă Ină practice,ă translationă isă oftenă replacedă byă copywriting.ă Ină theă sameăchapterăIăincludedătheăconceptsăofăsemasiologyăandăonomasiologyătoă demonstrateăthată findingă theă commonă referentsă foră theă sameă conceptsă isă essentială ină communicatingă acrossă websiteă content.ă Thisă isă theă maină orientationă ofă theă thesis,ă focusingă onă theă web-users,ă byă analyzingătheirăinputăinăsearchăengines. 135 Thisă chapterăalsoă discussedătheăinterconnectionsăbetweenă globalization,ăinternationalization,ă localizationă andă translation,ă Iă willă confirm,ă drawingă onă Cadieux’să andă Esselink’să formulaă (2004)ăthatăglobalizationăisăinternationalizationăplusătheănumberăofătargetălanguages,ăplusătheă numberăofătargetălocales,ăwhereălocalesăcanăbeăeitherăofăculturalăorălegalănatureăorăofătechnicală conditionsăandălimitations. Consideringă otheră industries,ă ifă Iă wereă toă suggestă aă successfulă localizationă modelă toă beă followedăinăwebălocalization,ătheămostăefficientăandămostăsuccessfulălocalizationăproductsăareă thoseăforăTVăshowsăsuchăasăWho wants to be a millionaire,ăThe Voice, theăGot Talent seriesă andăotherăsimilarăproductions.ăTheseăproductionsăsuccessfullyăcoverăallăaspectsăpertainingătoă localizationăandăareăculturallyăandălinguisticallyăoptimized. Glocalizationă isă anotheră keyă conceptă whichă emphasizesă theă interdependenceă ofă globală andă locală factorsă ină shapingă theă methodsă byă whichă onlineă andă traditională communicationă areă achievedăwithinătheăglobalăvillage. Ană importantă partă ofă theă chapteră dealtă withă theă importanceă ofă searchă enginesă ină onlineă marketingăandăsearchăengineăoptimizationă(SEO).ăAnotherămajorăcontributionătoătheăfieldăofă websiteălocalizationăisătheădiscussionăonătheăimportanceăofăSEOătranslation,ăthroughăkeywordă researchăandăoptimalăkeywordăselection. 136 3. Specificăissuesăinătranslatingăelectronicătext 3.1 Introduction The ever increasing content available online and its complexity as a form of communication, be that information on products, services, general knowledge, or just picture galleries and videos has led to cross-cultural communication issues. This chapter discusses the particularities involved in the translation of electronic text in the context of website localization, required to overcome language, culture, customs and other features of the target locale. Therefore I will analyze such issues as the particularities of the electronic text versus the traditional text, the online environment and how it influences the communicated text, how text is displayed, read and understood, how co-text, context and reference are rebuilt during the translation and localization processes. The appropriate transfer of presuppositions and implicatures is also important for an accurate communication of the original message. Last but not least, I will analyze how visual aids are used to make the communication more effective and more efficient as a subset of elements in localization.44 The discussions on e-text involve technical issues as well, issues that affect the manner in which the information communicated is built. TheăgrowthăofătheăWebăhasăbroughtăalongăaăhugeăgrowthăinătheănumberăandătypesăofătext,ătheă soăcalledăe-textsă(electronicătexts)ăorăweb-texts.ăElectronicăbooksăandăelectronicănewspapersă areăveryămuchă alikeătheirăcounterpartsăprintedăonăpaper.ăUsually,ătheseătextsăareăinăAdobe’să pdfăformat.ăOtherăformatsăareăepub,ăkindleăamazonăetc.ăHowever,ătheătextăofăwebpagesăisăaă newăkindăofătext.ăComparedătoătheăotherătypesăofăe-text,ăwebpagesăareădifficultătoăcategorizeă accordingătoătheătraditionalăclassificationsăofătext.ăWebpagesăareăirregularăandăindividualized,ă atătimesăpersonalized,ăasăusersăcanăchangeălayoutăelementsăorăaddătheirăcontributionăunderătheă formăofăfeedback,ăcommentsăorăreviews.ăTraditionalătextăisăratherălinearăwhileăwebpagesăhaveă ană interruptedă discourse,ă withă explanationsă oră in-textă advertisementsă thată takeă theă useră toă aă 44 ăSeeă4.5,ăBooking.com: a comparative study from a website localization perspective 137 completelyă newă webpageă thată mayă beă fromă theă sameă websiteă oră fromă ană entirelyă differentă website,ăpossiblyăwithăaădifferentătextătypology. Webpagesă relyă veryă muchă onă interactivityă andă multi-functionality:ă menus,ă sitemaps,ă searchă boxesă etc.ă Textualityă isă alsoă perturbedă byă extensiveă useă ofă images,ă animations,ă videosă andă otheră graphicală elements.ăAă webpageă usuallyă containsă moreă informationă thană requiredă andă textă isă notă onlyăuserăfriendlyăbută alsoă searchă engineăfriendly.ăAccordingătoă Nielsen’săonlineă article,ă webpagesă areă usuallyă skimmedă oră scanned,ă usersă readingă informationă selectivelyă (2008).ăAlso,ătheăreceiverăofătheăinformationăwillăreadăaăwebpageăinăanăF-ăpatternă(Nielsen,ă 2006).ă Searchă engineă resultsă pagesă (aă specială typeă ofă webpage)ă areă clustersă ofă textsă thată illustrateă theă intertextuală characteră ofă allă textsă foundă onă theă Web.ă Theă largeă varietyă ofă textă typesăonătheăWebămakesăitădifficultătoăcategorizeăe-texts.ă 3.2 TextăandăE-text Text is a rather self-explanatory term. However, I will consider several definitions both from a more general linguistic perspective and from a more specific point of view. For instance, I will examine pragmatic issues of communication from the standpoint of Translation Studies, but I will also consider it as a key element in the process of localization. As I have already demonstrated in 2.4.8, Content marketing – delimitations in the sphere of localization, text as web content is the most important factor in terms of effectiveness and degree of difficulty in its production. The term text most often refers to the written form of a communication. Text is defined by HallidayăandăHasanăasă“anyăpassage,ăspokenăorăwritten, of whatever length, that does form a unified whole.”(1976:1) The written form of a discourse versus the spoken form of a communication has a rather monological characteristic especially in the more traditional manner of distributing written text: printed books, booklets, magazines, fliers, and so on. I will not discuss here the differences of form, structure, style, register and other specificities between the written and the spoken types of discourse but rather focus on the dissimilarities between the traditional (printed) text and the online text or e-text. Text, as a monological string of written language shows coherence. Coherence “depends on the mental activity of the reader or listener, on their capacity to understand the message that 138 the producer of the text is trying to convey. The text can be thought of as providing clues as to what the message is, but the reader has to use these cues.”ă (Sanford 2006:585) Hence, coherence is not intrinsic to the text itself but to the emitter and receiver of the text who build up mentally the communication as a product of the communication process. Cohesionăisă “aă semantic relation between an element in text and some other element that is crucial to the interpretation of it.”ă (Halliday and Hasan 1976:8) They distinguish 5 types of cohesive elements: reference, ellipsis, substitution, conjunction, and lexical cohesion. While the purpose of the thesis is not to provide further details on these subtypes, cohesive elements may be obvious and present in a text, but this is not always the case. Cohesive elements may be missing (for instance a list of features of a product on a webpage) and the user can still understand the text; transform it in a meaningful message. So, meaning takes precedence over form in the communication process, however, the form helps in clarifying the meaning. At the strict level of text formation, cohesion is what makes a text well formed. However, cohesion does not always involve coherence, i.e. a text may be well formed from the point of view of using cohesive elements, yet incoherent. On websites, however, there will often be cases when the text is not monological but several people express their view on something. This may be the case of product feedback, where the same product can be seen by one user as perfect and given a positive review while someone else may see it as a product not worth buying. Also, comments to articles may be positive or negative. Although in such cases texts are not produced by one person, at topic level there is still a sense of coherence, at least at a general level. Coherence is a suprasegmental textual element, whereas cohesion is inherent to text. Sanders and Sanders assert that textual coherence is achieved through (2006:599): - Referential coherence (at lexical level, namely anaphoric, cataphoric and exophoric reference) - Relational coherence (cohesive or linking devices, either at phrase or sentence level: conjunctions, sentence connectors and subordinators) These are the only necessary conditions to consider a range of sentences to be a text. Strorrer distinguishes in the case of websites 2 types of coherence: “locală coherence”ă andă “hyper-textuală globală coherence”ă (cited in Jiménez-Crespo 2013:61).While local coherence 139 broadly corresponds to coherence in the traditional text, global coherence applies in general to any webpage and even more so in the specific case of SEO localization, as primary focus of my thesis. SERPs are collections of hyperlinks sharing global coherence and thus function as hyperdocuments. Kallmeyer (cited in Nord, 2005:16) defines text asă “theă totalityă ofă communicativeă signalsă usedă ină aă communicativeă interaction”.ă Basedă onă this definition, Nord further distinguishes between verbal and non-verbal elements (ibid.), in the case of printed text (illustrations) or online texts (illustrations and media files such as background music, animations). Beaugrande and Dressler (1981) add to cohesion and coherence further important aspects to be considered: intentionality, acceptability, informativity, situationality and intertextuality (to be discussed later on). Neubert (1996) applies all these elements to his textual approach to translation. Also, the text as a communication process is determined both by the sender on the one hand, and the receiver(s) on the other. However, while the sender is the initiator of a communicativeă interactionă “theă textă asă aă communicativeă actă isă completedă byă theă receiver”. (Nord, 2005:18) Based on the characteristics of a text, Darwishă (2008:155)ă definesă ită asă aă “multi-layered packageăofăinformation”ăandăclaims that there are 6 layers encapsulated into a text: - textual layer - contextual layer - cultural layer - temporal layer - intentionality layer - intertextuality layer Above these layers I propose the addition of the anticipation layer, as a pre-screening step of the main text.45 45 ăMoreădetailsăonăthisăinăsubchapteră3.4.7 140 As a specific type of e-text, Pym defines localized text as follows: “A localized text is not called on to represent any previous text; it is instead part of one and the same process of constant material distribution, which starts in one culture and may continue in many others.”ă (original emphasis, 2004a:5). E-text is thus perceived as a unitary distribution entity. Jiménez-Crespo, on the other hand, considers e-textsă “unitsă ofă production,ă informationă andă storage, or as units thatăareădevelopedăandăpresentedătoătheăuserăasăsuch” (2013: 47). Thus, I consider that e-text characteristics are determined by the medium through which they propagate (distribution) and the technology (CAT or MT) and strategy employed for its production (translation or copywriting). It is also interesting to see that the same author considers as the minimal textual unit (in website localization), a complete hypertext or website. He claims that a particular webpage cannot function as a complete text but as a subtext of the website that encloses it along other subtexts (webpages). (Jiménez-Crespo 2013: 50-51) Generally speaking this is a valid assertion, for instance, when the web-user regularly visits a website such as New York Times (www.nytimes.com) and relates to a great extent to brand loyalty. There are some faults with this reasoning, however. As shown in 2.4.8, Content marketing – delimitations in the sphere of web localization, the current marketing trend in marketing is the so called inbound marketing. Also, brand loyalty is no longer the driving factor behind buying decision-making. [34] The most effective inbound channel is through search engines. Thus, the most important interface for information retrieval is represented by search engines. When using search engines it is often the case that the user is neither aware nor interested in the whole website, especially when consideringă theă earlyă stagesă ofă aă customer’să lifecycleă (reach).ă Website analysis software such as Google Analytics [48] shows that unique visitors may reach a certain webpage of a website, stay on the page for several seconds or minutes and then exit the website from the same webpage (entry page). They return to the search results page and either access the other listed webpages or modify the search term. This is typical user behaviour for those searching for information. Within websites, publishers often use the so called landing pages. A landing page is any webpage from a website that a search engine user “lands on”ă afteră clicking on a result on a SERP. From an online marketing perspective a landing page functions as a standalone webpage, created for a specific objective, for instance, focusing on a single long tail keyword (second-hand car body parts). Furthermore, each webpages within a website shares global cohesion elements with other webpages from the sae website. 141 However, Jiménez-Crespo’sădefinitionăonăe-text accurately describes web pages either if seen asă individuală unitaryă textsă oră asă subpartsă ofă aă website:ă “A digital interactive entity that is developed and presented to the user as a unit in the WWW and is coherent within itself. It represents a thematic and functional unit that has a hyperlinked multilineal structure made up of subtexts. It comprises linguistic, graphic, visual, typographic, aural and interactive components.”ă (2013:ă 51) If we look at the source code of a typical HTML546 page, all the above mentioned elements are included in the syntax of this programing language. However, as I have shown previously, a webpage can function as the minimal coherent textual unit, while the website as a whole can function as context. On the other hand, webpages and websites may contain non-related content, such as 3rd party advertising banners. (Janoschka cited in Jiménez-Crespo 2013: 56). Also, website text is characterized by open-endness. Text can be modified, information added or removed in a more dynamic way and it allows numerous revisions involving little technical effort or costs. The following table summarizes the main features of traditional and electronic text: Feature Traditional text Electronic text Dynamicity static highly dynamic Direction linear disrupted Cohesion local local and global Interactivity limited highly interactive Reading volume dissemination gisting Communication pattern monologue dialogue Interlinking NA hypertextual Multimodality limited hypermodal Open end-ness NA content can be further added Semiotic complex limited audiomedial mix Productivity and distribution asynchronous synchronous Accessibility limited Highly distributable Table 4: Traditional versus electronic text features 46 ăhttps://developer.mozilla.org/en-US/docs/Web/Guide/HTML/HTML5/HTML5_element_list 142 3.3 Classificationăofătexts Thereăareăthreeămainăcriteriaăaccordingătoăwhichătextsămayăbeăclassified: -ăformal -ăfunctională -ăbyătype Formală typologiesă areă basedă onă textă register.ă Registeră mattersă areă discussedă ină 3.4.2.3.ă Functională typologiesă labelă textsă asă expressive,ă informativeă andă vocative.ă Ină termsă ofă type,ă textsă areă categorizedă accordingă toă theiră communicativeă scope.ă Theyă mayă beă expositive,ă argumentativeăandăinstructional.ăReissăfoundătheăfollowingăcharacteristicsăforăeachătextătypeă (citedăinăMunday,ă2008:ă72): (1)ă‘Plainăcommunicationăofăfacts’:ă information,ă knowledge,ăopinions,ăetc.ăTheălanguageă dimensionă usedă toă transmită theă informationă isă logicală oră referential,ă theă contentă oră ‘topic’ăisătheămainăfocusăofătheăcommunication,ăandătheătextătypeăisăinformative.ă (2)ă ‘Creativeă composition’:ă theă authoră usesă theă aesthetică dimensionă ofă language.ă Theă authorăoră ‘sender’ăisă foregrounded,ă asăwellăasătheăformă ofătheămessage,ă andătheătextă typeăisăexpressive.ă (3)ă‘Inducingăbehaviouralăresponses’:ătheăaimăofătheăappellativeăfunctionăisătoăappealătoăoră persuadeă theă readeră oră țreceiver‘ă ofă theă textă toă actă ină aă certaină way.ă Theă formă ofă languageăisădialogic,ătheăfocusăisăappellativeăandăReissăcallsăthisătextătypeăoperative.ă (4)ă Audiomedială texts,ă suchă asă filmsă andă visuală andă spokenă advertisementsă whichă supplementătheăotherăthreeăfunctionsăwithăvisualăimages,ămusic,ăetc.ă[...] On commercial websites, we find all four types of text. As Reiss admits that there are no pure text types, similarly, on the Internet, there is no pure type either, but there is always a predominant one. One online business will make use of all four types. The informative text typeăwillăpredominantlyăbeăusedăonătheă“Aboutăus”,ă“Contactăus”ăandăproductăpages.ăProductă pages would contain images to supplement the written text. The home page usually displays theăcompany’săvisionăandăsloganăandăthe expressive text is the predominant type. In the case of banners leading to the company’s websites, the predominant text type is the operative one, but expressive elements can also be present to a great extent. On product pages, even if the operative text may be short, it may still bear much persuasive force, by using visually 143 enhanced graphic elements, such as larger font size with a flashing background and with a compellingătextă(“Buyănow!ăLimitedăstock!”).ăTheăindustryăcallsăsuchăappellativeăorăoperativeă expressions call-to-action words47. Audiomedial texts are specific to the semiotic mix of websites. Web content is composed of text, graphics, video and sound. Text can or not be predominant on a webpage, but can be also of secondary importance, for instance, figure captions for an image gallery, or a short description for a video gallery. Nida foresaw the importanceă ofă complexă semiotică communicationă ină theă earlyă 1980’s: “theă significanceă ofă those signs which carry the message to receptors under particular circumstances”. (cited in Dimitriu, 2009:30) This applies even more so to websites and website localization. Figure 31: Munday’s functional characteristics after Reiss By considering the above table, we can notice the type of Audiomedial elements that can be used to supplement the three main types of texts when considering commercial webpages. Forătheăinformativeătextsăthereăwillăbeăsimpleăgraphicăelements,ăinătheăcaseăofă“aboutăus”ăandă “contactăus”ăpages,ăprimarily delimiting lines, and a simple page layout. On product pages, there will be strictly images and videos demonstrating the features mentioned by the text itself. 47 See also 3.4.8 Call to action (CTA) and persuasive words in SERPs. 144 In the case of expressive text types the graphics used will enhance the look of the webpage. Banners and homepages will use graphics and animations that elicit the desired effect. Figure 32: Audiomedial supplement in operative texts 3.4 Theălayersăofătheătext 3.4.1 The textual layer Darwish (2008:157) correctly definesătheătextualălayerăasătheă“mainăinterfaceăofătheătextăthată providesă accessă toă theă restă ofă theă otheră layers.”ă Indeed,ă theă text,ă graphică elements,ă organization of the text, drawings, pictures and other illustrations are the very first contact of the reader with the information communicated. The textual layer can provide some background information about what the text refers to in general and hints at the subject matter. If we consider letters then they have a certain format. Also, the layout of a business letter will be different from the layout of an informal letter. Another textual layer that I regard as very important is the shell of the text, something like “judgingă aă bookă byă itsă covers”.ă Foră instance,ă ifă weă consideră aă bookă asă comparedă toă aă magazine, a book is more voluminous, very frequently more valuable, it contains more information, it is printed on different paper sizes, and different editions are launched every few years, with updated information here and there, whereas a magazine would have around 145 50 pages (depending on how often it is issued, weekly, twice a week or monthly), and the information contained would be new with every issue. Also, magazines contain substantially more graphical elements. So, the very first contact that provides important clues about the text as information is the shell of the text. By merely choosing a book over a magazine a reader already has some expectations as to what information s/he will find. This is the first potential common background knowledge between the author(s) of the information and the receiver. While for web-users quantity of information is not predictable, search engines do consider the general size of a website, when listing certain webpages in the SERPs. The format and layout of the text could represent a class by which a text is categorized. It could also be further sub-classified according to text typology. Text typology also adds to the anticipation processes that occur in the reader’sămind. The four main text types, acknowledged by the traditional schools, are narration, description, exposition and argumentation. Hatim and Mason propose a fifth type of text, that is, instruction, further subdivided into instructions with option (advertisements) and instructions without option (product description pages, help and frequently asked questions pages). (1990: 153-158) Beaugrande and Dressler (1981) defined seven text types: descriptive, narrative, argumentative, scientific, didactic - equivalent with instructional, literary and poetic. Of interest to my thesis isă Biber’să textă typologyă (1995:ă 102) consisting of the following elements: involved production, informational production, narrative concern, explicit reference, situation-dependent reference, overt expression of persuasion, abstract information, and online informational elaboration. In the case of involved production the main features are the usage of private verbs (e.g. feel, believe), present tense, first and second person pronouns, wh- questions, that-omissions, contractions, action verbs. On websites this type of text is found on banners and in meta description (description for the results listed in search engine results pages). Informational production has the following characteristics: large percentage of nouns, lengthier words, more formal register, and passive constructions (product pages). Features of narrative concern: past tense verbs, third person pronouns, perfect aspect verbs, public verbs, synthetic negation, and present participial clause (not specific for e-commerce websites). Features of explicit reference: WH relative clauses, nominalizations, and phrasal coordination (any type 146 of webpage). Features of situation-dependent reference: time adverbials, place adverbials, and adverbs (call-to-action pages). Features of overt expression of persuasion: infinitives, prediction modals, conditional subordination, and necessity modals (call-to-action pages). Features of abstract information: conjuncts, passives, past participial clauses, and other adverbial subordinators (product specifications). Features of online informational elaboration: THAT clauses, and demonstratives (marketing content). These text typologies can be applied effectively to a wide range of online texts. The simplest and most comprehensive categorization would be dividing texts into Nominal and Verbal types. Nominal texts are characteristic of formal texts; they are concise, compact and information-rich. They can prove to be difficult to read. Online, this type of text can be found especially on official sites of governments, city halls, academic and corporate sites. Verbal texts are specific to e-commerce websites especially in the case of operative or appellative texts. As with the traditional texts, text typology for webpages is not a simple clear-cut matter, but there is a dominant type of text. Also, one particular website may require several dominant text types, depending on the meaning it intends to convey to the web-user on each of the various webpages. Moreover,ă webpagesă haveă aă visuală layoută thată permitsă theă inclusionă ofă severală textsă withă differentăcommunicativeăpurposesăonăaăsingleăwebpage. Textuality on webpages is rich in non-linguistic elements that can support the main verbal text or can be distracting elements (banners and links towards other sites). In the localization process, these website elements will either be removed or replaced with advertisements of interest for the target users. For instance, Yahoo localized for Romania not only localized the user interface, but also the advertisements and the news in order to be of potential interest to Romanians. In comparison with printed texts, webpages are more complex and more mixed in terms of layout and text (both linguistic and non-linguistic). Textuality of webpages is unlike textuality of linear documents. As webpages are an ever-increasing pool of texts, further 147 discussions and research is required both in establishing a suitable text typology that reflects text types on the Web and in how it can be used as a framework for translators and localizers. 3.4.2 Contextual layer The contextual layer is composed of context and co-text. 3.4.2.1 Co-text Co-text is also called physical context and refers to the linguistic aspect of the text. It is definedăasă“the other lexical items that occur before and after a word.”ă(HatimăandăMundayă 2004:337) Darwish defines it in a more general manner asă “theă relationshipă ofă words,ă sentencesăandăparagraphsătoăoneăanotherăwithătheătextualălayer.”ă(2008:158) Yuleă(2010:ă285)ădefinesăităasă“theăsetăofăotherăwordsăusedăinătheăsameăphraseăorăsentence”.ă Indeed, in the case of homonyms in writing and homophones in speaking, the neighboring words help the receiver quickly decide on the meaning of a certain word. For instance, the English words bow of a boat and bow as a weapon or there, their, they’re are at times used incorrectly even by natives. Co-text limits the possible meanings of a certain text. Therefore, co-text is important in opting for the semantic meaning of a word within the surrounding text. Meaning at co-text level can be determined also in case of polysemous words used literally or metaphorically (Mother of the child versus Mother Earth). I consider that prefixes and pseudo-prefixes are also part of the co-text. Adding the pseudoprefixă“e”ăoră“i"ătoăaăcertainăword,ăforăexampleăe-text, ebook, ecards.com, email, I-explorer, make it clear that the topic refers to “electronic”ă oră “Internet”ă oră meantă foră theă Internet or electronic usage. 3.4.2.2 Context Contextă refersătoă theă“setting”ăină whichătheăcommunicationă occurs.ăContextă canăbeădividedă intoăphysicală(spatial)ăcontextă–ăwhereătheăcommunicationămaterializes,ăandătemporalăcontextă –ăwhenătheăcommunicationăhappens.ăBiberă(1995:30)ăaddsătheă“superordinateăactivityătype”,ă 148 whichă refersă toă theă parentă activityă withină whichă theă textă isă transferred.ă Foră instance,ă ină writing,ăaăparagraphăinăaăbookăhasăasăitsăparentăactivityătheătextăonătheăpage,ăwhichăinăturnăisă partăofăaăchapter,ăwhichăisăpartăofăaăbook,ăwhichăisăpartăofăaăliteraryătrend,ăwhichăisăpartăofăaă culturalătrend. Theăsameăauthorăspeaksăaboută“theăextentătoăwhichăspaceăandătimeăisăsharedăbyăparticipants”ă (ibid).ă Theă receiversă mayă beă presentă oră notă ată theă timeă whenă theă senderă isă buildingă theă message.ăInăwritingătheyădoănotăshareătheăsameăspaceăandătime.ăOnătheăcommercialăwebsiteă mostă ofă theă interactivityă aroundă aă textă isă asynchronousă asă well.ă Thereă areă stillă justă aă fewă commercialăwebsitesăofferingăliveăsupport.ăSo,ăoftenăaăpotentialăbuyerăhasătoămakeăaădecisionă basedăonătheăinformationăfoundăonătheăwebsite,ăwhichăveryăoftenăisăincomplete.ă Schirato and Yell (cited in Darwish, 2008:155) define contextă“asătheăparticularăenvironmentsă in which communication, texts and meaning making occur, and in which they function as meaningful. The context is the situation in which we make sense of a text or a meaningful practice.”ă Theă virtuală environmentă ofă e-commerce websites relies on human-computer interaction. The context is often predetermined by the search engine through which the user accessed the webpage. Eugeneă Nidaă underlinesă theă importanceă ofă contextă foră determiningă theă meaningă ofă theă constituentă elements:ă “ină determiningă theă meaningă ofă wordsă theă roleă ofă theă contextă isă maximizedă andă theă roleă ofă anyă focală elementă isă minimized,ă whichă meansă thată theă contextă actuallyă providesă moreă distinctivenessă ofă meaningă thană theă termă beingă analyzed”ă (citedă ină Dimitriuă 2009:29).ă Onă websites,ă theă multitudeă ofă allă semiotică elementsă constitutesă theă context.ăParticularăelementsăareărecognizedăandăinterpretedăbasedăonătheăcontext. Some scholars perceive context as a three-phased process: initial state, intermediary state and final state. Halliday and Hasan (cited in Schäffner, 2002) extend the context to all extralinguistic factors that influence the text itself. There has been significant debate as to what falls under the extratextual factors concept and the importance of each of these factors in the translation process (ibid). It is an issue scholars are yet to agree on. 149 Weaver (1988) found four subcontexts: grammatical context, semantic context, pragmatic context, schematic context. However, these are broadly equivalents of register, co-text, time and space and background knowledge, respectively. 3.4.2.3 Register Longman Dictionaries Online provides for register theă followingă definitionă „ă theă words,ă style, and grammar used by speakers and writers in a particular situation or in a particular typeăofăwriting”ă[78].ăăCambridgeădictionariesădefineăităas:ă“the style of language, grammar, and words used in particular situations:ă aă formal/informală register”ă [79]ă andă Oxfordă Advancedă Learnerțsă Dictionaryă definesă ită asă „theă levelă andă styleă ofă aă pieceă ofă writingă oră speech,ăthatăisăusuallyăappropriateătoătheăsituationăthatăităisăusedăin.”[80] Hallidayă (citedă ină Devittă 2008:17)ă describesă registeră asă “theă configurationă ofă semantică resourcesă thată theă memberă ofă aă cultureă typicallyă associatesă withă aă situationă type.ă Ită isă theă meaningăpotentialăthatăisăaccessibleăinăaăgivenăsocialăcontext”. Registeră denotesă semantică resourcesă relatedă toă situationă types.ă Ită representsă theă potentială meaningă withină aă rangeă ofă aă givenă socială context.ă Onă theă otheră hand,ă registeră tendsă toă beă associatedă withă theă styleă ofă languageă usedă ină aă certaină situationă andă culturală environment.ă Officialăe-commerceăoră companyăwebsitesăuseăaăstrictăandăformalăregisterăwhereasăblogsăofă theăsameăcompaniesăuseăanăinformalăfriendlyăregister.ăAffiliateăwebsitesăuseăaăratherăinformală register.ă Onă Englishă websitesă theă differencesă betweenă formală andă informală registersă areă notă alwaysă perceivable.ă However,ă thereă areă targetă marketsă thată areă sensitiveă toă theseă issues.ă Ină Romaniaă theă usageă ofă formală oră informală registeră onă websitesă isă aă matteră ofă marketingă strategy,ă dependingă onă theă typeă ofă users/buyers.ă Someă companiesă useă formală registeră whileă othersăinformal. 3.4.2.4 Genre Longman Dictionaries Online define genre asă “aă particulară typeă ofă art,ă writing,ă musică etc,ă whichă hasă certaină featuresă thată allă examplesă ofă thisă typeă share”ă [78].ă Oxfordă Advancedă Learnerțsă Dictionaryă definesă ită asă “aă particular type or style of literature, art, film or music thatăyouăcanărecognizeăbecauseăofăitsăspecialăfeatures”ă[80]. 150 Genre,ăapartăfromătheăliteraryăgenreăcanăalsoăreferătoănon-literaryătexts.ăThereăareătwoămaină typesăofătext:ăliteraryăandănon-literary.ăTheseăareăseenăasăsuper-genres.ăNon-literaryăisămadeă upă ofă textsă fromă allă fieldsă ofă humană activity:ă administrative,ă legală andă officială documents,ă businessă texts,ă scientific,ă technicală bută alsoă publicistă textsă (advertisements,ă news,ă recipes,ă etc.).ă Non-literaryă textsă requireă preciseă andă unambiguousă termsă andă inappropriateă associationsă areă removed.ă Withă respectă toă structureă andă form,ă non-literaryă textsă useă stereotypicalăstructuresăandălinguisticăclichésăbeingămoreăformalizedăthanăliteraryătexts.ăNonliteraryătextsăareăpreciseăinădescribingăreality,ădrawăonăreasonăandăareăcharacterizedăbyălogicală argumentation.ăAlso,ănon-literaryătextsăareăwrittenătoăbeăskimmedăorăscanned,ă areăoftenălaidă oută asă listsă andă useă ană objectiveă pointă ofă view.ă Non-literaryă textsă fulfilă aă certaină pragmatică functionăwithăaăspecificăpurpose. Onă theă Internet,ă mostă ofă theă textsă areă non-literary,ă andă theă mostă commonă genreă isă theă technicală one,ă dueă toă theă multitudeă ofă onlineă businessesă whichă eitheră presentă oră sellă theiră productsăorăboth.ăTranslationăandălocalizationăofănon-literaryătextsăallowănoăinterpretationăandă requireă completeă accuracy.ă However,ă veryă often,ă becauseă ofă differentă lawsă andă regulations,ă evenă technicală translationsă requireă explicitation.ă Web-textsă whenă producedă oră localizedă shouldăincludeăconventionsătoăwhichăusersăareăaccustomedăto. Linguists have provided more specialized definitions for the concept of genre. David Russell (cited in Devitt, 2008:13) defines genres asă“typifiedăwaysăofăpurposefullyăinteractingăinăandă amongăsomeăactivityăsystem(s)”; while Carol Berkenkotter and Thomas N. Huckin (cited in Devitt, 2008:13) sees ită asă “dynamică rhetoricală formsă thată areă developedă fromă actors’ă responses to recurrent situations and that serve to stabilize experience and give it coherence andă meaning”.ă Hallidayă defines genre as:ă “aă dynamică responseă toă andă constructionă ofă recurringăsituation”ă(cited in Devitt, 2008:13) Oneă cană seeă thată theseă definitionsă coveră someă commonă traits:ă thată genreă involvesă typicală commonă featuresă andă typifiedă actions.ăTheseă areă developedă basedă onă repetitiveă limitations.ă Theărepetitiveăorărecurringăsituationsăareădeterminedăbyătheăsocialăcontextăinăwhichătheyăareă performed.ăTheă genreă ofă theă Internetă isă aă supra-genreă thată encompassesă andă combinesă bothă literaryă andă non-literaryă genres.ă Dependingă onă theă profileă ofă theă websiteă thereă isă alwaysă aă dominantă genreăandăseverală complementaryă genres.ăEvenăonăvariousăwebpagesăofătheăsameă 151 websiteăthereăcanăbeă genreădifferences.ăWhileăthereăisă someăsortăofăcohesionă ată theălevelă ofă websiteă structureă (Engebersten’să “structurală coherence”ă citedă ină Jiménez-Crespoă 2013:61)ă thereăcanăbeăgenreăagreementăorănot.ă Beneficială foră theă scopeă ofă thisă thesisă isă theă applicabilityă ofă Jiménez-Crespo’să proposedă categorizationăofă webă genresă(2013:97-99),ăespeciallyăreferencesătoă e-commerceă andăsearchă engines.ăThus,ătheăauthorăplacesăsearchăenginesăandăonlineăstoresăunderăInstrumental genres,ă andătheirăprimaryăfunctionsăisăexpositiveăorăreferential.ăWhileăconsideringăsearchăenginesăandă whată constitutesă theă highestă volumeă ofă globală search,ă searchă enginesă maină focusă isă theă referentială function.ă However,ă ifă searchesă areă e-commerceă driven,ă otheră function,ă areă similarlyă significantă andă areă languageă andă cultureă dependent.ă Thus,ă onă SERPsă theă ratioă betweenătheăappellativeăandăreferentială functionsăisămoreăbalanced.ăTheăcaseăstudyăină3.4.8,ă Call-to-action (CTA) and persuasive words in SERPs,ă demonstratesă thată SEOă professionalsă makeă useă ofă CTAă toă standă out.ă Ifă weă consideră nominală versusă verbală languagesă (sameă caseă study)ă oră Hall’să Lowă contextă versusă Highă Contextă culturesă theoryă (1976:ă 105-116)ă theă expressiveăfunctionămayăalsoăbeămoreăprominentăinătheăcaseăofăHCăcultures.ă 3.4.2.5 Style Styleăcanăreferătoăeitherătheăstyleăofăaăparticularăpersonăorătheăstyleăofăaăgroupăofăpersonsăthată shareătheăsameălinguisticăhabitsăatăaăcertainătimeă-ătheăparticularăwritingăstyleăofăaănovelăwriteră asăcomparedătoătheăwritingăstyleăofăseveralăwritersăthatăshareăaăsetăofăcommonăcharacteristică writingăstyle,ătheăRomanticsăforăinstance.ă Styleăalsoărefersătoălinguisticăappropriateness;ăusingăaăcertainătypeăofălanguageăbehaviorăbasedă onă theă situatională context.ă McIntoshă definesă ită asă “aă matteră ofă theă selectionă ofă particulară grammaticală patternsă andă sequencesă ofă patterns,ă andă ofă particulară itemsă ofă vocabularyă andă sequencesăofăitems,ăandătheăavoidanceăofăothers”ă(1963:ă248).ă Styleăcanăalsoăbeădefinedăinătermsăofătheăattentionăofătheătextăproducerăgivenătoătheăobservingă ofăhis/herăspeechăorăwriting.ăBellă(1984)ăspeaksăofă“audienceădesign”,ăwhichămeansămakingă changesă ină theăstyleăofătheătextă producedă accordingătoă theăaudience.ăWhileăină speakingăthisă canăbeărealizedăquickly,ăaccordingătoătheăfeedbackăofătheăaudience,ăinăwritingăthisădependsăonă theă speedă ofă exchangingă informationă backă andă forth.ă Ină caseă ofă letters,ă responseă andă thusă 152 potentială styleă changesă dependă onă theă post.ă Ină emailsă speedă isă moreă rapid.ă Ină caseă ofă advertisementăbillboardsăinătheăstreetăorăbannersăonline,ăchangingăofăstyleăwillăhappenăorănotă inărelationăwithătheăincreaseăorădecreaseăofăsalesăofăgoodsăorăservicesăbeingăadvertised.ăOrăină contentămarketingătextăcouldăbeăchangedăbasedăonăwebpageăstatistics. ItăisăimportantătoăconsiderăBell’săobservationsăaboutăstyle,ămoreăpreciselyătoăverifyăwhetheră theyă areă validă andă applicableă toă websites.ă Bellă summarizesăAudienceă Designă frameworkă ină nineăparagraphs: (1)ă Styleă isă whată ană individuală speakeră doesă withă aă languageă ină relationă toă otheră people.ă Styleă isă essentiallyă interactiveă andă social,ă markingă interpersonală andă intergroupărelations.ă(Bellă2007:97)ă Theămostăappropriateăexamplesăofăwebsitesăthatămayăbeăreferredătoăasăindividualăspeaker,ăor,ă toă aă largeră extent,ă text-producers,ă areă theă ownersă ofă personală blogsă oră anyă holderă ofă aă Facebookăaccountăthatăinteractsăandăsocializesăwithăhis/herăsocialăgroup.ăTheăonlyădifferenceă isăthatăinteractionăisădelayedăandăasynchronous.ă Ină caseăofăcommercialăwebsites,ăalthoughătheăproductionăofătheătextă involvesătheăworkăofăaă groupăofăpeople,ătheă companyă willăstillăhaveăaă consistentăstyleăonă allăitsăwebpages,ăbothă ină termsă ofă wordingă andă complementaryă elementsă onă theă webpages.ă Ină theă caseă ofă webpages,ă consistencyă isă aă moreă obviousă dueă toă theă navigationă menu,ă alwaysă presentă onă allă itsă webpages. (2)ăStyleăderivesăitsămeaningăfromătheăassociationăofălinguisticăfeaturesăwithăparticulară socialăgroups.ăTheăsocialăevaluationăofăaăgroupăisătransferredătoătheălinguisticăfeaturesă associatedăwithăthatăgroup.ăStylesăcarryăsocialămeaningsăthroughătheirăderivationăfromă theălanguageăofăparticularăgroups.ă(Bellă2007:97) Inătheăcaseăofăpersonalăwebsitesătheăstyleăcorrespondsătoătheăgroupătoăwhichătheăblogăowneră belongs.ăAăteenagerăwouldăuseăaăvocabularyăspecificătoăthatăageăgroup,ălotsăofăborrowingsăandă coinedăterms,ă oftenă grammaticallyăincorrectă structures,ăaămotherăwithă aăchildăwillă primarilyă useă diminutives,ă shortă sentencesă bută grammaticallyă well-structuredă sentences,ă whileă aă 153 technicală person,ăwillă provideăpreciseăinformationăonăaăproductă orăserviceăusingăspecializedă vocabulary. Onătheăotherăhand,ăcompaniesăwillăadoptăaăstyleăspecificătoă theăageă groupătargetedăbyătheiră products.ă Ifă theă targetă isă teenagers,ă slangă specifică toă themă mightă beă usedă bută ină termsă ofă grammarăstructureătheătextăwillăalwaysăbeăcorrect.ăBasedăonătheătargetăageăgroups,ăaăwebsiteă couldăbeălocalizedăaccordingătoăthisătypeăofăparameter. (3)ăTheăcoreăofăAudienceăDesignăisăthatăspeakersădesignătheirăstyleăprimarilyăforăandă ină responseă toă theiră audience.ăAudienceă Designă isă generallyă manifestedă ină aă speakeră shiftingăherăstyleătoăbeămoreălikeăthatăofătheăpersonăsheăisătalkingătoă—ă‘convergence’ă inătermsăofăaccommodationătheory.ă(Bellă2007:97) Ină speaking,ă theă changeă ofă styleă isă moreă dynamică dueă toă theă face-to-faceă communicationă channel.ăOnăaăblog,ăstyleămayăchangeăinătimeăandăearlierăpostsăwrittenăinăaăfriendlyăstyleămayă coexistăwithămoreăformalăstyles,ăwrittenăatălaterăstages. Commercială websitesă willă baseă theiră styleă andă operateă styleă modificationsă strictlyă basedă onă theă feedbackă ofă theiră websiteă users.ă Feedbackă cană beă translatedă ină sellingă moreă products,ă accessingă moreă webpages,ă spendingă moreă timeă onă certaină webpages.ă Styleă cană ină thisă caseă rangeă fromă familiară toă official/formală styles.ă Useră interactionă withă aă websiteă cană alsoă beă recordedăbyăinteractionătrackingăsoftware. (4)ă Audienceă Designă appliesă toă allă codesă andă levelsă ofă aă languageă repertoire,ă monolingualăandămultilingual.ă(Bellă2007:97) Thisăisăparticularlyătrueăinătheăcaseăofămultilingualălocalizedăwebsites.ăWebsitesăareăcomplexă semioticăsystems. (5)ă Variationă onă theă styleă dimensionă withină theă speechă ofă aă singleă speakeră derivesă fromă andă echoesă theă variationă whichă existsă betweenă speakersă onă theă ‘social’ă dimension.ă Thisă axiomă claimsă thată quantitativeă styleă differencesă areă normallyă lessă thanădifferencesăbetweenăsocialăgroups.ă(Bellă2007:97) 154 Variationsă withină theă targetă groupă cană beă easilyă tracedă andă recordedă withă theă assistanceă ofă onlineătoolsăandăareămostăoftenăbasedăonăsearchesăthatăareăenteredăinăsearchăenginesăandăonă theă websitesă themselves;ă theyă mayă beă alsoă basedă onă useră behavioră onă websites.ă Trackingă theseăvariationsăhelpsămaintainingătheăstyleăofătheăwebsite,ăbothăasătextăandăasăgraphicsăcloseă toă whată theă targetă groupă embracesă asă style.ă Foră instance,ă websitesă thată doă notă consideră theă issueăofăbannerăblindnessă(Nielsenă2007)ăfailătoăchangeătheăstyleăofătheirăcommunication. (6)ăSpeakersăshowăaăfine-grainedăabilityătoădesignătheirăstyleăforăaărangeăofădifferentă addressees,ăandătoăaălesseningădegreeăforăotherăaudienceămembersăsuchăasăauditorsăandă overhearers.ă(Bellă2007:97) Inăonlineămarketingăităisăimportantătoăfindătheăoptimumărangeăofăaddresseesăinăorderătoătakeă fullă advantageă ofă ană advertisingă campaign.ă Targetingă ană exceedinglyă wideă audienceă willă causeăfinancialălossesăinătermsăofăclicksăonăaăbannerăthatăisănotăproperlyădesigned,ăbutăalsoăofă potentialăbuyersăthatăwillăerroneouslyăclickătoăfindăsomethingăelseăthanăwhatătheyăexpected.ă (7)ăStyleăshiftsăaccordingătoătopicăorăsettingăderiveătheirămeaningăandădirectionăofăshiftă fromătheăunderlyingăassociationăofătopicsăorăsettingsăwithătypicalăaudienceămembers.ă (Bellă2007:98) Inătheăcaseăofăwebsitesătheăstyleămayăvaryădependingăonăvariousăpartsăofăaăwebsite.ăAăproductă presentationăpageăwillăvaryăinăstyleăfromăaăwebpageăthatăoffersăsupportăforătheăsameăproduct.ă Theălatterăwillăbeămoreăfriendlyăandăverbală(preponderantăusageăofăverbs)ăasăcomparedătoăaă moreărigidănominală(preponderantăusageăofănouns)ăstyleăofătheăfirst. (8)ăAsăwellăasătheă‘Responsive’ădimensionăofăstyle,ăthereăisătheă‘Initiative’ădimensionă whereăaăstyleăshiftăitselfăinitiatesăaăchangeăinătheăsituationăratherăthanăresultingăfromă suchă aă change.ă Sociolinguistsă haveă drawnă attentionă toă thisă distinctionă ată leastă sinceă Blomă andă Gumperz’să proposală ofă situatională versusă metaphoricală stylesă (1972).ă Ină responsiveă styleă shift,ă thereă isă aă regulară associationă betweenă languageă andă socială situation.ă Initiativeă styleă tradesă onă suchă associations,ă infusingă theă flavoură ofă oneă settingă intoă aă differentă context,ă ină whată Bakhtină hasă calledă ‘stylization’ă (l98l).ă Languageă becomesă ană independentă variableă whichă itselfă shapesă theă situation.ă (Bellă 2007:98) 155 Thereăareăoftenăcasesăofăwell-establishedăbrandsăwhichăinventăneedsă(productsăorăservices)ăforă theiră potentială clients.ăThoseă needsă mayă requireă aă specifică styleă ofă linguistică codingă soă thată targetăaudienceăisăapproachedăappropriately.ă (9)ă Initiativeă styleă shiftsă areă ină essenceă ‘Refereeă Design’,ă byă whichă theă linguistică featuresă associatedă withă aă groupă cană beă usedă toă expressă affiliationă withă thată group.ă Theyă focusă onă ană absentă referenceă groupă ratheră thană theă presentă audience.ă Thisă typicallyă occursă ină theă performanceă ofă aă languageă oră varietyă otheră thană one’să own.ă (Bellă2007:98) Onă commercială websites,ă oftenă thereă areă situationsă whenă websitesă collectă feedbackă fromă users,ă feedbackă leftă thereă foră othersă toă beă seen.ă Interestă foră aă certaină productă cană createă aă groupăaffiliation.ăTheăreferenceăgroupăconsistsăofătheăpeopleăwhoăpreviouslyăleftăfeedback. Actionădesignăisăaăquasi-equivalentăforătenorăofădiscourse.ăAnnaăTrosborgă(2002:11)ădefinesă ităasă“anăabstractătermă forătheărelationshipăbetweenăpeopleătakingăpartăinătheădiscourse.ăTheă languageăthatăpeopleăuseăvariesădependingăonăinterpersonală roleă relationships,ă whetherătheyă areăsymmetrical,ăi.e.ătheătwoăpartiesăareăequal,ăorăasymmetrical,ăi.e.ăoneăpartyăhasăaăsuperioră orăinferiorăstatusăinărelationătoătheăotherăparty.ăStatus,ăageăandăknowledgeăareăcrucialăfactors.ă Anăasymmetricalărelationshipăisăsaidătoăholdăbetweenăspecialistăandălayman,ăas,ăforăexample,ă betweenăaădoctorăandăhisăpatient,ăaăparentăandăhisăchild,ăaăteacherăandăhisăpupil,ăanăemployeră andăanăemployee,ăetc.ăAsăsuch,ătenorărefersătoătheăroleărelationshipăbetweenătheăaddresserăandă theă addresseeă andă mayă beă termedă ‘interpersonală tenor’.”ă However,ă asă Trosborgă definesă it,ă thereăisănoăsenseăofădynamicity,ătheărelationshipăisăstaticăduringătheăcommunicationăprocess.ă Also,ă theă relationshipă betweenă theă partsă isă alsoă culturallyă determined.ă Accordingă toă Hofstede’să studiesă (Hofstedeă etă al.ă 2010)ă thereă isă aă significantă differenceă betweenă theă employeră andă employeeă workingă relationshipă ină ană EUă countryă asă comparedă withă theă relationshipă withină ană Indiană company.ă Fromă myă ownă experienceă Iă cană assertă thată USă employersă wouldă generallyă consideră youă theiră equal,ă theiră workingă partner,ă whileă Indiană companyăownersăwouldăratherăadoptăaăparent-childărelationship.ăHCăculturesă(Hallă1976:ă105116)ămayă employă evenă moreălevelsă ofăinterpersonală roleă relationships.ă Usingă formalăversusă informalăaddressingăpatternsăonăwebsitesăalsoădeterminesătheătypeăofărelationship,ăsymmetrică orăasymmetric.ă 156 Tenorăofădiscourse,ăseeminglyădefinesăanăunchangingăcommunicationărelationshipăfromătheă initiationă phaseă toă theă finală state.ăWebsitesă withă establishedă authorityă ină aă certaină fieldă willă subconsciouslyăestablishăaăparent-childăasymmetricală relationship.ă However,ătheăstyleăneedsă toăbeăchangedăandăactionădesignăisămoreăappropriateătoădescribeătheărelationshipăbetweenătheă partsăinvolved.ăThereăhaveăbeenăsomeăstudiesădemonstratingăthatăaămale’săvoiceăandăfigureăină hisă mid-40să ină aă videoă advertisementă isă moreă likelyă toă beă trustedă andă thusă generateă moreă sales.ăOnăwebsites,ăoftenăthereăwillăbeăusedăaăpictureăofăaăsimilarămaleăwithăhisănameăandătitleă toăcreateătheăsameărelationship. Asă oneă cană seeă thereă isă someă overlappingă amongă register,ă genreă andă styleă ină someă respects.ă However,ă categorizingă textsă accordingă toă certaină commonă features,ă suchă asă form,ă semantică characteristics,ă socială recurringăhabits,ăorăsocialăcontextădeterminedă meaning,ă assistsătheătranslatorăandăimplicitlyă theă localizeră toă moreă accuratelyă comprehendă andătranslateătheămessageăformăoneălanguageă toăanother. Figure 33: Authority-like figure Source: http://www.drymat.de/de/ 3.4.3 The cultural layer Theăculturalălayerăisăseenăbyănumerousăscholarsăasăpartăofătheăcontext,ăanăextra-lingualăfactoră ofă importanceă ină theă economicsă ofă comprehendingă andă translatingă correctlyă theă intendedă meaningăfromătheăsourceătextătoătheătargetătext. Herskovitsă definesă cultureă asă “theă human-madeă partă ofă theă environment”ă (citedă ină Grossă 2013:ă254).ăNidaăperceivesăităasă “theătotalityăofăbeliefsăandăpracticesăinăaăsociety”.ă(citedăină Dimitriu,ă 2009:31)ă Hofstedeă etă al.ă defineă ită asă “theă collectiveă programmingă ofă theă mind”ă (2010:ă6).ăCultureăisă aătermăthatăencompassesăaăbroadărangeăofă objectsăandăconcepts:ăfromă 157 man-madeă toolsă andă objectsă fromă theă naturală environmentă toă values,ă ideas,ă languagesă andă philosophies. Cultureăisăaăsharedăphenomenonăcommonătoăpeopleăfromătheăsameăsocialăenvironmentăandăisă madeă upă ofă writtenă andă unwrittenă rules.ă Ită isă learntă andă notă transferredă throughă genes.ă Hofstedeă (2010:6)ă makesă aă cleară distinctionă amongă theă followingă concepts:ă humană nature,ă cultureăandăpersonality. Figure 34: Three Levels of Uniqueness in Mental Programming (Hofstede et al. 2010:6) Inter-culturală communicationă wouldă notă beă possibleă unlessă thereă wereă thisă universală andă inheritedăpackageăallăhumans,ăallăoverătheăworld,ăshareă-ăunderătheăconceptăofăhumanănature.ă Thisă commonă packageă allowsă usă toă partiallyă understandă humană beingsă fromă completelyă differentăcultures.ăHowever,ăasălanguageăisătheămostăimportantăpartăofăaăcultureăităisăaăbarrieră ină communication.ă Hence,ăină orderătoă beă acknowledgedăasă aăskilledătranslatorăandălocalizeră oneăneedsătoăgraspăallătheătraitsăofătheăsourceăandătargetăculture.ăTheăprimaryăextra-linguistică featureă isă culture.ă Personalityă isă whată makesă ană individuală unique.ă However,ă ifă thisă uniqueă individualăisăacceptedăasăanăauthorityăinăaăcertainăfieldăs/heăwillăsignificantlyăinfluenceăhis/heră socială environmentă andă ifă acceptedă asă exceptională s/heă willă setă trendsă withină aă culture.ă ShakespeareăorăEminescuăareăsuchăexamples.ăInătechnologyăandăbusinessăsuchăaăfigureăwasă 158 SteveăJobs.ăFurtherăon,ăweăshouldădifferentiateăbetweenăformalăleadersăandăinformalăleaders.ă Formală leadersă willă influenceă societyă byă “force”ă whileă informală leadersă areă embracedă naturallyă asătheyăareă feltălikeătrueăleaders.ăAăresearcherămayă beăbothăaă formală andăinformală readeră ifă heă isă recognizedă asă suchă bothă byă theă academică communityă (formally)ă andă foră exampleăbyăcontentăconsumersăandăbloggersăthatăwriteăcontentăonătheăsameătopic.ăAăblogger,ă isăalmostăalwaysăanăinformalăleaderăasăs/heăisănotăofficiallyăacknowledged48.ă2.6.3 Accordingă toă Hofstedeă etă al.ă (2010:8),ă thereă areă foură manifestationsă ofă culture:ă symbols,ă heroes,ă rituals,ă andă values.ă Theă closeră toă theă centeră aă typeă ofă manifestationă isă theă moreă problematicătoăacceptăchanges.ă Figure 35: Manifestation of culture (Hofstede et al. 2010: 8) Ină theă translation/localizationă processă ită isă importantă notă toă contradictă theă layersă ofă cultureă manifestation.ăWebsiteălocalizationăinvolvesăcommunicatingăacrossăcultures,ăandăwhileăsomeă productsă oră servicesă characteristicsă mayă beă acceptableă withină oneă cultureă theyă mayă beă completelyă unacceptableă foră theă membersă ofă anotheră culture,ă especiallyă ifă theyă contradictă valuesăorărituals.ăForăinstance,ămanyăEuropeansăwouldăassumeăthatătheămostăsuitableăclothesă toăwearăinătheăharshădesertăconditionsăisălight,ăwhiteăclothes.ăTheyăwouldăavoidăbuyingăblackă clothesăevenăifătheyăareăbyăfarămoreăefficientăatăclosingăoutădangerousăUVăandăIRăradiations.ă 48 ăSeeăalsoă2.6.3,ăInvisibility of the translator: when authorship matters 159 (Darkă colorsă attractă warmthă andă blackă ină Europeană culturesă isă associatedă withă mourning.)ă Resistanceătoăchangeăalsoăexistsăwithinătheăsameăculture.ăOnceămore,ătheăcloserătoătheăcenteră ofă theă circlesă ofă culturală manifestationsă theă moreă difficultă toă changeă theă culturală layer.ă Foră instance,ăwashingămachinesăwereăacceptedăwithădifficultyăbyăRomaniansăsomeă30ăyearsăago.ă Theăsameăisănowăhappeningăwithădishwashers.ăTheămostăcommonăexcuseăisăthatădishesăhaveă alwaysăbeenăwashedăbyăhand.ăToăchangeăthisăperspectiveăcompaniesăinvestăinăeducatingăandă demonstratingătoă potentială buyersătheăbenefitsă ofăsuchăapparel:ăreducedă wateră consumption,ă timeăandăeffortăsaving,ăenergyăsaving,ăadequateăhygiene,ăetc. 3.4.3.1 Cultural management Mostă researchersă ină culturală managementă andă cross-culturală managementă referă toă ită ină theă contextă ofă businessă organizations,ă managerială activitiesă andă businessă globală networking.ă Ină myăstudyăIăamămoreăinterestedăinăhowăcommunicationăisăachievedăacrossăvariousăculturesăandă especiallyăonline,ăwithăaăspecialăfocusăonăe-commerceăwebsites. Writtenăcross-culturalăcommunicationăisădeterminedăbyăconventionsăfromă theăsourceăcultureă andă targetă culture.ă Whileă Westernă culturesă wouldă useă mainlyă directă illocutionsă ină communication,ăEasternăculturesăwouldăuseăindirectăillocutionsăbecauseătheyăcanăofferămoreă optionsăforătheăaddresseeă(Hall’săHCăandăLCăcultures,ă1976:ă105-116).ăHowever,ăthereăisănoă automaticăsystemătoăcompletelyăattendătoăsuchăsophisticatedăculturallyăissues. Consideringăculturalămanagementăinătheăcontextăofăwebsiteălocalizationărequiresăconsideringă aă systemă thată recognizesă theă origină ofă aă siteă visitoră andă displaysă theă culturallyă determinedă featuresă suchă asă locală lawsă andă regulations,ă norms,ă customs,ă taxes,ă VAT,ă etc.ă accordingly.ă Thisămeansăthatătheălocalizerădoesănotăneedătoăworryăaboutătheseăconventionsăasătheăsystemă wouldă handleă themă automatically.ă Likeă theă navigationă menus,ă whichă needă toă beă localizedă onlyă onceă foră theă entireă websiteă andă notă individually,ă foră eachă ofă theă webpages,ă theă implementationă ofă localizationă elementsă intoă theă contentă managementă systemă andă theă databaseăneedsătoă beă attendedătoăonlyăonceă (yetă updatedăwheneverăthereă areă changesăină theă aboveămentionedăissues). 160 Theseăculturalămanagementăsettingsăcanăalsoăbeăappliedăandăautomatizedătoătheălayoutăofătheă website,ătoătheăformatăofănewslettersăorăpressăreleasesătoăbeăsentăout,ăorătoătheăformatăofătext,ă etc.ă Theseă areă theă elementsă ofă localizationă thată cană beă controlledă andă operatedă automatically,ă whereasă moreă refinedă aspectsă ofă translationă andă localization,ă namelyă culturală aspects,ă willă haveătoăbeădealtăwithăindividually.ă Otherăculturalăissuesăaălocalizerăshouldăbeăawareăofăareăbasedăonăvariousăculturalăstudies.ăVeryă often,ă evenă ifă theă localizeră isă nativeă ină theă targetă culture,ă s/heă mayă beă deficientă ină culturală skills.ă Suchăskillsărequireăspecialătraining.ă Inăwhată follows,ă Iăwillăbrieflyăcoverăsomeăofătheă moreăimportantăculturalăaspectsăofăwebsiteălocalization. Oneă ofă theă mostă importantă conceptsă which,ă ină myă opinion,ă mayă beă appliedă toă andă haveă ană impactă onă websiteă localizationă isă Hall’să high-contextă cultureă –ă low-contextă cultureă dichotomy.ăIăhaveădiscussedăatălengthătheăHCăandăLCăcontextsăinăpreviousăsubchapters49. Whileă Hall’să proxemicsă (theă useă ofă spaceă onă interpersonală communication)ă (1963)ă mayă beă viewedă asă applicableă onlyă toă traditională communication,ă ină myă opinionă communicationă distanceă cană alsoă beă impliedă andă determinedă linguistically.ă Ină Englishă aă directă styleă ofă communication,ăinătheăactiveăvoiceăandăwithămuchăusageăofătheă2ndăpersonăsingularăpronoună youă denotesă differentă proxemicsă thană theă usageă ofă passiveă voiceă oră thirdă personă pronoună singularăorăplural.ăFirstăpersonăpluralăpronounăweăshowsătheăhighestădegreeăofăclosenessăasăită involvesă thată everyoneă presentă isă includedă withină theă sameă spaceă andă involvesă theă sameă participatoryădegree.ăYouă-secondăpersonăpluralăpronoun,ădenotesătheăhighestădistanceăasăitănotă onlyă involvesă you asă aă separateă entityă bută youă asă aă distinctă groupă fromă ours.ă You -ă secondă personăsingular,ăisăspatiallyăcloserăasăităisăaădistinctăpersonăthatămayănotăbeăaffiliatedăwithăaă differentăgroupăandămayăadhereătoăwe.ă Anotheră dichotomyă introducedă byă Hallă andă Hallă (1990:ă 14)ă isă polychronică versusă monochronicătimeăandărefersătoăpeople’săpreferenceăforăusingătime,ăeitherăforăseveralătasksăată aătimeăorăforăonlyăoneătaskăthenăforăanotherăinăaăsuccessiveămanner.ăThisăisăoftenăreferredătoăasă ăSee,ăforăinstance,ă2.3.2ăInternationalization and text length, 2.5.2ăTranslation and business discourseăandă 2.6.4ăWeb content translation and search engine optimization (SEO) 49 161 Chronemics.ăOnăwebsitesăthisăwouldătranslateăbyăhavingăcontentăwithănumerousăhyperlinksă andănavigatingăbackăandăforthăthroughătheăinterconnectedătextărelatingătoătheăsameătimeăor,ăonă theăcontrary,ăfinishingăoneătextăandăthenăjumpingătoăaănewăpageăinătheăcaseăofăthoseăinclinedă toăbehaveăinăaămonochronicămanner. Anotherăsetăofăcultureărelatedăconceptsăthatămustăbeăaccountedăforăinăwebsiteălocalizationăisă Hofstede’sădimensionsăofăcultureă[81]:  Power distance index (PDI)ărefersătoă“theăextentătoăwhichătheălessăpowerfulămembersăofă institutions and organizations within a country expect and accept that power is distributedăunequally.”ăInătheăcaseăofăoldădemocraciesăPDIăisălow,ăwhereasăinătotalitariană countries PDI is high. PDI can be implied linguistically. Low PDI content consumer would be more responsive to a marketing strategy built on choices. For example compare the three best priced car insurance on the market and pick the one that suits you best. High PDI users may be manipulated in a different manner by using command and power words like must, or the imperative forms of verbs (often used as call-to-action words) –  not corroborated empirically. Individualism isă “theă degreeă ofă interdependenceă aă societyă maintainsă among its members”.ăTheăhigherătheăvalueăonătheăindividualismă scaleăisă theămoreăappreciatedăareă standards like personal achievements and individual rights. The lower the Individualism value, the more inclined a culture to collective sharing. Individual achievement is more dependent on the group to which the individual belongs to. Transposed into linguistic marketing communication low Individualism may use second person singular pronouns and verb forms while high Individualism (Collectivism) may use first person plural  pronouns and verb forms. Masculinity measures competition, achievement, ambition and power to the detriment of relationships and quality of life. Marketing content in the case of high masculinity value should be based on stressing competitive edge whereas in the case of feminine societies (with low masculinity index) the stress should be on improving quality of life (i.e.  advertisements for BMW versus IKEA). Uncertainty avoidance index (UAI):ă“Theăextentătoăwhichătheămembersăofăaăcultureăfeelă threatened by ambiguous or unknown situations and have created beliefs and institutions thatătryătoăavoidăthese”.ăInăcontentăproduction,ăwhetherăsourceăcontentăorătargetăcontent,ă high UAI content consumers expect explicit, exact and extensive information that leave no room to ambiguity and unexpected situations, whereas low UAI cultures feel more 162 secure and accept risk taking easier. High UAI cultures should have more verbose  webpages. Pragmatism refersătoă“howăeveryăsocietyăhasătoămaintainăsomeălinksăwithăitsăownăpastă whileădealingăwithătheăchallengesăofătheăpresentăandăfuture”.ăInăwebsiteălocalizationăthisă may be implemented as openness to novelty both in design, layout, and content  presentation, but also in convincing conservative cultures to purchase new products. Indulgence refersă toă “theă extentă toă whichă peopleă tryă toă controlă theiră desiresă andă impulses” Cultures scoring high on this dimension are more likely to allow themselves leisure time while thoseă scoringă lowă restraină themselvesă fromă “havingă fun”ă andă areă characterized by cynicism and pessimism. Content producer should be aware of this dimension as low indulgence scoring cultures may require extra effort to indulge themselves with seemingly unnecessary products or services. Whileă theseă culturală dimensions,ă bothă Hall’să andă Hofstede’s,ă mustă beă accountedă foră whenă translating and localizing, the translation and localization strategy must be adjusted to the product or service advertised. Considering the two approaches to cultural dimensions mentioned above, some of the findings may be contrasting when applied to certain cultures. Cultures, like languages, are subject to changes and are the result of unique historical, social and, recently, technological factors. 3.4.3.2 Romania’s cultural profile There is a significant amount of useful information about Romania on both official and research sites, and on general knowledge websites such as Wikipedia (the English and Romanian versions are different). From a cultural perspective the most useful website is The Hofstede Centre (http://geert-hofstede.com/index.php). 163 Figure 36: Hofstede's 6 cultural dimensions retrieved for Hall’s HC-LC countries Theă chartă aboveă illustratesă Hofstede’să 6ă culturală dimensionsă (source:ă http://geerthofstede.com/countries.html)ă ofă theă 10ă countriesă fromă Hall’să HC-LCă axis.ă Iă alsoă addedă Romaniaă andă Russia.ăAsă oneă cană see,ă whileă Romaniană isă ofă Latină origină itsă organizatională culturală profileă isă veryă closeă toă Russia’s.ă Romaniaă alsoă sharesă theseă organizatională culturală valuesăwithăBulgariaăandăSerbia.ăThereăisănoăinformationăonăUkraineăandăMoldova,ăbutăbothă haveăbeenăpartăofătheăUSSRăsharingăorganizatională valuesăunderătheăcommunistăregime.ăAsă forăMoldova,ăapartăfromăsharingăinătheăpastăsimilarăcommunistăregimesăwithăRomania,ămostă ofă itsă populationă speaksă Romanian.ă Theă Orthodoxă Church,ă theă maină religionă ină allă theseă countries,ă mayă alsoă playă aă crucială roleă ină theă culturală set-upă ofă theseă nations.ă Hungary,ă Romania’săwesternăneighborăhasăaădifferentăculturalăprofileă(sameăsource).ă Evenă thoughă Romaniaă isă closeă toă Slavică nationsă onă Hofestede’să 6ă dimensională culturală profile,ădueătoăcenturiesăofăhistoricăinfluenceăofăSlavonicănationsăonăRomaniaăandăRomanian,ă onăHall’săHCă-ăLCăaxisăRomaniaăisălocatedănextătoăItalyăandăSpaină[82][83].ăThus,ăRomaniaăisă classifiedăasăaăhigh-contextăculture. FromăHall’săproxemicsăpointăofăviewătheăusageăofănoi (we),ăwhereăapplicable,ăshouldăbeăusedă insteadă ofă theă secondă personă singulară tuă (you)ă oră dumneavoastrăă (theă politeă formă ofă you,ă secondăpersonăsingular),ăallătheămoreăsoăasăităcorroboratesătheăcollectivistăprofileăofăRomaniaă (lowă Individualismă index),ă seeă theă chartă above.ăWhileă tuă isă closeră thană dumneavoastrăă andă companiesă wouldă beă temptedă toă useă ită overă theă politeă form,ă notă allă theă clientsă oră potentială customersăareăpreparedătoăbeătalkedătoăinăanăinformalămanner,ăasătu isăperceivedăbyătheămoreă 164 conservativeă population.ă However,ă numerousă Romaniană companiesă useă theă secondă personă singulară onă theă Internetă bută theă politeă formă overă theă phoneă (myline-eon.roă –ă multinatională powerăcompany,ăemag.roă –ăonlineăretailă store,ăorange.roă –ămultinationalătelecommunicationă company);ă onă theă otheră hand,ă moreă conservative,ă usuallyă stateă ownedă companiesă useă theă politeăformă ină bothă casesă(efts.roă –ăRomanianăelectricăpowerăcompany).ă Bitdefender.com,ă aă Romanianăantivirusădeveloperăusesătheăformalămanerăofăaddressing,ăexpressedăthroughăverbală forms.ăThereăisănoăgeneralăruleăasătoăwhichăaddressingăformăisămoreăsuitableăinătheăRomaniană onlineăenvironment.ăAsăaăgeneralăruleăIăsuggestăusingătheăinformalătuăwhenătheătargetămarketă isă agedăupătoă 30,ăandătheăpoliteăformă dumneavoastrăăwhenătargetingă clientsă overă30.ăWhenă usingătheăpronounăproxemics,ăinăgeneralăoneăcanăuseătheăsameărulesăasăthoseăthatăapplyătoătheă Frenchă tu/vousă pair.ă Ină termsă ofă spaceă andă layoută usingă informală addressingă requiresă lessă spaceă asă theă pronominală formă isă impliedă ină theă verbală form.ă Whenă specificallyă usingă theă pronounsă ină theă message,ă theă Romaniană dvs.ă isă oftenă usedă insteadă ofă theă longă formă dumneavoastră. Furtherăon,ătheăRomanians’ăviewăonătheăissueăofătimeăisăpolychromic.ăTheyădoăseveralăjobsă simultaneously,ă familyă andă relationshipă isă moreă importantă thană punctuality.ă Whenă talkingă aboută Hall’să polycronică andă monocronică timeă ină websiteă localization,ă thisă couldă involveă havingă moreă linksă onă aă Romaniană webpageă than,ă foră instance,ă onă ană Englishă webpage.ă However,ăinăpractice,ălocalizationăwillămostălikelyăalwaysăinvolveătheăsameănumberăofălinksă toăexternalăpagesăonăbothătheăsourceăandătheătargetătext.ă AsăforăHofstede’săculturalădimensionsăappliedătoăRomanianăwebsites,ăIă considerăthatăfirstăofă allăoneăshouldăbeăawareăofătheăfactăthatăonlyă52.8șăofătheătotalăpopulationăresidesăinăurbană areasă ([84]ă 2011),ă andă Internetă penetrationă (Internetă Romaniană usersă onă Decă 31,ă 2013ă wasă 49.8șăaccordingătoăInternetworldstatsă[85])ăisăveryăhighăinăcitiesăandălowăinătheăruralăareas.ă Whenăconsideringăe-commerce,ăHofstede’săfindingsăappliedătoăurbanăinhabitantsă(theăhighestă potentialăcustomerăpoolăinăRomania)ămightădifferăquiteăsignificantly.ăTheyăareăwellăinformed,ă theyă areăwellăawareăofătheirărightsă andăareănotăafraidătoă speakătheirămind.ăAllăthisăisădueătoă closeăcontactăwithătheăWesternăEuropeanăcountriesăandătheăUSăoverătheălastă20ăyears.ăInămyă opinion,ăRomanianăurbanăcultureăisămuchăcloserătoăwesternăculturalădimensions,ăespeciallyăifă weăconsiderăpeopleăunderă40.ăThroughămultinationalsăandătheămediaăasătheămainăpromotersăofă westernă culturală valuesă ită isă safeă toă estimateă thată theă culturală profileă ofă Romaniansă willă continueătoăchangeăinătheăyearsătoăcomeăsoăasătoăcloserămatchăthatăofăwesternersăandătheăUS.ă 165 InătheăcaseăstudyăchapterăIăanalyzeăseveralăwebsitesăinăorderătoădetermineăifătheăculturalălayeră isăconsideredăasăpartăofătheăcommunicationăprocessăandătoăwhatăextent.ă Ină conclusion,ă Iă wouldă assertă thată reviewingă websiteă componentsă fromă theă perspectiveă ofă translationăandăe-localizationăisăusefulăinăaidingăaătranslatorătoăbecomeăaălocalizer.ăAăskillfulă translatoră needsă toă beă ană excellentă translator,ă advertiser,ă andă copywriter,ă beă awareă ofă theă economicăbackgroundăofătheăspecificătargetămarketăandămanageăculturalăissuesăconsistently. Theălocalizer’săskillsăneedătoăencompassăbothătranslationăandătechnicalăskillsăsuchăasăknowingă howătoăuseăaăCMS.ăCreativityăandăout-of-the-boxăthinking,ăawarenessăregardingătheăeconomică realitiesă ofă theă targetă marketă andă ofă locală andă globală trendsă areă assetsă thată transformă anyă translatorăintoăaăsuccessfulălocalizer. 3.4.4 Temporal layer TheătemporalăandăspatialăaspectsăofătheăcontextăareăseenăbyăDarwishăasăaăseparateălayerăfromă theă contextuală layeră (2008:164).ă Temporală andă spatială layersă cană beă sharedă oră notă byă theă initiatorăofătheăcommunicationăandătheăreceiver.ă Onăcommercialăwebsites,ătheătemporalăandă spatială layersă areă lessă important,ă asă theă textă isă ină agreementă withă theă currentă productsă oră services.ăTextsăaboutăproductsăareăwithdrawnăwhenătheăproductsăareăoutdated.ăAsăforăspatială aspects,ă ită isă importantă toă provideă referencesă thată areă pertinentă ină theă targetă culture.ă Foră instance,ă ină aă showă aboută touristică servicesă ină Egypt,ă comparingă theă heightă ofă theă Gizaă PyramidăwithăthatăofătheăStatueăofăLibertyăisătotallyăirrelevantăforăaăRomanianătourist.ăSuchăaă comparisonăwouldăbeăirrelevantăforăanyăpersonăthatăhasănotătravelledătoăNewăYorkăyet.ă 3.4.5 Intentionality layer Theă intentionalityă layeră encompassesă theă communicationă initiator’să intentions.ă (Darwishă 2008:165)ăThisălayerăcorrespondsătoătheăpurposeăorăskoposăofătheăcommunicatedătext.ăSearleă (1983:1)ădefinesăintentionalityăasă“directednessăorăaboutness,”ăoră“mind-to-worldădirectionăofă fit”.ă Intentionsă areă inferredă fromă context.ă Intentionalityă ofă theă produceră ofă theă textă mayă notă correspondă withă theă representationsă fromă theă mindă ofă theă receptor.ă Intentionalityă cană beă explicităasăpartăofătheămessageăorănotă(similarlyătoăLCăvsăHCăculturesăorăonăwebsitesăproductă descriptionăpageăversusăhomepage). 166 AccordingătoăLittlejohnă(citedăinăDarwish,ă2010:165),ăduringăaăcommunicationăprocessăthereă areă twoă levelsă ofă intent:ă informativeă intentionă andă communicativeă intention.ă Theă informativeăintentionămeansămakingăawareăofăsomething.ăTheăcommunicativeăintentionăisătoă haveăsomeoneăactăinăaăcertainădirectionă(inăaccordanceăwithătheăoperativeătypeăofătextă–ăReissă citedă ină Mundayă 2008:ă 72).ă Efficientă commercială websitesă produceă onă theiră productă pagesă bothă theă informativeă intentionă byă givingă theă featuresă andă technicală specificationsă ofă thată certaină pageă andă alsoă conveyă theă communicativeă intentionă withă textsă suchă as:ă “Addă toă theă basket”,ă “Viewă customeră review”,ă “Compatibleă accessories”ă andă otheră appellativeă constructions50.ă Ată theă communicativeă levelă theseă examplesă areă perceivedă moreă likeă informatională production,ă whereasă ană exampleă likeă theă headeră ofă theă followingă siteă impliesă involvedăproduction. Figure 37: Involved production at communicative level Source: creamandfudge.com Theăcommunicativeălevelăofătheăintentionalityălayerăinătheăcaseăofăcommercialăwebsitesăcanăbeă transposedărelativelyăeasilyăifătheătextăisăratherăofătheăinformationalăproductionătypeă(TMăoră MTăspeedăupătheăprocess).ăInvolvedăproductionăwillărequireămoreăattentionăfromătheăpartăofă theătranslatorăasăităinvolvesăcultural-specificăissues. 3.4.6 Intertextuality layer Fairclough (1992:101) describes intertextuality as the explicit juxtaposition of texts within each other. It means that a text is made up of several other texts, for instance by using citations, rephrasing, or reordering ideas. 50 ăSeeăcall-to-actionăconstructionsăină3.4.8ăCall-to-action (CTA) and persuasive words in SERPs 167 On websites, users are often allowed to comment on a news article or add a question on a support page, which can be answered by several other visitors. The original question and some of the answers can be cited several times as well. On commercial websites product pages allow comments which can lead to fervent debates. Intertextuality is a common feature on websites. Intertextuality can be intralingual (relating texts from the same language) or interlingual (involving texts from different languages – the relationship between source text and target text). Multilingual websites are the most common examples. One of the major problems is that a brand name can already carry a significant amount of information through its name, while in the target language, due to branding, the translator can transcode it only through explicitation. Slogans need to reflect the source text but maintain the intentionality aspect as well. Results in search engines are a distinct type of intertextuality as the search engine quotes chunks of texts based on the keyword(s) the user inputs. Jiménez-Crespo considers search engineăresultsăpagesă“theăsumăofăpartsăthatădoănotămakeăupăaăwhole.”ă(2013:ă49).ă However, this is not necessarily the case. The results, first of all, show unity through the common theme, reflected in common keyword usage, statistically calculated by search engine algorithms. Then we can consider the style particularities of such pages: specific structure, an ordered list, specific highlighting, and specific paragraph structure. The results create a possible context of what the user is interested in. A particular search result can function as the anticipation layer. The user will form his/her expectation based on the text displayed in the search engine results page. Intertextuality can also refer to chunks of text relating to other chunks of text from the same whole, being thus similar to a type of co-text. 3.4.7 Anticipation layer As already mentioned, above the 6 layers listed by Darwish (2008:155), I propose an additional one. The anticipation layer is a textual pre-reading component. The anticipation layer is based on subjectivity and first impression because the user accesses a certain type of content guided by emotions and feelings, as most often the information reaches the user for 168 only a limited period of time. Examples of such information are advertisements on TV, radio or in a magazine. Regardless if the advertisement is in a high-context culture or a low-context culture, the information supplied through advertisements is limited in time and space and there is no time for a verboseă presentation.ă Thisă isă aă typeă ofă “settingă theă scene”ă communication and it is anticipatory in nature. ThisăcorroboratesăNord’sădistinctionăbetweenă the addressee and the receiver (as recipients of the text)51. The anticipatory layer is intended for the addressee. In the case of search engines, unlike in traditional media and advertising, the users engage in receiving information (content produced with marketing intention in mind), by typing in keywords. This is non-intrusive marketing from the perspective of users, who wish to receive only the information they are interested in and only when they require it. From the perspective of companies this is called inbound marketing as they do not need to reach out directly. Potential customers are the ones to reach out for company products or services52. The results of their search is displayed in SERPs, as small snippets of text that may or may not convince the user that a particular webpage is the content the user is looking for. As with traditional advertising, the information provided by the search engine is limited in space, to a few lines of text. However, anticipatory information in search engines must resonate with the potential content consumer. The text snippet must contain either the keywords or compelling call-to-action words or both. The text of the snippets can be produced by the SEO engineer or automatically extracted by the search engine. This is what is called meta description by the industry. There have been numerous talks about the importance of the meta description and if it should be provided or left to be generated by search engines. Matt Cutts, a Google official cited by searchengineland.com [86], claims that the most important factor is to have unique meta descriptions. In case you cannot produce unique meta descriptions, it is advisable to let Google extract the snippet. Thus, meta descriptions can be an issue of duplicate information on the same website, and then the website is prone to penalizations. In translating and localizing the meta descriptions, the SEO unaware translator may be translating different meta descriptions from the source website using the same wording. For instance, assuming ăSeeăsubdivisionă2.4.9.3 ă Moreă detailsă onă thisă mayă beă foundă ină 2.4.8,ă Content marketing – delimitations in the sphere of web localization. 51 52 169 that on the source website there are three articles on the topic of skin care. For each of them the meta description is different. However, for each of the three webpages, the translator, for various reasons, may choose to use exactly the same meta description for each of the translated webpages. For example find out how to take care of your skin, take care of your skin with these tips, get tips on how to best care for your skin, could all be translated into Romanian as metode de îngrijire a tenului. TheăsameăsourceămentionsăthatăonăMattăCutts’ăblogănoămetaădescriptionăisăused.ăFurthermore,ă in practice, it appears that strictly from the point of view of search engines, most websites that target the Anglophone market use meta description on their pages. Out of 100 results on google.com for car insurance only 32 provided no snippet. On the other hand, Romanian companies prefer not to provide a meta description. Out of the first 100 results on google.ro for the asigurari auto keyword 68 websites provided no meta description for their car insurance related webpages53.ă Automatică snippetsă useă ellipsisă (…)ă to indicate that information is extracted from content residing on the targeted webpage. So, when targeting Romania for localization purposes, it might be safer to let Google generate the snippets. Letting Google generate them may also prove useful in finding unplanned keywords for which a website ranks in SERPs. On the same website, each webpage further needs to be unique both in terms of title and URL. As issues of duplicate content can appear in the case of meta description so the principles mentioned previously need to be applied for page titles as well. These issues are related to the previously discussed issues of originality in 2.5.6 Translation through parallel corpora and problem of originality. Considering the above mentioned concerns, in general, the anticipation layer is the consumer’săexpectationăorăprojectionăwithăregardsătoătheăadvertised/listedăproductăorăservice.ă If the anticipatory information is not in accordance with what the consumer anticipates upon accessing the webpage, the user will leave the page. Inconsistencies between the snippets in the SERP and the webpages that they link to can be either deliberate or accidental (lack of knowledge about Google guidelines for example). 53 ăSeeăAppendixăDăandăAppendixăE. 170 Deliberate inconsistencies (delivering content that is not in accordance with the keywords for whichăaăwebsiteăisărankedăinăSERPs)ăareădealtăwithăbyăGoogle’săantispamăalgorithms,ăwhileă forms of accidental inconsistencies may be displaying the content under the fold (requiring scrolling), because of a wide header or advertisements at the top, requiring sign-up, or prompting a pop-up, so that the user is delayed and forced to click. A similar tactic is used when leaving a webpage. The webpage traces the position of the mouse, and when heading towards the back button of the browser, it will serve a pop-up that will block the back button. So, going back on a previous page requires closing the pop-up first. Time spent on a webpage is measured by the search engines and thus is a factor that influences ranking in the SERPs. Someă websitesă evenă forceă useră toă “like”ă contentă onă socială networksă prioră toă actuallyă accessing the content. This is usually true in case of video content and is a rather widely used practice on Romanian websites. While such practices may fool Google bot (the search engine crawler) Google is also using a team of human raters that eventually will penalize such websites. If such practices are not employed on the source website, in theory there should be no such issues. However, often when a company outsources the localization of its website for a certain market the team to which the contract is awarded may need to impress the commissioner immediately. This is when the contractor team may use black SEO tactics (methods that are not in accordance with search engine optimization regulations). In the case of Romania there is a multitude of such cases, so companies wishing to outsource localization of their websites should choose carefully. While in the case of search engines most clicks are directed towards the top results on the first page, roughly 20% to the 1st position54, regardless of the text snippet, crafted or automatic, on the SERP, combining correctly keywords and call-to-action words increases the CTR (click-through rate). Therefore, in what follows, I am going to consider the most common call-to-action and persuasive words both in English and Romanian, with the purpose of making the Romanian list useful for companies that plan to localize their website for Romania. 54 ăPleaseărevisită2.6.4ăonăWeb content translation and search engine optimization (SEO). 171 3.4.8 Call-to-action (CTA) and persuasive words in SERPs Call to action and persuasive words may appear both on websites and in the meta description of the results from the SERPs. Such words are meant to determine a user to take a desired action: sign up for a newsletter, share his contact details, test a product or look inside a book. As the purpose of this subchapter is to consider SERPs and CTA as information found in SERPs form the anticipation layer, my focus will be on call-to-action words used in meta descriptions. Case study I built a list of the most popular words used in advertising from three different sources: 1. Aă BBCă programă citedă onă http://virtuallinguist.typepad.comă analyzingă theă mostă commonăwordsăusedăbothăinăpoetryăandăcopywriting.ăSeeăappendixăA. 2. WilliamăWellsăretrievedăfromă http://www.frankwbaker.com/persuasive.htmlăcollectedă theă listă ină appendixă B.ă Hisă researchă wasă basedă onă advertisementsă (Source:ă Wells,ă WilliamăD.ă(1965)ă"CommunicatingăwithăChildren."ăJournalăofăAdvertisingăResearch:ă 2-14.)ăSeeăappendixăB. 3. Mostăcommonăadjectivesăandăverbsăusedăinăadvertisingăaccordingătoălinguarama.com,ă aămulti-languageălearningăserviceăprovideră[87]ăSeeăappendixăC. First, I removed the duplicate words and there resulted a list of 74 words. They are displayed in alphabetical order. Second, I verified if in SERPs these words are as popular as with other types of advertising. The SERP against which I analyzed the popularity of the words is made up of the first 100 result for the car insurance keyword on Google.com. The following step was to translate the English list into Romanian, by using Google Translate, followed by adjusting the automatic translation by eliminating diacritics, for an easier analysis. Then I changed some of the words in order to highlight the Romanian words that were more likely to be used - some of them correctly suggested by Google translator, by clicking on the provided translation alternatives. I have not used any keyword research tools for the Romanian translation as the English words were not compared against any search engine statistics either and the list is quite general. Next, I modified the Romanian verbs to reflect the imperative mood which is the default verbal form used as call-to-action message. Unlike in English, the Romanian imperative verb form is different from the indicative mood. The feminine or masculine forms were also added for adjectives and nouns. Then I used the search capability of the browsers (by simultaneously pressing CTRL and F keys on my keyboard) to track the 172 number of occurrences for each of the 74 words. I repeated the same procedure for asigurari auto on Google.ro. I opted for car insurance and for asigurari auto, respectively, as it is a highly competitive market both in the Anglophone world and Romania. As for the Romanian keyword I opted for the plural form as it is the dominant keyword in Google Trends (http://www.google.com/trends/). By choosing this domain I further diminished my list (I removed words such as taste, fresh), retaining only those words that may relate to the car insurance industry, 67 words. The results can be seen in the table below: English words Google Count Romanian translation Adjusted Romanian count Romanian list alternative 0 Total count amazing 2 uimitor uimitor 0 announcing 0 anun ând anunta 2 bargain 0 afacere afacere 0 best 10 cel mai bun cel mai bun 0 better 1 mai bine mai bine 0 big 0 mare mare 0 bright 0 luminos inteligent 0 buy 17 cump ra cumpara 5 challenge 1 provocare provocare 0 choice 2 alegere alegere 1 selectie 0 1 choose 10 alege alege 7 selecteaza 1 8 come 0 vin vino 0 hai 0 0 compare 40 compara compara 11 11 don't 0 nu nu 8 8 easy 2 u or usor 0 0 extra 2 în plus in plus 2 supliment 0 2 first 2 primul primul 0 prima 1 1 free 21 liber gratis 2 gratuit 6 8 full 1 complet complet 5 get 74 ob ine obtine 0 give 7 da da 0 go 0 du-te du-te 0 mergi 0 0 good 0 bun bun 0 buna 2 2 173 uimitoare count 0 2 chilipir cea mai buna mai bun/a inteligenta/e 0 0 3 3 6 6 0 0 1 1 5 5 0 5 obtinerea 2 2 0 great 13 mare grozav 0 excelent 0 0 have 6 au ia (a lua) 0 ai 3 3 hurry 0 gr be te-te grabeste-te 0 0 îmbun t ire imbunatatire 0 introducerea prezinta 0 improvemen t 0 ameliorare 0 0 2 as introducing introdu 0 ctory keep 3 p stra pastreaza 0 know 4 tiu tii 1 like 0 ca sa iti placa 0 look 6 uita-te uita-te 0 0 love 2 dragoste (sa) iubesti/e 0 0 magic 0 magie magie 0 magia 0 0 make 0 face fa 0 sa faci 1 1 miracle 0 miracol extraordinar 0 minune 0 0 more 9 mai mult mai mult/e 3 3 most 3 cel mai cel mai 9 9 need 9 nevoie nevoie 1 new 13 nou nou/a 3 3 now 11 acum acum 2 2 offer 16 oferta oferta 17 17 only 2 numai numai 0 doar/decat 3 3 our 17 nostru nostru/noastra 3 nostri/e 0 3 people 3 oameni oameni 1 persoane 12 13 quality 1 calitate calitate 1 quick 4 rapid rapid 6 real 0 real real/a 0 remarkable 0 remarcabil remarcabil 0 0 revolu ionar revolutionar 0 0 rich 0 bogat (de) lux(os) 0 0 safe 7 inăsigurant 2 2 revolutionar y înăcondi iiădeă siguran 174 mentine 0 0 1 cum ar fi necesitate/tr ebuie 0 2 0 3 1 repede 0 6 0 extraordinar 0 0 see 6 vezi vezi 2 sensational 0 senza ional senzational 1 1 special 555 special special 2 2 start 2 Start start 0 incepe 1 1 startling 0 uimitor uimitor 0 surprinzator 0 0 suddenly 0 dintr-oădat instantaneu 0 0 0 sure 0 sigur sigur 0 0 0 take 3 ia ia 0 use 2 utilizare foloseste 2 wanted 2 a vrut dorit 1 we 9 noi noi 4 wonderful 0 minunat minunat 0 world 0 lume lume(a) 0 you 55 te te 1 your 42 dumneavoastr 2 dumneavoastra /dvs afla pe neasteptate de incredere 3 5 0 utilizeaza 1 3 1 ne 0 4 0 global/ 0 0 tu 1 2 tie 0 2 mondial Theă firstă columnă listsă theă mostă frequentă Englishă wordsă usedă ină advertising.ă Theă secondă columnălistsătheănumberăofăoccurrencesăforăeachăofătheăwordsăinătheăfirstăcolumn.ăTheărestăofă theă columnsă displayă informationă relatedă toă Romanian:ă Googleă Translator’să Romaniană translation,ăadjustedălistăfollowedăbyătheănumberăofăoccurrences,ăalternativeălistsăfollowedăbyă numberăofăoccurrencesăandătheătotalăoccurrences,ăsummingăupătheăadjustedălistăandăalternativeă listăoccurrences. English SERP Romanian SERP count count % symbol 24 7 1 digit no. 8 9 2 digit no. 6 1 3 digit no. 2 2 4 digit no. 1 21 Phone no. 2 3 Total no. 19 36 55 ăasăspecialist/specialise 175 Iăfurtherăaddedătoătheălistătheăpercentageăsymbolă(ș)ăandănumberăusage,ădividedăintoă1,ă2,ă3ă andă4ădigitănumbers,ăandăphoneănumbers.ăItăhasăbeenădemonstratedăinăseveralăonlineăstudiesă thatăusingăexactădataăisămoreăcompellingăthanăapproximateădata.ăForăinstance,ă usingămost of the people isăratherăvagueăasăităcanăbeăpresupposedăasăanythingăfromă(51șătoă99ș).ăWithoutăaă doubt,ă linguisticallyă thisă cană beă furtheră tunedă toă reflectă whată mostă means,ă fromă only slight majority forătheălowăpercentagesătoăvast majority forătheăhighăvalues.ăHowever,ăcomparedătoă usingă 80% of the peopleă theă linguistică expressionsă leaveă roomă foră interpretation.ă Foră thisă reason,ăusingănumbersăinăadvertisingăgeneratesămoreăleadsă(potentialăcustomers).ăPeopleăareă moreăwillingătoăacceptăoffersăthatăindicateăclearădelimitationăeitherăinătimeă(3 minutes to sign up versusă sign up in no time),ă stagesă (5 steps to whiter teethă vsă how to whiten your teeth),ă authorityăandăexperienceă(1st insurance company with 20 years of experienceăvsă one of the top company with many years of experience),ă oră choicesă (tens of insurance companies to choose from vsă top 10 insurance companies to choose from).ă However,ă usingă numbersă requiresă attention:ă …ă comparing cheap cară insurance quotes from over 120 insurance brands...ă oră We compare quotes from 139+ companies ... (actuală snippetsă fromă google.comă SERPăforăcar insurance, AppendixăD).ăProbablyătheăfirstăreactionăwouldăbeăthată“Wow,ătheseă guysă reallyă didă theiră homework!”ă Nevertheless,ă thereă areă studiesă provingă thată ană overwhelmingă numberă ofă optionsă cană blockă theă useră fromă makingă aă decisionă [88].ă Usingă smalleră valuesă wouldă inferă thată theă oneă providingă theă informationă savesă theă useră fromă theă burdenăofăhavingătoăscrutinizeăadditionalăinformationăhim/herself,ăthusăsavingăhim/herătime.ă Inăpractice,ăcompaniesăthatăfunctionăasăaffiliatesăchooseătoă advertiseătheă companiesăthatăareă theămostărewardingătoă them,ăasătheyă receiveăaăcertaină amountă forăleadsăorăsignăupăformsă oră otheră typesă ofă useră action.ă Anyway,ă theă useră isă moreă likelyă toă clickă onă aă resultă thată saysă compare the top 5 UK insurance companies. Lookingăbackăatătheăstatisticsăinătheătableăabove,ăanalyzingătheăEnglishăresultsăpage,ăoutăofătheă 67ăwordsăonlyă42ăappearăatăleastăonce,ăwhileăforăRomanianătheăvalueăisăslightlyăsmalleră–ă39.ă Inătermsăofăpercentageăofăpersuasiveăwordsăasăcomparedătoătheătotalăwords,ăforăEnglishăthereă areă 468ă oută ofă 3704ă -ă 12.63%,ă foră Romaniană 180ă oută ofă 3890ă -ă 4.62%.ă Whileă theă totală numberăofăwordsăareăroughlyătheăsameă(theădifferenceăisăprimarilyădeterminedăbyătheăhigheră numberă ofă paidă adsă ină theă caseă ofă theă Romaniană SERP),ă theă Romaniană percentageă ofă persuasiveăwordsăisăapproximatelyă3ătimesăsmaller.ăThisămayăindicateăthatătheătranslatedăandă adjustedă wordsă mayă notă beă theă mostă frequentlyă usedă foră theă Romaniană marketă oră thată theă 176 RomanianăSEOăprofessionalsăareănotăawareăofăcall-to-actionă(CTA)ăwords.ăSortingătheătableă byătheănumberăofăEnglishăwordăoccurrencesă-ăfromăhighestătoălowest,ărevealsăthatăwordăusageă inăEnglishăandăRomanianăisăvariable.ă English words Count Adjusted Romanian Romanian list alternative Total count tu you 55 te /dumneavoastra/ 4 dvs get 48 obtine/primi obtinerea 2-2 tau/ta 0 a(l) your 42 dumneavoastra/ dvs compare 40 compara 11 free 21 gratis buy 17 cumpara our 17 nostru/noastra nostri/e 3 offer 16 oferta oferte 26 great 13 grozav excelent 0 new 13 nou/a 3 now 11 acum 2 choose 10 alege selecteaza 8 best 10 cel mai bun cea mai buna 3 need 9 nevoie we 9 noi gratuit 8 5 necesitate/trebui e ne 3 4 TheătopăEnglishăwordăis,ăasăexpected,ătheăpronounăyouăwithă55ăoccurrences.ăTheăRomaniană equivalentătu/te/dumneavoastra/dvsăappearsăonlyă4ătimes.ăTheăpossessiveăform,ăyour,ăappearsă 42ătimes,ăitsăRomanianăequivalentsăareănotăused.ăTheălowăusageăofătheăRomaniană2ndăpersonă singulară pronoună isă standardă asă theă pronoună isă inferredă fromă theă formă ofă theă verbs.ă Whenă used,ăitsăpurposeăisătoăstressăorătoăinduceăaction.ăTu cu cine votezi?ă(Whoăareăyouăvotingăfor?)ă 177 isăsemanticallyăsimilarăto Cu cine votezi?.ăUsingătuămeansăthatăyouăpinpointătheăreaderătoătakeă action,ăinăthisăcaseătoăshareăhis/herăchoice.ă Theăsecondămostăusedăwordăisătheăverbăto getă–ă48ăoccurrences.ăForăRomanianăIăconsideredă twoăverbsăa obtine andăa primiăbutăIăalsoăconsideredătheănounăobtinere.ăOtherătranslationsălikeă prinde (prinde oferta), profita (profita de oferta) didă notă occur.ă Theă totală numberă ofă occurrencesă wasă 4.ă Thisă isă proofă thată get,ă aă veryă powerfulă (CTA)ă wordă isă usedă veryă frequently,ă whereasă Romaniană equivalentsă areă usedă scarcely;ă ină absoluteă numbersă 6ă timesă less,ă ină percentagesă 9.5șă ofă theă totală persuasiveă Englishă wordsă versusă 2.22șă ofă theă totală Romanianăpersuasiveăwords.ă Theă forthă entryă compareă andă comparaă evenă ifă ină absoluteă numbersă appeară toă beă quiteă divergent-ă40ăvsă11,ăinăpercentagesătheănumbersăareămoreăbalancedă–ă8.54șăvsă6.11ș.ăVeryă closeăinăpercentagesăareătheăvaluesăforăfreeă(4.75ș)ăandăgratis (4.44ș),ăandăweă(2.03ș)ăandă noiă (2.22ș).ă Foră mostă ofă theă otheră wordsă theă discrepanciesă areă ratheră high.ă However,ă Romaniană hasă itsă favoriteă CTAă wordsă thată outperformă asă occurrencesă theiră Englishă equivalents.ăThus,ăofferă(3.62ș)ăisăusedă4ătimesălessăasăcomparedătoăofertaă(14.44ș).ăOfertaă andăitsăpluralăformăoferteăisăalsoătheătopăRomanianăwordăwithă26ăoccurrences.ăalegeă(4.44ș)ă isăusedătwiceăasămuchăasăchooseă(2.22ș)ăandăităisătheăsecondămostăusedăRomanianăverb. Also,ăwhenăconsideringătheătopăwordsă(highestătoălowestăoccurrence)ătheăEnglishălistăisămadeă upăofă5ăverbsă(124ăoccurrences),ă4ăpronounsă(123ăoccurrences),ă1ănoună(16ăoccurrences),ă5ă adjectivesă andă adverbsă (68ă occurrences)ă oneă cană claimă thată theă snippetsă areă CTAă oriented.ă Theă similară numberă ofă verbsă andă pronounsă isă standardă ină Englishă asă onlyă imperativeă sentencesăcanăimplyăaăsubjectă(aăpronounăinătheăsecondăpersonăsingular).ăTheăsmallănumberăofă nounsăisăanăindicationăthatătheămainăpurposeăofătheămessageăisătoăinfluenceă(communicativeă intention)ă andă notă toă informă (informativeă intention).ă Onă theă otheră hand,ă Romaniană isă dominatedăbyănounsăandămodulatoryăwordsă(adjectivesăandăadverbs). 178 Adjsuted Romanian Romanian list alternative Total count oferta oameni 26 persoane 13 compara 11 cel mai 9 gratis gratuit 8 alege selecteaza 8 rapid repede 6 mai bine mai bun/a 6 Tableăcontinuedăfromăpreviousăpage Adjsuted Romanian Romanian list alternative Total count cumpara vezi 5 afla 5 complet 5 noi ne 4 nostru/noastra nostri/e 3 nou/a cel mai bun 3 cea mai buna 3 Ofertaă (offer)ă andă oameni/persoaneă (people)ă areă usedă 39ă times.ă Thereă areă 4ă verbsă (29ă occurrences),ă 7ă timesă firstă personă plurală pronounsă (noi, ne, nostru, etc.),ă 7ă adverbsă andă adjectivesă (40ă occurrences).ă Thisă suggestsă thată Romaniană marketingă isă moreă informationă oriented.ăTheăhighănumberăofăadverbsăandăadjectivesăsuggestăthatăRomanianăisăaăhigh-contextă culture,ăwhereămoodăisămoreăimportantăthanăinformationăitself.ăAăfullăanalysisăofăallătheăCTAă wordsăfurtherăsupportsămyăideas. 179 Theă twoă chartsă illustrateă theă preferencesă ofă advertisersă foră theă useă ofă certaină speechă parts.ă Whileă foră Englishă theă verbsă andă pronounsă couldă beă addedă upă asă theyă formă aă unity,ă ină Romanianătheăpronounsăcanăbeăusedăforăemphasis.ăTheăusageăofăadverbsăandăadjectivesăareă quiteăsimilar.ăNoteăthatăIăconsideredăforăeachăofătheăwordsăusageătype.ăForăinstance,ăqualityăină advertisingăisăoftenăusedăattributivelyăandănotăasăaănoună(quality cars).ăTheăusageăofă23șăofă nounsă foră Romaniană vsă 5șă ofă nounsă foră Englishă revealsă Romaniană marketersă orientationă towardsăinformationăwhichăisăratherăspecificătoălow-contextăcultures. 180 Asă foră numberă usage,ă oneă couldă claimă thată theă resultsă areă biasedă becauseă foră Romaniană Iă consideredăallătheănumbers,ăincludingăthoseărepresentingăyears.ăApparently,ăforătheăRomaniană web-userăităisăimportantătoăhaveătheăinformationăcontextualizedăforătheăcurrentăyear,ăbecauseă ofă theă cară insuranceă relatedă almostă yearlyă changesă ină lawsă andă regulations,ă henceă ită isă ană importantă elementă ofă theă ad.ăTheă useră doesă notă searchă foră informationă onă cară insuranceă foră 2012ă oră 2013ă bută foră 2014.ă Asă foră theă otheră numbers,ă theyă areă usedă foră providingă exactă information,ăbothăinătheăcaseăofăEnglishăandăRomanian.ă Telephoneă numbersă wereă alsoă includedă ină thisă caseă study.ă Addingă telephoneă numbersă indicatesăthatăthoseăbehindătheătextăsnippetăareăaărealăbusiness,ăbothăfromătheăperspectiveăofă potentialăbuyersăandăGoogle.ăCredibilityăisă builtă moreărapidlyăifătheămessageăfromăbothă theă SERPsăandătheăwebsitesăcanăbeăverifiedăandăreinforcedăoverătheăphoneăasăwell.ăSubsequently,ă Iăconsiderăthatăphoneănumbersăcanăalsoă beăeffectiveăCTAs.ăAllă inăall,ănumberăusage,ăthatăisă usingămoreăpreciseăsignifiers,ăappearsătoăbeăusedămoreăinăcaseăofătheăRomanianămarket. Asăforănegationăwords,ădon’tăorănot wasănotăusedăatăall.ăInstead,ătheăadverbănoăwasăusedă4ăoută 8ătimesăasăadvertisingăelementăandă4ătimesăasăpartăofăaăcompoundănounăusedăattributivelyă–ăno claimădiscount.ăForăRomanianănuăwasăusedă8ătimesăbutăonlyătwiceăwithăCTAăpurpose. ConsideringăthatătheăRomanianăpercentageăofăpersuasiveăwordsăresultedăfromămyăinitiativeătoă translateătheăEnglishăwordsăcausedăaăratherăsmallăpercentageă(4.62ș)ăoutăofătheătotalăwordsă fromătheăSERPsăasăcomparedătoăEnglishă(12.63ș),ăIădecidedătoălookăupăaălistăofăpersuasiveă andă CTAă Romaniană wordsă collectedă eitheră byă academicsă oră byă theă industry.ă Unfortunatelyă informationăonăsuchăwordsăisăeitherălimitedă-ăPsihosociologia publicităţiiă(AăPsychosociologyă ofă advertising)ă writtenă byă Septimiuă Chelceaă ină 2012,ă oră simplyă translatedă fromă English:ă http://romaniancopywriter.ro/15-cuvinte-care-vand/,ă gasim-cuvintele-care-vand-mai-bine/,ă http://romaniancopywriter.ro/cum- http://marketingprofitabil.com/cuvinte-care-vand-2/,ă http://www.wmm.ro/mesaj-publicitar/.ăForăRomanian,ătoărevealăsuchădata,ănoăactualăresearchă wasăconducted. AsăIăhaveăshownăpreviously,ătheătranslationăapproachăisănotătheămostăappropriateăor,ăifăităis,ăită mustă beă corroboratedă byă furtheră research.ă Theă objectiveă isă toă synthesizeă aă listă ofă provenă persuasiveă wordsă toă beă usedă ină metaă descriptionsă byă translatorsă andă localizers.ă Therefore,ă Iă decidedătoăuseăaădifferentămethodăappliedătoătheăsameădata,ăwhichăisătoădetermineăwhichăareă 181 actuallyă theă mostă usedă wordsă onă bothă theă Englishă SERPă andă theă Romaniană one.ă Oneă drawbackăwithătheădataăconsideredăforăanalysisăisăthatăităisăcarăinsuranceăspecific. TheătoolăusedăforăextractingătheăwordsăforătheătwoăsearchăengineăresultsăpagesăisăSide-by-Sideă SEOăComparisonăToolă[16].ăItărevealsătheăspecificăwordsăusedăbyăbothătheăAnglophoneăandă Romaniană websitesă fromă theă insuranceă industry.ă Beforeă providingă furtheră details,ă Iă haveă toă mentionăthatătheătoolădoesănotăaddressăappropriatelyătheăRomanianăcharactersăwithădiacriticsă ( , , ,î,â).ă Therefore,ă someă 1-wordă keywordsă mayă appeară ină theă tableă foră 2ă oră 3.ă Anotheră drawbackăisăthatătheălistăofăRomanianăstopăwordsăisănotăcomplete;ăhence,ămanyăofăthemăareă includedăinăthisăresultăpage.56ăHoweverătheătoolăisăstillăuseful. Google.com 100 results SERP analysis for Google.ro 100 results SERP analysis for car insurance asigurari auto There are 2907 words on this page 160 of There are 2685 words on this page 63 of these words were filtered out as "stop words" these words were filtered out as "stop words" 56 So there are 2747 words that are being So there are 2622 words that are being analyzed analyzed Word Count insurance 29 car Density Word Count 1.13% asigurari 124 4.84% 19 0.74% auto 122 4.76% quote 16 0.62% rca 88 3.43% auto 11 0.43% mai 33 1.29% online 7 0.27% casco 32 1.25% search 7 0.27% asigurare 32 1.25% nationwide 7 0.27% pentru 20 0.78% save 6 0.23% care 18 0.7% coverage 5 0.2% online 18 0.7% discounts 5 0.2% asigur ri 18 0.7% 2 Word Phrase Count 2 Word Phrase Count car insurance 17 0.66% asigurari auto 44 1.72% insurance quote 6 0.23% auto rca 23 0.9% auto insurance 6 0.23% asigurari 18 0.7% quote online 3 0.12% asigurari rca 13 0.51% Density ăForătheăcompleteăresultsăpageăseeăAppendixăF 182 Density Density quote today 3 0.12% rca casco 10 0.39% press enter 3 0.12% �ri auto 9 0.35% online quote 3 0.12% civila auto 9 0.35% enter search 3 0.12% cele mai 9 0.35% insurance quotes 3 0.12% raspundere civila 9 0.35% state farm 3 0.12% cel mai 8 0.31% 3 Word Phrase Count Density 3 Word Phrase Count car insurance quote 6 0.23% asigurari auto rca 12 0.47% press enter search 3 0.12% raspundere civila 9 0.35% asigurari auto 9 0.35% auto rca casco 6 0.23% asigurarea 4 0.16% asigurari auto casco 4 0.16% despre asigurarile 4 0.16% cea mai buna 4 0.16% unui accident auto 3 0.12% urma unui accident 3 0.12% Density auto raspundere civila auto Byăconsideringătheă1-wordătableăabove,ăweăcanăseeăthatămostăwords,ăbothăinătermsăofăabsoluteă numbersăandăinăpercentages,ăreflectătheăindustryăfromăwhichătheădataăwasăextracted.ăAăwordă thatăhasănotăbeenămentionedăinătheăinitialăEnglishăwordăisătheătermăonline.ăWhileăonăitsăownăită isă notă perceivedă aă persuasiveă oră aă CTAă word,ă itsă simpleă usageă presupposesă thată oneă ofă theă featuresăofătheăproductăorăserviceăisăthatăyouădoănotădependăonăaăperson,ăonăopeningăhours,ăită isăfaceăsavingă(oneămightăbeăafraidănotătoăaskă“aăstupidăquestion”),ătimeăsavingă(oneămustănotă waităonătheăphoneăline)ăandăfurtherăsimilarăexamples.ăWhileăallătheseăareăalreadyăimpliedăbyă usingătheăInternetăitself,ănotăallăonlineăcompaniesăuseăautomatedăsystemsăthatăhelpăweb-users.ă Foră instance,ă theă insuranceă industryă shouldă haveă aă systemă thată calculatesă theă priceă ofă theă insuranceăbasedăonăuserăinput.ăSo,ăonlineăisăaăpowerăwordăthatăcanăincreaseăCTR.ăOnline ină traditionalămarketingăisăreplacedăbyămoreăspecificăinformation,ăi.e.ăbyăusingătheăexactăInternetă address.ăOnăpostersăorăbillboardsăoneăcanăoftenăseeăitămentionedăeitherăasăaădirectăpageătoătheă productă oră serviceă advertisedă oră asă aă socială networkă page.ă Theă two-wordă andă three-wordă tablesăconfirmăthatăonlineăisăusedăintensivelyăinătheăonlineăadvertisingămedium.ăItă wouldăbeă 183 valuableătoădetermineăifăinăpracticeăusingătheăwordăonlineăactuallyăincreasesăCTR,ăconsideringă itsăinconstantăusageăamongăweb-users.ă Figure 38: Google Trends chart for car insurance related keywords Usageă variabilityă dependsă onă theă maturityă ofă theă industryă onă theă Internet.ă Online car insurance wasăusedăbyăweb-usersăintensivelyăfromă2004ăthroughă2008,ăwhenătheăindustryăwasă functioningă moreă likeă providingă informationă andă collectingă dataă fromă potentială clientsă throughăforms.ăWithătheămaturingăofătheăindustry,ăorăofăusers’ăinformationăawareness,ăaănewă typeă ofă toolă appeared,ă insuranceă comparisonă tool.ă Fromă 2009ă toă 2011ă car insurance comparisonăoutperformedăOnline car insurance.ăFromă2012ăonward,ăallătheseătermsăappearătoă haveă beenă balanced.ă Ină realityă allă ofă themă haveă beenă takenă overă byă aă moreă CTAă locution:ă compare car insuranceăwhichăstartedătoăbeăusedămoreăsignificantlyăatătheăbeginningăofă2008.ă Ină myă opinion,ă theă usageă of onlineă foră insuranceă industryă relatedă searchesă hasă decreased,ă becauseăonlineăbecameăanăinherentăfeature. Figure 39: Google Trends chart showing the dominant compareăcarăinsurance keyword 184 Thus,ă ită wasă replacedă byă aă differentă termă thată fulfilsă anotheră need,ă i.e.ă beingă adequatelyă informedă onă priceă andă featuresă ofă similară productsă oră services,ă ină thisă caseă cară insurance.ă Takenă individuallyă andă oută ofă contextă onlineă isă alwaysă onă theă increaseă andă surpassesă byă aă largeămarginăcompareăandăevenăadultăindustryărelatedăterms.ăHowever,ăusedăforăe-commerce,ă compare isă aă CTAă wordă thată corroboratesă theă findingsă fromă theă firstă approach.ă Theă sameă appliesăforăRomanianăbothăinăcaseăofăonlineăandăcompara. quote(s)ă andă price haveă ină thisă caseă aă veryă highă frequency,ă totalingă 55ă occurrence.ă Theă Romaniană equivalentă ină theă contextă ofă cară insuranceă isă pret(ul/uri) –ă 17ă occurrences.ă Synonymicătoăthemăareăcostă–ă8ăcounts,ăcosta (asăaăvb.)ăandăcosturi (asăaănounăinătheăpl.)ă–ă2ă countsăinăall. FurtherăanalyzingătheădataăinăAppendixăFăaăwordăthatăcanăbeăaddedătoătheăEnglishălistăofăCTAă isăsaveă–ă21ăoccurrences.ăItsăRomanianăequivalentăisăeconomisesteă–ă2ăoccurencesă(withinătheă sameă textă snippet),ă andă thereă areă noă otheră equivalentsă withină theă dataă analyzed.ă Anotheră RomanianăCTAăthatăisăequivalentăwouldăbeăfa economie.ăAnotherăterm,ărelatedătoăsaving,ăisă discount(s) –ă9ăoccurrences.ăInăRomanianătheătermăisăusedăasădiscount (usedăonceăinăaăURL)ă orăreducere(i)ă–ă10ăoccurrences. low (9ă occurrences)ă andă cheap (8ă occurrences)ă areă alsoă frequentlyă used.ă Itsă Romaniană correspondentă isă ieftin(e)ă –ă 12ă counts.ă Otheră wordsă areă minutes usedă toă conveyă thată theă processăsupposesăonlyăaălimitedătimeătoăaccomplish.ăjustăcorroboratesăonly andătheăRomaniană doarăandănumai. Theă completeă listă ofă wordsă foundă throughă analyzingă wordă densityă (numberă ofă wordă occurrencesăappliedătoătheătargetăextractedădata)ăisădisplayedăbelow: 185 Corroboratedălistăofă Furtherăsuggestedă Corroboratedălistăofă Furtherăsuggestedă CTAăEnglishăwordsă listăofăEnglishăCTAă CTAăRomaniană listăofăRomaniană (exactămatch) words wordsă(exactămatch) CTAăwords best call mutual alege afla buy cheap online anunta calculeaza choose check plans buna c uta compare cost price cel mai cel mai ieftin easy discounts provideă cel mai bun gasesti good enter quote compara importante need fit rates cumpara instant offer guide results dvs ofera people instantly right gratis oferim seeă just save gratuit online start learn selection mai buna persoane want low tip oferta poti minutes top pret visit reduceri tarifele Theă firstă columnă showsă aă listă ofă wordsă commonă withă theă initială list,ă synthesizedă fromă theă threeă listsă (Appendicesă C,ă D,ă E).ăTheă secondă columnă listsă myă CTAă suggestionsă ofă Englishă words,ăbasedăonătheăanalysisăofătargetădata.ăTheăthirdăcolumnădisplaysătheăwordsăthatăvalidateă theă adjustedă listă ofă wordsă fromă theă initială methodă ofă analysis.ă Theă lastă oneă containsă myă suggestionsăforăRomanianăCTAăwords.ă Theăpurposeăofătheăexperimentăwasănotătoăprovideăaăfull,ădefiniteălistăofăCTAăandăpersuasiveă wordsă bută toă buildă aă modelă foră furtheră analysis.ă Theă listă cană beă furtheră enlargedă andă finetunedăbyăincreasingătheădataătoăbeăanalyzedăbothăasănumberăofăresultsăinătheăSERPsăandăbyă usingăseveralămainăkeywordsăfromădifferentăindustries.ăThisăwouldăcreateăaăusefulălistăinăanyă language,ăsourceăorătarget.ăForăcreatingăaăspecificăCTAăforăvariousăindustriesăoneăshouldăuseă severalăkeywordsăforătheăsameăindustry. 186 Ită isă importantă toă underlineă thată CTAă andă persuasiveă wordsă areă NOTă similară toă keywords.ă TheyăcanăcoincideăforămoreăgeneralăsearchesăbutăasăIămentionedăinătheăsubchapterăonă Longtail keywordsă(2.6.7)ătheămostălucrativeăkeywordsăareămadeăupăofă4ăorămoreăwords.ăCTAsăcană beăpartăofăkeywords.ăTheămoreămatureăaămarket,ătheăsmallerătheălikelihoodătoăuseăsubjectiveă terms.ă Similarlyă toă keywords,ă theă contentă producersă andă theă contentă consumers,ă influenceă eachăotherăintoăfindingătheăcommonăgroundăterms.ăWhileăthereăwillăalwaysăbeăaăcoreăofăCTAă wordsăforăallătheăindustries,ătranslatingăforăaăcertainăindustryăwillărequireămoreăspecificăCTAă andăresearchămayăbeărequired.ăForăinstance,ă freshăcannotăbeăusedăforăanyăindustry.ăAlsoătheă usageă ofă CTAsă cană beă dictatedă byă theă company’să purpose.ă Cheapă willă neveră beă usedă foră aă Mercedes Benz.ă Generallyă speakingă theă cară companiesă willă notă useă suchă aă termă aboută theiră ownă productă bută aă positiveă synonymă likeă affordable.ă Instead,ă cheap couldă beă usedă byă affiliatesă thată planătoă attractă customers.ăTheămainădifferenceăbetweenăCTAăandăkeywordsă isă thată CTAsă areă addressedă toă theă emotionsă ofă theă web-useră (appellativeă function)ă whileă keywordsă toă theiră intellectă (informative/referentială function);ă CTAsă areă ratheră subjective,ă keywordsăratherăobjective.ăCTAsăratherăaddressăweb-users,ăwhereasăkeywordsăaddressăsearchă engineă robots.ă However,ă thereă isă noă clear-cută usageă betweenă CTAsă andă keywordsă andă theyă oftenăoverlap. Theătranslatorănotăonlyăneedsătoăhaveăadequateălocalizationăskillsăandătoăknowăhowătoăfindătheă suitableăkeywords,ăs/heăalsoăneedsătoădemonstrateăstrongăsalesăandămarketingăskillsăandăthusă becomeăaăcopywriterăinătheătargetălanguage. All in all, the textual layers suggested by Darwish (2010:155) can aid the translator and localizer in creating a framework for the process of translating and localizing. However, there is not always a distinguishable boundary between each of the layers. Some of the suggested layers are considered by other scholars as part of the context or of the co-text. The usefulness of such a framework can be tested in website localization. 3.4.9 The text as a communication process Pragmatică approachesă toă textă shouldă beă seenă asă complementaryă methodsă ofă analyzingă textsă andătheirămeaning.ăCrystală(2003:ă364)ădefinesăpragmaticsăas:ă 187 “…ătheăstudyăofăLANGUAGEăfromătheăpointăofăviewăofătheăusers,ăespeciallyăofătheă choicesă theyă make,ă theă CONSTRAINTSă theyă encounteră ină usingă languageă ină socială interaction,ăandătheăeffectsătheirăuseăofălanguageăhasăonătheăotherăparticipantsăinăanăactă ofăcommunication.ă“ă(originalăemphasis)ă Textăfocusesănoălongerăonăitselfăbutăonăitsăusersăandăhowătheyăinteractăbyăusingălanguageătoă influenceăoneăanother.ăInăe-commerce,ăwhileăcompaniesăstillăhaveăaămajorăroleăină“educating”ă potentialăcustomers,ăsocială interactionăshiftedăfromă B2Că communicationătoăC2Bă(customerto-business)ăcommunication,ăcalledăbyătheăindustryăinboundămarketing.ă 3.4.10 Presupositions Pragmatică presuppositionă “refersă toă theă variousă aspectsă ofă theă pragmatică meaningă ofă aă passageăthatăitsăwriterăassumesăareăpreviouslyăknownătoătheăhearerăorăreceiveră(…)ăTheyăhaveă toă beă establishedă prioră toă theă utteranceă ofă theă messageă andă mustă necessarilyă beă trueă foră theă messageătoămakeăsenseăandăflowănormally”(AlcarazăcitedăinăDimitriu,ă2002:ă43).ă Caffiă(2009),ăonătheăotherăhand,ădoesănotăagreeăwithăAlcarazăandăotherăscholarsăonătheăqualityă ofătheămessageăină termsă ofătrue/falsehoodă –ăthisăisă ratherăfromă aăsemanticăperspective,ă asăaă necessaryăcondition,ăandădefinesăpresuppositionăasădeterminedăbyăfelicityăconditions.ă Ină theă advertisementă foră Samsungă Galaxyă Note,ă ată theă endă thereă isă theă followingă text:ă “Phone?ăTablet?”ăTheăreceiverăofătheămessageăwouldăpresupposeăthatăthisădeviceăwouldăhaveă theăfeaturesăofăbothăaăphoneăandăaătablet,ăorăthatătheădeviceăisăfinallyăonăsale,ăetc. Presuppositionăisăalsoădiscussedăinăsubchapteră3.5.2,ăThe non-verbal component (non-text),ăină theă intratextual factorsă subdivision.ă However,ă becauseă onă theă Internetă pragmatică presuppositionăisăpartăofătheăAnticipation layeră(3.4.7)ăasăwell,ăinătheăSERPsăwhenăusersăenteră forătheăfirstătimeăintoăcontactăwithătheămainăcontentăthroughătheămetaădescription,ăIăincludedăită hereă asă well.ăTheă metaă descriptionă confirmsă theă user’să presuppositionă uponă searchingă byă aă keywordă onă aă searchă engine.ă Theă contentă andă metaă descriptionă mustă beă writtenă soă thată ită assumesăpreviouslyăknowledgeăofătheăuser.ăEstablishingăthemăpriorătoăproducingătheăcontentă andătextăsnippetsăisăachievedăthroughăkeywordăresearch.ăAsădemonstratedăpreviously,ăsettingă 188 upătheăkeywordsătransformsătheămessageăintoăaăpotentiallyăappropriateăcommunicationătool.ă Determiningăuserăpresuppositionăisăbasedăonăstatisticsăandăresearch.ă 3.4.11 Implicature The concept of implicature is often used in accordance with Grice’săconversationalămaxims.ă Implicature refers to what implications can be understood from an utterance or text. Implicature contrasts with the term explicature, which refers to what is explicitly communicated. Robyn Carston defines implicature asă “ană ostensively communicated assumption that is not an explicature; that is, a communicated assumption which is derived solely via processes of pragmatic inference.”ă(citedăinăMeibauer, 2006:576) On commercial websites we can assume the situation when, next to a product on sale, the followingămessageăappears:ă“Stockălow”. It implies that there are several products on stock or thereăisăatăleastăoneăproduct.ăExplicatureăinăsuchăaăcaseăwouldăbeămoreăappropriate:ă“Stock:ă7ă items”.ăTheăimplicature would still be that there are only 7 items left: “Hurry up if you want to get one”.ă“Outăofăstock”ăandă“Stock:ă0ăitems”ăimplyăbothăthatăaăparticularăitemăwasăonăsaleă and stock may be renewed. In the case of the SERPs, implicature is even more so worth considering. The few lines of texts imply the content of the webpage to which the user is taken to. Figure 40: SERP result The above result and the stressed words (in bold) imply that the explicature is that one can purchase cell phone accessories. Not naming the accessories, due to lack of space, one will assume that any type of accessory is available. Motorola, Samsung involves that these are the top brands for which there are accessories. The usage of ellipsis implies that there are significantly more brands. This is further emphasized by almost every mobile phone. However, almost denotes that not all the phone accessories for all the brands are on sale but 189 most of the brands. The localization of such a site may be difficult because the success of various brands depends on the purchasing power of various countries, so the localizer would need to have access to information such as top cell phones sales in the USA versus Romania if the site is to be localized for Romanian users. 3.5 Textăandănon-textăonăcommercialăwebsites.ă This subchapter is a further development to previous research. (Lakó 2012c) In 3.2, Text and E-text, I have reviewed several definitions for the term text and the one that accurately describes web-text or e-text as well is that of Kallmeyer (cited in Nord, 2005:16) who defines text asă“theătotalityăofăcommunicativeăsignalsăusedăinăaăcommunicativeăinteraction”. This definition considers text the entire semiotic amalgam used for communicating. In the case of websites text can then be seen as the totality of elements that make up a given webpage, including, on the one hand, the text itself (the verbal component of the website) and graphics, animations, videos and webpage layout on the other. Nord further differentiates between two main types of elements that may form a text and defines text asă “aă communicativeă actionă whichă can be realized by a combination of verbal and non-verbală means.”ă (2005:ă 16)ă Non-verbal text are further divided into elements of composition and layout, suprasegmental features (punctuation, using underline or italics), omission, irony, intentionality etc., while illustrations, diagrams, drawings (and sound and video files in the case of websites) as supllements or additional non-verbal means. Nord also states that non-verbal may transmit more information than the verbal component of a text (2005). As seen previously this is especially true in the case of HC cultures. This is in accordance with the saying - usedă veryă oftenă ină marketingă andă massă media:ă “Aă pictureă isă worthăaăthousandăwords.”ăandă“Seeingăisăbelieving.” For instance, a brand logo will communicate significantly more about who is behind the communication (who the tenor of discourse is). In Lakó (2010b), I discussed about the impact a brand name convey on potential customers. A certain brand may fall (or not) in the category of background knowledge context due to previous branding processes. However, a brand involves a significant amount of prior investment in terms of communicating (called branding by the industry). Some websites try to build reputation on pictures that imply 190 extensive information.ăOnăwebsites,ăoftenăthereăisăană„Ourăclients”ăsectionăwhichădisplaysăasă pictures the logos of the most important clients and nothing else. The communicated message andăintentionalityăisă“Buyăourăservicesăasăprestigiousăcompaniesăalreadyăhave.ăIfăwe are good enoughăforăthem,ădefinitelyăweăareăaăgoodăfităforăyouăasăwell.”ăLetăusăanalyze the following webpage: Figure 41: Example of "Our clients” webpage Source: bazaarvoice.com It is obvious that apart from the interface text, at the top, which is translated only once for the entire website, so it is not particular to this one webpage, the translator does not need to translate anything. Still, localization is required. If bazaarvoices.com were to localize their webpage for the Romanian market, most logos from above would be irrelevant. Relevant are only the logos of the companies that are globally established: Panasonic, Infinity, Timex and About.com, for instance. However, this isărelativeăasătheă“reader”ăofăthisăpageămayăbeăawareă or not about a particular car, watch or information service brand, so, the message is understood only if the web-user previously showed interest in those types of products or services. The localizer, in order to facilitate for the target user the comprehension of the webpage, would delete irrelevant company logos and replace them with brands that are wellknown in Romania, that is, only if the company established a business relationship with the 191 respective Romanian or multinational. On the other hand, there are numerous websites that indiscriminately display several previous clients just to make a first good impression. For the sake of simplicity and clearer delimitations I will refer to the text itself as the verbal component while all of the other website elements (non-text) will be enclosed under the term non-verbal component. These verbal and non-verbal components are often referred to as content. This division is similar to Nielsen suggestion (cited in Santini et al. 2010a :16), regarding the classification of websites based on two-dimensional genre model: in terms of traditional text characteristics and in terms of the communication channel - the Web; and includes website menus, sitemaps and internal links. Bateman (ibid) uses a tri-dimensional axis for categorizing websites: content, form, functionality. Functionality is what Nielsen includes under the navigation features of a website. I found this distinction useful and will refer to it in my thesis as the functionality component (3.5.3). A forth main component of a website, I consider important, is its technical component. Whereas the first two components can refer both to webpages and websites, the functionality and technical components are more practical to be used at website level only. 3.5.1 The verbal component Web.ăBruceă(citedăinăSantiniăetăal.ă2010aă:7)ăproposedătwoămainăgenresăforătext:ăsocialăgenreă andăcognitiveăgenre.ăOtherăbi-dimensionalăcategorizationsăofăwebpagesăcanăbeănominalăandă verbalătypes,ăandăIăproposeăaăthirdătwo-dimensionalămodelăthatăofăcorporateătypeăandănoncorporateătype. The cognitive and social genre is in consensus with Santini (2010b:97) who calls them rhetorical genres, and web genres, respectively. The cognitive/rhetorical genres have the features of the traditional printed text typology. As for the social genre, Santini (ibid) proposes 7 other genres which correspond more or less to actual types of websites on the Web: 192 Figure 42: Santini's genres(2010b:97) In what follows, I will study the types of websites acknowledged by most web-users. Most websites fall into one of the three following categories: - content-based sites (offering information – cognitive genre) - product- or service-based sites (commercial websites – mixed genre) - social sites (social genre) Other classifications of websites list thirteen types of websites, which in terms of numbers is pretty close to that of genres found by Santini. These website types are (Shelly 2010: 67-70): - portal (offers several services under the same brand(email, calendar, news, search) - news (the online equivalent of mass media entities(newspapers, radio, Television) - informational (websites that acknowledge information of general interest such as taxes, decisions made at governmental or local level, research result etc.) - business/marketing (websites belonging to businesses) - blog (hybrid diary/guide website; content may be of one individual or content can be contributed by users as well) - wiki (websites created primarily by the collective effort of the site visitors) - online social network (websites that allow users to create public profiles that can be used to access other peoples public profile, to share opinions, use chatrooms games etc.) 193 - educational (websites that teach the user on various subjects) - entertainment (online equivalent of television, but with extra interactivity added) - advocacy (lobby, cause, opinion websites) - web-application (online software functioning directly from a web browser, for instance a photo editor) - content aggregator (websites that gather Web content from different online sources for reuse or resale) - personal (websites created and maintained by individuals) Search engine sites may be considered a specialized type of web-application. Like genres, websites fallă intoă severală ofă theă aboveă categories.ă Lookingă backă ată Santini’să genres I propose two more important genre types: - administrative pages - navigation/interconnection structure (sitemaps) Furthermore, I consider that the administrative genre can be subcategorized into user administration pages and corporate administration pages. User administration pages are those pages when a user, for instance, signs up for a certain service (a free email service) by filling in the necessary fields of the sign up form. Corporate administration pages are pages from which a non-technical person can directly input/edit/change/delete content. These are also known as CMS (Content management systems). Administration pages function as an interface for the content manager. Using a CMS requires training. The interface of the CMS may be multilingual, thus it can operate in several languages. CMSs are also useful tools for handling all the multilingual content. The translator/ localizer can work directly in a CMS and instantly preview the page with the translated text. I consider the navigation structure of a website as a separate genre, and while it can be found on all of the pages of a website, it is a unitary structure, delimited from the rest of the site both in terms of functionality and visually. Navigation structure consists of the menu, sitemap and main outgoing links. In one word, everything that is recurring on each of the 194 webpages within the same website. Navigation structure is only one part of the user interface (UI). Also,ăstillălookingăbackăatăSantini’sătable,ăI do not agree with delimiting the front page genre to newspapers only. The term front page would be much more permissive especially that each website strives to be perceived as unique (sometimes even between localized versions of the same source website.) 3.5.2 The non-verbal component (non-text) Asă partă ofă theă non-verbală componentă Nord’să delimitationă betweenă intratextuală andă extratextuală(alsoăcalledăsupplementsăbyăNord)ădimensionsămustăalsoăbeăconsidered.ăSomeă scholarsă wouldăreferătoă thisă asăcontext,ă ină general.ă Nordă(2005:ă41)ăproposedătheă followingă schemeăofăanalyzingătheăintratextualăandăextratextualăfactors: Figure 43: Intratextual and extratextual factors First,ăIăwillăanalyzeătheăextratextualăfactors: - senderă –ă Ină theă caseă ofă websites,ă andă especiallyă commercială websites,ă ită isă veryă importantă foră theă receiveră toă knowă whoă exactlyă isă behindă theă business,ă asă theă communicationă processă initiator.ă Whenă translatingă andă localizing,ă ită isă importantă toă maintainătheăsameăvaluesăaroundăaăbrandăforăinstance.ăThereăareănumerousăbrandsăthată alreadyăcontainăaăcommunicationămessageăthroughătheănameăchosen.ăInălocalization,ă brandănameăwillăneverăbeătranslated.ăAsătheyăareănotătranslatedătheătranslatorăwillăhaveă toă explicitateă onă theă messageă carriedă byă theă companyă nameă ină theă initială 195 language/culture.ă Ifă localizationă precedesă traditională brandingă marketingă theă roleă ofă theătranslator/marketerăasăanăintermediaryăisăcrucial.ăIfălocalizationăisăaăpost-brandingă processă theă companyă nameă andă brandă denoteă theă sameă messageă asă ină theă sourceă language/culture.ă Ină localizationă althoughă theă translatoră playsă aă majoră role,ă s/heă isă irrelevantă toă theă receiveră asă theă web-useră isă interestedă ină theă initiatoră ofă theă textă (websiteă owneră oră theă brandă ină oură case).ă Ifă brandingă precedesă theă localizationă process,ăthatăis,ătheăbrandăisăalreadyăwell-knownătoătheătargetăculture,ălocalizationăisăaă significantlyăeasierăprocess.ăIfăbrandingăandălocalizationăareăachievedăsimultaneouslyă theătwoăprocessesăareălinkedăthroughăaăcoherentămessage. - intentionă–ăTheămainăpurposeăofăcommercialăwebsitesăisătoăinformătheăusersăaboutătheă productsă onă saleă andă toă closeă aă deal.ă Ită isă oftenă theă caseă whenă theă localizeră hasă toă changeă theă messageă toă reflectă productionă capacity.ă Globală mobileă phoneă manufacturersăwillăfirstădeliverătheirăproductătoătheăaffluentămarketsă(withătheăhighestă purchasingă power),ă andă then,ă ină stagesă theyă willă deliveră ită toă allă theă marketsă ofă theă world.ă So,ă whileă ină theă USă aă productă mayă haveă aă firmă productă launchă date,ă ină Romaniaătheălaunchingădateăwillăbeăjustăestimative.ăThisăwillăbeăreflectedăinătheătextă translated/produced.ăInătheăcaseăofăwebăservices,ăthereăisătheăadvantageăofăscalability.ă Webă servicesă cană beă virtuallyă extendedă toă anyă numberă ofă users.ă Foră instanceă givingă accessă toă anăinfiniteănumberăofăreadersăforăaăpaidă linguisticsămagazineăisă automatedă andă requiresă noă extraă productionă effort.ă Also,ă thereă areă marketingă strategiesă andă pricingăpoliciesăthatămayădifferăfromămarketă toă marketă –ăinfluencedăbyă crossăborderă taxes,ă currentă andă predictedă exchangeă rates,ă etc.ă Foră instance,ă Adobeă sellsă theiră softwareăatăaăcertainăpriceăinătheăUSăbutăforăEuropeătheăpriceăisănotătheăequivalentăină Eurosăbutănumericallyăidenticală(9.99€/monthăvsăUSȘ9.99/month).ăTheăadministrationă partă ofă anyă commercială websiteă willă mostă likelyă preserveă pricingă policiesă separateă fromătheărestăofătheăcontent. - receiveră –ă Theă receiveră isă theă mostă importantă factor,ă especiallyă ină e-commerce,ă asă his/herăresponseăcanăbeătracedăandăthusătheămessageă(fromătheămarketingăcampaignătoă productăspecifications,ăuserămanualăandăpricing)ăcanăallăbeăchangedăsoăthatăităfitsătheă largestăpossibleărangeăofăbuyers.ă - mediumă –ă Theă mediumă ină thisă caseă isă theă Worldă Wideă Web.ăAsă theă mostă complexă channelăofăcommunication,ătheămajorăproblemăwithătheăcommunicationăprocessăisăthată thereă areă variousă disruptions:ă linksă (externală –ă differentă websiteă oră internală –ă sameă website),ăadvertisements,ăanimatedăbanners,ăbackgroundăsound,ăetc.ăInălocalization,ăIă 196 haveă alreadyă shownă thată advertisingă willă consideră theă featuresă ofă theă locale.ăAsă foră links,ă thereă areă alwaysă linksă backă toă theă sourceă text,ă especiallyă ină theă caseă ofă theă globallyă establishedă companies.ă So,ă theă localizeră mayă needă toă addă ină severală moreă linksăthanăinătheăexistingăwebsite,ăasăanăalternativeătoăexplicitation,ăforăinstance. - ătimeă –ăThisă isă ratheră ană irrelevantă factoră ină caseă ofă commercială websitesă asă shownă previously.ăProductsănoălongerăonăsaleăneedănoătextătoăbeădisplayed. - placeă –ă Similarlyă toă time,ă becauseă ofă theă medium,ă thisă isă aă factoră thată doesă notă influenceăhowătheătextăisăreceived.ăVisitingăaăwebsiteăisănotălinkedătoătimeăandăspaceă andăusersăcanăreturnăasăoftenăasărequired.ă - motiveă–ăTheămotiveăofăanyăcommercialăwebsiteăisătoămakeăsales.ăWebsiteăvisitorsăareă awareăofătheămotiveăandăthisăisănotăaăfactorăofăsignificantăinfluenceăinătheămessage. Intratextualăfactors: “Intratextualăfactorsăareăanalyzedăbyăenquiringăaboutătheăsubjectămatterătheătextădealsăwithă(onă whatăsubjectămatter?),ătheăinformationăorăcontentăpresentedăinătheătextă(what?),ătheăknowledgeă presuppositionsămadeăbyătheăauthoră(whatănot?),ătheăcompositionăorăconstructionăofătheătextă (inăwhatăorder?),ătheănon-linguisticăorăparalinguisticăelementsăaccompanyingătheătextă(usingă whichă non-verbală elements?),ă theă lexicală characteristicsă (ină whichă words?)ă andă syntactică structuresă(inăwhatăkindăofăsentences?)ăfoundăinătheătext,ăandătheăsuprasegmentalăfeaturesăofă intonationăandăprosodyă(inăwhichătone?).”(Nord,ă2005:ă42) - theăsubjectămatteră–ăinăwebăcommerceăităgenerallyărefersătoăaăcategoryăofăproductsă onă saleă (laptops,ă tablets,ă etc.)ă oră toă aă brandă (meaningă allă typesă ofă productsă manufacturedăunderăaăparticularăbrand)ă - theăinformationă–ătheăexactătextualăcontentăaboutăaăproductăorăserviceăandăaboutăhowă ităcanăbeăpurchased - presuppositionsă–ăassumptionămadeăorănotăbyătheăauthor.ăForăexample,ănewăfeaturesă ofă aă productă cannotă beă presupposed,ă andă probablyă requireă lengthyă verbalization.ă Oră visitorsă toă theă websiteă cană beă attractedă byă “whată not?”ă isă foundă onă sale.ă Ită isă ană efficientătechniqueăforăchangingăaăpotentialăbuyer’sămind.ăTheăuserămightăbeălookingă foră“cheapăcellăphones”.ăStill,ătheălargerăco-textăofătheăverbalătextămayăbeă“…ă avoidă cheapă cellă phones”.ă Thisă doesă faultă Grice’să maximsă bută aă marketeră knowsă thată theă 197 mostă importantă factoră ină e-commerceă isă toă directă theă visitoră toă theă “shop”.ăValuableă andăconvincingăcontentămayătransformătheăpotentialăbuyerăintoăaăreturningăcustomer. - non-verbalăintratextualăelements: o punctuation:ă isă usedă toă showă variousă suprasegmentală unitsă fromă orală discourseătransferredăină writing.ăTheyăcanăbeăusedătoămakeăco-textăclearerăoră addă extraă meaningă –ă “foră (e)thought”.ă Punctuationă isă culturallyă specific.ă Foră instance,ăinăHungarian,ăinăaăletter,ătheăusageăofătheăexclamationămarkăafterătheă openingăofătheăletterăisătheănormă(TanárăÚr!ă–ăProfessor!),ăwhereasăinăEnglishă andăRomanianăităisănotăacceptable,ăasăcommaăisă used.ăTheăexclamationămarkă indicatesăaăfalseăchangeăofătheăasymmetricărelationshipăbetweenătheăsenderăandă theăreceiverăofătheăcommunicationă(seeătenorăofădiscourse). o fontă formatting:ă usageă ofă capitală andă lowercaseă letters,ă differentă fontă color,ă differentăfontăsizes o layout:ăblockăstyleăorăindentedăstyle,ăalignmentăofătext,ătextăspacing,ăusageăofă lists - lexicalăcharacteristicsă–ămostlyătechnicalăandăadvertisementărelatedălexicalăcreativity - syntacticăstructuresă–ăspecificătoăinformativeăandăoperativeăfunctionsăofătheătext - suprasegmentalăfeaturesă–ăforăinstance,ăfontăstyleă(bold,ăitalics,ăunderlined)ăcanăbeă usedătoăstressăimportantăideas. Non-verbalăorănon-textărefersătoăbothămarkersăwithinătheătextă(theăparalinguisiticăcomponentă asăseenăabove)ăandătoăvisualăenhancersăthatăareăaddedătoătheătextă(calledăsupplementsăbyăNordăă 2005) E-commerceăwebsitesăuseătheăinformationalăandăoperativeăfunctionsăofătheănon-textăelementsă extensively.ă Thisă translatesă usuallyă ină employingă multipleă picturesă ofă theă productă onă sale,ă professionalăorăcustomerăvideoăreviews,ăanimatedăuser-controlledă360ădegreeăview,ăusageăofă demoăillustrationsăorăvideo,ăadvertisementsăplayedăonătelevision,ăetc. Theălocalizerăwillăhaveătoătranslateătextsăonăpicturesăbutăalsoătheăvideosăandăprepareătheătargetă textăeitherăforădubbingăorăforăsubtitling.ăOrătheălocalizerămightă optăforăaădifferentăapproach,ă i.e.ătoăinviteăaăcustomerăfromătheătargetălanguageătoătestăandăreviewătheăproduct.ă(Videosăareă anotherăcomplexăsemioticămixăbutăinsteadăofăwrittenătextătheyărelyăonăoralătextăandăuseăwrittenă textăforăemphasis.) 198 Non-verbală elementsă onă websitesă cană beă aă challenge,ă especiallyă whereă theă appellativeă oră operativeă functionă ofă theă languageă isă used.ă Theă informativeă function,ă ină theă caseă ofă commercialăwebsites,ăwillănotăcauseăanyăproblemsătoătheătranslator/localizerăasălongăasăs/heăisă familiarăwithătheătechnicalăvocabularyăemployedăbothăinătheăsourceălanguageăandăinătheătargetă language. 3.5.3 The functional component Accordingă toă theă purposeă ofă theă webpageă Iă consideredă thată thereă areă threeă maină typesă ofă pages: - informativeă(Bruce’săcognitiveăgenre) - administrativeă(seeăabove) - socializingă (Bruce’să socială genre,ă whichă ină myă opinionă isă theă mostă appropriateă exampleăofătheă phaticăfunctionă ofăcommunication.ăProducedătext/contentăisă usedătoă establishăandăprolongăcommunication)ă Anotheră approachă toă theă functională componentă isă usabilityă ofă theă website.ăAlsoă calledă theă user-friendlyăcharacteristicăandăcanăreferătoătwoăaspects: - contentă qualityă (givenă byă affiliation,ă authorityă ină theă fieldă ofă activity,ă objectivity,ă accuracy) - easeăofăaccessingăcontentă(layout,ădesign,ănavigation,ăpagesădownloadingăfast) Also,ăaăthirdătypeă ofăcategorization,ăină termsă ofăfunctionality,ăisă byă analyzingătheăprofileăofă theătargetăreader/user57ă(specializedătexts,ăgeneralătexts,ăforăchildren,ăetc.). 57 ăSeeă2.4.9.3,ăClassification by recipient profile 199 3.5.4 The technical component The main differentiation worth noting here is between dynamic and static websites. We can consider it from two perspectives, that of the user and that of the website administrator. From the point of view of a user a static website is one that does not allow displaying content based on user input. User input can vary from changing font size or modifying the color scheme of a full webpage (from a range allowed by the website administrator) to obtaining results based on what is inserted in a search box. From a user perspective not all of a website isă dynamic.ă Pagesă likeă “Aboută us”ă andă “Contactă us”ă willă remaină theă same,ă whileă “Latestă news”,ăsupposedly,ăwillăchangeăveryăoften. From the perspective of a website administrator the webpages are static when content is not separate from design and layout. A dynamic website will involve an administration interface that is linked to a database (DB). All the webpages are dynamic when the website administrator can change easily any content on any and all of the webpages through the administration interface. The technical component allows for effortless administration both for translation and localization. An important tool for the localizer is a site statistics engine. The role of the statistics engine is to monitor user activity on the website and how the website performs. From the perspective of the localization of a website the following information provided by the analytics tool is vital in determining if the localization process is successful and to what extent. Based on the information provided, the localizer, also having the role of content administrator can operate changes to localize each of the webpages based on how they perform towards users. Details such as page views (with time spent on page), visits, unique visitors and repeat visitors over time, trace where exactly on a page users are clicking, browser type, track mouse movement, infer age of repeat visitors, etc. Based on the reading pattern the localizer should position the most important part of the content in the most visible areas. Viewing the actual click-string of each visitor, that is, the click history of a web-user through the entire site from the entry page to the exit page, is essential to verify the functional 200 component of the website, the intra-textuality of each webpage with the other webpages from the same website (website coherence) and the inter-textuality of webpages with external links (links to pages from other websites). There are also tools that indicate broken links: brokenlinkcheck.com, linkchecker.submitexpress.com, internetmarketingninjas.com/seo- tools/google-sitemap-generator/, etc. Clicks can also be tracked in case of pictures, videos, user manuals or animations by timing the periods spent on viewing the specific supplemental non-verbal element. In the case of pages predominantly using non-verbal elements it is useful to track which exact graphical elements are viewed. Considering Figure 40: Example of "Our clients” webpage, in the introduction of subchapter 3.5, the user will not spend time acknowledging all of the logos, yet the marketer and localizer should position at the top the brands most clicked on. I have demonstrated previously the importance of authorship58. Also, it is useful for the localizer to analyze the title, the anchor text, and the surrounding text of the inbound links from the referring webpages. It is even more useful to collect the keywords by which visitors reach a certain webpage and the search engine used. (2.6.6 Keyword research and website localization) The most important factor in assessing website performance is by the number of visitors who are then converted into buyers. 3.6 Theăelementsăofăaăwebsiteăandătheălocalizationăprocessă(aătechnicalăview) The components of a website are interconnected and all of them contribute to the success of the website. I will consider the typical parts of a commercial website. These considerations represent a continuation of my previous research on the subject. (Lakó 2012b) First of all, a website is composed of front-end elements and back-end elements. Frontend is what the website visitor sees or the website interface, while back-end refers to what happens behind the scenes. 58 ăSeeă2.6.3,ăInvisibility of the translator: when authorship matters 201 A third important factor is the human factor represented by all the persons involved in the localization process. 3.6.1 Front-end elements Front-end includes:  The domain name represents the very first contact of the visitor with the website. The domain name should be easy to remember and it can become the brand of the site. Domain names should be short and easy to spell and may contain keywords or  popular search terms. The logo. I have already shown the importance of the logo for branding. The logo bears a multitude of meanings and it is the most important non-textual element. The logo is positioned on all pages in the same location, usually in the top-left corner.  The navigation structure is similar to the contents of a book. It helps the user to quickly move from one page to another. It determines the hierarchy of the pages (main pages, subpages, sub-subpages). The navigation structure is represented on a site by one or several navigation menus.  Page layout refers to how text and non-text elements appear on the webpage. For instance the navigation menu can be positioned at the top, at the side or both at the top and side. When there is content under the fold as well, it is common to use a third navigation menu at the bottom of the page as well. Other layout features include delimitation of titles and subtitles, taking into account the reading patterns of visitors, determining the percentage of non-text elements versus text, consistency of layout design all over the website, etc. Usually, there is a layout design for the main page  design and a layout template for all of the other pages. Content is made up by text and non-text. o The quality of the text into achieving a certain language function is imperative in acquiring and retaining returning visitors. The text should contain keywords or popular search terms and equivalent expressions such as synonyms, elaborate sentences explaining those keywords, antonyms and antonymicăconstructionsă(“…ăavoid buyingăcheapăcellăphones”) 202 o Graphic elements. Backgrounds, images, delimiting lines, navigation bars, buttons, animations, videos, etc. are all very common on modern websites. All these front-end components are important for the localizer as each of them (except for the logo) may require to be changed for a successful localization. 3.6.2 Back-end elements Back-end elements are a method of making websites dynamic and stimulating for the visitors by providing more interactivity, quicker access to information (by adding a search box for instance), allowing retrieval of personalized information based on user input, allowing user to contribute content, etc. In what follows, I list the most important back-end components:  Content Management System (CMS) allows updating content without changing programming code. An appropriate CMS allows the administrator of the site to preview the updated content prior to pushing it live, update any pages of the website fromă“contactăus”ăandă“aboutăus”ăpagesătoăinformationalăpages,ăcreationăandăaddingăofă new pages in various areas of the website, or removing pages (in the case of products on sale, new pages must be created while for old products pages must be deleted.) The CMS, as already mentioned previously, is an important tool for the localizer due to its  features that allow control over the localized content. Newsletters are an important means of communicating novel information about the website or products on sale, discounts etc. The job of the localizer is to localize the content of the original newsletter. It may often be the case that newsletters are conceived directly in the target language by local marketers by means of  copywriting Website analytics software. I have previously mentioned its importance to the economics of efficient website localization Other website components with impact on the localizing process:  E-commerce functionality. Although there are items that are prone to manual localization, if the CMS benefits form advanced features the localizer needs not worry about changing currency symbols or using dot or comma for delimiting digits 203 of prices (1,23 versus 1.23). The system will automatically display the correct currency symbol based on the location of the potential buyer and, as shown previously pricing policy is separate from the information about a product. All of the features specific to a certain locale will be already established in the database (DB) and specific information retrieved accordingly (information on VAT, customs taxes,  delivery etc.) On-Site Search. Results will be displayed according to the locale of the user. A product that is on sale only in the US will not be displayed in Romania as well. Searches will be processed only based on the language of the visitor, content already  localized through the CMS. Blog feature, testimonials page or feedback pages are very common but the role of a localizer would be that of a censor only, that is, to eliminate inappropriate language, because everything would be in the language/culture of the target buyers. Modern systems can automatically disallow inappropriate language. Blogs are often used nowadays for inbound marketing as blogs are the genre that appropriately mirror the  SEO website localization type. Client support in the form of live chat room or through email. Information recorded and analyzed from these interactions can further improve the localization process. For instance, if the description about a feature of a product is not accurate (for instance, there are several questions regarding a certain feature of a particular service) the   localizer can operate changes (through explicitation) in the CMS. Referral forms and social networks are used for viral marketing. They are a form of measuring success of the site in general and of the localization process in particular. Online databases allow storing large amounts of information. They store information such as product information, comments, testimonials, feedback, specifics of a certain  locale, stocks, pricing policies etc. Downloadable files are useful in the case of user manuals, quick tips, new feature highlights, etc. Often they are replaced or accompanied by multimedia files. They   may be either localized or created from scratch for the target locale. Security is very important as there is a large amount of data that is confidential: client credit card numbers, trade secrets, proprietary programming, etc. Internal and external links on anchor text are also an important part of a website as they show the importance of a webpage. The more internal and external links to a 204 webpage, the more important the webpage. When translation and localization is applied this must also be reflected.59 3.6.3 The human factor One of the major factors in website localization is the human factor. The more people are involved in it, theămoreă likelyătheăpossibilityătoă imprintăeachăperson’să characteristicsăină theă output. The main players involved in the website localization process are the website owner and the web-user. While the role of the website owner is rather obvious, the role of the webusers is often ignored. I have shown in the previous chapters that the role of the end-user has increased significantly in how and when content and advertising is displayed.60 While the role of the web-users and their importance in the process of web localization has been extensively discussed in the previous chapters, the people involved in the actual process of localization has most often been referred to as the translator or localizer. For successful localization, however, most often there is an entire team involved. The number of the people involved depends on the site and type of the project. The more ample the project, the higher the number of people employed. Corporate owners and their website may hire hundreds of people, and have separate teams for each of the GILT processes. On the other hand, small companies or start-ups may function successfully with 2 or 3 people and extend as the number of target markets extend. For the sake of simplicity I will consider the smallest team possible and for one target market. For instance, in the case of a reverse-localization project a small team can offer a service both in their native language and in English. If the vocabulary involved in the localization process is highly technical translation can be achieved by a technical person. As for the other elements involved in localization, the team can either use information from other related sites (or from the competition) or use a localization-aware CMS. The website owner also referred to as the commissioner is at the top of the trophic chain. S/he outlines the purpose of the project (to localize it to a certain market), the budget and other administration and financial tasks. The project manager is the person who supervises a team of people in order to accomplish the desired outcome outlined by the website owner. 59 60 ăPleaseăreferăbackătoă2.6.5,ăSEO website localization ăForămoreăinformation,ăseeă2.4.8ăContent marketing – delimitations in the sphere of web localization 205 The most important person in the SEO oriented website localization is the SEO engineer as s/he is the person who does the keyword research and sets the direction and approach of the content marketing strategy for a specific locale. While the SEO might base keyword research on statistical data he/ she might be deficient in linguistic and cultural communication skills, evenă ifă theă targetă marketă isă theă engineer’să nativeă language.ă Otheră SEOă dutiesă includeă linkă building and exchanging links (getting other websites to link to the pages of the website, ensuring quality links and with a certain anchor text), guiding the programmer into making the localized website load as rapidly as possible on all types of devices, directing the designer optimizing the picture for fast loading, other optimization tweaks, sending out to directories articles that will link back to the website. While some of these tasks can be taken over by the localizer, especially that there are free SEO tools that analyze the page (forăexampleăGoogle’să PageSpeed), others, like link building, can be very time consuming, so, a lucrative solution is to maintain at least for the off-page SEO work a dedicated SEO engineer. If the SEO is important for determining the web-user visit the localized website through the search engines, the designer is responsible for adjusting user experience (UX) and user interfaceă(UI)ătoătheă“onlineăhabits”ăofătheătargetămarket.ăTheădesignerămustăfindătheăbalanceă between the requirements of the locale and branding requirements. There are numerous studies regarding UX and UI conducted by Nielsen Group (nielsen.com) and while there are differences in computer usage from market to market, the designer can rely on general human-computer interaction findings. However, numerous designers and companies change or adapt only the images, but as I have shown in 3.4.3, The cultural layer, the designer should be well aware of the target market. A high-context culture would require more images to be used as compared to a low-context culture. Most often, especially if the website is CMS or e-commerce driven, the programmer must not make any kind of changes as such systems are localization aware and written modularly. While a programmer may be required to develop a certain application or widget for a certain market, he needs be neither language nor culture aware. On the other hand, it is the programmer’săresponsibilityătoăfacilitateătheătranslationăandălocalizationăprocessesăinăsuchăaă manner that the translations and localization changes can be accomplished in a contextualized scenario. Delimiting textual strings from computer language, while very useful, can generate layout issues if the translated text cannot be visualized within the design elements. Building a 206 system in which the translator or localizer can preview their work can also significantly reduce quality testing. In this set-up I consider the localizer and the translator as one and the same person as the more technical locale related work must be implemented by the programmer, through appropriate programming. The localizer and translator concepts can thus be encompassed under the concept of localization-aware translator. Such a person will have to make the necessary culture specific changes so that it is web-user oriented. Furthermore, keywords have to be implemented in the content being produced so that it is relevant to search engines, and search engine optimized. The last but not least component of the team is the quality assurance (QA) tester. The QA tester needs to have a rather extensive set of skills: linguistic, cultural, detail oriented, plus verifyingăifăcontentăcompliesăwithăbothătheăcompany’sărequestsăandăusers’ădemandsăthroughă keywords on search engines. In website localization quality is often determined by financial factors:ă “qualityă assuranceă almostă alwaysă comesă secondă toă economică considerations”ă (Gouadec cited in Jiménez-Crespo 2013:108). This is one of the motives for introducing among the search engine algorithms tracing the time spent a specific webpage. Time spent on a webpage is indicative of the quality of a webpage, not necessarily from a linguistic perspective (for which a different set of algorithms is set up) but from a cognitive point of view: is the information useful for web users or not? Social networks are also important in evaluating cognitive value. The more shares the more valuable a webpage. Practice shows this is not always the case, but, ideally, the multitude of the algorithms used eliminates spam. Quality assessment in localization is achieved from a technical perspective and content perspective. The technical part verification can be automated by using tools such as World Wide Web Consortium’s http://validator.w3.org/ or various webmaster tools. Regarding content quality, companies use statistics tools which can rapidly generate reports with regard to user behavior and traffic. Decrease in traffic, quick exit from a certain webpage or other similar signals require attending to the factors that led to the problems. Analyzing the effect of webpage quality on user in website localization isăsimilarătoăNida’săreceptor-oriented and the functionalist approaches from TS (Dimitriu 2009:26). However, due to website complexity, it is difficult to create a quality assessment model. Jiménez-Crespo proposes as strategyătheă “qualityămadeătoă order”ămodel,ă whichăappliesă appropriatelyă toă aăwideă rangeă of websites and refers to expecting a certain degree of quality based on particular existing 207 conditions. (2013:131-132) However, this is not a new approach but a reenunciation of Nida’să viewă onă translation:ă “One cannot state that a particular translation is good or bad without taking into consideration a myriad of factors which, in turn, must be weighed in a number of different ways, with appreciably different answers. Hence there will be always a varietyă ofă validă answersă toă theă questionă ‘Isă thisă aă goodă translation?’ă ”ă (citedă ină Dimitriu,ă 2009:30). A further added factor to be considered is that, due to the high dynamicity of web content, the conditions change continuously. In summary, the three types of elements, front-end, back-end and the human factor make up the website complex that determines the efficiency and ROI of a localized website. 3.7 Conclusion Ină thisă chapteră Iă analyzedă issuesă regardingă textă andă electronică textă (similaritiesă andă differences).ă Thenă Iă reviewedă theă classificationă ofă onlineă textsă byă applyingă traditională textă classificationătheoriesătoă e-text.ă Ofăinterestă toăe-textă isă theăAudiomedială functională typologyă (Reissă citedă ină Munday,ă 2008:ă 72).ă Audiomedială textsă areă specifică toă theă semiotică mixă ofă websites.ăWebăcontentăisăcomposedăofătext,ăgraphics,ăvideoăandăsound.ă Next,ăIăreviewedăDarwish’sătextualălayersăandăhowătheyăapplyătoăwebsiteălocalization.ăWithină theă textuală layeră Iă proposedă aă newă element:ă theă shellă ofă theă text,ă whichă refersă toă theă macrostructurală organizationă ofă textsă (asă books,ă brochures,ă magazines,ă etc.).ă Basedă onă textă typologiesă Hatimă andă Masonă distinguishă instructională textsă asă aă distinctă typeă ofă textă (1990:ă 153-158).ă Onă websites,ă ită isă aă majoră textă typeă asă aă largeă amountă ofă theă informationă onă theă Internetă isă foră educatională purposes.ă Companiesă “educate”ă bothă theiră potentială andă existingă customers.ăThisăisăsimilarătoăBiber’săonlineăinformationalăelaborationă(1995:ă102). Aă two-dimensională classificationă ofă textsă intoă verbală andă nominală textsă revealsă thată mostă informationă onă theă Internetă isă nominal,ă thusă corroboratingă theă importanceă ofă instructională texts.ă Next,ă Iă analyzedă theă elementsă thată constituteă theă contextuală layeră ofă e-texts,ă highlightingătheăparticularitiesăthatăapplyătoăelectronicătexts. Regardingă theă culturală layer,ă Iă reviewedă majoră culturală trendsă appliedă toă text.ă Iă alsoă synthesizedă theă culturală profileă ofă theă typicală Romaniană web-user.ă Theă intentionalityă layeră 208 withă itsăsubdivisionsă–ăinformativeăintentionăandă communicativeăintention,ă showsăaăhighă degreeăofăapplicabilityăespeciallyăonăe-commerceăwebsites. Theăintertextualityălayerăofătextsăcanăbeăseenăbothăfromăanăintra-ăandăinterlingualăperspective.ă Intralinguală mayă referă toă coherenceă withină eachă ofă theă localizedă websites,ă whereasă interlingualărefersătoăcoherenceăamongătheăvariousălocalizedăwebsitesă(brandingăstrategy).ăAă specialătypeăofăintertextualityăcanăbeăaăsearchăengineăresultsăpage.ă Intertextualityăisăachievedă byătheăkeywordăenteredăbyătheăweb-user. Myămainăcontributionăinăthisăpartăofătheăthesisăisătheăadditionăofătheăanticipationălayerăasăană importantă partă ofă theă users’ă interactionă withă theă searchă engines.ă Theă anticipationă layeră constitutesăaăpre-readingăactivity.ăTheăanticipationălayerăalsoăinvolvesăconsideringătheăcall-toactionă wordsă fromă theă resultsă listedă ină theă SERPs.ă Iă analyzedă whată call-to-actionă refersă to,ă developedă aă methodă foră collectingă call-to-actionă expressions,ă builtă aă generală listă ofă call-toactionăwordsăforătheăRomanianămarket,ădifferentiatedăbetweenăcall-to-actionăwordsăandăwebuseră generatedă keywords.ă Theă comparativeă studyă onă Romaniană andă Englishă CTAsă alsoă revealedăthatăRomanianăusesăratherănominalătexts,ăwhereasăEnglishăverbalătexts. Subchapteră3.4.9,ăonăThe text as a communication process,ăcorroboratesătheămainăhypothesisă ofămyăthesis,ăbyăhighlightingăthatăe-textăevenămoreăsoăshouldăfocusăonăitsăconsumers,ăi.e.ătheă web-users.ă Interactionă shiftedă fromă B2Că communicationă toă C2Bă (customer-to-business)ă communication,ă calledă byă theă industryă inboundă marketing.ă Next,ă Iă alsoă discussedă howă pragmaticăpresuppositionăandăimplicatureătheoriesăapplyătoăwebsiteălocalization. Theăfollowingăpartăconsistedăofăanalysesăofătextualăandănon-textualăcomponentsăofăaăwebsiteă bothăfromătheăperspectiveăofăscholarsăinălinguisticsăandăTranslationăStudiesăonătheăoneăhand,ă andăfromăaămoreătechnicalăstandpointăonătheăother.ăWithinătheăverbalăcomponentăofăaăwebsiteă Iă proposedă theă additionă ofă theă administrativeă textă type,ă asă aă distinctă typeă ofă text,ă withă twoă subdivisions:ăuserăadministrationăpagesăandăcorporateăadministrationăpages.ă InătheăfinalăpartăofăthisăchapterăIăexaminedăallătheăfactorsăinvolvedăinătheălocalizationăprocess,ă fromătheăobjectiveăfactorsălikeătheăfront-endăandătheăback-endătoătheămoreăsubjectiveăhumană factor.ă Theă humană factoră isă theă mostă importantă oneă asă ită isă notă alwaysă aă factoră thată cană beă accuratelyădetermined. 209 4. Websiteălocalizationăanalysisăandăcaseăstudiesă 4.1 Introduction Inăthisă chapterămyăpurposeăisă toă applyătheătheoriesăandăverifyătheăclaimsăfromă theăpreviousă chaptersăonărealăwebsites.ăIăwillăbeginăwithăoneăimportantăstepăinăanyălocalizationăinitiative:ă benchmarking.ăWhileă benchmarkingă mayă beă seenă asă aă moreă generală process,ă ită hasă severală similaritiesătoătheăprocessăofăanalyzingătheăcompetitionăonăaăcertainămarket.ăConsideringătheă companiesă analyzed,ă theă topă globală companies,ă oneă mightă believeă thată thată theyă areă notă rightfullyă chosenă foră websiteă localizationă benchmarking.ă Ină myă opinion,ă theseă globală companiesăareătheămostăreliableăsourcesătoăfollowăforălocalizationăpurposesăasătheyăhaveătheă necessaryăfinancialăpower,ăexperienceăandăauthorityătoăsucceedăwhereăsmallerăcompaniesămayă fail. Whileă theă topă companiesă mayă beă ană excellentă modelă foră localizationă ină whată productsă andă servicesă offeredă areă concerned,ă theă legală issuesă andă relatedă topicsă foră theă EUă marketă andă implicitlyă foră Romaniaă are,ă ină theory,ă moreă suitablyă appliedă byă theă officială websiteă ofă theă EuropeanăUnion. Fromătheăperspectiveăofămajor-to-minorăandăminor-to-majorăculture,ăIăwillăalsoăconsiderătwoă maină typesă ofă internatională websites:ă websitesă thată employedă localizationă andă websitesă thată employedă reverse-localizationă (theă Romaniană Bitdefender.comă company,ă Daciaă andă variousă wineăproducers)ă Anotheră importantă partă ofă thisă chapteră willă showă thată localizationă doesă notă alwaysă implyă translationăandăinăpracticeănumerousăcompaniesămayăoptăforăcopywriting.ăWhileătheăapproachă ină theă caseă ofă theă twoă methodsă isă differentă theă purposeă ofă theă producedă textă shouldă beă theă same,ătoărankăhighăinăSERPsăandăgenerateăsales. Theă chapteră alsoă includesă someă experimentsă ofă producingă textsă eitheră throughă translatingă fromăSTăorăcopywritingăwhileăconsideringălocalizationăissues. 210 4.2 Benchmarkingăwebsiteălocalization,ăfromăanăSEOăperspective Numerous companies wish to promote their services and products on the global market. Websites are the most direct and less expensive persuasive methods addressed to potential buyers from around the world. Using only English is an adequate step but most often it is not sufficient, especially if planning to target the end-user. Consequently, I recommend researching leading practices and best-in-class localized websites that achieved global coverage, i.e. benchmarking localization. This case study is a continuation of previously published research. (Lakó 2013b) Replicating success often involves finding and going beyond what predecessors in a certain domain achieved. Online businesses adhere to the same principle of mimicking the accomplishments of similar websites. Google was not the first search engine on the market yet it is the most widely used [39], surpassing Lycos, Alta Vista, Magellan, Yahoo! and other search engines. The same applies in the case of website localization. One has to analyze websites that achieved measurable success in website localization. If, currently the most important web search engine is Google, any website should be localized considering search engine optimization (SEO). Prior to engaging in developing a multilingual geo-targeted website one should also conduct research on local competitors or multinationals that have already entered the target market. In the case of commercial website localization, there are as many locales as geographical target areas. While two or more different languages may not imply a localization process (city hall websites of bi- or multilingual cities), the same language used in different parts of the world requires localization as there are specific customs, laws, currency, interests, market specifics (for instance price policy) etc. A website for the UK should be different from that for the US or for Australia or Canada. Through website localization companies may show their specific interest in various regions of the world, by adapting their products and services to the specific locals. The more locales targeted, the more potential customers for a certain business. 211 businessdictionary.com defines benchmarking as: A measurement of the quality of an organization's policies, products, programs, strategies, etc., and their comparison with standard measurements or similar measurements of its peers. The objectives of benchmarking are (1) to determine what and where improvements are called for, (2) to analyze how other organizations achieve their high performance levels, and (3) to use this information to improve performance. In the case of website localization, when focused on SEO aware content, you must consider howăcontentăqualityăcanăbeămeasuredă andăimproved,ăandăhowăităcomparesătoă theăindustry’să best practices; it is a routine that has proved over time equally effective in achieving certain results by means of measurable Internet traffic, usage of a web service, online sales or visits to a particular article on a webpage. Benchmarking can be achieved internally as well for similar articles from the same website (internal benchmarking) or competitor webpage versus company page. Benchmarking can be applied both to start-up websites and existing websites. In the case of start-up websites it requires thorough research and planned modularity and adaptation to new industry standards, so that change policies also follow leading practices. In the case of existing websites a more suitable benchmarking definition would be somewhat similar to internal benchmarking mentioned previously but websites taken in general should always be compared with its competitors that show the highest performance in the same field of online activity. This type of benchmarking requires carefully organized change management. In the case of website localization the most appropriate benchmarking type is that of best-inclass benchmarking. Best-in-class benchmarking is defined as: The highest current performance level in an industry, used as a standard or benchmark to be equaled or exceeded. Also called best of breed. (http://www.businessdictionary.com) 212 With regard to online presence and website localization a simple and measurable method to find the leaders in this industry is by reading various statistics provided either by those companies themselves or by third-parties. One very specific technique for determining who the leaders in website localization are is to find the websites with content or services translated and localized in the highest number of languages. However, the number of languages is not necessarily an indicator of quality translation and localization for websites. As shown previously, there are several free translation tools that can automatically translate a website into several languages. Google covers 126 languages, but as with any automatic translation tools, there is a quality deficiency issue and localization is accidental or missing completely. Such automatic tools are based on websites crawled by various search engines. Google Translate does not use a rules-based approach but a statistical methodology. Google is based on a global corpus containing billions of words from monolingual texts in various languages around the world and also human aligned translated texts. Translations on Google often provide variants as well. Usability and quality of automatic translation is based on the number of webpages written in aă certaină languageă andă indexedă byă Google’să webă crawlers.ă Accordingă toă w3techs.com,ă English is the most used language to publish content online with 55.4% of the total WWW content where Russian is positioned second with only 6.1%. While Chinese comes second as number of users it is 7th in terms of content. Content by language does not reflect the number of non-English users. It reflects the status of English as a lingua franca (Crystal 2003, 2006). This means that there are numerous opportunities for website translation and localization. 213 Figure 44: The number of Internet users by language Considering the two graphics, translation, for some of the language pairs, is a more suitable strategy than in the case of others. English is often used as an intermediary language (pivot language) when translating from a less important language, in terms of numbers of speaker, to another minor language. However, automatic translation will improve once Google crawls and analyzes even more content. Figure 45: Internet content by languages[45] Generally speaking, automatic translation quality depends on the corpus available. The larger the indexed text corpus grows, the more accurate the translation and its quality. It is also useful to notice, that a more specific text is often more appropriately translated than a general text. Translation quality has also been improved significantly by indexing human translated texts from the EU and the UN. 214 While automated translations have improved significantly, there is still need to tackle the localization issues. The general localization issues can be usually addressed by using free ready-made solutions such as Drupal (drupal.org), Prestashop (prestashop.com) and other similar Content Management Systems. Such systems already handle some specific issues like setting currency, VAT, country areas, languages, counties etc. Content itself will still require to be localized and be in accordance with the marketing goals of a product or the entire website. Price policy, local context, local glossary, keyword research, local landing pages, cultural awareness, user interface and graphics localization, dubbing video or even recreating it, etc. are just some of the localization issues that require to be attended to manually. Whereas it is recommended to consider EU official translations as a corpus in case of legal terms and other law related texts on aă company’s website, the rest of the content must not only be useful but also unique, to provide new information to the reader. This is an important issue, in order to rank well in the search engine results pages. Google penalizes duplicate content, so, instrumental translation is the most appropriate strategy. An alternative to translation is copywriting in the target language, which, to some extent, is more localized as it is produced by native speakers. While translation and localization can be achieved using various tools, free or paid, there is still need to consider the best-performing websites within the industry or even adhere to their recommendations. The appropriate manner to initiate the benchmarking process for website localization purposes is to select the top global brands. One can study a general listing or companies from a certain industrial sector. Interbrand.com offers the list with the top 100 highest-performant global companies [89]. The obvious websites to consider are those of the top global companies. Apple is number one for 2013 followed by Google, CocaCola, IBM, and Microsoft. In what follows, I compare the top 25 global brands from the perspective of brand value to the Web Globalization Report Card provided by bytelevel.com. 215 Interbrand’sătopăglobală brands by brand value 2013 Rank Brand Name 1 Apple 2 Google 3 Coca-Cola 4 IBM 5 Microsoft 6 GE 7 McDonald's 8 Samsung 9 Intel 10 Toyota 11 Mercedes-Benz 12 BMW 13 Cisco 14 Disney 15 HP 16 Gillette 17 Louis Vuitton 18 Oracle 19 Amazon 20 Honda 21 H&M 22 Pepsi 23 American Express 24 Nike 25 SAP All these companies, either from the first top or the second are appropriate examples to follow when considering localization. If we compare the two listings and further limit the number of websites to be included for benchmarking the obvious brands to be followed when seeking both financial success and online global presence through globalization and implicitly localization are: Google, Microsoft, Samsung, Cisco, and American Express. 216 Please note that apart from American Express which offers financial services, the other four activate in the technology sector. Google is the obvious best-in-class as it offers all its services in 126 languages and it enjoys the largest user base through its multiple applications out of which the most important are: its Search Engine (see image below), Gmail, YouTube, and Google Plus When considering how content, services or products are to be provided to potential users, Google does not use the term to localize content, but touches the very essence of what localization means: particularization through unique, exclusive content. Anticipate differences in users' understanding of your topic and offer unique, exclusive content. [29: 14] A piece of information about a product or a service, even if not entirely new, will still have to reflect the information in the source text while being also part of a new, unique context, and being adapted to a locale. For instance, while Yahoo! is mostly used in Romania for its email service, and while there is not a completely localized version for Romanians, they offer news and weather information of interest for Romanians in the Romanian language. Google also provides google.com/webmasters/tools/ through which webmasters can fine tune their websites to comply with current best practices. Keyword Tool, Google Trends, and 217 Google Analytics are also useful tools for adjusting the layout, structure and content of a website. Google and its services is the best-in class for localization benchmarking purposes as it outperforms by far all its competitors. It is a valuable model for localization planning as it is both a hands-on example of successful localization in more than 100 languages. Through the free guides it provides it sets common sense standards both for general content marketing and localized content marketing. Localization benchmarking may include several companies for comparison, but Google with all its wide variety of services is a rather compelling model to be followed by any company that wishes to achieve global exposure. 4.3 IsăităGILTăallătheăway,ăwhenăităcomesătoăwebsiteălocalization?ă –ăAăcaseăstudy As seen in the first chapter, website localization researchers refer to three additional interrelated practices: globalization, internationalization and translation. Localization academics and the translation industry often integrate localization and translation with the other two processes. Is it always the case? Are the international companies always localizing and translating their content? Are localization and translation always mandatory in order to ensure the success of an online company? In this subchapter I will investigate the Privacy page of several international companies to weigh the extent to which its content is localized. There are two main reasons for choosing this particular page. First, there are legislative differences and other specific locales among various countries around the world. Second, there is a rising concern among online content consumers with regard to their privacy. There is also an EU directive about privacy, more specifically about using cookies, which can be noticed when you visit for the first time a website from a country that is an EU member61. ăhttp://searchdatamanagement.techtarget.com/opinion/Access-vs-privacy-Information-ethics-issues-confrontdata-pros 61 218 Figure 46: Cookie notification Consequently, companies should attach a greater importance to the content provided to their users through the privacy webpages as it is imperative for users to be notified about the data collected, how it is used and shared. Choosing the companies. The companies chosen for this research have been featured in various online business publications as being both among the top one hundred global brands interbrand.com [89] - and/or among the Top 25 Best Global Websites of 2013 according to [90]. While interbrand.com makes a classification of the companies based on their financial performance and global reach either by traditional means of communication or online through websites, bytelevel.com has a rather different approach in classifying companies, focusing itsăscoringămethodologyăonătheăcompanies’ăonlineăpresence.ăTheăfirstăcriterionăisătoă verify the evolution of the company websites in 2013 as compared to 2012. The following one is by the raw number of the languages in which content and services are available. Third, the bytelevel report takes into consideration the availability and accessibility of a global navigation. Forth, they analyze how websites are structured so that they can cover as many device screens as possible, in terms of content display adapted to various screen sizes. Last but not least, the degree of localization and the integration of social media. While the sites listed by interbrand.com and bytelevel.com are to be considered at a global scale, there are numerous national websites that are important to national and local communities. Therefore, I decided to consider statistical data as well, i.e. strictly listing websites based on web-user reach. Alexa.com [39] is a service that offers information on how company websites perform globally or at national level [91], strictly by recording and measuring user activity on the pages of websites. The table below shows how the three listings compare. I have chosen as the initial list the one provided by bytelevel.com as the list 219 is built up from the perspective of website localization practices. In theory, the Interbrand and Alexa lists should validate the initial list. Bytelevel.com Interbrand.com Alexa.com Combined listing[A] listing[B] Listing[C] Listing*[D] Google 1 2 1 4 Hotels.com 2 NA 656 1658 Facebook 3 52 2 57 Cisco Systems 4 13 1549 1556 3M 5 76 9638 9719 Philips 6 40 3652 3698 Booking.com 7 NA 138 1145 Samsung 8 8 264 280 Twitter 9 NA 11 1020 NIVEA 10 NA 65349 66359 Microsoft 11 5 39 55 Kayak 12 NA 779 1791 HP 13 15 286 314 Starbucks 14 91 2040 2145 Wikipedia 15 NA 6 1021 Yahoo! 16 NA 4 1020 LG 17 NA 1561 2578 Autodesk 18 NA 3107 4125 Intel 19 9 1003 1031 American Express 20 23 334 377 Merck 21 NA 63471 64492 Adobe 22 79 64 165 KLM 23 NA 4123 5146 Deloitte 24 NA 5859 6883 KPMG 25 NA 12153 13178 *The combined listing D is the sum of A+B+C. For each of the columns, the smaller the numeric value, the more valuable the brand from each particular perspective. Thus, the smaller the value in D, the higher the combined tripartite brand value. For companies that are not in listing B, in order to simplify the final output, yet accounting for financial value of 220 various global brands, for the sum in D, I defaulted NA to 1000 (only the top 100 are officially listed at interbrand.com). As can be seen, there is not always an agreement between what the leading global websites in terms of potential global reach and localization are (column A) or what their financial data holds (column B) on the one hand, and the actual number of users who use the company websites. Some of the highest performing global websites do not even qualify for the top one hundred of global brands. While the differences between A and B or B and C are understandable, as the older companies still employ various traditional marketing strategies, the variances between A and C show quite a significant number of discrepancies, considering that they are the results of statistical data, based on online presence. However, while in A the methodology is rather prescriptive, with the companies in focus, in C the direction is strictly from the point of view of website visibility, usability and number of users employing those websites. The differences between A and C for the various companies both confirm and invalidate the reliability of the findings of bytelevel.com. Thus, while hotels.com and booking.com listed in C corroborates the information in A, if we analyze NIVEA, Merck and KPMG, they show very low web-user traffic in column C. Google, Facebook, Microsoft, and Adobe show similar performances in all the three listings. They are all software companies which have all benefited from substantial growth and, hence, capital. Furthermore, in the case of software companies the products can be improved and tested without investing in raw materials, retail stores and production technology, etc. Also, moving from software to web applications, web services and web presence is only a natural step. The next group of companies which are relatively well-placed in all three lists are predominantly hardware companies: Samsung, HP, and Intel. They score high in the Combined list. American Express, from the banking industry, also scores high. The last group of companies that is listed in all three columns activates in various fields but the discrepancies are rather attributable to the manner in which marketing is conducted, online and traditionally, probably with more focus on the traditional channels (TV, radio, print media). 221 While hotels.com, booking.com and Twitter, Kayak, Wikipedia, Yahoo! are not listed in the top 100 leading brands (column B), they score high in the Alexa listing. All these companies offer web services, their business is software and they are Internet oriented. Therefore, I have chosen to limit the list of companies to be listed to the following 14 companies: Google, Facebook, Microsoft, Adobe, Samsung, HP, Intel, American Express, hotels.com, booking.com, Twitter, Kayak, Wikipedia, Yahoo! Analyzing the privacy page and related privacy information. In analyzing the privacy page(s) for all the above mentioned websites I verified if there is a translated privacy page and if the same page is also localized and to what extent. In the table below you can see the complete list of features inspected. The sector column states the primary industry of the company. The Number of languages column refers to the options offered by the website to choose from several language and country combinations. The Country/Language gate column verifies whether there is a separate page for opting for an alternative language or country. The Privacy page column confirms the existence of a Privacy terms page. The Cookie warning alert shows if this warning is displayed the first time you visit a website from an EU member country. As most of our personal details are transmitted through cookies, I examined if there is a separate page that explains what information the cookies collect. Next, I verified if there was a difference between the original text, usually American English, and the other English dialects. I also examined if there was any type of localization on the privacy page when English is used for several countries. The same was applied for German, French and Spanish. Last but not least, I analyzed how the localized site was integrated into the main website. That is, if it was set as a subdomain (country.thewebsite.com), as a national top level domain (www.thewebsite.country) or as a subfolder (www.thewebsite .com/country/) 222 Brand Sector Languages/ Country/ regions Country/ language language Auto- Privacy Cookie page gate Separate Trl. English German French URL warning cookie group group alert page TD L TD L group localization TD type L redirect 1 Google Technology 126 Y N Y Y Y Y Y N N N N N TLD 2 Facebook Technology 56 N Y Y Y Y Y N N N N N N Same URL 3 Microsoft Technology 96 Y Y Y Y N Y N N Y N Y Y TLD with redirection to folder URL 4 Adobe Technology 58 Y Y Y Y Y Y N N N N N N Folder URL 5 Samsung Technology 164 Y Y Y Y N Y Y Y Y Y Y 6 HP Technology 97 N Y Y Y Y Y N N N N N N Folder URL 7 Intel Technology 55 N Y Y N Y Y N N N N N N Folder URL 8 American Banking 93* N Y Y Y N Y Y Y Y Y Y Y Y Express 9 hotels.com Folder URL TLD/folder URL Tourism 91 N Y Y N Y Y N N N N N N subdomain 10 booking.com Tourism 42 Y Y Y Y N Y N N N N N N Same URL 11 Twitter Technology 36 N Y Y Y Y N N N N N N N Same URL 12 Kayak Tourism 29 N Y Y N N Y Y Y Y Y Y Y TLD/ subdomain 13 Wikipedia Information 51 N Y Y N 223 N Y N N N N N N subdomain 14 Yahoo! Technology 35 Y Y Y N N Y Y Y Y Y Y Y subdomain Trl= translation among various languages of the privacy page, TD=differences at textual level, L=localization, TLD=top level domain (i.e. .ro), folder URL= http://www.microsoft.com/ro-ro/ *One’s own language can be set upon accessing a specific country Data collection period: from 1st of March through 20th of March 2014 224 Google allows choosing from among 126 languages. While it uses a redirect script on accessing Google.com the very first time (accessing Google.com would redirect to the top level domain of the country from where it is accessed) it will also remember if you have opted for .com. It is important to notice that Google also offers under the top level domain of a certain country its services in the languages of the minorities. For example Google.ro can also be used in Hungarian or German. While Google allows you to select from among 126 languages, you would expect to notice some differences, especially if you can opt between English (global) and English (United Kingdom). The language differences between the .com EnglishăandăBritishăEnglishăareăminimal,ăadditionăofă“that”ăinătheă.co.uk,ă“suchăas”ăinsteadăofă “like”,ă spellingă withă “s”ă versusă spellingă withă “z”ă foră someă ofă theă words,ă slightlyă differentă usage of comma etc. Although there are some dialect specific differences, there are no other locale specific differences. Furthermore, the .com version is used for all the other Anglophone countries around the world. This is reasonable up to a certain degree, as all the Internet technology-related vocabulary originates from the U.S. but there are still some differences with regard to the law of each country. Next, I verified if there were any differences among the countries where German is the official language. The privacy policy page is identical for Germany, Austria and Switzerland. As they are all members of the EU, identical privacy policy is reasonable. It was even more challenging to observe if there were differences between the French of .fr and that of .ca. I have found that in terms of the language employed there were only some terminological differences: .fr: Règles de confidentialité, collectons ,informations ,cookies, un meilleur confort d’utilisation, etc. .ca: Politique de confidentialité, recueillons ,données ,témoins ,une utilisation plus conviviale, etc. Also, the .ca French used some first letter capitalization when naming services, similarly to the .com English. It is useful to notice that while .ca uses for the English term cookies a French term témoins, .fr uses the English borrowing cookies. All the differences are rather synonymous expressions and there is no information specific to any locale. 225 If we compare Spanish (European) and Spanish (Latin America) there are significantly more differences between the analyzed privacy policy pages. The differences at wording and expression level rise to 50%. This is much higher than in the case of .com English and .co.uk English. However, there are no specific privacy pages for each of the Latin American countries. Regarding Facebook, there is no language auto- redirect, and while the same URL is used, one can easily opt for his/her favorite language. In my case, Romanian is the first suggested alternative to English (US) the default. Facebook is offered in 56 languages and while the privacy page is translated into each of the obviously different languages, there are no differences between UK and US English; there is a unique page for German, and the same privacy pages for the following pairs (although they are marked as separate languages): Canada and France French, Spain Spanish and International Spanish, Brazil and Portugal Portuguese. Upon accessing Microsoft’să countries/regionsă pageă one may assume that they are offering their services in 96 languages. However, if we consider the considerable number of English variants for several regions/countries, apparently they only implement it to satisfy national pride. Whichever English variant is chosen, it directs to either the UK or the US privacy page. Also, there is no difference between the UK and the US page, with the exception of some omissions (Microsoft.com versus Microsoft) or minor typos. The same applies to the Spanish, French and Portuguese variants. If we consider the 3 German privacy pages, the differences are again minimal. All Spanish privacy pages are 100% identical at textual level. All French privacy pages are the same, for all of the francophone countries, with the exception of Canada. There are some differences at vocabulary and at sentence structure level among the different variants, to some extent similarly to the differences found on the Google French privacy pages. With regard to Adobe you may choose among 58 languages/regions. While there is a privacy page for all of the international languages, if we compare the Canadian and the US pages it is unexpected toă noticeă thată onă theă Canadiană pageă thereă isă ană extraă sentence:ă “Foră moreă informationă aboută whichă country’să lawsă applyă toă theă collectionă andă useă ofă youră personală information, please see the Information for non-U.S.ă usersă page.” Otherwise the pages are identical. The rest of the English pages are similar to the Canadian pages, a characteristic 226 which makes them different from the previously analyzed websites. If we further study Belgian and North Africa French, again there is no difference at all. Spanish variants are treated in a similar maner. As for the German variants, the privacy pages show some differences, probably due to the last updated date of the privacy policy (Austriaăpageăonă“7.ă Maiă 2012”ă whereasă theă Germanyă pageă onă “20.ă Dezemberă 2013”)ă So, apart from the difference that may be attributed to updating information, there are vocabulary differences such as Gültig ab versus Letzte Aktualisierung or Adobe-Kennung versus Adobe-ID. Investigating the Swiss German privacy page showed the same update page as that for Austria. This demonstrates that the German privacy pages actually follow the same policy as in the case of English, French and Spanish. When analyzing Samsung, one can see that it shows 164 languages and regions. If we examine the privacy pages, all of the privacy pages are particularized to the target country and language. However, there are discrepancies in the updating of the various privacy pages. This demonstrates that there are local branches responsible for the local websites. Furthermore, if we compare Samsung to all of the previous companies, their websites show superior levels with regard to both translation and localization. Not only are their marketing campaigns targeted to each of the locales, but they pay special attention to privacy issues as well. HP, much like Samsung, appears to have local teams that update content at regular intervals, yet not unitarily. For instance, the Australian privacy page is more elaborate with some 10% more information and updated in 2014, as compared to the US and the UK versions (updated in 2013). Regarding Intel, all the English variants of the privacy page requests are redirected to the default .com URL. Same is true for French. In the case of Spanish there is around a 10% difference between the language employed on the .es website and the general Latin American privacy page, used for all the South America Spanish speaking countries. In the case of German, there is no separate language for Austria, and the .ch domain was inaccessible at the moment of conducting this research. The achievement of American Express is superior when considering the translation and localization of privacy pages. This is probably a necessity considering the legal issues in the 227 domain it activates, that of financial services. While there are no significant differences between various countries using the same language there are differences linked to localization. Out of the remaining 6 websites Twitter uses the English privacy page for several countries. The other 5 have translated and localized privacy pages but only Kayak and Yahoo! use somewhat different privacy pages for the countries that share the same language. In conclusion, from the findings of this case study we can see that most companies fail to actually localize the privacy page content. Translation of the pages is not always necessary. The Romanian translation of the privacy page for the Romanian webpages is not perceived as a natural output and proper legal terminology is not used either. Therefore, I would assume that the same happens with all of the other languages to a higher or lesser extent. If the translator is not familiar with the legal terminology of the target language, we cannot discuss about localized versions of content. However, it is important to notice that, most of the companies comply with the EU regulation regarding the cookies used to track user activity on websites. On the other hand, some companies may not prompt you about the usage of cookies. For instance, as a non-EU citizen one may opt to access the default .com content from an EU country. The cookie law does not apply as s/he is not an EU citizen. This is probably standard as a tourist from the US or Australia might be visiting Europe and still use their own country’săwebsite. While there are raising concerns among Internet users with regard to their privacy, and steps have been taken into limiting what information is collected and shared (for instance, Google no longer provides the keywords to analytics software), some of the companies do not differentiate at all between the default English privacy page and the English variants of the other regions. Findings for Spanish, German and French are similar. While some companies respond actively to regional and national sensibilities, other strictly focusăonlyăonălanguageădifferencesă(Samsung’să164ăversus Intel’să55).ăSomeăcompaniesăuse flags while others don’t. In the case of the US market, one should notice that only few companies offer information in Spanish, although there is an important Spanish community within the US borders. 228 Considering the collected data, one can affirm that when talking about GILT, localization and translation is not always applied. English, Spanish, French, German variants of the privacy page are most often the same, so, there is neither a localization nor a translation process. Translation for minor languages is often implemented only at a later stage. For instance, while you may choose Romanian, you are redirected to the US privacy page. It is a standard procedure that webpages are localized and/or translated if they are of interest to the users. Privacy pages and other legal related webpages are most often skipped as there is no active interest in reading them. 4.4 WhichăWayăWebsiteăLocalization:ăTranslationăorăCopywriting? The purpose of this subchapter is to evaluate website localization practices of several important websites and to examine the similarities and differences between two approaches to text localization through translation and copywriting. This case study is a continuation of my initial research on the same topic. (Lakó 2013a) Website localization is a must for any company that intends to offer its products and services globally to an international market. Companies have been practicing localization at some level ever since organized and professional trade went beyond borders at the beginning of the 19th century, and I can only speculate that it began with the simple translation of the product type name and the quantity conversion (from royal to metric and vice versa). Most likely the translation was undertaken by the importer and not by the exporter of the product. Nowadays, well-established multinationals are in control of how their merchandise or services are marketed in various areas of the world, imposing on their business branches and importers strict rules which generally follow the processes of globalization and internationalization of any product or service. Also, franchise is another similar example of such practices. The mother company rules apply to marketing strategies as well. In traditional marketing, efficiency (measured in ROI – return on investment) of advertisements depends to a high degree on their placement. On a general audience broadcaster, a prime time ad will address a very wide audience, and results are less predictable, whereas the same ad on a sports channel, with shows addressed specifically to 229 men is more likely to yield significantly improved results. For example, a premium car brand with a price tag of over 30.000 Euros advertised on such a sports channel, during the transmission of a golf event, is more appropriately targeted, while in terms of costs the campaign may be similar. For such an advertisement, localization consists of translation and adaptation of the verbal part. In online marketing, although there are numerous similarities with traditional marketing, there are particularities that have to be taken into account. Online marketing covers several aspects, which are all important for the success of a business: websites, social media, digital branding, blogging, video, search engine optimization (SEO), emailă marketing,ă affiliateă marketing,ă onlineă PR,ă andă digitală advertising.ă Otheră marketingă professionals distinguish 14 types of marketing online: Search Engine Optimization, Pay per click, Email, Video Marketing, Blogging, Content Marketing, Social Media, Network Marketing, Community Building, Location Based Marketing, Contextual marketing, Affiliate (performance based) Marketing, User Experience Branding (UxB), Interactive Advertising [92]. Among these Internet-marketing subfields there is naturally some overlapping. What I am most interested in for the purpose of this thesis is Content marketing and Search Engine Optimization, because in the case of website localization content is the main part that is localized. As shown in chapter 2, localization, is part of a more complex endeavor that includes globalization and internationalization as processes that are achieved prior to localization. Most localizers consider translation as part of the localization. However, there are examples which articulately demonstrate that translation is not the only approach to content localization. Let us consider Eurosport (the former pan-European sports channel which is now global) and its website(s) [93]. The main site asks for user’s permission to set the country and language of his/her choice. 230 Figure 47: Eurosport in localized instances The six instances captured represent localization brought to personalization level for several local Eurosport websites. The content of each website demonstrates that localization is not about translation, but about content created specifically for the UK, Germany, Turkey, Russia, the Arab countries - primarily Egypt, and China. One can easily see that the main picture in the top news is different and it is about sport events specific to each country or region. For the Arabic version even the layout of the website is in accordance with the writing rules of Arabic, that is, right to left page viewing. On a closer inspection, one can also observe that above the Eurosport logo, on the grey bar, for some of the countries there are some extra services while for others there is none. This is 231 the result of the immersion of Eurosport and Yahoo! with third party services or with the physical and technical possibility of offering their own services. Also, while on some of these localized websites Eurosport allows web advertising from third parties, on others there is no advertisement. If we analyze the main menu and the submenu of the home page we can see that even those are localized. By tracking user activity, Eurosport can determine which sports are of interest in all of the countries they cover. Thus, the menus are different. The homepage button is not on all of the sites represented by a button with the textă“home”.ăOnăsomeăofătheăsitesătheătextăisăreplacedăby an icon, a stylized house that stands for the main page. While the submenus of the start page are very rich in diversity for the UK and Germany, for Turkey and China there are only two subcategories, for the Arab site the text is replaced by icons, and for Russia there is no submenu. Furthermore, we can see that football is the most popular sport with all nations but there is variation when considering other sports. F1 (Formula 1) is immediately after football in Germany, third in the UK and not to be found on the other versions (unless it is related to the “More”ăcategory).ăTennisăisăagainăpopular to various extents in various parts of the world and so on. This demonstrates that Eurosport is very much user oriented and not interested in translating but in offering original content that appeals to regional users, is adjusted to its consumers, while an international section for each of the pages and an international website are maintained. The Romanian version [94] preserves the general characteristics of the other versions although it looks barren both in content and design as compared to the other Eurosport websites. Yahoo itself is another successfully localized service provider that initially functioned as a search engine and a free email provider. Originally, it was provided in English only, first for the US and extended to be a services provider covering most ofătheăworld’săcountries.ăWhileă there is a common feeling to the entire localized versions of Yahoo services, again, similarly to Eurosport, content is different and of interest to the region it is targeted at. While yahoo.com stands apart both in terms of design and quantity of information offered, the other versions show the preference of the communities for various types of news. Advertisements displayed also vary according to the target country. 232 Figure 48: Languages and countries in which localized Yahoo websites are available[95] The examples above eloquently reveal that translation is often replaced by copywriting. Several other multinationals that offer electronics, for instance, will more likely use translation rather than content (re)creation, whereas small companies or importers of less known brands will opt for the latter. For instance, for fighting mold, the simple translation into Romanian of some of the original American sites might work to some extent, but an analysis of the Romanian market is a must, in order to learn what are the key terms used by Romanians, to find information on such products, and to learn about user behavior. In order to decide which method, translation or copywriting, is more efficient I did two experiments. Both involved producing articles to be published online in Romanian. For the first experiment, which began as a potential job for a Romanian importer, my purpose was to produce content about blooming tea. The first step was to observe if the Romanian market was aware of this type of product. Using tools such as Google KeywordPlanner, Google Suggest, Ubersuggest, I found out that the most widely used terms would be ceai infloritor and ceai blooming. However, KeywordPlanner showed that there was no search volume for either of the terms. Therefore, I had to look for more general concepts related to tea, concepts that would also resonate with the source texts. I found some common ground in the ingredients section: ceai alb (white tea), iasomie (jasmine tea), gălbenele (Calendula tea). Next step was to find long-tail keywords related to benefits of such teas to health. Some of the keywords were: contraindicatii ceai alb, proprietati ceai alb, efecte ceai alb, indicatii ceai iasomie, beneficii ceai iasomie, proprietati ceai iasomie, ceai de 233 iasomie mod de preparare, beneficii ceai galbenele, indicatii ceai galbenele, contraindicatii ceai galbenele. I also included more general tea related keywords such as ceaiuri curative and ceaiuri medicinale (herbal teas). Next, I asked 10 senior students and former students in translation studies to produce 10 articles on various blooming teas by translating 10 articles from a US website. The instructions provided were that the articles were supposed to be perceived as if produced in Romanian and include the keywords (ideally one or two keywords per article). They were also encouraged to translate the tea names as well so that they could be possibly used for marketing purposes (one of the students suggested ceai eflorescent – efflorescent tea. (This could function as a new marketing term or even for branding, as it is not commonly used, yet it can be understood by the potential niche market. However, as expected, it is not a suitable candidate as a main keyword.) They also had to time themselves. All articles were roughly 500 words long. Another set of 10 articles were to be produced through copywriting in Romanian, by students in PR and communication. They were instructed to use the same set of keywords, to write original content (verified using Copyscape) about blooming tea, also 500 words long. They were also instructed to monitor the time necessary to produce the texts. In terms of time necessary for producing the final output content, producing the final text took the translators almost 20% more time, mostly because of the requirement to include the specific keywords in the target text. During the non-SEO aware translation process the translation process would have been linear. The volume of translation would have been correlated using the same number of TUs (translation units). However SEO-aware translation requires not only transcreation or an instrumental approach, but also an inclusion of extra sentences or sections. Thisă corroboratesă Pym’să viewă onă producing text within localization considering that “addition and omission are legitimate strategies, to an extent not envisaged in classical theories of equivalence. Further, cultural adaptation may require degrees of transformation that go well beyond the classical limits of translation but can be justified within the purpose paradigm.” (2014:130-131) While it could be argued that this is similar to explicitation, the actual sections included do not provide further explanations to the reader but use related concepts used by web-user to find information on search engines. They need to be perceived as a natural part of the output. This requires from the part of the translators creative writing skills, thus extra effort to complete. On the other hand, the copywriters’ work was more straightforward and more rapid as they were not constricted in anyway by a source 234 text. The most time consuming stage in their case was the documentation stage. However, considering that the topic of tea and blooming tea is not very specific helped both translators and copywriters be more efficient. The use of TM and MT in the case of translators did not improve very much the process of producing the output text. In terms of language output naturalness (that is content written for users and not search engines), a separate team of 10 students in Romanian linguistics were asked to rate on a scale from 1 to 5 each of the 20 produced texts. 6 out of 10 translations were rated with values from 4.00 to 4.50, 3 with 3.50 to 3.90, 1 with 3.20. In the case of the copywriter produced texts 1 article was rated 4.10 while the other 9 were rated between 4.40 and 4.95. This demonstrates that copywriting is perceived as more natural than translation. One of the main reasons might be the influence of the source text on the translators: style, vocabulary, syntax, culture. The copywriters’ă styleă andă language,ă onă theă otheră hand,ă areă influencedă duringă theă documentation stage by other copywriters from the same cultural area. All the articles were published on an experimental website in August 2013. In terms of efficiency for ranking, after 6 months (it takes more on more competitive markets such as the US or the UK) all 20 articles ranked on positions ranging from 10 to 60. Only on-site SEO optimization was employed, so that ranking in SERPs was not biased by external factors such as number of back links and their quality. Analyzing the ranking performance of the two types of texts I found that there was no direct correlation between the content producing strategy and the ranking for various keywords. Some of the articles ranked even for keywords that were not even initially targeted. As a second experiment I asked several students to produce 12 Romanian texts on mold removing methods: 6 through translation and 6 through copywriting. One article was published as category page content. The category had 11 articles. The text producers were trained to use certain keywords. Articles on the more traditional mold removal methods were produced by copywriters. Prior to producing the articles, the students were instructed to gather information on the same topic from various Romanian websites. The more modern methods were taken over from a UK website and translated. One important requirement was, again, to produce original content in both cases (verified for original content using Copyscape). The articles were then posted on an experimental WordPress driven website on which I also installed an SEO extension. All articles were further modified according to the 235 directions offered by the SEO extension. It included changing title, URL, adding in several instances of the same keyword so that the suggested keyword density might be acquired, writing the meta description with the main keyword included (as for some more general articles I decided not to let Google select the text to be displayed in the SERP). All the articles were made public in February 2014 and I used only on-site optimization. After 6 months the copywriter produced articles ranked for various mold related keywords on positions ranging from the 15th to the 50th. More specific keywords, for the newer modern technology anti mold concepts ranked in top ten (tencuiala antimucegai and its similar tencuiala antiigrasie and tencuiala antiumezeala; batoane antimucegai, unde electromagnetice antimucegai). Regarding the speed of producing the texts, as compared to the tea articles experiment, the results were more balanced. Some translators were more rapid than their copywriter peers, others were slower. So, speed was determined by experience and familiarity with the topic. In terms of language naturalness results were similar to the first experience. Copywriters sounded more natural. As for ranking, all the articles ranked well considering that I did not undertake any off-site optimization of which the most important practice is to do link exchange or buy links to the most important webpages. However, the purpose of the experiments was to determine which is more effective: translation or copywriting. As long as keyword related instructions are followed by either the translator or copywriter, ranking in Google is not determined by translating or copywriting. In terms of naturalness copywriting outperforms translation. As for speed, I would affirm that it depends on background knowledge. From the examples above, one can see that translation is not always the strategy to employ during the localization process. Numerous companies will use copywriting services rather than translation. Copywriting and Christiane Nord’să instrumental translation (1997:81) are rather similar as shown in the theoretical chapters. While copywriting is based on research on several texts in the target language, instrumental translation will be based on a source text in a different language. From the perspective of SEO both strategies can use specific keywords, yet a copywriter will be more familiar with the target language and cultural codes. The translator even if a native speaker of the target language may still misunderstand the source text. Some may be against copywriting considering that by writing the content from scratch, there is no contact with the original text. 236 In some cases (as in the examples above), there is no need to have the same feeling as in the source text, as can be demonstrated by means of non-verbal elements. Also, by imposing the same keywords, even if localized for each market, the texts will still be based on a common reference with the specificity of each market being observed. To sum up copywriting is somewhat superior to translation but only in terms of naturalness. 4.5 Booking.com:ăaăcomparativeăstudyăfromăaăwebsiteălocalizationă perspective In this subchapter I will analyze the level of localization of two hotel pages. The study focuses on applying website localization theories from the theoretical chapters. The languages/cultures considered are American English, British English and Romanian. The pages analyzed refer to a hotel from London, UK: http://www.booking.com/hotel/gb/oxford-house.en-us.html http://www.booking.com/hotel/gb/oxford-house.en-gb.html http://www.booking.com/hotel/gb/oxford-house.ro.html and a hotel from Ia i, Romania: http://www.booking.com/hotel/ro/unirea.en-us.html http://www.booking.com/hotel/ro/unirea.en-gb.html http://www.booking.com/hotel/ro/unirea.ro.html Data was collected in April 2014. I considered both a UK and a Romanian hotel as, presumably, the main description content is provided on the website by the hotel owners themselves. The analysis consists of two main parts: interface and content. It is based on the website elements listed by Esselink (2000) yet limited to user visible content. Notation conventions used in the tables below: - white background for relevant findings - grey background for interpretation of the findings - black background for faulty elements 237 Interface analysis (applies to both hotels) Element analyzed American English British English General design and layout The same Color usage The same Top menu Romanian Numerous similarities - Currency Interchangeable, regardless of the language selected by the user; flag is used to differentiate between currencies with the same name While some may consider that currency should be set automatically, upon choosing a language, often, there are clients who have savings in several currencies. Thus, a client may opt for a currency over others, based on currency rates or the amount of money in a certain currency. - Dateăformatăandă month-day-year: day-month-year: day-month-year: 22- punctuation Oct-22-2010 22-Oct-2010 oct-2010 (theăsearchăboxăonătheă Month capitalized left) - Currencyăformată andăotherădigită No moth capitalization Usage of comma to Usage of period to delimit decimals formatting Usage of comma as thousands separator delimit decimals Usage of period as thousands separator Currency placed No space between currency symbol and amount before amount. RON used instead of Lei. Amounts rounded Capitalization Amounts with decimals Mixed approach: Usage of capital some sentences letters for all words in headings or buttons capitalize only the initial letter while other sentences 238 Amounts rounded Usage of capital letters for the first word in headings or buttons capitalize the initials of all the words Iconăusage Call-to-action Inconsistent: Highly Destinaţie/numele Destination/hotel inconsistent: hotelului, Top 5 - name, Your Top 5 Destination/hotel hoteluri văzute de Viewed Properties name, Your top 5 dumneavoastră viewed hotels etc. Inconsistency in versus Room article usage: Types, indefinite and Hotel Policies, definite article used Extra Facilities, (-le, -ului) in the Two-Bedroom same heading Penthouse (DestinaţieØ/numele Apartment, etc. hotelului) Consistent among the three localized version Lock in a great price for your upcoming Securiza i-v ăunăpreţă stay excelent pentru sejurulăurm tor! The verb a securiza in Romanian is used inappropriately. It is used to convey security, to make secure or to secure. Alternative translation suggested: Asigurați-vă că benficiați de cel mai bun preț! I removed pentru sejurul următor as it may imply for the next stay and not the current one. The page context already implies that the message refers to the current hotel Rezervaţiăcâtătimpă camerele mai sunt Book while rooms are still available, or disponibile, sau secure your reservation for these dates. asiguraţi-v ăc ăaveţiă o rezervare pentru aceste date. Usage of the verb to book instead of to 239 Same verb used a reserve rezerva both for the interface button and CTA text Preţurileăpotăcreşte,ă maiăbineăface iă Prices may go up, so secure your rezervarea chiar reservation today. ast zi. Addition of CTA mai bine Otherăinterfaceăelements Property Highlights Property highlights Punctele forte ale propriet ţii Capitalization Salva iăpentruămaiă Saveăforălater târziu Alternative:ăPune iă semn de carte Înapoi la început It could be Backătoătop understood as start over. An alternative translation could be Top/Început pagină Guestsăconsistentlyărateă… ReasonsătoăChoose reasonsătoăchoose Capitalization Clienţiiăapreciaz ăînă mod constant … motive pentru a alege BusinessăTwinăRoom Camer ătwină Business Camer ăbusiness dubl Suite Suit Term not used in Romanian to designate a home or 240 an area to host tourists Check-in Check-out While not standard terms in Romanian (there are no dictionary entries for these terms), the Romanian hotel industry makes extensive use of these terms instead of the Romanian expressions: (dată) sosire and (dată) plecare or Din and Până în, similarly to the French interface (Du and Au) Content analysis (applies to both hotels) Element analyzed American English Hotel address British English Romanian Romanian format English format and wording English wording Address format is based on the format of the country of origin Names relevant for Usage of first names for each of the hotels. Name input may feedback lack diacritics usage (user generated error): Abdulaziz instead of Abdülaziz. Names are also input using a different writing system, in this case Arab, ‫ ع لي‬, which may be confusing for most non-Arab speakers. It is interesting that within the English or Romanian text, the direction of the Arab text is correctly rendered, from right to left, to maintain the original text characteristic to the source user. This and other names are used on purpose as a foreignization strategy to underline the variety of clients that are hosted at various hotels. Telephone hotels number (provided of upon Format is based on the format of the country of origin reservation) Units of measure Imperial system Metric system The metric system is the official measurement 241 Metric system system, but the popular system is the imperial system Non-text elements Mainly pictures in this case, same pictures and same number of pictures. Romanian hotels tend to display somewhat more pictures (toă compensateă foră theiră defaultă customers’ă – Romanians – high Uncertainty avoidance index). Textual content Localized - Usageăofădiacritics Not applicable - Userăfeedback Text is displayed as entered by users Complying  Proneătoălinguistic,ăpunctuationăorăcapitalizationăerrors,ă informalăstyle,ăbutăusedăasăsuchătoăconveyătheămeaningăthată feedbackăisăfromărealăusers: - Sleeping matrailes wera cleaned Kitchen was well prepared. Good location; - I did not like the the way the staffs treated as I was been assured that the towels will be changed everyday but it got never canged and all those missguidance by the staff, - Great apartment in a good location. 1 block to the Tube and market. Nice kitchen. - Baia nu e nici p-aproape de patru stele.; - foarte curat liniste personal calificat cu adevărat restaurantul face toți banii - la cam.duble paturi f.inguste sub standar  NoădiacriticsăforăRomaniană  WhenăsetăonăRomanian,ătheăRomanianăhotelăprovidesăalmostă allăfeedbackăinăRomanianăfromăRomanianăguests.ăThisămightă implyăthatătheăareănoăforeignăguests.ăWhenăsetătoăEnglishă feedbackăisăprovidedăfromăguests,ăRomanianăorăinternational,ă whoăusedăEnglish  TheăUKăhotelăprovidesăonlyăfeedbackăinăEnglishăfromă previousăguests.ăIfălanguageăisăsetătoăRomanian,ănoă translationăisăavailable 242 Content linguistic analysis (UK hotel page) - Maină textuală content Similar but localized texts for each of the three targets The differences are of spelling, lexical and syntactic nature examples Situated just 10 from minutes’ walk from feature offer oferă theaters theatres teatre comes equipped comes with a DVD dotat cu DVD with a DVD player player player an oven The fully equipped kitchen comes with an oven underground subway station Wordăcount Situat la doar 10 A 10-minute walk The full kitchen has - NA station minute de mers pe jos de Bucătăria utilată complet are cuptor Staţia de metrou a 10-minute walk a 10-minute walk 10 minute de mers from the away from the pe jos de apartments apartments apartamente 130 134 142 Content linguistic analysis (Romanian hotel page) - Maină textuală content Similar but localized texts for each of the three targets The differences are of spelling, lexical and syntactic nature examples NA Featuring Offering oferă the Unirea Hotel & Unirea Hotel & Unirea Hotel & Spa Spa Spa 243 in the center of Iasi at Unirea Square situat în centrul centre of Iasi oraşului Iaşi at Union (Unirea) Square în Piaţa Unirii vedere unică la views of the city views over the city specialties specialities aerobic classes aerobic available on site programmes on site just under 1 mile 1.5 km 1.5 km just over half a mile 1 km 1 km local train station - located in the Wordăcount 152 local railway station 147 oraş specialităţi programele de aerobic oferite de hotel gara locală 149 Further findings and observations: - Theăusageăofăreserveăină Englishăforătheăinterfaceăbuttonăinsteadăofăusingătheăverbă to book,ăconsideringătheănameăofătheăcompanyă(booking.com).ăTheăexplanationăwouldăbeă thată mostă internatională non-nativeă usersă wouldă thinkă ofă bookă asă theă noună andă notă considerăităasătheăimperative/appellativeăformăofătheăverb.ăToăavoidăconfusionătheyăuseă reserve evenăifătheălanguageăchosenăisăAmericanăorăBritishăEnglish. - TheătranslationăofătheăReserveăbuttonăinăRomanianăisărenderedăonătheăsiteăasăRezervați acum (reserveănow).ăTheăusageăofătheăCTAăadverbăacumămayăbeăusedăinăRomanianăasă comparedătoăEnglishăasăaăcompensationăforăRomanians’ăhighăPowerădistanceăscore,ă - Paperăsizeăforăprintingăisănoălongerăanăissueăasăită isă aătechnicală problemă thatăcanăbeă solvedă byă coding.ă Booking.comă replacedă theă usageă ofă paperă withă aă smartă phoneă applicationăthatăcanăbeăusedăinsteadăofătheăprintedăvoucher - Regardingă pronoună proxemicsă foră Romanian,ă theă politeă addressingă formsă areă predominant,ă however,ă thereă areă 2ndă personă informală imperativeă formsă asă well:ă Schimbă căutareaă(Changeăsearch) - ForătheăUKăhotelădistanceăisăexpressedăthroughătimeăexpressionsăasătimeănecessaryătoă reachăvariousălandmarksă(A 10-minute walk from…),ăwhereasăforătheăRomanianăhotelă 244 distanceăisăexpressedăină kilometers.ăUsingătimeătoăexpressădistanceăisăanăappropriateă strategyătoăavoidăusingăeitherătheămetricăorăimperialămeasurementăsystem,ăespeciallyă forăcustomersăfromă theă UK.ăOnătheăotherăhand,ă forătheăRomanianăhotels,ă theămetrică systemăisăconvertedătoăimperialăthroughăapproximation.ăjust under 1 mileăinăAmericană EnglishăisădisplayedăonătheăRomanianăandăBritishăEnglishăpagesăasă1.5ăkm.ăAlso,ăjust over half a mile isăconvertedătoă1ăkm. - Apartment isăcorrectlyăusedăinătheăuserăinterfaceăindiscriminatelyăonăbothătheăUKăandă theă USă pagesă asă ină Britishă Englishă ită refersă toă “Aă flat,ă typicallyă oneă thată isă wellă appointedăorăusedăforăholidays”[105] - Amountsăareăroundedăforămostăcurrenciesăandăproperties.ăThisă mayăbeăunnaturală foră marketsăusingăpowerfulăcurrencies,ăsuchăasătheăȘUSăorătheăEuro.ăForăinstance,ăGermană websitesăwouldădisplayăpricesăusingăaăcommaăregardlessăofătheăprice,ăwithădecimalăoră not:ăEURă1.299,00.ăOnătheăotherăhand,ăRomaniansămayăfeelăcomfortableăwithăroundedă amounts,ăespeciallyăolderăgenerations,ăthatăwereăusedătoătheăoldăweakăLeuăcurrency. - Usageăofă“limitedăoffer”ăstrategy:ă o Last chance! We have 1 apartment left! VS.ă Ultima şansă! Mai avem 1 cameră! o Most recent booking for this property was 27 minutes ago from Belgium. VS. Cea mai recentă rezervare pentru această proprietate a fost făcută acum 27 minute din Belgia o There are 26 people looking at these apartments.i VS.ă26 persoane se uită la aceste apartamente. - ă Interfaceăcontradictoryămessage:ăTheăheadingăsaysăthatătheălanguageăofătheăvisitorăisă spoken,ăbutăactuallyăonlyăEnglishăisălisted. Generally speaking, localization aspects of the pages under observance are appropriately attended to. While numerous websites do not differentiate between the various English speaking countries, Booking.com does seem to handle cultural differences properly. The two variants show spelling, lexical and syntactic variances. Regarding punctuation and style, while the culture specific elements are predominant for each variants (U.S. or U.K.), there are 245 some minor inconsistencies, namely capitalization. As for the metric system used for the U.K., it is interesting that they opt for the official metric system versus the popular imperial system. As for the text in Romanian, it contains untranslated English text elements in the address, which may be involuntary, or interface text, check-in and check-out, used as such on purpose. In general, Romanian texts convey the same message as in English. However, as highlighted in the tables above, there are some awkward sounding Romanian expressions such as Suită for Suite, vedere unică la oraş for views of the city/views over the city and Securizați-vă un preţ excellent…for Lock in a great price… To sum up, in general, the website fulfils its purpose in all three localized versions. The errors or the inconsistencies do not obstruct the communication process and the general context of each of the pages is determinant in achieving an appropriate communication. For instance, the translation of …your upcoming stay as …pentru sejurul următor (which can be interpreted as the next stay and not as the current offer on the page) is clear due to the context within the page. Also, the English stay is translated as sejur. The English term may refer to any period of time, from a few hours to a few days, whereas the Romanian one is rather used for a period of time that covers several days. These communication errors have a small impact on the customers as there is a highly controlled step-by-step guidance interface that clearly reinforces the information about a certain hotel, before actually accomplishing the booking process. Furthermore, a confirmation email is sent. However, one major communication error is acknowledging the user through the header interface text that their language is spoken at the hotel while in the description the language of choice from the interface is different. Considering that most users scan content – they read only bolded characters, and do not read the text in full, many will be misled into believing that the personnel speaks French or even Romanian. This case study corroborates Jiménez-Crespo’să findingsă withă regardă toă qualityă ină websiteă localization (2013:131-132). Regular improvements to the website, in terms of interface and content updates, show that the localization process is a continuous maintenance process. (Pym 2014:126) As for keyword usage, the texts are optimized to be displayed in search engines for general keywords such as london hotel, hotel iași or as long-tail keywords such as the following format: country name - city name - hotel name 246 4.6 WebsiteălocalizationăforătheăRomanianămarketă–ăaăculturalăapproach Foră thisă caseă studyă Iă consideredă theă topă twentyă websitesă fromă theă perspectiveă ofă theă sitesă localizedă foră Romania.ăAsă aă startingă pointă foră theă comparativeă studyă Iă consideredă emag.ro-ă launchedă ină Decemberă 2001,ă theă largestă onlineă Romaniană retailer,ă [96][97][98],ă withă aă cumulatedă experienceă ofă 13ă yearsă ină theă e-commerceă business.ă Dueă toă itsă successă ită wasă boughtă byă Naspersă (aă South-Africană company)ă ină 2012,ă and,ă whileă originallyă ită begană byă sellingă ITȚCă goods,ă ită hasă extendedă itsă productă rangeă continuously;ă thereforeă ită shouldă beă consideredăbyăcompaniesăthatăwishătoălocalizeătheirăwebsiteăforăRomaniaăasăwell.ăTheăreasonă foră includingă aă Romaniană companyă isă thată aă companyă developedă withină itsă culturală andă natională economică environmentă isă closeră toă theă expectationsă ofă theă locală marketă andă ită isă viewedăasăaălandmarkăinăitsăfield. Anotheră successfulă Romaniană companyă bothă onă theă locală andă internatională marketsă isă bitdefender.ro.ăItăoffersăonlineăsecurityăservices. Ină myă analysisă Iă consideredă high-contextă cultureă characteristics,ă linguisticsă proxemics,ă polychronică features.ă Thisă translatesă intoă usageă ofă moreă imagesă bută lessă text,ă usageă ofă noi (we)ă vsă informal tu (you)ă vsă formală dumneavoastră (you).ă Otheră characteristicsă consideredă areăpowerădistanceăindex,ăindividualism,ămasculinity,ăuncertaintyăavoidance,ăpragmatismăandă indulgence. emag.ro:ă Ină termsă ofă textă andă imageryă usageă theă websiteă isă muchă likeă anyă highă profileă internatională e-commerceăwebsiteă (amazon.com,ă tesco.com,ă ebay.comă andăsoăon).ă Fromătheă perspectiveăofăproxemicsătheyăoptedăforătheăinformalăaddressingăthroughăusageăofăpronounsă andă verbală formsă whichă implyă closeness.ă Theyă areă usedă foră informativeă purposesă (Recomandări pentru tine din…, Istoricul tău de navigare)ă oră asă CTAsă (Câștigă 2000 lei Card Cadou!,ă Scapă de cabluri!,ă Bucură-te…,ă Atinge noi obiectiveă …,ă Performanțe incredibile, adaugă în coș …ăetc.)ăTheăpresenceăofătheăuser’sănameăorănickănameăatătheătopăofă theăpageămayăalsoăimplyăcloseness.ăHowever,ădisplayingătheăfullănameăofătheăuserămayătriggeră someăreserveăasăRomaniansăareăaddressedăbyăfullănameă(orălastănameăonly)ăusuallyăinăofficială circumstances.ă 247 Asă foră poweră distanceă index,ă emag.roă offersă theă web-useră aă wideă rangeă ofă choicesă withă severalăfiltersăsoăthatătheăuserăcanăoptăforătheăproductăorăserviceăbasedăonăhisăorăherăjudgment.ă Regardingă individualismă versusă collectivism,ă thereă isă aă tendencyă towardsă individualism,ă illustratingă variousă degreesă ofă ită asă Iă consideră thată ită muchă dependsă onă theă productă type.ă LinguisticallyătheseăareăexpressedăasăAtinge noi obiectiveă…,ăPerformanțe incredibile …inătheă caseăofămobileăphonesăandăotherăgoodsămeantăforăpersonalăuse.ăSuchăCTAsăalsoăimplyăaăhighă masculinityă degree.ă CTAă suchă asă Scapă de cabluri!ă impliesă pragmatismă whileă Bucurăte…denotesăindulgency. Asă oneă cană see,ă theă Romaniană onlineă e-commerceă websitesă reflectă throughă theă usageă ofă certainăwordingsăaădifferentăsetăofăculturalădimensionsăthanăthoseăseenăatăHofstede.ăFromăthisă perspectiveăităappearsăthatăRomaniaăisămovingăcloserăandăcloserătoătheăculturalădimensionsăofă theă Westernă societies.ă However,ă apparently,ă oneă dimensionă remainedă unchanged.ă Theă uncertaintyăavoidanceăindexă(UAI)ăisăreflectedăonăemag.roăinămultipleăwaysăinăanăattemptătoă diminishă Romaniană web-users’ă fearsă thată theyă mayă beă deceived.ă Thus,ă emag.roă allowsă 30ă daysă moneyă backă guaranteeă uponă returningă aă productă (20ă daysă moreă thană theă legală timeframe),ă allowsă severală typesă ofă payment,ă oută ofă whichă theă mostă populară seemsă toă beă paymentăuponăreceiptă–ăevenăifăităisănotătheămostăeffectiveătypeăofăprocessingăpayment.ăTheyă advertiseăthisăpaymentămethodăonătheămainăpage. Figure 49: Pay upon receipt emagăfurtherăoffersă(andăIăwouldăsayă “educates”)ăitsăclientsătoăopenătheăparcelăuponăreceipt.ă TheseăareăallădirectămethodsătoăcompensateăforătheăhighăUAIăandăareănecessaryăasăthereăhaveă beenăreportedănumerousăproblemsăonăvariousăwebsitesăthatăbuyersăwereădeceivedăinăvariousă ways:ătheyădidănotăreceiveătheăgoods,ătheăgoodsăwereănotăasădescribedăorăfaulty,ăorăreplacedă withăworthlessăitems.ăTheăimplementationăofăsocialănetworksăbothăreducesă“userăanxiety”ăandă impliesă “takeă youră friendsă andă acquaintancesă asă ană example”.ă Ită isă aă commonă marketingă strategyă usedă toă convertă reluctantă usersă intoă customers.ă Anotheră characteristică thată compensatesă foră Romaniansă highă UAI,ă whichă isă lessă directă andă concernsă eachă individuală product,ă isă allowingă useră commentă onă theă productsă theyă alreadyă used.ă Thisă isă commonă 248 practiceăonănumerousămultinationalăe-commerceăwebsites.ăCategoryăpagesăareăquiteăsimilarătoă theămainăpage,ămadeăupăpredominatelyăofăimagesăandăshortădescriptions.ăProductăpagesăuseă moreăverboseăcontentăandăpicturesăandăareăoftenămoreăobjective,ăreflectingătheăfeaturesăofătheă productădescribed.ăThisăisăalsoăcommonăpracticeăonăinternationalăwebsites. bitdefender.ro:ăFromătheăperspectiveăofăHall’săLC-HCădimensionătheăusageăofăimagesăonăthisă websiteăisăbalanced.ăHowever,ăoneăcanănoticeămoreăwhiteăspaceăonătheăpage,ăinformationăisă notă crammed,ă soă theă feelingă isă oneă ofă relaxation.ă Ită followsă theă generală patternă ofă modernă websites.ăAsăforăproxemics,ătheyăuseătheămoreădistant,ămoreăformal,ăyetămoreăpolite,ăformsăofă addressing,ăreflectedălargelyăinăverbalăforms:ăReinnoiți-vă,ăActualizați-vă,ăBine ați venit la…,ă Vizualizați toate produsele,ăAflați mai multe,ăTransmiteti o solicitare,ăAveți nevoie de…,ăVă rugăm folosiți …etc.ă Thisă isă ină contrastă withă emag.ro,ă bută bitdefender’să maină targetă ină Romaniaăareăbusinessăclientsăandănotăhomeăusers,ăasămostăhomeăusersăpreferăbasicăandăfreeă antivirusăsoftware.ăThisămightăexplainătheăusageăofăformalăaddressing.ăHowever,ăonătheăsameă page,ă thereă areă Romaniană verbă formsă thată cană beă interpretedă bothă asă theă secondă personă singularăimperativeămoodăandătheăthirdăpersonăsingularăindicativeămood:ăElimină toți virușii cu un singur click,ă Învinge amenințările,ă Îmbunătățește performanța,ă Asigură controlul.ă Furthermore,ă thereă isă aă buttonă usingă ană informală verbală form:ă Testează-l gratuit.ă Regardingă theă Englishă versionă (bitdefender.com)ă showsă thată onlyă theă veryă firstă sentenceă isă usedă descriptivelyă whileă theă othersă areă ină theă imperativeă mood:ă Eliminates All Viruses With A Single Click,ă Defeat Threats,ă Drive Performance,ă Deliver Control.ă Aă closeră lookă revealsă anotherăinconsistencyă(inătheăpictureăbelow,ăhighlightedăwithăorangeăframes).ăTheătextăonătheă banneră makesă useă ofă secondă personă singulară imperativeă mood 62,ă whileă theă textă ină theă greyă areaăusesătheădebatableămeaning,ădiscussedăabove.ă ăTheseăverbalăformsăareăsimilarăinătheăsingularăandăplural,ăbutăIăconsideredăthisătoăbeăasăaăsingularăformăbecauseă computersăorămobileădevicesăareămeantăforăpersonalăuse. 62 249 Figure 50: Formal vs. informal inconsitency Theă sameă discrepancyă cană beă seenă onă theă Englishă versionă ofă bitdefender.com:ă eliminateă versusăeliminates.ăTheăoffersăonăbitdefenderăareănotăthatăvariedăasăthoseăofăemag,ăbutăthisăisă dueă toă theă stricteră domaină ofă activityă ofă theă companyă –ă computeră technology.ă Asă foră Hofstede’să culturală dimensions,ă individualismă andă masculinityă [81],ă theă discourseă isă moreă balanced.ăThereăareănoăspecificăwordsăthatăinfluenceătheseădimensions.ăRegardingăuncertaintyă avoidance,ă theă antivirusă productsă themselvesă areă meantă toă compensateă foră highă UAI.ă However,ăcomparedătoăemag.ro,ăthereăareăfewerăfactorsăthatăwouldăincreaseătrust.ăStill,ăusersă canătestătheăsoftwareăforăfreeăforă30ădays,ăwhichăisăquiteăsimilarătoăusingăaăproductăatăhome.ă Socialămediaăimplementationăisăalsoălessăobvious.ăAsăforăpragmatismăandăindulgenceă[81],ăIă considerătheyăareănotăapplicableătoăthisătypeăofăproductăandătoămarketingăsuchăproducts. Ină orderă toă localizeă aă websiteă oră aă webpageă foră Romania,ă oneă shouldă analyzeă theă topă companiesăfromătheăsearchăresultsăforăvariousăkeywordsăthată theăproductămarketingăstrategyă willătarget.ăInătheăcaseăofăRomaniaărankingăonătheăfirstăresultăpageădoesănotămeanăqualityăandă accuracyă ofă information,ă especiallyă ină theă areasă ină whichă thereă isă lowă competition.ă Onlineă marketingă effortsă mayă adhereă toă theă practicesă ofă companiesă thată areă generallyă wellă appreciatedă andă acceptedă onă theă Romaniană market.ă Analyzingă companiesă listedă ină 4.2,ă Benchmarking website localization, from an SEO perspective,ă mayă beă usedă asă ană example.ă However,ădependingăonătheăsizeăofătheătargetămarket,ăinăthisăcaseăRomania,ăconsideringătheă informationă foundă onă highestă valueă brandsă doesă notă implyă thată translationsă oră lawă relatedă informationăcanăbeătakenăforăgranted.ăForăinstance,ăwhileăGoogleăasăaăsearchăengineămayănotă appearăproneătoălinguisticăerrorsădueătoătheăsimpleăinterfaceăităemploysăallăaroundătheăworld,ă andăbecauseămostăofătheăusersăneverăuseăitsămoreăadvancedăsearchăfacilities,ămanyăRomaniansă mayăhaveănotănoticedăthatăsomeătranslationsăareăincorrectăifăcomparedătoătheăEnglishăversion. 250 Figure 51: Google interface translation error Săptămâna trecută, Luna trecută andă Anul trecută areă theă Romaniană translationsă foră Past week,ă Past monthă andă Past year,ă respectively.ă Back-translatingă quicklyă highlightsă theă problem:ăLast week,ăLast month,ăLast year.ăWhileăpastăandălastămayăbeăusedăbyăsomeănativeă speakersă interchangeably,ă Merriam-Websteră clearlyă definesă pastă asă havingă existedă ată aă timeă beforeătheăpresentă:ăfrom,ădone,ăorăusedăinăanăearlierătime;ăusedătoăreferătoăaătimeăthatăhasăgoneă byărecently. Henceă theă Romaniană translationă shouldă haveă beenă similară toă theă translationsă foră Past hour andă Past 24 hour.ă Thus,ă (din) ultima săptămână, (din) ultima lună, (din) ultimul an wouldă haveă beenă theă appropriateă translation.ă Verifyingă Google’să functionalityă confirmsă thată theyă referătoătheăpreviousăperiodăofă7ădays,ă30ădays,ăandă365ădays,ărespectively.ă Therefore,ă ină orderă toă buildă ană accurateă localizationă model,ă oneă shouldă analyzeă bothă nativeă companiesă thată areă successfulă onă theă targetă marketă andă internatională companiesă thată areă alreadyăonătheătargetămarket.ăHiringăaălocalăSEO-awareălocalizationăcompanyăisăalsoăaămust.ă Fromătheăperspectiveăofătheătranslatorăandălocalizerăofăaăwebsite,ăwhileăităisăusefulătoăcreateăaă corpusătoăbeăused,ădataăcollectedămustăbeăconsideredăcritically.ă 251 4.7 Reverseălocalization:ăRomanianăcompaniesătargetingăforeignămarkets.ă InăthisăsubchapterăIăwillăanalyzeăseveralăRomanianăwebsitesăfromătheăperspectiveăofăhowătheyă targetăforeignămarkets.ăTheăfirstăproductătypeăisăwineăasăRomaniaăisăoneăofătheămostăprolifică wineăproducersăinăEuropeăandăinătheăworldă–ă21stăpositionă(ină2012)ăaccordingătoăFoodăandă Agricultureă Organizationă [101].ă Howeveră itsă wineă productionă isă ină declineă asă comparedă toă 2008ă whenă ită wasă theă 12thă largestă wineă producer.ă 2012ă sawă aă roughlyă 4ă timesă decreaseă ină productionăasăcomparedătoă200863,ăwhileăbeerăproductionăwasămaintainedăconstantly.ăInă2012ă beeră productionă wasă byă almostă 15ă timesă higheră thană wineă productionă [101].ă Theseă areă sufficientă reasonsă toă studyă howă Romaniană wineă exportersă couldă increaseă theiră brandă awarenessă ină theă worldă andă ină Europe.ă Theă mostă accessibleă marketsă areă anyă ofă theă UEă countriesăbecauseăRomaniaăcanăbenefităfromătheăadvantageăofăbeingăanăEUămember;ăRomaniaă andătheăotherăEUăcountriesăshareătheăsameămarketăregulationsă(forăinstance,ănoăextraătaxes).ă Consideringăthată Iăamă mostăinterestă ină Englishăwebsites,ămyă focusă forăthisăcaseăstudyă isăonă UKă websites.ă However,ă theă methodsă andă findingsă cană beă appliedă toă anyă otheră countryă andă furtherăgeneralizedătoăotherăRomanianăproductsăorăservices. Firstă Iă analyzedă http://www.romanian-winegrowers.com/,ă aă websiteă ofă 8ă Romaniană wineă producers,ă membersă ofă theă Romaniană Associationă ofă Wineă Producersă andă Exportersă (Asociaţiaă Produc toriloră şiă Exportatoriloră deă Vinuriă dină România)ă [102]ă whichă ină 2010ă visitedă theă UKă foră promotingă Romaniană wines.ă romanian-winegrowers.comă isă aă supportă website,ă toă furtheră sustaină exportă efforts.ă Iă willă alsoă analyzeă theă websitesă ofă theă followingă wineă brands:ă Casaă Panciu,ă Crameleă Recas,ă Jidvei,ă Murfatlară Romania,ă Vinconă Vrancea,ă BuciumăIasi,ăCasaădeăVinuriăHusi,ăCasaădeăVinuriăPietroasa,ăCotnari. TheănextăstepăisătoăanalyzeăbrandăawarenessăforăeachăofătheseăspecificăRomanianăbrandsăandă thenă ină generală theă UKă inhabitants’ă awarenessă ofă Romaniană wines.ă Iă usedă Google’să KeywordPlanneră setă toă Unitedă Kingdomă asă targetă countryă andă productă typeă toă wine.ă Theă results,ăi.e.ăsearchesăconductedăbyăUKăusersăonăGoogle,ăareăveryălow.ă 63 ăPleaseăseeăAppendixăGăforăfurtherădetails. 252 Keyword Avg. monthly Keyword searches Avg. monthly searches romanian wine 210 panciu 20 husi 170 jidvei wine 10 vincon 90 wine from romania 10 murfatlar 90 wine in romania 10 recas 50 wines of romania 10 cotnari 40 wine romania 10 jidvei 30 pietroasa 10 Thisă meansă thată theă targetingă ofă onlyă theseă wordsă wouldă yieldă lowă traffic.ă Theă obviousă strategyă isă toă generalizeă searchă keywordsă byă allowingă Googleă toă suggestă someă relatedă keywords.ăThereăresultedăaălistăofă800ăkeywordsăwithăatăleastă720ăaverageămonthlyăsearches.ă Then,ăIăexcludedăone-wordăkeywordsă(whichăgenerateătrafficăbutădoănotăconvertăintoăbuyers)ă andă keywordsă thată wereă tooă specifică (containedă competitoră brandsă oră retaileră name).ă However,ăoneăcanăchooseăasăaăcontentămarketingăstrategyătoăuseăgeneralăcompetitorăkeywordsă suchă asă french wine, bordeaux wine, italian wine, spanish red wine (lackă ofă capitalizationă intendedăasătakenăoverăfromăKeywordPlanner) whileăexpressingăinătheămessageăthatăthereăareă Romanianăalternativesătoăthoseăcompetitorăwines. Next,ă weă needă toă corroborateă thisă listă withă topă rankingă websitesă ină SERPsă foră variousă keywords.ă Foră thisă analysis,ă Iă willă consideră onlyă theă topă tenă keywordsă byă averageă monthlyă search,ăbutăoneăshouldăconsider,ăideally,ăallătheăkeywordsăforăanalysis.ă rose wine 9900 wine box 4400 new wine 8100 cabernet sauvignon 3600 sauvignon blanc 6600 wine offers 3600 white wine 4400 cheap champagne 2900 sparkling wine 4400 wine mixed cases 2900 Next,ăIăusedăCyberăGhostă(http://www.cyberghostvpn.com/)ătoăchangeămyăcomputersăIPătoăaă UKăIP,ăsoăthatăGoogle.co.ukăwouldăprovideămeăwithăasăaccurateădataăasăpossible.ăForăeachăofă theăkeywordsăaboveăIăconsideredătheăfirstă10ăresults.ăIăhadătoăremoveănew wine asăinătheăUKăită hasăaăreligiousăconnotationăandăitărefersătoăaămovementăofăchurches.ăTheătopăsitesăwereăthenă 253 analyzedă withă aă Keywordă DensityăAnalysisăToolă [103].ăThisă stepă revealedă furtheră potentială keywordsă foră buildingă long-tailă keywords64:ă reviews, fine, finest, dry, sweet, cheap, deals, virgin, sparkling.ăUsingă Ubersuggestă [104]ă assistsăinăfurtherăbuildingălong-tailă keywords.ăAă listă ofă long-tailă keywordsă isă alsoă suggestedă byă Google,ă ată theă endă ofă theă SERPs,ă underă theă “Searchesărelatedătoă…”ăheading. howăisăroseăwineămade personalisedăwineăbox roseăwineătypes personalisedăwoodenăwineăbox bestăsauvignonăblanc wineăgiftăbox sauvignonăblancăcharacteristics cabernetăsauvignonăwine sauvignonăblancăbrands cabernetăsauvignonădescription sweetăwhiteăwine wineăoffersăonline dryăwhiteăwine goodăcheapăchampagne whiteăwineăvarieties mixedăwineăcasesăuk typesăofăwhiteăwine mixedăwineăcasesăonline whiteăwineăcalories wineămixedăcaseădeals typesăofăsparklingăwine redăwineămixedăcase sparklingăwineăoffers fineăwineămixedăcase bestăsparklingăwine Asă shownă ină 2.6.7,ă theă moreă wordsă weă useă toă makeă upă theă keywordă phrase,ă theă higheră theă possibilityăofăconvertingăaăuserăthatăusesăsuchăaălongăkeyword.ă2-wordăandă3-wordăkeywordsă shouldă beă extendedă toă ată leastă 4ă wordsă toă increaseă conversionă rate.ă Theă listsă aboveă cană beă combinedăintoăsuchăkeywordsăasăsweet white wine deals,ădry white wine deals, good cheap white wine, rose wine gift set,ăetc. Inătheory,ăvariousătypesăofăwinesăshouldăhaveătheirăownălandingăpagesă(aăsingleăwebpageăthată isă displayedă toă theă web-useră whenă clickingă onă aă resultă fromă aă SERP);ă ină practiceă ită oftenă happensăthatătheăuserăisătakenătoătheăhomepageăorătoăotherăirrelevantăpages. Eachă ofă theă pagesă ofă theă http://www.romanian-winegrowers.com/ă websiteă rankă ină Google.co.ukă onlyă foră keywordsă thată areă notă usedă byă UKă web-usersă (eitheră B2Că oră B2B):ă 64 ăForătheăimportanceăofălong-tailăkeywordsăseeă2.6.7,ăLong-tail keywords 254 romanian winegrowers,ă halewood romania, andă specifică Romaniană wineă typesă suchă asă fetească alba (whiteă maiden), fetească neagrăă (blackă maiden), fetească regală (royală maiden), tămâioasă româneascăă(Romanianăfrankincense)ă asă Britishăusersă areănotă awareăofă suchăproducts.ăSearchingăbyăRomanianăbrandăwinesăyieldsăsatisfactoryăresults,ăbutăasăshownă earlier,ăthereăisănoăbrandăawarenessăonătheăUKămarketăforăthem. Usingă 2-wordă keywordsă yieldsă noă resultsă foră anyă ofă theă Romaniană wineă brandă websites.ă However,ă addingă romaniană toă theă setă ofă 2-wordă keywordsă showsă highă improvementă ină ranking:ăromanian pinot noir, romanian white wine,ăetc.ăUnfortunatelyăwineărelatedăsearchesă thatăinclude romanianăareăveryălowă(aroundă200ăaămonth).ăFurthermore,ăsuchăsearchesăusuallyă directăusersătoăonlineăresellerăwebsites.ăHowever,ăităwasăunexpectedăthatăromanian pinot noiră rankedă theă Romaniană andă Britishă http://www.halewood-romania-pinotnoir.co.ukă websiteă onă theăsecondăpositionăinătheăSERPăandăinăitsătextăcontentăpinot noirăoccursă22ătimesăoutăofă5939ă totală words,ă whichă isă aă 8.33șă density.ă Generalizingă theă searchă toă pinot noiră positionedă theă sameă websiteă onă ană honorableă placeă 58,ă consideringă theă competitionă foră thată keyword.ă However,ămostăoftenăusersăneverănavigateăbeyondătheăfirstăresultsăpage. Consideringă theseă findings,ă theă contrastă betweenă keywordă usageă ofă UKă usersă andă keywordă densityăonăRomanianăwineăproducersă(whichăstatisticallyămeasuresătheăimportanceăofăaăpageă fromă theă perspectiveă ofă Googleă algorithms)ă Iă consideră thată Romaniană contentă mustă beă generalized,ătoăuseăhypernyms,ăsoăthatăatăsomeălevel,ăUKăusersănamingăofăwineăconceptsăisă metă byă contentă producers.ă Thisă alsoă showsă thată translationă isă notă aă viableă solutionă ină suchă casesă (youă cannotă simplyă translateă theă Romaniană webpagesă targetedă ată Romaniansă thată areă somewhatăbrandăorăproductătypeăaware)ăintoăEnglish,ăforăaătargetămarketăthatăisăhardlyăawareă ofătheseăproducts.ăItărequiresăcopywritingăwhichăinvolvesăaămoreăproactiveăapproachătowardsă theătargetămarket. Whenăthereăisănoăbrandingăawareness,ăeitherăveryăspecificăsuchăasăaăRomanianăbrandăwine,ăită isăadvisableătoăbrandămoreăgeneralăconceptsăsuchăasăRomanian wine.ăWhileăităisăcompulsoryă toăstudyătheădirectăcompetitors,ăităisăoftenăbeneficialătoăalsoăincludeăinătheăanalysisăRomaniană companiesăthatăareăalreadyăsuccessfulăonătheătargetămarket. Thus,ăRomanianăwinegrowersămayă studyă theăUKăwebsitesăofăcompaniesăsuchăasăDaciaă andă Bitdefender:ăwww.dacia.co.ukăandăbitdefender.co.uk.ăWhileătheăprofileăofătheătwoăwebsitesăisă 255 veryă different,ă thereă areă websiteă localizationă elementsă thată cană beă consideredă foră theă wineă growers’ăwebsitesăasăwell: - legalăissues - technologiesăusedă(forăinstance,ăseveralăRomanianăwineăgrowersăstillăuseăfullăAdobeă Flashătechnologyăwebsites,ăwhileătheăstepătoăHTML5ăhasăbeenălongăpastădueă –ăevenă DaciaăonăitsăhomelandăwebsiteăstillăusesăFlashătechnologyăforăanimations), - usageăofăanimationsă(notăusedăonătheăUKăwebsiteăwhichăisăthatăofăaăLCăculture), - usingăinformationăproducedăinătheătargetăcountryă(reviewsăandăawards), - facilitatingă directă buyingă (onă Dacia.co.ukă oneă cană buyă aă cară onlineă –ă whichă isă acceptableăforăculturesăwithălowăUAIăevenăifăpurchasingăhighăvalueăproducts),ă - orăliveăhelpă(communicatingăonlineăwithăsupportăstaff,ăwhichăwouldăbeăveryăusefulătoă implementăforăcountriesăwithăhighăUAIăasăwell),ă - videoăcontentăpresentedăbyănativeăspeaker,ă - registeringă aă domaină nameă ină theă UKă (foră instanceă asă http://www.cotnari.co.ukă –ă whichă unfortunatelyă everă sinceă 2007,ă whenă ită wasă bought,ă remainedă unchanged,ă offeringăonlyăUKăbasedăcontactădetails,ăwhichăisăaăratherăstrictlyăB2Băorientation).ă Mostă Romaniană wineă producersă (http://www.wineromania.com/)ă haveă developedă contentă bothăinăRomanianăandăEnglish,ăbutătheyăareăallăratherătargetingăbusinessă clientsăconsideringă thată Englishă contentă targetsă outeră markets,ă throughă usageă ofă keywordsă addressingă internationalăend-buyersăorăbusinessăbuyers.ăSomeăofăthemăevenăuseă.comădomainsăandăonlyă twoăofăthemă.co.ukă(CotnariăandăHalewood).ăUsingă.co.ukăforătheăUKăwillăsubconsciouslyăbeă ană indicativeă foră UKă web-usersă thată theă companyă usingă ită isă aă culturallyă andă sociallyă UKă integratedă company.ă Apparentlyă Googleă noă longeră listsă resultsă ină SERPsă dependingă onă aă domain’săcountryăextension,ăbutăitălaysăemphasisăonăuserăinputăandălocation. Asăforăbitdefenderăitălistsăinătopătenăforătheăgeneralăkeywordăantivirus onăgoogle.co.uk, evenăifă apparentlyă theiră approachă isă deficientă ină locală elements.ă Exceptă foră Malaysia,ă allă theă otheră Englishămarketsăuseătheăsameătextăandăimageăcontentăandălayout.ă ComparingătheătwoăRomanianăwebsites,ăinătermsăofăSinghăandăPereiraăclassificationă(2005)65ă Dacia’săUKăwebsiteăhasătheăcharacteristicsăofăaăCulturallyăcustomizedăwebsiteăwithăaăhighă 65 ăPleaseărevisită2.4.9.5,ăClassification by level (degree). 256 degreeă ofă beingă perceivedă asă such.ă bitdefender.co.ukă hasă aă mixedă approachă andă doesă notă clearlyă fallă intoă anyă ofă ă Singhă andă Pereira’să categories.ă Itsă maină characteristicsă pointă toă theă Localizedă categoryă bută dueă toă theă usageă ofă countryă specifică TLDă oneă couldă claimă ită isă Culturallyălocalized.ăApparently,ăforăthatătypeăofăproduct,ăităfunctionsăefficiently.ăFurtherăon,ă oneăcouldăconsiderăhowătopăcompaniesălocalizedătheirăownăwebsitesăforăvariousăcountries66.ă WhileăinăthisăsubchapterăIăanalyzedăhowăRomanianăwinegrowersăareăperceivedăandăseenăonă theăinternatională markets,ăandămoreăspecificallyă onătheăUKămarket,ă reverseălocalizationăcană beăaccomplishedăbyăanyăcompanyăbelongingătoăaă“minorăculture”ăandăforăanyătypeăofăproduct.ă Regardlessăofătargetingăbusinessăorăend-buyers,ăcontentămustăbeăproducedăandăimplementedă consideringătheănewestătechnologiesăandămarketingătrends,ăwhileăadaptingătheămessageătoătheă potentială customers.ă Simpleă translationă intoă aă targetă languageă isă noă guaranteeă ofă success.ă Inefficientăonlineămarketing,ăthatămayăorămayănotăbeădoubledăbyătraditionalămarketingăefforts,ă translatesă intoă lowă sales.ă However,ă theă degreeă ofă localizationă mayă dependă onă theă productă type.ăCulturallyălocalizedăwebsitesăshowătheăhighestăpotentialăinărankingăhighăinăSERPsăandă inăconvertingăvisitorsăintoăbuyers. 4.8 Conclusions Ină thisă chapter,ă theă maină purposeă wasă toă testă theă claimsă andă theoriesă fromă theă previousă chapters.ă Oneă ofă theă importantă topicsă coveredă benchmarkingă websiteă localization.ă Ină thată particularăsubchapteră(4.2)ăIăanalyzedăseveralăinternationalăbrandsăbothăfromătheăperspectiveă ofă theiră brandă valueă (Interbrands),ă localizationă degreeă (Johnă Yunker’să bytelevel.com),ă andă userăusageă(Alexa.com).ă Anotheră importantă concernă wasă toă demonstrateă thată localizationă doesă notă alwaysă involveă translation.ăAnglophoneăcountries,ăjustălikeăFrancophoneăorăGermanăspeakingăcountriesămayă shareătheăsameăcontentăor,ăonătheăcontrary,ătheăcontentămayăbeăpersonalizedătoăsuchăaădegreeă thatăităisăproducedătakingăintoăaccountăonlyălocalăinterests.ă(Eurosport)ă Manyă companiesă applyă localizationă strategiesă bută notă necessarilyă translate.ă Thereă areă twoă maină directionsă ină producingă localizedă content:ă translationă andă copywriting.ă Ină theă thirdă 66 ăSeeăsubchapteră4.2,ăBenchmarking website localization, from an SEO perspective 257 subchapteră Iă conductedă twoă experimentsă thată involvedă keywordă researchă andă on-siteă SEOă optimizationăappliedătoătheătwoătypesăofăstrategies.ăWhileătheămethodsădifferătheăoutcomeăisă ratheră similară asă longă asă keywordă researchă andă keywordă implementationă areă achievedă methodically. Fromăaăratheră“traditional”ăTranslationăStudiesăandălocalizationăperspective,ătheăcaseăstudyăină subchapteră4.5ăimpliedătheăanalysisăandă comparisonăofălocalizedăwebpagesăwithinătheăsameă website. Ină theă followingă subchapteră (4.6),ă Iă developedă aă strategyă foră foreignă companiesă onă howă toă localizeă theiră websitesă foră theă Romaniană market,ă pinpointingă culturală aspectsă thată mayă beă applicableătoăthisăparticularămarket. TheăfinalăsubchapterăanalyzedăRomanianăcompanyăwebsitesăandătheirăperformanceăonăforeignă marketsă(namelyătheăUK).ăIăcomparedăRomanianăwinegrowers’ăwebsitesăwithătheăwebsitesăofă twoăofătheămostăsuccessfulăRomanianăcompanies:ăDaciaăandăBitdefender. 258 5. Finalăconclusions The main purpose of this thesis was to demonstrate that the academic community and website localization industry are not always aware of the end-user (that can be an individual or a business). Website localization, in most cases, follows the traditional strategy of intrusive marketing and does not appear to openly and actively embrace the potential benefits of inbound marketing, namely content marketing. Content marketing, as demonstrated, is the most rewarding strategy in comparison with all the other types of internet related advertising. Content marketing in the case of localized websites cannot be accomplished without conducting a separate research on the end-users from the target market. Search engines are the main force behind inbound marketing. They function by serving results based on user input – keywords. Therefore, keywords become a matter of shared knowledge between website owners and end-users. In Translation Studies this idea has been around since the introduction of reader-oriented translation, derived from pragmatics, rhetoric and readerresponse criticism. Theăresearchăinvolvedăanalyzingătheămainăconceptsăassociatedăwithătheălocalizationăprocessăină general,ă andă theă relatedă processesă appliedă toă theă specifică fieldă ofă websiteă localization.ă Initially,ă Iă reviewedă theă moreă generală concepts,ă globalizationă andă internationalization,ă analyzingă specifică issuesă sometimesă overlookedă byă theă academică community,ă suchă asă theă internationalizationăorăstandardizationăofătheăsourceătextă soăthatătheăefficiencyăandăspeedăofă theătranslationăintoătheătargetătextsămayăbeăincreased.ă However, from a management perspective, I consider that decomposing the transfer of a source website to several other target market websites into four distinct processes – globalization, internationalization, localization and translation (GILT) – is a most rewarding endeavor. They are consecutive processes that derive from one another. While globalization tends to be a management specific process, internationalization requires minute preparations. Some of the ideas linked to internationalization which were omitted by the academic community include: text standardization and setting a global translation strategy. As for localization itself, I have shown that various processes can be automated either by using advanced content management systems (CMS) or by collecting information from such systems. Thus, the translator can easily become a localizer him- herself. Also, I considered 259 that localization should be considered from a slightly different perspective. I defined it as adapting the communication and marketing strategy to the customers’ă expectations, based on a locale mix. All previous definitions focus on the product or service that must be changed. In my opinion it is not the products themselves that need to be changed but the interface. Modifying the interface is not changing the product. Thus, human interaction makes localization a communicative issue and not (only) a product related one. Furthermore, my definition of localization is also valid when it comes to e-commerce websites. The website owner does not change the products on sale but the communication and marketing strategy. Localization could also be seen as a personalization process. If localization is compared to personalization, the latter appears to be a specific type of localization as it targets a smaller community or even a group of people. A niche can be regarded as a specific type of localization comparable to personalization. However, a community targeted through personalization shares the same locale, while a niche shares common characteristics (like the academic community interested in localization) while being part of different locales. ByăfurtherăanalyzingătheăGILTărelatedăconceptsăIăhaveăshownăthat,ăinăpractice,ătranslationăisă notă alwaysă necessary.ă Whileă translationă isă moreă appropriateă foră informativeă textă types,ă copywritingă isă moreă efficientă foră expressiveă andă operativeă texts.ă Onă websitesă thereă areă noă clear-cutătextătypesăandăoneătypeămayăprevailăoverătheăothers.ăHowever,ăifăthereăareăelementsă ofăexpressiveăandăoperativeătexts,ăcopywritingămayăbeăaămoreăeffectiveă strategy.ăIăhaveăalsoă shownă thată copywritingă cană easilyă includeă relatedă keywordsă (foră casesă ină whichă theă targetă marketăisănotăawareăofăaăcertainăproductăorăservice).ăRegardlessăofătheătranslationăstrategyăthată isă used,ătranslatorsăareărequiredătoă addăină extraăinformationă soăthatătheăcontentămayăincludeă relatedă keywords.ă Consideringă thată translationă isă oftenă replacedă byă copywriting,ă theă twoă competingăconceptsăshouldăbeăencompassedăunderăaăbroaderănotion,ăthatăofătextăproduction.ă Fromăaăcultural,ăsocial,ăpoliticalăandăeconomicăpointăofăview,ăaămoreăappropriateătermătoăbeă usedăinsteadăofăGILTăisăthatăofăglocalization.ăGlocalizationăemphasizesătheăinterdependenceă ofă globală andă locală factorsă ină shapingă theă methodsă byă whichă onlineă andă traditională communicationă areă achieved.ă Ită suggestsă moreă appropriatelyă theă ideaă thată globală andă locală forcesă influenceă ină aă synchronousă manneră theă inhabitantsă ofă theă “globală village”.ă Iă haveă demonstratedă thată whileă GILTă cană beă consideredă aă setă ofă ratheră finite,ă consecutiveă andă unidirectională processes,ă glocalizationă isă aă continuous,ă simultaneousă andă multidirectională 260 process.ă Glocalizationă mirrorsă ină aă moreă accurateă manneră theă realityă ofă theă currentă stateă ofă affairs. Glocalization also suggests the fact that the perspective on localization and text production from the point of view of the academic community should not be an exclusively prescriptive one but should include descriptive approaches as well. In a similar manner to adapting translations for children based on their age and level of knowledge, or old language texts to modernized ones (diachronic intra-translation), so should localizers adopt the descriptive approach to translation. Prescriptive translation products shouldă beă “imposedă on”ă theă enduser in stages. Standard language should be suggested and not imposed. The prescriptive approach may also disregard an important share of potential customers. My suggested descriptive approach means further focusing on the end-user and finding a common linguistic base (onomasiological or semasiological approaches) even if this means using foreign elements. It also involves using language standard and non-standard keywords. Thus, the e-text is closer to its purpose: efficient online communication. Focusing on the end-user translates into higher traffic on company websites, i.e. a higher base for conversion rate into buyers. Thus, including keyword research in the website localization process is of utmost importance. In the thesis I proposed that keyword based website localization be called SEO localization. With the multitude of available online tools, this type of localization is an achievable goal. Keyword research highlights the importance of search engines in online marketing and search engine optimization (SEO). SEO localization implies a subsequent specialized type of translation – SEO translation. In practice, keyword research and selection is applied to textual content. Theă thesisă alsoă analyzedă issuesă regardingă textă andă electronică textă (similaritiesă andă differences).ă Iă reviewedă theă classificationă ofă onlineă textsă byă applyingă traditională textă classificationă theoriesă toă e-texts.ă Theă Audiomedială functională typologyă (Reissă citedă ină Munday,ă 2008:ă 72)ă isă ofă particulară interestă toă e-texts.ăAudiomedială textsă areă specifică toă theă semioticămixăofăwebsites.ăWebăcontentăisăcomposedăofătext,ăgraphics,ăvideoăandăsound.ă Darwish’sătextualălayersă(2010:155)ăandăhowătheyăapplyătoăwebsiteălocalizationăcanăbeăusefulă inădeterminingătextăcomprehensionăandătextăproductionăstrategies.ăWithinătheătextualălayerăIă 261 proposedă aă newă element:ă theă shellă ofă theă text,ă whichă refersă toă theă macrostructurală organizationăofătextsă(asăbooks,ăbrochures,ămagazines,ăetc.).ăBasedăonătextătypologiesăHatimă andă Masonă distinguishă instructională textsă asă aă distinctă text-typeă (1990:ă 153-158).ă Onă websites,ăităisăaămajorătextătypeăasămostăofătheăinformationăonătheăInternetăisăforăeducatională purposes.ă Companiesă “educate”ă bothă theiră potentială andă existingă customersă -ă similară toă Biber’săonlineăinformationalăelaboration.ă(1995:ă102) Aătwo-dimensională classificationăofătextsă intoă verbalăandănominalătextsă revealsă thată aăhighă percentageă ofă theă informationă foundă onă theă Internetă isă nominal,ă thusă corroboratingă theă importanceă ofă instructională texts.ăAnă analysisă ofă theă elementsă thată constituteă theă contextuală layeră ofă e-textsă wasă alsoă conducted,ă highlightingă theă particularitiesă thată applyă toă electronică texts. Regardingă theă culturală layer,ă Iă reviewedă someă majoră culturală trendsă appliedă toă texts.ă Iă alsoă synthesizedă theă culturală profileă ofă theă typicală Romaniană web-user.ă Theă intentionalityă layeră withă itsăsubdivisionsă–ăinformativeăintentionăandă communicativeăintention,ă showsăaăhighă degreeăofăapplicabilityăespeciallyăonăe-commerceăwebsites. Theă intertextuală layeră ofă textsă cană beă seenă bothă fromă ană intra-ă andă interlinguală perspective.ă Intralinguală aspectsă mayă referă toă coherenceă withină eachă ofă theă localizedă websites,ă whereasă interlingualăonesăreferătoăcoherenceăamongătheăvariousălocalizedăwebsitesă(brandingăstrategy).ă Aă specială typeă ofă intertextualityă cană beă aă searchă engineă resultsă page.ă Intertextualityă isă achievedăbyătheăkeywordăenteredăbyătheăweb-user. Myămainăcontributionăinăthisăpartăofătheăthesisăisătheăadditionăofătheăanticipationălayerăasăană importantă partă ofă theă users’ă interactionă withă theă searchă engines.ă Theă anticipationă layeră constitutesă aă pre-readingă activity.ă Thisă layeră alsoă involvesă consideringă theă call-to-actionă wordsă fromă theă resultsă listedă ină theă SERPs.ă Iă analyzedă whată theă call-to-actionă refersă to,ă developedă aă methodă foră collectingă call-to-actionă expressions,ă builtă aă generală listă ofă call-toactionăwordsăforătheăRomanianămarket,ădifferentiatedăbetweenăcall-to-actionăwordsăandăwebuseră generatedă keywords.ă Theă comparativeă studyă onă Romaniană andă Englishă CTAsă alsoă revealedăthatăRomanianătendsătoăuseănominalătexts,ăwhereasăEnglishăusesăverbalătexts. 262 Theă resultsă ofă theă analysesă clearlyă showă thată theă E-textă isă perceivedă asă aă communicationă process,ăthusăcorroboratingătheămainăhypothesisăofămyăthesis,ăi.e.ăe-textsăshouldăfocusăonătheiră consumers.ăConsideringătheăcurrentăinboundămarketingătrend,ătheăproposedăSEOălocalizationă approachă andă theă keyword-basedă communicationă mediatedă throughă searchă enginesă (alsoă supportedă byă empiricală evidence),ă ită cană beă assertedă thată interactionă shiftedă fromă B2Că (business-to-customer)ăcommunicationătoăC2Bă(customer-to-business)ăcommunication. InăorderătoăverifyătheăcorrectnessăofătheăconceptualăclaimsăIăconductedăseveralăanalysesăofătheă textuală andă non-textuală componentsă ofă aă websiteă bothă fromă theă perspectiveă ofă scholarsă ină linguisticsăandăTSăonătheăoneăhand,ăandăfromăaămoreătechnicalăstandpointăonătheăother.ăWithină theăverbalăcomponentăofăaăwebsiteăIăproposedătheăadditionăofătheăadministrativeătextătype,ăasă aă distinctă typeă ofă text,ă withă twoă subdivisions:ă useră administrationă pagesă andă corporateă administrationăpages.ă Theăresearchăalsoăinvolvedăexaminingăallătheăfactorsăassociatedăwithătheălocalizationăprocess,ă fromătheăobjectiveăfactorsălikeătheăfront-endăandătheăback-endătoătheămoreăsubjectiveăhumană factor.ăTheă humană factoră isă theă mostă importantă oneă asă ită isă notă alwaysă aă factoră thată cană beă accuratelyă determined.ă Therefore,ă theă roleă ofă keywordsă ină textă productionă isă ofă paramountă importance. Theămainăpurposeăofătheăfinalăpartăofătheăthesisăwasătoătestătheăclaimsăandătheoriesăfromătheă previousăchapters.ăOneăofătheăimportantătopicsăcoveredăbenchmarkingăwebsiteălocalization.ăInă thată particulară subchapteră (4.2)ă Iă analyzedă severală internatională brandsă bothă fromă theă perspectiveă ofă theiră brandă valueă (Interbrands),ă localizationă degreeă (Johnă Yunker’să bytelevel.com),ă andă useră usageă (Alexa.com).ă Iă showedă thată theă findingsă ofă theă threeă perspectivesă doă notă alwaysă match.ă Localizationă isă thusă notă alwaysă theă onlyă factoră thată determinesă theă popularityă ofă aă website.ă Usabilityă mayă stillă beă theă primordială factoră ină determiningă theă successă oră failureă ofă aă website.ă Usabilityă mayă implyă localization,ă bută localizationă doesă notă necessarilyă implyă usability;ă foră instance,ă aă user-friendlyă iconographică onlineă editingă softwareă withă directedă andă controlledă stepsă foră usageă wouldă requireă internationalizationă throughă standardization,ă i.e.ă graphicală resemioticizationă ofă theă textuală interface. 263 Anotheră importantă aspectă wasă toă demonstrateă throughă observationă andă analysisă thată localizationădoesănotăalwaysăimplyătranslation.ăAnglophoneăcountries,ăjustălikeăFrancophoneă oră Germană speakingă countries,ă mayă shareă theă sameă contentă or,ă onă theă contrary,ă theă contentă mayăbeăpersonalizedătoăsuchăaădegreeăthatăităisăproducedătakingăintoăaccountăonlyălocalămarketă interestsă(Eurosport).ă Asăalreadyădiscussedăinătheătheoreticalăpartăofătheăthesis,ămanyăcompaniesăapplyălocalizationă strategiesăbutănotănecessarilyătranslate.ăThereăareătwoămainădirectionsăinăproducingălocalizedă content:ă translationă andă copywriting.ă Iă conductedă twoă experimentsă thată involvedă keywordă researchă andă on-siteă SEOă optimizationă appliedă toă theă twoă typesă ofă strategies.ă Whileă theă methodsămayădiffer,ătheăoutcomeăisăratherăsimilarăasălongăasăkeywordăresearchăandăkeywordă implementationăareăachievedămethodically. Theăresearchăhasăalsoăincludedăaăcompleteăanalysisăofătwoăhotelăpagesăfromăbooking.com,ăină orderă toă verifyă theă extentă toă whichă localizationă isă appliedă ină theă caseă ofă oneă ofă theă mostă successfulă globală brands.ă Theă findingsă cană beă appliedă toă bothă localizationă andă reverseă localization. Furthermore,ă Iă developedă aă strategyă foră localizingă websitesă foră theă Romaniană market,ă pinpointingă culturală aspectsă thată areă applicableă toă thisă particulară marketă thată couldă beă ofă interestă forăinternatională companies.ă Ită isă worthă mentioningăthatămyă findings,ăappliedătoă theă onlineă Romaniană users’ă profile,ă doă notă alwaysă corroborateă Hall’să (1963,1976,ă 1990)ă andă Hofstede’să (2010,ă [81])ă conclusionsă withă regardă toă theă high-contextă vs.ă low-contextă cultureă dichotomy,ă proxemicsă andă chronemics,ă onă theă oneă hand,ă andă poweră distanceă index,ă individualism,ă masculinity,ă uncertaintyă avoidanceă index,ă pragmatism,ă indulgence,ă onă theă other. IăalsoăanalyzedăwebsitesăofăRomanianăcompaniesăandătheirăperformanceăonăforeignămarketsă (i.e.ătheăUK).ăIăcomparedăRomanianăwinegrowers’ăwebsitesăwithătheăwebsitesăofătwoăofătheă mostăsuccessfulăRomanianăcompanies:ăDaciaăandăBitdefender. Theă websiteă localizationă industryă isă likelyă toă continueă toă followă aă sharpă ascendingă trendă asă thereăareăstillănumerousăcompaniesăthatăplanătoăapproachătheăglobalăonlineămarkets.ăAsăforătheă technologiesăinvolvedăinălocalization,ătheyăareădevelopingăexponentially,ătogetherăwithăotheră 264 communicationă relatedă technologies.ă Ină theă notă soă distantă future,ă theă informativeă textă translationă willă beă highlyă acceptableă bothă ină termsă ofă outpută qualityă andă productionă speed.ă Forăsomeătypesăofătextsăandăforăcertainălanguageăpairsăităalreadyăfunctionsăsatisfactorily.ăTheă combinationăofăalignedăcorporaăandăstatisticămachineătranslationă(basedăonăanăever-increasingă quantityă ofă indexedă contentă byă Googleă andă otheră searchă engines)ă willă furtheră improveă machineă translationă technologiesă andă CATă software.ă However,ă theă extensiveă usageă ofă automatedă translationă toolsă willă leadă toă theă homogenizationă ofă textsă –ă accidentală (self-)ă plagiarismăandăduplicateăcontent).ăTheărapidăspreadăofăCMSăplatformsăwillăfurtherăfacilitateă theă increaseă ofă cross-borderă onlineă salesă and,ă thus,ă ofă theă demandă foră localizedă contentă production. Consideringă theă rapidă advanceă ofă theă localizationă andă translationă technologiesă ită couldă beă predictedăthatătheăroleăofătheătranslatorsăwillădecline.ăNevertheless,ăwhileămachinesăwillătakeă overă significantlyă moreă ofă theă translationă process,ă aă humană translatoră willă stillă beă heldă responsibleăforătheăfinalăoutput.ăTheăminimumărequirementăforăaătranslatorăwillăbeătoădoăposteditingă work.ă Ată theă sameă time,ă asă shownă previously,ă machineă translationă willă beă reliableă onlyă foră informativeă textă types.ă Foră expressiveă andă appellativeă texts,ă MTă willă neveră completelyăreplaceătheăworkăofăaăhumanătranslator. Thisă thesisă hasă attemptedă toă offeră newă perspectivesă onă aă topică thată isă nowadaysă widelyă debated,ă i.e.ă localization.ă Throughă theă variousă researchă methodsă Iă employedă Iă haveă triedă toă demonstrateăthatătheăinitialăhypothesesăofătheăthesisăwereăcorrect.ăTheăcaseăstudiesăalsoăhelpedă ină theă demonstrationă ofă itsă maină andă secondaryă objectives.ă Weă evenă ventureă toă predictă thată keywordăbasedă orăSEOă translationăandălocalizationăcouldă soonărepresentă aă newădirectionă ină TranslationăStudies,ăespeciallyăforănon-literaryătexts.ăTheăcontinuousăuseăofătechnology,ăbothă inătrackingăuserăbehaviorăandăinătextăproductionăwillăeventuallyădetermineăshiftingăfocusăontoă theăend-user.ăThisăisăaănecessaryăstepăinăorderătoăachieveăhigherăglobalăexposureăandăsuperioră performanceăonălocalămarketsăthroughăkeywordăbasedăwebsiteălocalizationăstrategies. 265 Bibliography 1. Arnold,ăDouglas,ăLornaăBalkan,ăSietyăMeijer,ăR.ăLeeăHumphreys,ăLouisaăSadleră(1994)ă Machine Translation: an Introductory Guide.ăLondon:ăNCCăBlackwellăLtd 2. Austermühl,ă Frankă (2011)ă “Onă Cloudsă andă Crowds:ă Currentă Developmentsă ină Translationă Technology” ină Translation in Transition,ă Trier:ă Wissenschaftlicheră Verlag,ă retrievedăfromăhttp://www.t21n.com/homepage/articles/T21N-2011-09-Austermuehl.pdfă 3. Barber,ă Wendyă andă Albertă Badreă (2001)ă “Culturability:ă Theă Mergingă ofă Cultureă andă Usability.”ă4th Conference on Human Factors and the Web.ăNewăJersey:ăBaskingăRidgeă 4. Barker,ăChrisă(2004)ăThe SAGE Dictionary of Cultural Studies,ăLondon,ăThousandăOaks,ă CA,ăNewăDelhi:ăSAGEăPublications 5. Bastin,ăGeorgesăL.,ă(2009)ă“Adaptation”ăinăMonaăBakerăandăGabrielaăSaldanhaă(eds.)ă Routledge Encyclopedia of Translation Studies.ăNewăYork:ăRoutledge,ă2ndăedition,ăpp.57 6. Bell,ăAllană(1984)ă“Languageăstyleăasăaudienceădesign”ăinăLanguage in Society,ăVolumeă 13,ăIssueă02,ăJune,ăCambridge:ăCambridgeăUniversityăPress,ăpp.ă145-204ă 7. Bell,ăAllană(2007)ă„Styleăandătheălinguisticărepertoire.”ăInăLlamas,ăCarmen,ăLouiseă MullanyăandăPeterăStockwellă(Eds.).ăThe Routledge Companion to Sociolinguistics.ă Abingdon:ăRoutledge,ăpp.ă95-100 8. Bell,ă Rogeră T.ă (1998).ă “Psychological/cognitiveă approaches”ă ină M.ă Bakeră (ed.),ă Routledgeă Encyclopedia of Translation Studies.ă Londonă Ță Newă York:ă Routledge,ă pp.ă 185-189 9. Beaugrande,ă Robert-Alaină deă andă Wolfgangă Dressleră (1981)ă Introduction to text linguisticsăLondonăandăNewăYork:ăLongman 10. Biber,ăDouglasă(1995)ăVariation across Speech and Writing.ăCambridge:ăCambridgeă UniversityăPress 11. Brooks,ă Davidă (2000)ă “Whată Priceă Globalization?ă Managingă Costsă ată Microsoft.”ă In:ă Robertă C.ă Sprungă (ed.)ă Translating Into Success. Cutting-edge strategies for going multilingual in a global age.ă Americană Translatorsă Associationă Scholarlyă Monographă SeriesăXI.ăAmsterdamă-ăPhiladelphia:ăBenjamins,ăpp.ă43-58 12. Cadieux,ă Pierre,ă Bertă Esselink,ă (2004)ă “GILT:ă Globalization,ă Internationalization,ă Localization,ăTranslation”,ăretrievedăfromăwww.i18n.ca/publications/GILT.pdf 266 13. Caffi,ăClaudia.ă(2009)ă„PragmaticăPresupposition.”ăJacobăL.ăMey,ăConcise Encyclopedia of Pragmatics.ăOxford:ăElsevier,ăpp.ă759-767 14. Cardey,ă Sylviane,ă Peteră Greenfieldm,ă Xiaohong Wu,ă (2004)ă “Designingă aă Controlledă Languageă foră theă Machineă Translationă ofă Medicală Protocols:ă Theă Caseă ofă Englishă toă Chinese”,ăinăRobertăE.ăFrederkingăKathrynăB.ăTayloră(Eds.)ăMachine Translation: From Real Users to Research, Berlin-Heidelberg:ăSpringer,ăpp.ă37-47 15. Chesterman,ăAndrewă(1997)ăMemes of Translation: The Spread of Ideas in Translation Theory.ăAmsterdam:ăJohnăBenjamins 16. Chriss,ăRogeră(2006)ăTranslation as a ProfessionăMorrisville,ăNC:ăLuluăPress 17. Crystal,ăDavidă(2003)ăEnglish as a Global Language.ă2ndăedition.ăCambridgeăUniversityă Press. 18. Crystal,ăDavidă(2006)ăLanguage and the Internet.ă2ndăedition.ăCambridgeăUniversityă Press 19. Curran,ăKevinăandăJudeăMcGIincheyă(2009)ă“Verticalăsearchăengines”ăinăKevinăCurrană (ed.) Understanding the Internet: A Glimpse into the Building Blocks, Applications, Security and Hidden Secrets of the Web (Chandos Information Professional Series)ă Oxford:ăChandosăPublishing, pp.ă275-282 20. Cutroni,ăJustină(2010)ăGoogle Analytics.ăSebastopol:ăO’Reilly 21. Cyr,ăDianneă(2003)ă“LocalizationăofăWebăDesign:ăAnăEmpiricalăComparisonăofăGerman,ă Japanese,ăandăU.S.ăWebsiteăCharacteristics”ăJournal of the American Society for Information Science and Technology, 55(33),ăpp.1-10,ăretrievedăfromă http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.85.8765Țrep=rep1Țtype=pdf 22. Darwish,ăAliă(2008)ăOptimality in Translation.ăVictoria:ăWritescopeăPtyăLtd. 23. DePalma,ă Donaldă A.,ă Benjamină Bă Sargent,ă Renatoă Să Beninatto,ă (2006).ă Can't Read, Won’t Buy: Why Language Matters on Global Websites An International Survey of Global Consumer Buying Preferences,ăLowell,ăMA:ăCommonăSenseăAdvisory,ăretrievedă fromă http://www.commonsenseadvisory.com/portals/_default/knowledgebase/articleimages/06 0926_r_global_consumer_preview.pdf 24. Devitt,ăAmyăJ.ă(2008)ăWriting Genres.ăIllinois:ăSouthernăIllinoisăUniversity 25. Dimitriu,ăRodicaă(2002)ăTheories and Practice of Translation.ăIa i:ăInstitutulăEuropean 26. Dimitriu,ăRodicaă(2006)ăThe Cultural Turn in Translation Studies.ăIa i:ăInstitutulă European 267 27. Dimitriu,ăRodicaăandăKarl-HeinzăFreigangă(2008)ă(eds.)ăTranslation Technology in Translation ClassesăIa i:ăInstitutulăEuropean 28. Dimitriu,ăRodica.ă(2009)ă“KeyăwordsăandăconceptsăinăE.ăA.ăNida’săapproachătoă translationăandătheirăfurtherădevelopmentăinăTranslationăStudies”ăinăRodicaăDimitriuăandă MiriamăShlesingeră(eds.)ăTranslators and Their Reader,ăBruxelles:ăLesăÉditionsăduă Hazard,ăISBNă:ă2-930154-23-3,ăpp.ă23-41 29. Esselink,ăBertă(2000)ăA practical guide to localization,ăAmsterdam:ăJohnăBenjaminsă Publishing 30. Fairclough,ăNormană(1992)ăDiscourse and social change.ăCambridge:ăPolityă 31. Fishkin,ăRandăandăThomasăHøgenhavenă(2013)ăInbound Marketing and SEO: Insights from the Moz Blog,ăIndianapolis,ăIN:ăWileyăPublishing 32. Friedman,ăThomasăL.ă(2000)ăThe Lexus and the Olive Tree.ăNewăYork:ăAnchorăBooks 33. Gale,ăWilliamăA.ăandăKennethăW.ăChurchă(1993)ă“Aăprogramăforăaligningăsentencesăină bilingualăcorpora”.ăComputational Linguistics,ă19(1),ăpp.75–90 34. Gouadec,ăDanielă(2003)ă“Leăbagageăspecifiqueăduălocaliseur/localisateur.ăLeăvraiă <<nouveauăprofit>>ărequis”.ăInăMETA,ăXLVIII,ăretrievedăfromă http://www.erudit.org/revue/meta/2003ă/v48/n4/008724ar.ăhtml 35. Grzega,ăJoachimăandăMarionăSchöneră(2007)ăEnglish and General Historical Lexicology, Materials for Onomasiology Seminars,ăretrievedăfromăhttp://www1.kueichstaett.de/SLF/EngluVglSW/OnOnMon1.pdf 36. Grzega,ăJoachimă(2012)ă“Lexical-SemanticăVariables”ăinăJ.ăM.ăHernández-Campoy,ăJ.ăC.ă Conde-Silvestreă(eds.)ăThe Handbook of Historical Sociolinguistics,ăWiley-Blackwell,ă pp.ă271-292 37. Grice,ăH.ăPaulă(1975),ă“Logicăandăconversation.”ăinăPeterăColeăandăJerryăL.ăMorgană (eds.).ăSyntax and Semantics,ăVol.ă3:ăSpeech Acts.ăNewăYork:ăAcademicăPress,ăpp.ă41-58 38. Gross,ăRichardă(2013)ăBeing Human: Psychological and Philosophical Perspectives,ă NewăYork:ăRoutledge 39. Hall,ăEdwardăT.ă(1963).ă“AăSystemăforătheăNotationăofăProxemicăBehavior”.ăAmerican Anthropologistă65ă(5),ăpp.1003–1026,ăretrievedăfromă http://onlinelibrary.wiley.com/doi/10.1525/aa.1963.65.5.02a00020/pdfă 40. Hall,ăEdwardăT.ă(1976)ăBeyond Culture,ăNewăYork:ăRandomăHouse 41. Hall,ă E.ă T.ă andă Hall,ă M.ă (1990).ă Understanding cultural differences: Germans, French and Americans.ăYarmouth:ăInterculturalăPress 268 42. Halliday,ă Michaelă A.ă K.ă andă Ruqaiyaă Hasană (1976).ă Cohesionă ină English.ă London:ă Longman. 43. Hatim,ăBasilăandăIanăMason.ă(1990)ăDiscourse and the Translator.ăLondon:ăLongman 44. Hatim,ăBasilăandăJeremyăMundayă(2004)ăTranslation: An advanced resource book.ăă LondonăȚăNewăYork:ăRoutledge 45. Hofstede,ăGeert,ăGertăJanăHofstedeăandăMichaelăMinkovă(2010)ăCultures and Organizations. Software of the Mind. 3rdăedition.ăNewăYork:ăMcGraw-Hill.ă 46. Hutchins,ăJohnă(2005)ă“Currentăcommercialămachineătranslationăsystemsăandăcomputerbasedă translationă tools:ă systemă typesă andă theiră uses”,ă International Journal of Translation,ă17,ăpp.ă5-38.ăretrievedăfromăhttp://www.hutchinsweb.me.uk/IJT-2005.pdf 47. Jakobson,ăRomană(1992).ă“OnăLinguisticăAspectsăofăTranslation”ăinăJohnăBiguenetăandă RainerăSchulteă(eds.).ăTheories of Translation: An Anthology of Essays from Dryden to Derrida.ăChicagoăandăLondon:ăTheăUniversityăofăChicagoăPress,ăpp.ă144-151 48. Jiménez-Crespo,ă MiguelăA.ă (2013)ă Translation and Web Localizationă Londonă Ță Newă York:ăRoutledge 49. Kelly,ăNatalyăandăDonaldăA.ăDePalma,ă(2012)ăThe Top 100 Language Service Providersă,ă Lowell,ă MA:ă Commonă Senseă Advisory,ă retrievedă fromă http://www.commonsenseadvisory.com/Portals/0/downloads/120531_QT_Top_100_LSP s.pdfă 50. Kenny,ă Dorothyă (2009).ă “Unită ofă translation”ă ină Monaă Bakeră andă Gabrielaă Saldanhaă (Eds.),ă Routledgeă Encyclopedia of Translation Studies.ă 2ndă edition.ă Londonă Ță Newă York:ăRoutledge,ăpp.ă304-306 51. Kress,ă Guntheră andăTheoă vană Leeuwenă (2001)ă Multimodal Discourse: The Modes and Media of Contemporary Communication.ăOxfordăUK:ăOxfordăUniversityăPress 52. Lakó,ăCristiană(2007a)ă“Facetsăofăadvertisingăonătheăinternet”ăStudia Universitatis Petru Maior. Philologia (6),ăTârgu-Mureş:ăEdituraăUniversit ţiiă„PetruăMaior”,ăpp.ă231-235 53. Lakó,ăCristiană(2007b)ă“BeyondăOnlineăAdvertising”ă The Proceedings of the European Integration - Between Tradition and Modernity Congress (2),ă Târgu-Mureş:ă Edituraă Universit ţiiă„PetruăMaior”,ăpp.357-363 54. Lakó,ă Cristiană (2008)ă “Emailă advertisingă asă metadiscourse”ă Studia Universitatis Petru Maior. Philologia (7),ăTârgu-Mureş:ăEdituraăUniversit ţiiă„PetruăMaior”,ăpp.ă235-239 55. Lakó,ă Cristiană (2009a)ă “Referenceă Erroră Occurrencesă andă Howă Theyă Affectă Communication”ăStudia Universitatis Petru Maior. Philologia (8),ăTârgu-Mureş:ăEdituraă Universit ţiiă„PetruăMaior”,ăpp.ă221-227 269 56. Lakó,ă Cristiană (2009b)ă “Translatingă foră theă Webă -ă Theă Keywordă Orientedă Translationă Process”ă The Proceedings of the European Integration - Between Tradition and Modernity Congress (3),ăTârgu-Mureş:ăEdituraăUniversit ţiiă„PetruăMaior”,ăpp.761-767ă 57. Lakó,ăCristiană(2010a)ă“Anăonomasiologicalăapproachătoăonlineămarketing”ăProceedings of the International Conference Communication, context, interdisciplinarity (1),ăTârguMureş:ăEdituraăUniversit ţiiă„PetruăMaior”,ăpp.ă417-421 58. Lakó,ă Cristiană (2010b)ă “Companyă Namingă (Whenă namesă speakă volumes)”ă Studia Universitatis Petru Maior. Philologia (9),ă Târgu-Mureş:ă Edituraă Universit ţiiă „Petruă Maior”,ăpp.ă337-ă343 59. Lakó,ă Cristiană (2011)ă “Teachingă Localizationă ină Romania”ă Studia Universitatis Petru Maior. Philologia (10),ăTârgu-Mureş:ăEdituraăUniversit ţiiă„PetruăMaior”,ăpp.ă266-269 60. Lakó,ăCristiană(2012a)ă“LocalizationăandăTranslationăStudies”ăStudia Universitatis Petru Maior. Philologia (12),ăTârgu-Mureş:ăEdituraăUniversit ţiiă„PetruăMaior”,ăpp.ă202-208 61. Lakó,ă Cristiană (2012b)ă “Theă Elementsă ofă theă Websiteă Localizationă Process”ă Studia Universitatis Petru Maior. Philologia (9),ă Târgu-Mureş:ă Edituraă Universit ţiiă „Petruă Maior”,ăpp.ă358-ă362 62. Lakó,ăCristiană(2012c)ă“Non-verbalăCommunicationăonăCommercialăWebsitesă–ăIssuesăină Websiteă Localization”ă Proceedings of the International Conference Communication, context, interdisciplinarity (2),ă Târgu-Mureş:ă Edituraă Universit ţiiă „Petruă Maior”,ă pp.1111-1118 63. Lakó,ăCristiană(2013a)ă“WhichăWayăWebsiteăLocalization:ăTranslationăorăCopywriting?”ă inăIulianăBoldeaă(ed.)ăMemory Identity and Intercultural communication,ăRoma:ăEdizioniă NuovaăCultura,ăpp.ă433-440 64. Lakó,ăCristiană(2013b)ă“Benchmarkingă Localization,ăaăContentăMarketingăPerspective”ă The Proceedings of the European Integration - Between Tradition and Modernity Congress (5),ăTârgu-Mureş:ăEdituraăUniversit ţiiă„PetruăMaior”,ăpp.706-712 65. Ledford,ă Jerriă L.ă (2009)ă SEO Search Engine Optimization Bible, 2ndă editionă Indianapolis,ăIN:ăWileyăPublishing 66. Meibauer,ă Jörgă (2006)ă “Implicature”.ă Ină E.ă K.ă Brownă etă ală (eds.),ă Encyclopedia of language and linguistics.ă2ndăeditionăBoston:ăElsevier,ăpp.ă568-581 67. Mclntosh,ă Angusă (1963)ă A new approach to Middle English dialectology.ă Englishă Studiesă44,ăpp.1–11,ăreprintedăinăLassă1969,ăpp.392–403. 270 68. Malmkjær,ăKirstenă (1998).ă“Unită ofătranslation”ă inăMonaăBakerăandăKirstenăMalmkjæră (Eds.),ăRoutledgeăEncyclopedia of Translation Studies.ăLondonăȚăNewăYork:ăRoutledge,ă pp.ă286-288 69. Massion,ăFrançoisă(2011)ă“Localizationăofămachineăsoftware”,ăMultiLingual,ă4/5,ăIdahoă:ă MultiLinguală Computing,ă Inc.ă pp.40-44,ă retrievedă fromă http://www.multilingual.com/downloads/coreFocus119.pdf 70. Mazur,ăIwonaă(2007)ă“TheăMetalanguageăofăLocalization”,ăTargetă(2),ăpp.ă337–357. 71. Mehler,ăAlexander,ăSergeăSharoff,ăMarinaăSantiniă(2010)ăGenres on the Web: Computational Models and Empirical Studies.ăDordrecht:ăSpringer 72. Mendis,ă Patrickă (2007)ă Glocalization: The Human Side of Globalization as If the Washington Consensus MatteredăMorrisville,ăNC:ăLuluăPress 73. Munday,ăJeremyă(2008)ăIntroducing Translation Studies: Theories and Applications,ă2ndă edition.ăNewăYork:ăRoutledge 74. NederveenăPieterse,ăJană(2009).ăGlobalization and culture: Global mélange,ă2ndăeditionă Lanham,ăMD:ăRowmanăȚăLittlefield. 75. Neubert,ăAlbrechtă(1996)ă“TextlinguisticsăofăTranslation:ătheăTextualăApproach”ăină MarilynăGaddisăRoseă(ed.),ăTranslation Horizons- Beyond the Boundaries of Translation Spectrum,ăTranslationăPerspectivesăIX,ăStateăUniversityăofăNewăYorkăatăBinghamton:ă CenterăforăResearchăinăTranslation,ăpp.ă87-106 76. Newmark,ăPeteră(1981)ăApproaches to translation.ăOxford:ăPergamonăPress 77. Newmark,ăPeteră(1988)ăA Textbook of Translation,ăHertfordshire:ăPrenticeăHallă 78. Newmark,ăPeteră(1993)ăParagraphs on Translation,ăClevedon:ăMultilingualăMatters 79. Nida,ăE.A.ă(1959).ă“PrinciplesăofătranslationăexemplifiedăbyăBibleătranslation.”ăInăR.A.ă Broweră(ed.),ăOn translation,ăCambridge:ăHarvardăUniversityăPress,ăpp.ă11-31 80. Nida,ă E.A.ă (2000)ă “Principlesă ofă correspondence”ă ină Lawrenceă Venutiă (ed.)ă The Translation Studies Reader.ăLondonăȚăNewăYork:ăRoutledge,ăpp.141-155 81. Nielsen,ăJakobă(1990)ăHypertext and Hypermedia.ăNewăYork:ăAcademicăPress. 82. Nielsen,ă Jakobă (1996,ă Januaryă 1)ă “Internatională Usabilityă Testing”,ă Nielsen Norman Group,ăretrievedăfromăhttp://www.nngroup.com/articles/international-usability-testing/ 83. Nielsen,ăJakobă(2006,ăAprilă17)ă“F-ShapedăPatternăForăReadingăWebăContent”,ăNielsenă Normană Group,ă retrievedă fromă http://www.nngroup.com/articles/f-shaped-patternreading-web-content/ 271 84. Nielsen,ăJakobă(2007,ăAugustă 20)ă “BannerăBlindness:ăOldăandăNewăFindings”,ă Nielsen Norman Group,ă retrievedă fromă http://www.nngroup.com/articles/banner-blindness-oldand-new-findings/ 85. Nielsen,ăJakob.ă(2008,ăMayă6)ă“HowăLittleăDoăUsersăRead?”ăNielsen Norman Group,ă retrievedăfromăhttp://www.nngroup.com/articles/how-little-do-users-read/ă 86. Nielsen,ă Jakobă (2011,ă Juneă 6)ă “Internatională Usability:ă Bigă Stuffă theă Same,ă Detailsă Differ”,ă Nielsen Norman Group,ă retrievedă fromă http://www.nngroup.com/articles/international-usability-details-differ/ 87. Nord,ăChristianeă(1997)ă Translating as a Purposeful Activity. Functionalist Approaches Explained.ăManchester,ăUK:ăSt.ăJeromeăPublishing. 88. Nord,ăChristianeă(2005)ăText Analysis in Translation: Theory, Methodology, and Didactic Application of a Model for Translation-Oriented Text Analysis,ăAmsterdam:ăRodopi 89. O’Hagan,ăMinakoăandăDavidăAshworthă(2002)ăTranslation-mediated Communication in a Digital World.ăBristol:ăMultilingualăMattersăLtd 90. OțHagan,ăMinakoăandăCarmenăMangironă(2013)ăGame Localization: Translating for the global digital entertainment industry.ăAmsterdam,ăPhiladelphia:ăJohnăBenjamins 91. Pucci,ăTommaso,ăChristianăSimoniăandăLorenzoăZanniă(2012)ă“Countryăofăoriginăeffect,ă brandăimageăandăretailămanagementăforătheăexploitationăofă‘MadeăinăItaly‘ăinăChina”ăină GiuseppeăBertoliăandăRiccardoăRescinitiă(Eds.)ăInternational Marketing and the Country of Origin Effect.ăCheltenham,ăUKăȚăNorthampton,ăMA:ăEdwardăElgarăPublishing 92. Pym,ă Anthonyă (1999)ă “Localizingă Localizationă ină Translator-Trainingă Curricula”.ă retrievedăfromăhttp://usuaris.tinet.cat/apym/on-line/translation/localization.html 93. Pym,ă Anthonyă (2004a),ă The Moving Text. Localization, Translation, and Distribution,ă Amsterdam,ăPhiladelphia:ăJohnăBenjamins 94. Pym,ă Anthonyă (2004b)ă “Localizationă fromă theă Perspectiveă ofă Translationă Studies:ă Overlapsă ină theă Digitală Divide?”ă ă retrievedă fromă http://www.elda.org/en/proj/scalla/SCALLA2004/Pymv2.pdfă 95. Pym,ăA.,ă Perekrestenkoă andă B.ă Starinkă (eds.)ă (2005a)ă “Translationă Technologyă andă itsă teachingă (withă muchă mentionă ofă localization)”.ăTarragona:ă Interculturală Studiesă Group.ă retrievedăfromăhttp://isg.urv.es/publicity/isg/projăects/2002-2005ăbff/index.html 96. Pym,ăAnthonyă(2005b)ă“Localization:ăOnăitsănature,ăvirtuesăandădangers”ăretrievedăfromă http://www.rut.es/~apym/ăon-line/Localizationăbergen.ădocă 272 97. Pym,ăAnthonyă (2009)ă “Allă thingsă toă allă people.ă Onă Nidaă andă involvement.”ă Ină Rodicaă Dimitriuă andă Miriamă Shlesingeră (eds.)ă Translators and Their Readers. In Homage to Eugene A. Nida.ăBrussels:ăLesăEditionsăduăHazard,ăpp.ă317-332. 98. Pym,ă Anthonyă (2010)ă “Websiteă localization”ă The Oxford Handbook of Translation Studies,ă Oxford:ă Oxfordă Universityă Press,ă 2011.ă retrievedă fromă http://usuaris.tinet.cat/apym/on-line/translation/2009_website_localization_feb.pdf 99. Pym,ăAnthonyă(2014)ăExploring Translation Theories, 2ndăedition LondonăȚăNewăYork:ă Routledge 100. Ray,ă Rebeccaă andă Natalyă Kellyă (2010)ă Reaching New Markets through Transcreation,ă Lowell,ă MA:ă Commonă Senseă Advisoryă retrievedă fromă http://www.commonsenseadvisory.com/Portals/_default/Knowledgebase/ArticleImages/1 00330_R_Transcreation_Preview.pdfă 101. Ritzer,ăGeorgeă(2011)ăThe McDonaldization of Society 6,ăSAGEăPublications 102. Robertson,ă Rolandă (2010)ă “Glocalizationă Time-Spaceă andă Homogeneity— Heterogeneity”ă ină Georgeă Ritzeră andă ZeynepăAtalayă (Eds.)ă Readings in Globalization: Key Concepts and Major Debates, Indianapolis,ăIN:ăWileyăPublishing,ăpp.ă334-343 103. Robinson, Douglas (2007) Becoming a Translator 2nd edition London & New York: Routledge 104. Sager,ăJuanăCarlosă(1983)ă“QualityăandăStandards.ăTheăEvaluationăofăTranslations”,ă in C. Picken (ed.) The Translator's Handbook, London: Aslib, pp. 121-128 105. Sanders,ăT.,ăandăSanders,ăJ.ă(2006)ă“TextăandăTextăAnalysis”.ăInăE.ăK.ăBrownăetăală(eds.),ă Encyclopedia of language and linguistics.ă2ndăeditionăBoston:ăElsevier,ăpp.ă597-607 106. Sandrini,ă Peteră (2005)ă “Websiteă Localizationă andă Translation”ă retrievedă fromă http://www.euroconferences.info/proceedings/2005_Proceedings/2005_Sandrini_Peter.p df 107. Sandrini,ă Peteră (2008)ă “Localizationă andă Translation”ă retrievedă fromă http://homepage.uibk.ac.at/~c61302/publik/localiz.pdfă 108. Sanford,ăA.ă(2006).ă“Coherence:ăPsycholinguisticăapproach”.ăInăE.ăK.ăBrownăetăală(eds.),ă Encyclopedia of language and linguistics.ă2ndăeditionăBoston:ăElsevier,ăpp.ă585-591 109. Santini,ăMarina,ăAlexanderăMehlerăandăSergeăSharoff.ă(2010a)ă“RidingătheăRoughăWavesă ofăGenreăonătheăWeb.”ăInăAlexanderăMehler,ăSergeăSharoff,ăMarinaăSantiniă(Eds.)ăGenres on the Web: Computational Models and Empirical Studies.ă Dordrecht:ă Springer,ă pp.ă 332. 273 110. Santini,ă Marinaă (2010b)ă “Cross-Testingă aă Genreă Classificationă Modelă foră theăWeb.”ă Ină Alexanderă Mehler,ă Sergeă Sharoff,ă Marinaă Santiniă (Eds.)ă Genres on the Web: Computational Models and Empirical Studies.ăDordrecht:ăSpringer,ăpp.ă87-128. 111. Saussure,ă Ferdinandă deă (2011)ă Course in General Linguistics, translatedă byă Wadeă Baskin,ăNewăYork:ăColumbiaăUniversityăPress 112. Schäffner,ăChristinaă(2002)ăThe Role of Discourse Analysis for Translation and in Translator Training.ăClevedon:ăMultilingualăMatters 113. Schäler,ă Reinhardă (2002),ă “Reverseă Localization”ă The International Journal of Localisation, Vol.6, Issue 1, Localisation Research Centre, CSIS Dept, University of Limerick, Limerick Ireland, pp. 39-48 114. Schäler,ăReinhardă(2009),ă“Localization”ăinăMonaăBakerăandăGabrielaăSaldanhaă(eds.)ă Routledge Encyclopedia of Translation Studies.ăNewăYork:ăRoutledge,ă2ndăedition,ăpp.ă 157-161 115. Searle,ăJ.ă(1983)ăIntentionality: An Essay in the Philosophy of Mind.ăCambridgeă UniversityăPress.ă 116. Shea,ăDavidă(2005)ă“InterviewăwithăPeterăNewmark,ăpioneeringătheoreticianăinăscientifică translation”,ăPanace@ăVol.ă6,ăIssueă21-22,ăOpenăEducationalăResources,ăpp.ă392-394ă Retrievedăfromăhttp://www.medtrad.org/panacea/IndiceGeneral/n_2122_semblanzas_Shea.pdfă 117. Shelly,ăGaryăB.ăandăMistyăE.ăVermaată(2010)ăDiscovering Computers 2010: Living in a Digital WorldăBoston:ăCengageăLearningă 118. Singh,ăNitishăandăArunăPereiraă(2005)ăThe Culturally Customized Web Site: Customizing Web Sites for the Global Marketplace,ăBurlingtonăUS,ăOxfordăUK:ăElsevier,ăButterworth,ă Heinemann. 119. Sprung, Robert C. (ed.) (2000) Translating Into Success. Cutting-edge strategies for going multilingual in a global age. Amsterdam - Philadelphia: Benjamins. 120. Trosborg,ăAnnaă(2002)ă“DiscourseăAnalysisăasăPartăofăTranslatorăTraining.”ăInăChristinaă Schäffneră(ed.)ăThe Role of Discourse Analysis for Translation and in Translator Training.ăClevedon:ăMultilingualăMattersăLtd,ăpp.ă9-52. 121. Tufi ,ăR.ăIon,ăA.ăCeau uăandăD.ă tef nescuă(2006)ă“Improved lexical alignment by combining multiple reified alignments”. In Proceedings of the 11th Conference of the European Chapter of the Association for Computational Linguistics (EACL2006), pp. 153–160 122. Ulrych. Margherita. (2009) “Translationăandăeditingăasămediatedădiscourse:ăfocusăonă 274 theă recipient.”ă Ină Rodica Dimitriu and Miriam Shlesinger (eds.) Translators and Their Readers. In Homage to Eugene A. Nida, Brussels: Les Editions du Hazard, pp. 219-234. 123. Venuti,ă L.ă (1998).ă “Strategies ofă translation”.ă Ină M.ă Bakeră (ed.), Encyclopedia of translation studies. London & New York: Routledge, pp. 240-244 124. Venuti,ăLawrenceă(1995)ăThe Translator’s Invisibility: A history of translation,ăLondonă andăNewăYork:ăRoutledge 125. Vermeer,ăHansăJ.ă(2000)ă“SkoposăandăCommissionăinăTranslationalăAction”ăinăLawrenceă Venutiă(ed.)ăThe Translation Studies Reader.ăLondonăȚăNewăYork:ăRoutledge,ăpp.ă191202 126. Weaver,ăC.ă(1988)ăReading Process and Practice from socio-linguistics to whole language,ăPonsmouzh,ăNH:ăHeinemannăEducationalăBooks.ă 127. Wells,ăWilliamăD.ă(1965)ă“CommunicatingăwithăChildren.”ăJournalăofăAdvertisingă Researchă(5)ăpp.ă2-14 128. Yule,ăGeorge.ă(2010)ăThe Study of Language,ă4thăed.ăCambridge:ăCambridgeăUniversityă Press 129. Yunker,ăJohnă(2003)ăBeyond borders: web globalization strategiesăBerkeley,ăCA:ăNewă RidersăPublishing 275 Internet sources [1]http://www.commonsenseadvisory.com/default.aspx?Contenttype=ArticleDetȚtabID=64Ț moduleId=392ȚAid=2909ȚPR=PRă [2]http://www.capital.ro/ins-exporturile-si-importurile-romaniei-in-crestere.html [3]https://www.paypalmedia.com/assets/pdf/fact_sheet/PayPal_ModernSpiceRoutes_Report_Final.pdf [4]http://ec.europa.eu/dgs/translation/programmes/emt/index_en.htm [5]http://ec.europa.eu/dgs/translation/programmes/emt/key_documents/emt_competences_tra nslators_en.pdf [6]http://ec.europa.eu/dgs/translation/workwithus/staff/profile/index_en.htm [7]http://ec.europa.eu/dgs/translation/programmes/emt/key_documents/translator_trainer_pro file_en.pdf [8]http://www.edu.ro/index.php/articles/12704 [9] http://www.acclaro.com/assets/files/downloads/whitepapers/lisa_globalization_primer.pdf [10]http://ifk-kurse.fh-flensburg.de/ecolotrain/?L=1 [11]http://www.merriam-webster.com/dictionary/globalization [12]http://dictionary.cambridge.org/dictionary/british/globalization [13] https://www.multilingual.com/downloads/2011RD.pdf [14]ăhttp://msdn.microsoft.com/en-us/library/h6270d0z.aspx [15]ăhttp://download.oracle.com/tech/blaf/specs/ssnls.html [16]ăhttp://www.internetmarketingninjas.com/seo-tools/seo-compare/ă [17]ăhttp://www.multilingual.com/downloads/coreFocus139.pdf [18]http://www.citibank.com/transactionservices/home/sa/cab/meetings/documents/ben_ensor .pdf [19]http://www.ft.com/cms/s/0/d79ffff8-cfb7-11e3-9b2b-00144feabdc0.html#axzz33khifoU3 [20]ăhttp://data.worldbank.org/indicator/PA.NUS.PPPC.RF/countries/1W?display=default [21]http://ico.org.uk/for_organisations/privacy_and_electronic_communications/the_guide/co okies [22]http://www.merriam-webster.com/dictionary/cookie [23]ăhttp://www.apple.com/legal/privacy/en-ww/cookies/ [24]ăhttp://www.multilingual.com/downloads/screenSupp103.pdf [25]ăhttp://www.businessdictionary.com/definition/web-personalization.htmlă 276 [26]http://www.oracle.com/us/products/applications/atg/advanced-personalizationecommerce-333296.pdf [27]ăhttp://searchengineland.com/google-worlds-most-popular-search-engine-148089 [28]ăhttps://support.google.com/webmasters/answer/35769?hl=en [29]http://static.googleusercontent.com/media/www.google.com/ro//webmasters/docs/searchengine-optimization-starter-guide.pdfă [30]ăhttp://www.nngroup.com/ [31]ăhttp://nichcy.org/dissemination/tools/webwriting/reading [32]ăhttp://www.craigbailey.net/content-is-king-by-bill-gates/ă [33]ăhttp://contentmarketinginstitute.com/ [34]http://www.marketingcharts.com/wp/traditional/consumers-no-longer-associating-namebrands-with-higher-quality-36877/ [35]http://www.marketingcharts.com/wp/online/google-search-ads-impact-brand-awareness43404/ă [36]https://www.marketingsherpa.com/data/public/reports/special-reports/SR-A-TacticalApproach-to-Content-Marketing.pdfă [37]http://images.prsoftware.vocus.com/Web/Vocus/ș7B14ac27fd-fcb5-4f16-a5fbdaa04a663b57ș7D_Vocus_-_State_of_Marketing.pdfă [38]ăhttp://www.alexa.com/topsites/countries [39]ăhttp://www.alexa.com/topsitesă [40]http://europa.eu/legislation_summaries/information_society/legislative_framework/l2412 0_en.htmă [41]ăhttps://translate.google.comă [42]ăhttp://www.bing.com/translator/ă [43]ăhttp://www.prestashop.com/ [44]ăhttp://www.prestashop.com/en/translations [45]ăhttp://w3techs.com/technologies/overview/content_language/all [46]ăhttp://ontology.buffalo.edu/smith/varia/snow.html [47]ăhttp://www.copywriting.comăă [48]ăhttp://www.google.com/analytics/ă [49]ăwww.eamt.org [50]ăhttp://blog.lionbridge.com/translation/2013/09/19/rise-machines/ [51]ăhttp://content.lionbridge.com/translation-memory-vendor-use-one/ [53]ăhttp://itranslate4.eu/en/ 277 [54]ăhttp://patentscope.wipo.int/translateUN/translateUn.jsf [55]ăhttp://www.oxforddictionaries.com/us/definition/american_english/semasiology [56]ăhttp://www.oxforddictionaries.com/us/definition/american_english/onomasiology [57]ăhttp://www.oxforddictionaries.com/us/definition/american_english/autumn?q=autumnă [58]ăhttp://www.oxforddictionaries.com/definition/english/voucher [59]http://lege5.ro/Gratuit/ge4tmnru/legea-nr-142-1998-privind-acordarea-tichetelor-de-masa [60]ăhttp://www.afm.ro/main/programe/psipan/2014/comunicat_psipan_2014.pdf [61]ăhttp://www.google.ro/trends/ [62]ăhttps://adwords.google.com/ko/KeywordPlanner/ [63]ăhttp://www.oxforddictionaries.com/definition/english/glocal [64]ăhttp://www.oxforddictionaries.com/definition/english/search-engine?q=search+engine [65]ăhttp://www.merriam-webster.com/dictionary/searchș20engine [66]ăhttp://www.google.com/edu/training/tools/search/level1.html [67]ăhttp://moz.com/beginners-guide-to-seo [68]http://www.searchenginepromotionhelp.com/m/articles/promotion-encyclopedia/searchengine-optimization.php [69]ăhttp://searchengineland.com/seotable [70]ăhttp://searchengineland.com/ [71]ăhttp://moz.com/blog/mission-imposserpble-establishing-clickthrough-rates [72]ăhttp://www.lionbridge.com/files/2013/01/Lionbridge_eBook_WebLocalization2.pdf [73]ăhttp://moz.com/beginners-guide-to-seo/keyword-research [74]ăhttp://www.calculatemarketing.com/blog/techniques/benefits-of-long-tail-keywords/ [75]ăhttp://www.bing.com/toolbox/keywords [76]ăhttp://www.bing.com/webmaster/help/keyword-research-tool-628070b5 [77]ăhttp://www.copyscape.com/ [78]ăhttp://www.ldoceonline.com/dictionary/ [79]ăhttp://dictionary.cambridge.org [80]ăhttp://oald8.oxfordlearnersdictionaries.com/dictionary/register_2 [81] http://geert-hofstede.com/dimensions.html [82] http://www.intersperience.com/article_more.asp?art_id=34 [83]http://www.romania-insider.com/explicit-implicit-communication-and-the-barbarianreflex-where-do-romanians-fit/36389/ [84] http://www.indexmundi.com/romania/demographics_profile.html [85] http://www.internetworldstats.com/europa.htm 278 [86]http://searchengineland.com/googles-matt-cutts-dont-duplicate-your-meta-descriptions177706 [87]http://www.linguarama.com/ps/marketing-themed-english/the-language-ofadvertising.htm [88] http://www.columbia.edu/~ss957/articles/Choice_is_Demotivating.pdf [89] http://www.interbrand.com/en/best-global-brands/2013/top-100-list-view.aspx [90] bytelevel.com [91] http://www.alexa.com/topsites/countries/RO [92] http://tommy.ismy.name/and-these-are-my-thoughts/14-types-of-online-marketing/ [93] http://www.eurosport.com/ [94] http://tv.eurosport.ro/ [95] http://everything.yahoo.com/world/ [96]http://www.marketwatch.ro/articol/5942/Ce_tehnologie_foloseste_eMAG_o_afacere_onl ine_cu_potential_de_100_de_milioane_de_euro_in_2010/ [97][http://www.zf.ro/zf-24/grupul-naspers-si-iulian-stanciu-raman-singurii-actionari-aiemag-12733708 [98] http://www.capital.ro/emag-lanseaza-o-aplicatie-pentru-dispozitive-mobile.html [99]http://www.zf.ro/business-hi-tech/talpes-bitdefender-veniturile-grupului-au-crescut-cu26-anul-trecut-tinta-pentru-2014-plus-30-13109726 [100]http://www.marketwatch.com/story/bitdefender-enhances-privacy-security-with-applock-for-android-2014-05-20 [101] http://faostat.fao.org/site/636/DesktopDefault.aspx?PageID=636#ancor [102] http://www.wineromania.com/index.jsp?page=apev [103] http://www.internetmarketingninjas.com/seo-tools/keyword-density/ [104] http://ubersuggest.org/ [105] http://www.oxforddictionaries.com/definition/english/apartment 279 AppendixăA: List of BBC call to action words you, your, we, world, best, more, good, better, new, taste, people, our, first, like, don't, most, only, quality, great and choice Source: http://virtuallinguist.typepad.com/the_virtual_linguist/2008/11/the-twenty-mostpopular-words-in-advertising-slogans.html Retrieved in March 2014. 280 AppendixăB: WilliamăWells’ălistăofăcallătoăactionăwords suddenly miracle now magic announcing offer introducing quick improvement easy amazing wanted sensational challenge remarkable compare revolutionary bargain startling hurry Cited on http://www.frankwbaker.com/persuasive.html Retrieved in March 2014. 281 AppendixăC: Most common adjectives and verbs used in advertising The twenty most frequent verbs: The twenty most frequent adjectives: 1.ănew 1. make 2.ăgood/better/best 2. get 3.ăfree 3. give 4.ăfresh 4. have 5.ădelicious 5. see 6.ăfull 6. buy 6.ăsure 7. come 8.ăclean 8. go 8.ăwonderful 9. know 10.ăspecial 10. keep 11.ăcrisp 10. look 12.ăfine 12. need 13.ăbig 13. love 14.ăgreat 14. use 15.ăreal 15. feel 16.ăeasy 15. like 16.ăbright 17. choose 18.ăextra 18. take 18.ăsafe 19. start 20.ărich 20. taste Source: http://www.linguarama.com/ps/marketing-themed-english/the-language-ofadvertising.htm Retrieved in March 2014. 282 AppendixăD: Google.com SERP, listing 100 results (unedited) for the car insurance keyword. Retrieved in March 2014. 1. CompareăCheap Car Insurance QuotesăatăGocompare.com™ www.gocompare.com/car-insurance/ Gocompare.com See if you could save by comparing cheap car insurance quotes from over 120 insurance brands at Gocompare.com today! Women's car insurance - Young drivers - Car insurance guides - Gap insurance 2. Auto Insurance –ăStateăFarm®ă-ăStateăFarmăInsurance https://www.statefarm.com/insurance/auto State Farm Insurance Auto insurance quotes from State Farm are fast and easy, with discounts calculated automatically. Compare your options and customize coverage to fit your ... 3. CompareăCheap Car Insurance Quotesă-ăMoneySuperMarket www.moneysupermarket.com/car-insurance/ Moneysupermarket.com - 2,255 reviews You could save up to £227 on your car insurance with MoneySuperMarket. We compare quotes from 139+ companies, so see if you could save. Car Insurance for Women - What are the Cheapest Cars ... - Young Drivers - Guides 4. Auto Insurance -ăProgressive www.progressive.com/auto/ Progressive Corporation Get the coverage you need, for the price you want. Start a car insurance quote today with the details you know offhand. 5. Car Insurance | Auto Insurance Quotesă|ăFarmersăInsurance www.farmers.com/auto/ Farmers Group, Inc. Get a car insurance quote online or contact a local Farmers agent to learn about ourauto insurance discounts. 6. Car Insurance -ăCheap Car Insurance Quotesă-ăTescoăBank www.tescobank.com/insurance/carins/index.html Tesco Bank Get a Car Insurance quote online. Guaranteed discount available to Clubcard customers plus no claims discount and 24/7, UK based helpline. 7. Car Insurance -ăLiverpoolăVictoria www.lv.com/car-insurance Rating: - 376 reviews 10% of new LV= Liverpool Victoria customers with 75% no claim discount paid £169 or less (May-June '14). 75% NCD with five years claim free. 5 Star rated car ... 8. Car Insurance -ăCompareăCheapăQuotesăatăConfused.com www.confused.com/car-insurance Confused.com - 356 reviews 1 MILLION GIVEAWAY! Buy car insurance and enter our comp at Confused.com for your chance to win £1 million pounds! Compare our cheapest quotes today. 283 9. Car insurance |ăGetăyour car insurance quoteă|ăAA www.theaa.com › Insurance The Automobile Association - Review by Defaqto Get award-winning car insurance from a name you can trust. New AA car insurancecustomers get free breakdown cover for a year, plus AA Rewards discounts. 10. Allianz Car Insurance - Car Insurance QuotesăAustralia www.allianz.com.au/car-insurance/ Allianz Save Up To 15%* When You Buy Online Car Insurance From Allianz Australia! Get A Fast Online Car Insurance Quote Today At Allianz. 11. Car Insurance QuotesăOnlineă|ăswiftcover.com www.swiftcover.com/carinsurance/ Swiftcover Voted best value for money car insurer by lovemoney.com readers for 3 years in a row! Get a quote with 100% online car insurance. Get a life, get Swiftcovered. 12. Car Insurance |ăCheap Car Insurance Quotesă|ăRAC www.rac.co.uk › Insurance RAC Limited At the RAC we're committed to finding our customers a great deal on their car insurance. We search a panel of carefully selected insurers to give you our best ... 13. Car Insurance,ăAffordable Auto Insurance |ăLibertyăMutual www.libertymutual.com › ... › Auto Insurance Products Liberty Mutual When you buy car insurance online, you're in the driver's seat. Here's a review of the basics you need to know to get an online quote and purchase a policy. 14. Cheap Car Insurance Quoteăinăjustă2ăminsă-ăBudgetăDirect™ www.budgetdirect.com.au/ Budget Direct Australia's CHEAPEST Car Insurance. Awarded by Money magazine 'Best of the Best'. Save 20% on Comprehensive Car Insurance bought online! Quote now! 15. Car insurance fromă£191ă-ăCheapăonlineăquotesă-ăAviva www.aviva.co.uk › Insurance › Motor insurance Aviva - 5,173 reviews Car insurance from £191 and get an extra 10% off when you buy online ... We're the only car insurer to share our customer claim reviews on our website, so now ... 16. YouăCouldăSaveăOveră$500ăon Car Insurance -ăGetăa ... -ăGeico https://www.geico.com/getaquote/auto/ GEICO Get a free car insurance quote in 15 minutes or less and see how much you could save today on auto insurance. Start your free car insurance quote now. 17. EsuranceăOnline Car Insurance —ăGetăYourăQuoteă&ăSave ... www.esurance.com/ Esurance See how much you can save on reliable, affordable car insurance. Get your free quote online or over the phone and compare auto insurance rates in minutes. 18. Auto Insurance & Car Insurance Quotes—Allstate www.allstate.com/auto-insurance.aspx 284 Allstate Customized car insurance coverage options, discounts galore and amazing features. Get a free online quote and speak to a dedicated Allstate agent. 19. Vehicle insurance -ăWikipedia,ătheăfreeăencyclopedia en.wikipedia.org/wiki/Vehicle_insurance Wikipedia Vehicle insurance (also known as , GAP insurance, car insurance, or motor insurance) is insurance purchased for cars, trucks, motorcycles, and other road ... 20. CompareăCheap Car Insurance Quotesă|ăCompareăTheăMarket www.comparethemarket.com/car-insurance/ comparethemarket.com Buy car insurance and get your free exclusive Meerkat toy! Compare car insurancenow to save up to £230! 21. iSelectă- Car Insurance |ăCompare Car Insurance Quotes www.iselect.com.au/car/ iSelect Compare and save on Car Insurance from a range of competitive options matched to your needs and budget. 22. Car Insurance & Car Insurance Quotesă-ăNRMAăInsurance www.nrma.com.au/car-insurance Get a Car Insurance quote online today from NRMA Insurance, SGIO or SGIC. Great value cover to give you peace of mind. 23. Car Insurance -ăOnlineăinsurancesăUKă|ăAdmiralăUKăInsurance www.admiral.com/ Admiral Insurance Compare Admiral car insurance & home insurance quotes. Great insurance prices crammed with benefits. Insuring 2+ cars? Get MultiCar discounts on ALL the ... 24. Cheap Car Insurance -ăMoneyăSavingăExpert www.moneysavingexpert.com › Insurance MoneySavingExpert.com Combine comparison sites to compare 100s of cheap car insurance quotes, and then get hidden cashback deals too. 25. Car Insurance -ăQBEăInsuranceăAustralia www.qbe.com.au › Home › Personal Insurance QBE Australia 20% off online car insurance quotes. Comprehensive, budget, third party, nominated driver and more options with QBE Insurance Australia. 26. Auto Insurance QuotesăfromăErieăInsurance www.erieinsurance.com/auto-insurance Erie Insurance Group Since 1925, people have been choosing ERIE to insure their cars and other vehicles because of our fair and competitive prices, but ERIE's much more. We also ... 27. Car Insurance -ăAAMI www.aami.com.au/car-insurance Get an Online Comprehensive or Third Party Car Insurance Quote from AAMI. Receive $100 off a New Comprehensive Policy Today. 28. Upătoă80%ăoff Car Insurance |ăChurchillăInsuranceăUK www.churchill.com/car-insurance 285 Churchill Insurance Company Get up to 80% off car insurance with Churchill. Valid for new customers with 8+ years' No Claim Discount. 29. Cheap Car Insurance Informationă-ăNoăNonsenseăInsurance ... www.nononsense.ie/cheap-car-insurance/ No Nonsense Insurance Simply choose the level of cover you want and we have a Ready Made package for you! Get an instant quick quote online right now! 30. Car Insurance: Auto Insurance QuotesăinăCanadaă|ăTD ... www.tdinsurance.com/products.../auto-insurance/tdin-auto-index.jsp TD Insurance offers auto insurance coverage for all of your vehicles at affordable prices. Get a free car insurance quote online today! 31. Just Car Insurance: Car Insurance Quotesă-ăOnlineăRates ... www.justcarinsurance.com.au/ Just Car Insurance offers specialist car insurance for drivers and owners of high performance, imported and modified cars in Australia. Get a quote online now! 32. Car Insurance -ăAAăIreland www.theaa.ie/AA/Insurance/Car-Insurance.aspx AA Car Insurance Ireland gives you great perks when you take our motor ... Our aim is to keep you moving so we'll give you a courtesy car while yours is out of ... 33. Car Insurance fromăTheăCo-operativeăInsurance www.co-operativeinsurance.co.uk/carinsura... The Co-operative Insurance Looking for cheap car insurance? Get a motor insurance quote from The Co-operative & save up to 70% on your car insurance when you buy online today. 34. Car Insurance QuotesăOnlineăNSWăAustraliaă-ăMotorăVehicle ... www.gio.com.au/car-insurance GIO General Be confident your car is protected with GIO's great range of car insurance cover. With a choice of benefits and optional extras, we pride ourselves on providing ... 35. 6ăReasonsătoăChooseăNationwideăfor Car Insurance www.nationwide.com/car-insura... Nationwide Mutual Insurance Company Protect what matters most to you with car insurance from Nationwide. We provide online auto insurance rates and the coverage you need to keep safe on the ... 36. Car Insurance -ăCompare Car Insurance QuotesăandăSave www.carinsurance.com/ Finding the right car insurance just got easier. CarInsurance.com lets you comparecar insurance quotes side by side so you can choose the best quote instantly. 37. Cheap Car Insurance:ăLookingăforătheăCheapestăCar ... www.progressiveonline.com.au/cheap-car-insurance.aspx Looking for cheap car insurance in Australia? 100% online car insurance means you save without sacrifice. 38. Car Insurance Quotesă|ăLaăCapitale www.lacapitale.com › Home › Individuals › Insurance Discover the benefits of insurance for your car/automobile and take advantage of La Capitale's rapid online insurance service to get a quote. 39. Car Insurance &ăMotorăInsuranceăQuotesăOnlineă-ăFBD ... 286 www.fbd.ie/car-insurance/ With Online discount on all new car insurance policies, get a QUICK Car InsuranceQuotes Online. Look For The Car Insurance Company That's Right For You. 40. Car Insurance Irelandă–ăcheapăquotesăonline,ăsaveă20%ăwith ... www.aviva.ie/online/driving/car-insurance/ Aviva Excellent car insurance with cheap quotes and great discounts for safe and experienced drivers. Get 8 weeks free and 20% off when you buy online with Aviva ... 41. Car Insurance -ăSuncorpăInsurance www.suncorp.com.au/insurance/car Suncorp Group Car Insurance with more flexibility. With the choice of different cover options, Suncorp Insurance gives you the flexibility to choose the level of cover that is best ... 42. Car Insurance Irelandă|ăAIG.ieă-ăAIGăEuropeăLimited www.aig.ie/car-insurance_3192_508433.h... American International Group As well as offering you low cost car insurance from o ly €260**, we'll also give you a 10% online discount when you quote with us online or up to 25% with ... 43. Car insurance -ăGetăyourăquickăquoteăonlineăatăState.co.nz www.state.co.nz/personal-insurance/car-insurance/.../car-insurance.aspx Compare State's three most popular car insurance policies for NZ in a simple to use comparison table. Read policy details or get a free quote now. 44. Car Insurance -ăAsdaăMoney money.asda.com/car-insurance/ Asda Stores Limited 5 Star Cover Car Insurance at great value. Get 5 star quality and great value with breakdown and legal cover as standard – ASDA Money. 45. Car Insurance Irelandă-ăMotorăQuotesăOnlineă-ăLiberty ... www.libertyinsurance.ie/car-insurance/ Liberty Insurance Liberty Insurance offer competitive Car Insurance rates with unrivalled service and excellent benefits. Purchase your Car Insurance o l e d s ve up o €100! 46. Car Insurance -ăCoverăForăYourăVehicleă|ăPostăOffice www.postoffice.co.uk/car-insurance Post Office Ltd Buy a tailored car insurance quote with competitive benefits including free breakdown cover worth £49 for the first 12 months. Get an online quote today! 47. Car Insurance -ăCommBank https://www.commbank.com.au/.../insurance/car-in... Commonwealth Bank We have three levels of car insurance to choose from: Comprehensive, Third Party Property Damage, Fire and Theft, and Third Party Property Damage. 48. Car Insurance Quotesă-ăCompare Car Insurance Rates ... www.dmv.org/car-insurance.php The Unofficial Guide to the DMV Get car insurance quotes online fast, from multiple providers. Buy car insuranceonline. Fill out quotes, compare and save. 49. esure www.esure.com/ 287 Esure Home, Travel, Pet and Car Insurance from esure. Excellent service and competitive prices. Get an insurance quote online in minutes. 50. Car Insurance |ăUSA.gov www.usa.gov › ... › Car Buying and Maintenance USA.gov The requirements for auto insurance vary from state to state. Check with your state insurance regulator to learn more about individual requirements, as well as ... 51. 21stăCentury Auto Insurance: Car Insurance QuotesăOnline www.21st.com/ Get a car insurance quote free online from 21st Century Insurance. Find advice on the right auto insurance policy for your vehicle and coverage needs. 52. Car Insurance -ăMoneyăEssentials,ăLessonă22ă-ăCNNăMoney money.cnn.com/magazines/moneymag/money101/lesson22/ CNNMoney Money's guide to car insurance shows you how to understand your auto coverage and picking an insurer. 53. ReadăMoreă-ăAMIă::ăProductsă:: Car insurance –ăGetăanăonline ... www.ami.co.nz/products/car-insurance/ AMI Insurance AMI car insurance provides cover for private cars, vans and small trucks. Choose from three levels of car insurance cover. Visit AMI.co.nz to get a quote. 54. Car Insurance |ăOveră50săInsuranceă|ăGetăaăQuoteăfromăSaga ... www.saga.co.uk › Insurance Saga Group As the leading specialist over 50s car insurance provider, Saga aim to offer excellent value for money without compromising on cover. Find out more. 55. Car -ăAllianz Insurance www.allianz.ie/Car-Insurance/ Allianz Get 14% OFF a new car insurance policy online! Allianz Ireland offers you competitive quotes on your Car Insurance Online! 56. Zurich Car Insurance Irelandă| Car Insurance Quotes www.zurichinsurance.ie/car-insurance/ Why Zurich Car Insurance? More and more people are moving their car insuranceto Zurich to avail of our award winning customer service and these ... 57. Kwik-Fit Insurance:ăOfferingăCheap Car,ăHome,ăBikeă&ăVan ... www.kwik-fitinsurance.co.uk/ Provides cover for all leading UK brands including Ford, Vauxhall and Peugeot. 58. Car Insurance QuotesăOnlineă|ăTheăGeneral Auto Insurance www.thegeneral.com/ The General The General Car Insurance offers online auto insurance quotes for most drivers, even if you need high risk insurance. Try our Speedy Quote now, and receive ... 59. MercuryăInsurance: Auto Insurance & Car Insurance Quotes www.mercuryinsurance.com/ Mercury Insurance Group 288 Auto Insurance and Car Insurance Quotes from Mercury Insurance. Find Local Agents or Discover Our Low Rates on Auto Insurance, Car Insurance & More! 60. Car Insurance -ăInsuranceăBureauăofăCanada www.ibc.ca/en/car_insurance/ Insurance Bureau of Canada Car Insurance - Whenever you get behind the wheel of a car, it is possible that you may cause damage to other people's property, or injure – or even kill ... 61. Car insurance optionsă-ăBupa www.bupa.com.au › Other insurance Bupa We have a great range of Bupa car insurance options to suit your needs. You can choose from three levels of cover: Comprehensive Cover, Third Party Property ... 62. Cheap Car Insurance |ăMotorăInsuranceăIrelandă|ăOneăDirect www.onedirect.ie › Home › General Insurance OneDirect.ie. Your One Stop Car Insurance Website. Great cover at a great price! With competitive prices that include lots of great benefits, One Direct offers ... 63. HowStuffWorksă"How Car Insurance Works" auto.howstuffworks.com/buying-selling/car-insurance.htm Car insurance shopping can be an intimidating task. Learn about the various types of car insurance and how to find the best car insurance rates. 64. Car Insurance Quotesă|GetăYour Vehicle Insured withăiWYZE www.oldmutual.co.za/personal/insurance/car-insurance.aspx Old Mutual For affordable, reliable car insurance in South Africa from Old Mutual iWYZE, call 0860 93 94 93, leave your details & we'll call you or get a quick quote online. 65. Car Insurance QuotesăUKă|ăNectarădiscountă|Sainsbury'săBank www.sainsburysbank.co.uk/insuring/car-insurance.shtml Sainsbury's Bank 25% off car insurance for Sainsbury's shoppers with a Nectar card. Introductory offer for new customers. Minimum premiums apply. 66. Car Insurance -ăGetăa Car Insurance Quoteă|ăAXAăUK www.axa.co.uk/insurance/personal/car/ AXA car insurance. Click here to get an instant quote from AXA car insurance and see how much you could save today! 67. Car Insurance -ăRBCăInsurance www.rbcinsurance.com/car-insurance/ RBC Insurance Services Inc. Call 1-877-749-7224 or request a car insurance quote online today. RBC Insurance offers affordable car insurance coverage for all your personal vehicles. 68. Car Insurance |ăVirginăMoney virginmoney.com.au › Insurance Save 20% off Virgin Car Insurance Price Saver when you get a comprehensive quote and buy online today! Park your car insurance in a better place with Virgin ... 69. SafecoăInsurance:ăSafeco Auto Insurance, Car Insurance ... www.safeco.com/ Safeco Insurance Safeco Insurance offers car insurance, home insurance and other personal insurance through independent insurance agents. Get free auto and home insurance ... 289 70. Car Insurance -ăBhartiăAXAăGeneralăInsurance www.bharti-axagi.co.in/car-insurance - 14,718 reviews Bharti AXA offers car insurance online. Buy or renew car insurance policy and get NCB up to 50%, PA cover and Roadside Assistance cover. Get a free car ... 71. Car Insurance -ăTOWERăInsurance www.tower.co.nz › Insurance Tower Insurance Get car insurance online for your vehicle with TOWER. Choose from a range of cover for auto insurance and receive a competitive quote. Call us now on 0800 ... 72. Car insurance |ăMoneySmartăbyăASIC www.moneysmart.gov.au › Managing your money › Insurance Describes the different types of car insurance available and how to choose the right one for your needs. 73. DirectăGeneral Auto Insurance: Car Insurance |ăMotorcycle ... www.directgeneral.com/ Get a free Direct Auto Insurance quote online or call 1-877-GO-DIRECT to save money on your car insurance and motorcycle insurance. 74. Compare Car Insurance Quotes.ăGetăVehicleăCoveră|ăHippo https://www.hippo.co.za/car-insurance-quote/ Rather enjoy peace of mind by getting car insurance that will cover you in these situations. Fill in your details below to instantly compare multiple competitive car ... 75. Auto Insurance fromăMetLifeăAutoă&ăHome® https://www.metlife.com/individual/insurance/auto-insurance/ MetLife Inc Auto insurance from MetLife Auto & Home offers coverage that protects you from financial losses. Get a car insurance quote online or call 1-888-675-6610! 76. Comprehensive Car Insurance |ăOnlineăQuotesăfromăApia www.apia.com.au/car-insurance Save 5% online for Comprehensive Car Insurance with Apia, the leading insurer for Australian Seniors over 50 that rewards experience - get a quote now. 77. Car Insurance Comparisonă| Car Insurance Compareă|ăHBF https://www.hbf.com.au/...insurance/car-insurance/cove... HBF Health Fund Get a car insurance rate in minutes with HBF Australia. Compare insurance quotes and find the best deal online now! 78. Car Insurance| Car Insurance Renewală|ăMotorăInsurance www.bajajallianz.com/...insurance/car-insura... Bajaj Allianz Life Insurance - 14,464 reviews Buy or renew car insurance policies online. Buy car insurance policy in easy steps. Get 24x7 spot assistance cover with Bajaj Allianz Car Insurance policy. 79. Car Insurance Irelandă|ăCheap Car Insurance https://www.25plus.ie/ Looking for cheap Irish car insurance? We specialise in cheap car insuranceIreland for over 25s. Complete our Irish car Insurance Online Form and you will get ... 80. ElephantăNo-fuss Car Insurance |ăElephantăUKăInsurance www.elephant.co.uk/ 290 elephant.co.uk Specialise in providing cheap motor car insurance - customers can apply for a quote online. 81. LloydsăBankă–ăUKăInsuranceă– Car Insurance www.lloydsbank.com/insurance/home/car.asp Get an online car insurance quote with Lloyds Bank Car Insurance. Free added benefits you won't find on price comparison sites. 82. Car Insurance -ăGet Car Insurance QuotesăOnlineă-ăDirectăLine www.directline.com/car-insurance Direct Line 20% of consumers could save up to £269. Get an introductory offer of 10% discount when you buy online. Take the direct line to see how much you could save. 83. Car insurance forăWomenă-ăDiamondăUK www.diamond.co.uk/ Get car insurance specifically shaped for women at diamond.co.uk. Visit Diamond for car insurance, home insurance & break down cover for women. 84. AISă- Auto Insurance Specialistsă| Auto Insurance https://www.aisinsurance.com/ Get competitive quotes from multiple Auto insurance companies Instantly. New AIS customers save an average of $600 per year. 85. Car Insurance -ăCompareă&ăGetăAăQuoteă-ăAustraliaăPost auspost.com.au › Money & insurance › Insurance Australia Post Australia Post offers great value on a range of car insurance products. From Gold Comprehensive to Third Party Property, we have a policy to suit your needs. 86. Car insurance -ăuSwitch.com www.uswitch.com/car-insurance/ uSwitch Switching your car insurance with uSwitch is simple and speedy - try our car insurance comparison tool and you could save money in just a couple of minutes. 87. Auto Insurance -ăYourăMoneyăGuidesă-ăNYTimes.com topics.nytimes.com/your.../insurance/auto-insuranc... The New York Times Articles and videos from The NY Times on auto insurance, including information on liability, the cost of car insurance and comparing automobile or vehicle ... 88. Vehicle insurance -ăGOV.UK https://www.gov.uk/vehicle-insurance/overview United Kingdom Jun 27, 2014 - You must have motor insurance to drive your vehicle on UK roads. ... causing damage or injury to any other person, vehicle, animal or property. 89. Infinity Auto Insurance savesăyouămoneyăon car insurance ... https://www.infinityauto.com/ Infinity Property & Casualty Corporation Infinity Insurance provides low cost car insurance. We deliver first-class customer service through our agencies, web services and call center. Infinity consistently ... 90. Car Insurance Quotesă|ăCompareătheăMarket www.comparethemarket.com.au/car-insurance/ comparethemarket.com 291 Looking for car insurance? Compare the Market lets you compare affordable Car Insurance quotes from some of Australia's trusted online car insurance brands. 91. Auto Insurance Quotesă| Car Insurance Ratesă|ăTravelers ... www.travelers.com › For Individuals The Travelers Companies Auto insurance policies protect you and your loved ones. Check out our rates and get a car insurance quote online from Travelers Insurance. 92. Car Insurance -ăICICIăBank www.icicibank.com/.../insurance/...insurance/...insurance/car-... ICICI Bank A comprehensive Car Insurance policy for your car that keeps it secure against damage caused by natural and man-made calamities, including acts of terrorism. 93. Car insurance |ăMoneyă|ăTheăGuardian www.theguardian.com/money/car-insurance The Guardian Latest news and comment on Car insurance from the Guardian. 94. Compare Car Insurance Ratesă|ăNerdWalletăFinance www.nerdwallet.com/insurance/compare-car-insurance-r... NerdWallet.com NerdWallet's rate comparison tool allows users to compare personalized auto insurance quotes from major carriers in every state for free. No phone calls or ... 95. Auto Insurance - Auto Insurance Guideă-ăPortal.state.pa.us www.portal.state.pa.us/portal/.../auto_insurance/.../auto_insurance.../536... Simply stated, automobile insurance is a contract between you and your insurance company that protects you against financial loss if you are in an accident. 96. Car -ăIntact Insurance www.intact.ca/car-insurance Types of car insurance coverage. No matter what kind of vehicle you are insuring you can choose various levels of coverage, depending on how much ... 97. Compare Car Insurance |ăOnlineăQuotes,ăReviews,ăRatings car-insurance.findthebest.com/ Listings 1 - 10 of 189 - Find and compare the best auto insurance providers based on expert ratings, credit ratings, financial strength, sample premiums, user ... 98. Car Insurance -ăNationwide www.nationwide.co.uk/.../insurance/car-insu... Nationwide Building Society Car insurance. No claims in 5 years? Get 75% no claim discount. Nationwide car insurance is provided by Liverpool Victoria Insurance Company Limited (LV=). 99. FreeăOnline Car Insurance Quotesă& Auto Insurance Advice autos.aol.com › Hot Topics AOL Get free car insurance quotes and auto insurance advice at AOL Autos. 100. Auto insurance - Car insurance |ăDesjardins www.desjardins.com/ca/.../insurance/car-insurance/ Desjardins Group Desjardins General Insurance has made-to-measure car insurance to suit your ... Member advantage: $50 annual rebate per insured auto and free Roadside ... 292 Why this ad? 1. Used Car ForăSaleă-ăHugeăSelectionă&ăAmazingăPricesă Adwww.audichantilly.com/ Buy Your Next Used Car Here! Searches related to car insurance cheap car insurance the general car insurance car insurance quotes car insurance companies car insurance comparison progressive car insurance car insurance calculator car insurance reviews 293 AppendixăE: Google.ro SERP, listing 100 results (unedited) for the asigurari auto keyword. Retrieved in March 2014. 1. AsigurariăAuto 201 - Cele m Adasigurariauto r o o o o l e s ro C ef e asigurariăauto 0372 100 100 Comanda online asigurari auto 2014 Broker de asigurare Comanda RCA Online Plata la livrare Livrare gratuita 2. Asigurare RCA Auto Ieftină-ărcamaiieftin.roă Ad l ul or-r r m ef ro 20% Reducere La Toti Asiguratorii. Calculeaza Cel Mai ieftin RCA 2014 3. Asigurare Lo u Adwww o h er ro C s o o o O l e - o h er ro e Asigura-ti bunurile & locuinta cu cel mai complet pachet.Vezi oferta! Daune Cumpara acum Contact Search Results 1. RCAăieftină2014ă-ă2015ă|ăComparatorăpreturiămiciăonline ... www.24rca.ro/ ... polita la domiciliu! Aici compari online tarife asigurari RCA de la toti asiguratorii cu plata prin card. ... Oferte asigurari auto RCA online ieftine. Despre 24RCA ... Vezi tarife asigurari RCA 2014 - Asigurari de bunuri - Asigurari de persoane 2. RCAă2014. Asigurari RCAăieftinăonlineă-ăeMAG.ro www.emag.ro/asigurari-rca Compara ofertele de asigurari RCA 2014 online si alege-o pe cea mai buna! Iti livram asigurarea RCA gratuit, oriunde in tara! Politele RCA sunt emise automat. 3. Asigurari-auto-o l e www.asigurari-auto-online.ro/ Asigurarile auto sunt necesare oricarui conducator care doreste sa se stie in siguranta. In afara de acest lucru, asigurarile auto protejeaza si atunci cand ... 4. Asigurariăauto ef e- www.rca-ieftin.ro/asigurari-auto-ieftine C Ief Mai mult, in functie de cerintele si necesitatile noastre, putem gasi printre ofertele asiguratorilor polite de asigurari auto. In principal, companiile de asigurare se ... 5. Asigurare C - Asigurari 2 www.asigurari24.com/asigurare-rca/ Cumpara o asigurare rca ieftina de la brokerul Asigurari 24. ... persoanele din a caror vina s-a produs accidentul auto nu vor fi despagubite; aceasta polita va ... 6. Cascoăieftină|ăCalculatorăonlineăpentruăCASCOăsiăRCA 294 www.casco-ieftin.ro/ ... sa afli care sunt reducerile de care poti beneficia la incheierea unei asigurari ... faci in caz de Dauna; Cum platesc asiguratorii Reparatiile auto; Ce este si de ... 7. C rp s - AsigurariăAuto www.carpaticaasig.ro/Asigurari-Auto.aspx Asigurari Auto s ur re Obl or e de ă spu dere C v lă Auto - Rca. Produsul de asigurare RCA este dedicat tuturor detinatorilor de autovehicule. In urma ... 8. 2 C l ul or C - EU OINS www.euroins.ro/calculator-rca.html ASIGURARI INTELIGENTE. Engleza EN. -Euroins insurance ... Calculator asigurarecalatorie · Calculator asigurare locuinta ... CNP/CUI. Serie sasiu. Marca auto. 9. www burs C Ief 2013 - Ce m s ur r lor ro › RCA asigurare C ef Asigurarea RCA - este o asigurare prin care persoanele prejudiciate in urma unui accident auto, produs din vina conducatorului auto asigurat, primesc ... 10. RCAăieftină|ăRCAăONLINEă2014ă|ăRCA AUTO | Asigurari ... www.onerca.ro/ Pretul cel mai mic la asigurare RCA 2014 si livrare gratuita a politei RCA in toata tara.Comanda online acum o asigurare rca ieftina si beneficiezi pe loc de ... 11. Asigurariăauto - s r s ur r https://www.astrasig.ro/ro/asigurari/.../asigurari-auto Asigurare de op, omplexă ş flex b lă pe ru o e pur le de u oveh ule Pro e e omple ă ş perso l z ă pr pos b l e le er u or oper r mul ple ... 12. RCAăieftină| Asigurare RCAăieftinăonlineă2014ă|ăCalculator ... www.romasig.ro/ Economiseste timp si bani si incheie o asigurare rca ieftin online.Compara ... Doresti sa economisesti bani si timp la incheierea asigurarii auto obligatorii? 13. www burs C l ul or C - Com s ur r lor ro › RCA d o l e s ur re C Asigurarea de Raspundere Civila Auto - denumita si RCA - este o asigurare prin care persoanele prejudiciate in urma unui accident auto, produs din vina ... 14. s ur re www.asigurareamasinilor.ro/ s lor asigurari,masini,bonus,malus,bonus malus,legislatie,carte verde,rca,asigurarea de raspundere civila auto,centre tractari auto,asiguratori RCA. 15. Asigurariăauto C www.asigurari-auto-rca.ro/ Aici gasesti totul despre asigurarile auto RCA din Romania! In plus, cea mai buna oferta si parerea celorlalti vizitatori! 16. AsigurariăAllianz-Tiriac: asigurariăauto,ăviata,ălocuintaăsi ... https://www.allianztiriac.ro/ Allianz-Tiriac Asigurari - omp um rul 1 pe p s ur r lor, pro e e financiara prin ... Asigurari Allianz-Tiriac: auto, locuinta, viata, calatorii, generale. 17. Asigurareăauto - ll z D re https://www.allianzdirect.ro/asigurare/asigurare-auto.ht... 295 A description for this result is not available because of this site's robots.txt – learn more. 18. Auto - Ge er l www.generali.ro/asigurari/persoane-fizice/auto Generali Asigurari: Asigurari Auto - Casco, RCA si Carte Verde. Asistenta rutiere si a persoanelor transportate. 19. AsigurariăAuto O l e - pre ur s ofer e - O o Broker www.ottobroker.ro/asigurari-auto.html Alegeti o asigurare auto online si beneficiati de reduceri de preturi. Otto Broker oferaasigurari auto personalizare, pentru orice autoturism din Romania. 20. RapidAsig.md asigurari onlineăRCAăCASCOăCarteăVerde ... rapidasig.md/ Asigur ri online: RCA, CASCO, Carte VERDE, Asigurare ed lă pes e ho re Doi parlamentari liberal-demo r er l ber l z re p e e s urăr lor auto ş ... 21. Asigur riăauto | C y I sur cityinsurance.ro/persoane-fizice/asigurari-auto/ Pre um b e ş , ex s ă două pur mpor răspu dere v lă ş C SCO L C y I sur 22. e e de asigur riăauto. Cea obligatorie de e, be ef ez de mbele pur ... Group m Asigurariăauto www.groupama.ro/.../asigurari-auto-@/index.jspz?id... Asigurarea auto f ul vă C SCO I de e î r f se po î âmpl î or e l pă Ev ă-le cu asigurarea auto CASCO de la Groupama Asigur ri. 23. RCAăORAVITA,ASIGURARI ORAVITA,ASIGURARI IEFTINE ... rca-oravita.com/ BROKER DE ASIGURARI. Asigurari Auto. RCA,CASCO. Banner Maker. Asigurari locuinte. Banner Maker. Asigurari mopede. Banner Maker. Asigurari medicale ... 24. Celeămaiăieftine asigurari dinăBulgariaă-ăEKOăBroker ... rcabulgaria.ro/ Cele mai ieftine si rapide asigurari RCA pentru autoturismele inmatriculare in Bulgaria, ... Asigur riăAuto Bul r fără intermediari / NON-STOP / 0723.83.62.83 ... C Ief 25. 201 - Ef ro asigurari.efin.ro/rca-ieftin-2014 RCA Ieftin 2014 - puteti comanda cea mai ieftina asigurare RCA in functie de ... Asigurarea de Raspundere Civila Auto - denumita si RCA - este o asigurare prin ... 26. CS www.c... Romania - Insurance Supervisory Commission Asigur riăauto subscrise, fapt 27. s urăr le u o e ere ză m mul de 60% d pr mele bru e re r ă ă es e su ele m ău e forme de s urăr î ... Asigurare C - ING om https://www.ing.ro/ingb/.../asigurari/asigurare-rca.html Polita de asigurare RCA este oferita de societatile de asigurare. Politele de raspundere civila auto sunt intermediate prin Amsterdam Broker de Asigurare, ING ... 28. Asigurariăauto,ăcomparaăsiăcomandaăonline asigurareăauto ... 296 www.asigurari-sanatate.ro/auto/ by Adrian Obreja - Asigurarile auto se pot imparti in doua categorii distincte. Asigurarea de raspundere civila auto, care este o asigurare obligatorie pentru toti soferii care detin un ... 29. UNIQ - C www.uniqa.ro/home/02.../01_Auto/02_RCA/ UNIQA - s ur re obl or e de răspu dere v lă auto RCA. ... pot fi posibile prin asigurarea de raspundere civila obligatorie RCA de la UNIQA Asigurari! 30. Asigurari Mokaă|ăDacaăaltiiăsuntăieftini,ănoiăsuntemămoka asigurarimoka.ro/ Avantajele colaborarii cu un broker de asigurari sunt: afli preturi de la absolut toate societatile de asigurari din Romania; poti incheia ... contact asigurari auto. 31. C 201 C re su pre ur le pr www.avocatnet.ro/.../RCA-2014-Care-sunt-preturile-p... e de so e le de ... In ceea ce priveste societatile de asigurare autorizate de ASF pentru a practicaasigurari obligatorii de rapundere civila auto, acestea sunt in numar de 11:. 32. Asigurariăauto C SCO www.asigurari-auto-casco.ro/ Aici gasesti totul despre asigurarile auto CASCO din Romania! In plus, cea mai buna oferta si parerea celorlalti vizitatori! 33. Asigurariăauto ( C _C SCO) o Tw er https://twitter.com/RCA_CASCO Asigurari auto @RCA_CASCO · 2 Jan 2010. Asigurarile casco nu se vor scumpi in 2010 datorita conditiilor economice precare si a concurentei existente pe ... 34. UE - V l b l e pol e de asigurareăauto l mu re î ... europa.eu/.../citizens/vehicles/.../resident_ro.ht... Europa Mar 7, 2013 - Dov d s urăr d r de or e ( s ură orul dvs es e re u os u î s ră ă e?) Ob ere u e o pol e de asigurare î s ră ă e 35. ASITOă-ăCompaniaăInternaţional ăde Asigur ri,ăRepublica ... www.asito.md/ Comp I er o lă de Asigur ri ASITO S.A. este prima companie deasigurare din Moldova, fiind fondata la 6 iunie 1991, în contextul demo opol zăr ... 36. AsigurariăAuto - s rom www.asirom.ro/asigurari-auto.html CASCO Protector. Unele lucruri nu se lasa la intamplare. Protejeaza-ti masina cu noul CASCO de la ASIROM. Servicii de calitate la un pret corect. Acesta este ... 37. Gh d Asigurari www.ghid-asigurari.ro/ Cel mai cunoscut tip de polita este asigurarea auto RCA, datorita obligativitatii fiecarui posesor de autovehicule de a contracta aceasta polita de raspundere ... 38. ASIGURAREăAUTO 201 www.realitatea.net/asigurare-auto-2014-c... Realitatea TV 297 Câ os ă o pol ă C î 201 ş ... Jan 2, 2014 - In ceea ce priveste societatile de asigurare autorizate de ASF pentru a practica asigurari obligatorii de rapundere civila auto, acestea sunt in ... 39. RcaăIEFTINă2014. Asigurari rcaăieftine.ăCalculatorărcaăieftin www.rcalatineacasa.ro/ by Rca Ieftin la tine Acasa - Cauti o asigurare RCA Ieftina? La noi asigurarea autoRCA este Ieftina si o calculezi Imediat! Online, simplu si ieftin! Puteti afla preturile asigurarilor de tip RCA ... 40. Asigur riăauto - Despre SF www.asfromania.ro/index...asigurari/asigurrari-auto Asigur riăauto T păr re Pe ru ve î î âmp re evo de form re s ur ț lor, po e ț l lor s ur ț ș pă ub ț lor, SF vă pu e l d spoz ț e o ... 41. Asigur riăauto - Gr we C r www.grawe.md/ro/asigurari_auto.htm s urăr le Auto oferite de Comp pe ru dum e vo s ră, p s erul ş s urăr S G WE C r prevăd oper re omple ă u oveh ulul dvs Or re d p he ele ... 42. Asigurari Brasov: Asigurari Ieftineă| Asigurari RCAăOnline www.asiguraribrasov.ro/ Daca ai accesat acest site si citesti acest articol inseamna ca esti unul dintre cei interesati sa isi incheie o asigurare auto obligatorie RCA si CASCO online. 43. Om www.omniasig.ro/ s Asigurari persoane juridice · Auto · Bunuri si cladiri · Angajati · Raspundere civila · Tehnice · Transport · Financiare · Maritime si aviatie · Agricole · IMM. 44. Asigurariăauto - Bus ess Serv es | F https://www.facebook.com/pages/Asigurari-auto/715921585106273?... ebook Asigurari auto. 3 likes · 1 was here. Business Services. 45. Asigurari Bul www.asiguraribulgaria.eu/ r Broker autorizat pentru s urăr C de Bul r , el borăm s urăr î or e zo ă de pe teritoriul ... BROKER AUTORIAZAT ASIGURARI AUTO BULGARIA. 46. Cum fu o e z s s emul bo us-m lus? | Asigurariăauto ... www o so ro › Ghiduri › Asigurari RCA Un asigurat nou, fara istoric in asigurare, este incadrat in clasa neutra B0. Clientii din aceasta clasa nu beneficiaza de nicio reducere fata de tariful standard, dar ... 47. Asigurariăauto C SCO s www.auto.ro/asigurari-auto C - u o ro Gaseste cele mai ieftine polite RCA si CASCO din Romania.Afla ce tipuri deasigurari auto poti cumpara. 48. AsigurareăAuto RCAă2014ă-ă2015ăpentruăAutovehicule ... www.autoasigurare.com/ Asigurare auto RCA 2014 - 2015, calculator tarife RCA, livrare gratuita si reducere pe loc. Cumpara acum polita RCA auto, RCA moto si RCA moped. 49. CASCOăIEFTINă|ăCeleămaiămiciăpreturiăsiăcelămaiărapidătimp ... www.24casco.ro/ ... tanara si experimentata de peste 20 de profesionisti cu ajutorul carora primesti cele mai bune oferte de asigurari. ... Oferte asigurari auto CASCO online ieftine ... 298 s ur re auto - C ze s I form 50. www.citizensinformation.ie/.../mo... Citizens Information o Dec 5, 2013 - În Irlanda, pentru a putea conduce un vehicul într-u sp ț u publ rebu e să veț obl or u asigurare auto. Pe ru obț ere xe u o ... 51. RCA.md: Asigurari online, asigurare ieftina,ărapid www.rca.md/ Compara online asigurari auto si asigurare de sanatate la pret mic, asigurare RCA, Carte verde, CASCO asigurare medicala pentru calatorie in strainatate. 52. Asigureanuă-ăRCAăNonăStopă/ Asigurari Acte Auto ... www.asigureanu.ro/ Asigurari non-stop si acte auto. Telefon: 0763 024 024.Inmatriculari Bulgaria,Asigurari Bulgaria, Cel mai ieftin RCA. 53. Comparaătarifeleărcaă- Asigurare RCAă-ăDespreărca-auto.ro www.rca-auto.ro/asigurare-rca Cu rca-auto.ro incheierea unei polite de raspundere civila auto tip RCA nu a fost ... Completati formularul de mai sus si alegeti varianta de asigurare care ... 54. Firma Auto Eoodăofera Asigurari Bulgaria,ăITPăBulgaria ... actebulgaria.ro/ Firma Auto Bulgaria Eood, birou de acte auto va ofera Asigurari Bulgaria, ITP Bulgaria, Inmatriculari Bulgaria. 55. Asigurare Ieftina.ăRCA,ăCasco,ăLocuintaă.ăPretăminimăpentru ... www.asiguri.ro/ Spre deosebire de asigurarea CASCO auto, care este o asigurare facultativa, ... Asigurarea R.C.A. - Raspundere Civila Auto – dupa cum ii spune si numele, ... 56. GhidăAsigă- Asigurari -ăTotulădespre asigurari -ăPiata ... www.1asig.ro/Ghid-Asig-13,125.htm Marea majoritate a oamenilor au, in tarile vestice, tot felul de asigurari, nu doar cea de raspundere civila auto, ci si de viata, de sanatate, de protejare a bunurilor ... 57. Companiiăde asigurare - Asigurare.mdă-ăPortalăde Asigurari www.asigurare.md/ro/content/105/ Companii de asigurare, Asigurare.md - Portal de Asigurari. ... ASITO SA · ASTERRA GRUP SRL · AUTO-SIGU NT S · C “Gr we C r Asigur ri” SA ... 58. www u Asigurariăauto - Tu ro › Auto ro Stiri Asigurari auto. Afla totul despre asigurarile auto. Criterii de calcul, preturi asigurari obligatorii RCA, conditii casco. Descopera ultimele promotii oferite de ... 59. U de po depu e o re l m el www.constatulamiabil.ro/165_unde-pot-depune-o-recl... dres omp lor de ... GHIDUL DE ASIGURARI AUTO ... DESPRE ASIGURARILE AUTO ... Unde pot depune o reclamatie la adresa companiilor de asigurari sau de brokeraj in ... 60. Asigurari pe ru o - C - C rrefour ro www.carrefour.ro/asigurari-pentru-toti.html Carrefour La birourile de vanzari Carrefour - Otto Broker poti sa alegi o asigurare RCA de la cele mai reprezentative companii de asigurari ! Gasesti cele mai mici preturi, ... 299 61. Asigurariăauto - e www.renault.ro/finantare/.../asigurari-auto/ Renault ul Asigurarile auto CASCO si RCA se incheie prin RCI Broker de Asigurare la alegere prin ALLIANZ, GROUPAMA, OMNIASIG sau UNIQA pentru contractele de ... 62. Cl ms Pro edures asigurariăauto - ST www.astrapoistovna.sk/en/claims/what-to-do-in-case-of POIS OVŇ Act 8/2009 statue of traffic and on amendments to certain acts deemed traffic accident such as incident on the road, which will be in direct relation to vehicle ... 63. Stiriădespre asigurariăauto inăsiteulăstirileprotv.ro.ăInăfiecareăzi ... stirileprotv.ro/stiri-despre/asigurari+auto Stiri despre asigurari auto in siteul stirileprotv.ro. In fiecare zi, cele mai importante evenimente, transmisiuni LIVE, analize, anchete si reportaje. 64. MoldCargoă-ăSocietateăde asigur ri. AUTO CASCOăin ... moldcargo.md/auto-casco.html Ob e l s urăr Auto Casco îl constituie interesele patrimoniale ale persoanelor ... Conform contractului de asigurare s ură orul h ă despă ub re de ... 65. Asigurari T m so r www.asigurari-timisoara.ro/ Asigurari Timisoara - Asigurare Timisoara. ... asigurari de cladiri si bunuri, asigurari auto RCA si CASCO), cat si pentru persoane fizice(medicale de calatorie in ... 66. Asigurariăauto RCAăsiăAsigurariăLocuinteăonline www.asiguratorul.com/ Asiguratorul.com furnizor online de polite de asigurari auto RCA si asigurari de locuinta la cele mai mici preturi. Aici gasiti Asigurari ieftine pentru masini si ... 67. Asigur riăauto C - mod de î rebu re - YouTube ► 29:23► 29:23 www.youtube.com/watch?v=fDz8l-4gNqY o Feb 25, 2014 - Uploaded by RFI România Asigurarea RCA este o asigurare obligatorie auto prin care pă ub ul es e despă ub de omp l re 68. Asigurari s r - C www.plata-rca.ro/astra s r - Pl C Asigurarea RCA de la ASTRA ASIGURARI este acel tip de asigurare prin care persoanele fizice sau juridice prejudiciate in urma unui accident auto produs din ... 69. AsigurariăAuto - Compex Tr d www.compexit.ro/asigurari/ Cluj-N po Prin divizia Compexit Asigurari suntem parteneri cu cele mai importante societati deasigurare: Allianz-Tiriac, Generali, Uniqa, Omniasig, Groupama, Porsche ... 70. Auto - BC Asigurari www.abcasigurari.ro/.../index.php?...auto... e s ur r S Informatii privind procedura de avizare (anuntare) si constatare a daunelor auto. A. AVIZAREA (ANUNTAREA). In cazul in care ati fost pagubit intr-un eveniment ... 71. www Asigurariăauto - C r ro › Asigurari 300 A. Asigurarea facultativa pentru avarii si furt a autovehiculelor/vehiculelor (intern si extern) tip CASCO. Asigurarea auto CASCO, cea mai populara asigurare ... 72. AsigurariăDiverseă-ăPolita AsigurareăAuto -ăRcaăsiăCasco www.asiguraridiverse.ro/ Noi iti oferim asigurare Casco si asigurare locuinta de care ai nevoie, la costuri ... in trafic fara a avea incheiata o asigurare obligatorie de raspundere civila auto ... 73. S B Broker de Asigurare www.sab-broker.ro/ Asigurari Auto CASCO ... Folositi calculatoarele noastre pentru a va calcula polita de asigurare care vi se ... Cum puteti beneficia de consultanta in asigurari? 74. ASIGURARI |ăRCAăIEFTINă|ăCASCOă|ăLOCUINTA ... https://www.vivendis.ro/ Comanda si tu asigurari ieftine online! Asigurare RCA. Aici cumperi cea mai ieftinaasigurare auto RCA! Compara ofertele si alege asigurarea auto RCA dorita. 75. Asigurari www.asigurarircaieftin.com/ C ef o l e RCA ieftin 2014. Compara asigurari RCA ieftine online - sute de oferte pentru asigurarea de raspundere civila auto folosind calculatorul online. 76. AUTO-SIGU de asigurare - xpr mm md NT - Comp www.xprimm.md/AUTO-SIGURANTA-asiguratori-pr... AUTO-SIGURANTA - Companii de asigurare - Totul despre asigurari! Piata asigurarilor din Republica Moldova! Companii, societati, brokeri, produse de ... 77. DeltaăSpeedăSolutions:ăInmatriculari auto,ăradieri auto ... www.delta-inmatriculari.ro/ Delta Speed Solution ofera servicii de inmatriculari auto, radieri auto, asigurari auto, asigurari casco, taxa speciala de inmatriculare, duplicat act pierdut, ... 78. C l ul or asigurariăauto r www.1rca-ieftin.ro/cel-mai-ieftin-rca/ ef o l e Orice proprietar de masina doreste sa cumpere cel mai ieftin RCA de pe piata, avand in vedere ca cel mai important criteriu este pretul asigurarii auto RCA. v Asigurari 79. www.activasigurari.ro/ Iti oferim cele mai bune oferte de asigurari auto la cele mai mici preturi de pe piata. Pentru clientii nostri studiem toti ofertantii, negociem constant ofertele de ... 80. Asigurari C Auto Ief e| www.agentie-asigurari.ro/asigurari-auto/asigurarea-rca/ e e asigurari Asigurarea RCA ofera despagubiri pentru daune materiale si/sau vatamari corporale suferite de o terta persoana si provocate prin accidente auto de catre ... 81. RCAăIeftină|ăRCA Auto |ăRCAăOnlineă|ăSelect-Asigurari.ro select-asigurari.ro/ Compara tarifele RCA utilizand calculatorul RCA si cumpara o asigurare rca ieftin. 82. s ur re C www.oferta-rca.ro/asigurare-rca.html Asigurariăauto Pre ur r m me Doar aici gasesti asigurari rca cu cele mai mari reduceri. Cele mai mici preturi ... Te ajutam sa afli pretul corect pentru asigurarea auto obligatorie . Doar 3 minute ... 301 83. Asigurare C o l e S mul or r fe C 201 www.i-asigurare.ro/asigurare-rca/ Compara tarife RCA de la toate companiile de asigurare si incheie o ... Pe iasigurare ai calculatorul de pret pentru RCA 2014. ... Selecteaza categoria auto. 84. RCAăieftină| Asigurareăauto obligatorieă|ăPret asigurareăauto ... www.rca-sigur.ro/ Oferta RCA ieftin. Calculator RCA si CASCO. Transport gratuit pentru polite mai mari de 165 lei. 85. RCAăIeftinăină3ăMinuteă-ăComandaăAsiguarareăRCA Auto ... www.rcaautoieftin.ro/ by Itexclusiv Constantin - RCA Auto Ieftin | RCA Ieftin | 15% Reducere | Reduceri Senzationale la un RCA Ieftin in 3 Minute | Asigurari RCA + Masina GRATIS la ACCIDENT. 86. Serv ul D u e Auto- VICTO I -Asigurari - victoria- s ur r md › Principala › Suport Clienti Lista actelor necesare pentru deschiderea dosarului de daune în caz de accident. RCA. Asigurat (vinovat). Lista actelor necesare pentru instrumentarea ... 87. Co s r d u e s asigurariăauto | dr s ro www.adras.ro/.../constatari-daune-si-asigurari-auto.ph... Constatari daune si asigurari auto. Adras are incheiate contracte de decontare directa cu urmatoarele societati de asigurare: Allianz-Tiriac; Groupama Asigurari ... 88. s rom asigurari - Asigurari lo u e www.asiguracasa.ro/asigurari_asirom.html Asigurari Asirom ONLINE, informatii asigurari casa Asirom, asigurari auto Asirom, asigurari Asirom rca. 89. CalculeazaăRCAăIeftinăcuăpanaălaă18șămaiăieftină-ăRcaPedia www.rca-pedia.ro/ Pentru accidentele auto soldate cu vatamari corporale sau decese, limita de ... si sunt aceleasi pentru toate societatile de asigurare indiferent daca incheiati un ... 90. Asigurari C SCO s C www.dacia.ro/.../asigurari-auto-si-trafic-accident Dacia u pre ur v jo se, spe l ... ASIGUR RIăAUTO. Prin serviciul de Asigur riăauto Dacia Finance pune la d spoz u uror e or lor de l e , re he e u o r de le s financiar ... 91. Despre o - AsigurariăAuto www.asigurariauto.org/despre-noi/ Ati ajuns in locul potrivit! La noi puteti gasi cele mai bune oferte pentru asigurarea obligatorie auto. Se incheie atat de persoane fizice cat si juridice. Comparativ ... 92. asigauto1.ro/ s u o1 - ASIGURARIăAUTO C Asigurari si inmatriculari Bulgaria. Inmatriculam masini in regim de urgenta pe Bulgaria cu preturi incepand de la 275 de euro. Asiguram ITP si transport pe ... 93. Asigurari î oldov , Comp totul.md/ro/company/Strakhovye-uslugi.html 302 de asigurari s ur re î omp e de s urăr , o e pur le de s urăr - casco, osago,asigurare auto obl or e, s ur re propr e e, să ă e, mob l î C Ief 94. 201 Asigurari C 201 - Auto Speed www.auto.autospeedmarket.com/rca-ieftin-2014-asigu... oldov rke RCA Ieftin 2014: Asigurari RCA 2014: Calculator asigurari RCA ieftin 2014 (cel mai ieftin RCA din Romania) | Auto SpeedMarket :: Stiri despre masini. 95. C l ul or C O l e | Asigurare C Ief www.rcaobligatoriu.ro/asigurarirca.html Asigurari RCA cu emitere instant si livrare in cel mult 5 minute pe e-mail. Telefon 0745380227. C Ief 96. www.rcabulgaria.eu/ - asigurari rom bul r Asigurari RCA bulgaria ieftine in Romania . Rapid ,corect si transparent.Transport gratuit in toata tara. 97. Forum asigurari RCAă/ăCASCO.ăProblemeălegateăde ... www.rca-asigurat.ro/forum/ Omniasig ofera servicii de asigurare - reasigurare: asigurari auto (RCA, CASCO), asigurare locuinta, asigurari raspundere civila profesionala, asigurari ... 98. Fr ude de ze de m de euro î asigur riăauto u ju orul ... www.sursazilei.ro/fraude-de-zeci-de-mii-de-euro-in-as... Jun 6, 2013 - Pontul ve de l m mul er re u î er să fr udeze u su e de m de euro do m r ju ă or de pe p s urăr lor auto: Alianz ... 99. baar.ro/ B de asigurare d s ră ă e se omple e ză formul rul să u l pse s ă documentul de asigurare de răspu dere v lă auto “C r e Verde” ş “Co s Amiabil ... 100. Asigurariăauto Bu ures - u o www.autom rke ro › Adrese utile ul rke Asigurari auto Bucuresti - Allianz Tiriac, ConceptCarService, ITP AUTO -MOTO, Agras, Ardaf, Asiban, Asito Kapital, Astra, BCR Asigurari, Certasig, Why these ads? 1. Asigurareăauto f r Adwww ll zd re rj ro C s o_d s ou o Allianz iti asigura prioritatea la rezolvarea dosarelor. Nu rata! 2. 15șăReducereăLaăRCAă-ăLivrareăGRATISăLaăTineăAcasaă Adwww.asigurari2 om o o o 021 319 1121 Masina La Schimb In Caz De Dauna. Reducere 15%rca+5%casco Contact Asigurari 24 303 o Searches related to asigurari auto astra asigurari asigurari auto carpatica calculator asigurari auto preturi asigurari auto anglia asigurari auto astra asigurari auto carpatica asigurari auto online asigurari auto casco 304 AppendixăF: Parallel comparison between the top 100 results for car insurance in google.com and asigurari auto in google.ro Keyword usage of non-linked words on-page https://www.google.com/search?num=100&clie nt=opera&q=car+insurance&oq=car+insu&gs_l =serp.1.0.0l10.21245.24967.0.28241.8.8.0.0.0.0. 110.754.5j3.8.0....0...1c.1.51.serp..0.8.752.gCCD aO5QVO8 https://www.google.ro/search?num=100&client= opera&q=asigurari+auto&oq=asigurari+auto&gs _l=serp.3..0l10.37706.40661.0.41168.14.8.0.6.6. 0.118.733.6j2.8.0....0...1c.1.51.serp..0.14.767.Ms KBe9ilVrw Keywo Keywor There are 2666 words on this page 68 of these words were filtered out as " stop words" So there are 2598 words that are being analyzed There are 2831 words on this page 156 of these words were filtered out as "stop words" So there are 2675 words that are being analyzed Word Count Density Word Count Density insurance 219 8.43% auto 123 4.73% auto 105 4.04% asigurari 95 3.66% car 78 3% rca 78 3% online 31 1.19% asigurare 46 1.77% quote 29 1.12% mai 41 1.58% quotes 26 1% pentru 38 1.46% coverage 21 0.81% asigurarea 32 1.23% save 21 0.81% este 24 0.92% rates 19 0.73% casco 23 0.89% https 17 0.65% care 21 0.81% auto-insurance 16 0.62% din 20 0.77% compare 16 0.62% prin 17 0.65% state 13 0.5% asigurari-auto 17 0.65% companies 13 0.5% bulgaria 15 0.58% policy 12 0.46% cele 15 0.58% 305 affordable 10 0.38% online 15 0.58% car-insurance 9 0.35% obligatorie 15 0.58% low 9 0.35% civila 14 0.54% money 9 0.35% asigur ri 14 0.54% discounts 9 0.35% ieftin 12 0.46% search 8 0.31% raspundere 12 0.46% cheap 8 0.31% cel 12 0.46% minutes 8 0.31% în 11 0.42% nationwide 8 0.31% preturi 11 0.42% cost 8 0.31% acest 11 0.42% learn 8 0.31% asigurarile 10 0.38% service 8 0.31% sunt 10 0.38% vehicle 7 0.27% ieftine 9 0.35% see 7 0.27% romania 9 0.35% just 7 0.27% afişeaz 9 0.35% offers 7 0.27% calculator 9 0.35% drivers 7 0.27% anunţ 9 0.35% today 6 0.23% civil 8 0.31% pay 6 0.23% de 8 0.31% york 6 0.23% polita 8 0.31% oneonta 6 0.23% unui 7 0.27% agent 6 0.23% polite 7 0.27% cars 6 0.23% r spundere 7 0.27% provides 6 0.23% şi 7 0.27% call 6 0.23% groupama 7 0.27% insuranceloading 6 0.23% livrare 7 0.27% best 6 0.23% reduceri 6 0.23% progressive 5 0.19% insurance 6 0.23% american 5 0.19% s 6 0.23% 306 company 5 0.19% broker 6 0.23% need 5 0.19% google 6 0.23% texas 5 0.19% unei 6 0.23% customer 5 0.19% sau 6 0.23% agents 5 0.19% https 6 0.23% carinsurance 5 0.19% php 6 0.23% purchase 5 0.19% gratuit 6 0.23% aspx 5 0.19% autovehicule 6 0.23% financial 5 0.19% urma 5 0.19% google 5 0.19% mici 5 0.19% options 4 0.15% produse 5 0.19% insure 4 0.15% despre 5 0.19% guide 4 0.15% c utare 5 0.19% right 4 0.15% societatile 5 0.19% visit 4 0.15% cea 5 0.19% people 4 0.15% oferta 5 0.19% life 4 0.15% uniqa 5 0.19% motor 4 0.15% asigurare-rca 5 0.19% business 4 0.15% obligatorii 5 0.19% liability 4 0.15% compara 5 0.19% customers 4 0.15% toate 5 0.19% comloading 4 0.15% allianz-tiriac 5 0.19% fast 4 0.15% aceasta 5 0.19% automobile 4 0.15% persoane-fizice 5 0.19% easy 4 0.15% oferte 4 0.15% department 4 0.15% ofera 4 0.15% customized 4 0.15% gasesti 4 0.15% mutual 4 0.15% aici 4 0.15% main 4 0.15% pana 4 0.15% 307 travelers 4 0.15% farmers 4 0.15% drive 4 0.15% safeco 4 autos orice 4 0.15% 4 0.15% locuinta 4 0.15% 0.15% incheie 4 0.15% 4 0.15% persoane 4 0.15% metlife 4 0.15% pret 4 0.15% portal 3 0.12% asigurari 4 0.15% tips 3 0.12% afla 4 0.15% family 3 0.12% informatii 4 0.15% ratings 3 0.12% doar 4 0.15% edmunds 3 0.12% omniasig 4 0.15% michigan 3 0.12% gratuita 4 0.15% amica 3 0.12% inmatriculari 4 0.15% indiana 3 0.12% accident 4 0.15% plans 3 0.12% allianz 4 0.15% california 3 0.12% generali 4 0.15% eligible 3 0.12% asf 3 0.12% top 3 0.12% in 3 0.12% insweb 3 0.12% euroins 3 0.12% moneysupermarket 3 0.12% calculeaza 3 0.12% citi 3 0.12% intrebari 3 0.12% cure 3 0.12% transport 3 0.12% yahoo 3 0.12% tip 3 0.12% access 3 0.12% facebook 3 0.12% general 3 0.12% tarifele 3 0.12% ameriprise 3 0.12% bcr 3 0.12% hagerty 3 0.12% itp 3 0.12% classic 3 0.12% piata 3 0.12% buying 3 0.12% contact 3 0.12% i 308 good 3 0.12% masina 3 0.12% arizona 3 0.12% asirom 3 0.12% pennsylvania 3 0.12% dauna 3 0.12% side 3 0.12% rca-auto 3 0.12% agency 3 0.12% str in tate 3 0.12% start 3 0.12% stiri 3 0.12% allstate 3 0.12% telefon 3 0.12% loading 3 0.12% daune 3 0.12% corporationloading 3 0.12% produs 3 0.12% want 3 0.12% oferim 3 0.12% price 3 0.12% dvs 3 0.12% fit 3 0.12% accidente 3 0.12% geico 3 0.12% c uta 3 0.12% results 3 0.12% tertii 3 0.12% enter 3 0.12% importante 3 0.12% press 3 0.12% prejudiciati 3 0.12% ads 3 0.12% companiile 3 0.12% based 3 0.12% anunţuri 3 0.12% selection 3 0.12% aceste 3 0.12% current 3 0.12% ofertele 3 0.12% esurance 3 0.12% emag 3 0.12% phone 3 0.12% buna 3 0.12% insurance 3 0.12% iti 3 0.12% check 3 0.12% mult 3 0.12% needs 3 0.12% servicii 3 0.12% infinity 3 0.12% enter 3 0.12% nerdwallet 3 0.12% ap saţi 3 0.12% health 3 0.12% privind 3 0.12% bankrate 3 0.12% totul 3 0.12% 309 advice 3 0.12% vina 3 0.12% erie 3 0.12% asigurari-auto-rca 3 0.12% claims 3 0.12% pune 3 0.12% the 3 0.12% poti 3 0.12% companyloading 3 0.12% acestea 3 0.12% choose 3 0.12% cumpara 3 0.12% instantly 3 0.12% index 3 0.12% providers 3 0.12% mic 3 0.12% grouploading 3 0.12% instant 3 0.12% arbella 3 0.12% pretul 3 0.12% ieftina 3 0.12% loc 3 0.12% anunţwww 3 0.12% alegeti 3 0.12% incheierea 3 0.12% Show/Hide more Show/Hide more https://www.google.com/search?num=100&clie nt=opera&q=car+insurance&oq=car+insu&gs_l =serp.1.0.0l10.21245.24967.0.28241.8.8.0.0.0.0. 110.754.5j3.8.0....0...1c.1.51.serp..0.8.752.gCCD aO5QVO8 https://www.google.ro/search?num=100&client= opera&q=asigurari+auto&oq=asigurari+auto&gs _l=serp.3..0l10.37706.40661.0.41168.14.8.0.6.6. 0.118.733.6j2.8.0....0...1c.1.51.serp..0.14.767.Ms KBe9ilVrw 2 word 2 word There are 2831 words on this page 156 of these words were filtered out as " stop words" So there are 2675 words that are being analyzed There are 2666 words on this page 68 of these words were filtered out as " stop words" So there are 2598 words that are being analyzed 2 Word Phrase Count Density 2 Word Phrase 310 Count Density auto insurance 80 3.08% asigurari auto 41 1.58% car insurance 66 2.54% auto rca 15 0.58% insurance quotes 21 0.81% cele mai 14 0.54% insurance quote 15 0.58% asigur �ri 13 0.5% quote online 8 0.31% civila auto 12 0.46% low cost 7 0.27% raspundere civila 11 0.42% insurance coverage 7 0.27% asigurare rca 11 0.42% insurance rates 7 0.27% cel mai 11 0.42% quotes online 6 0.23% acest anun 9 0.35% insurance companies 6 0.23% afi �eaz 9 0.35% insurance car 6 0.23% �eaz acest 9 0.35% customer service 5 0.19% asigurarea rca 8 0.31% insurance online 5 0.19% din romania 8 0.31% compare auto 5 0.19% asigurari rca 8 0.31% insurance auto 5 0.19% asigurarile auto 7 0.27% online quote 5 0.19% este asigurare 7 0.27% save money 4 0.15% rca ieftin 7 0.27% rates auto 4 0.15% �spundere civil 7 0.27% see save 4 0.15% rca este 7 0.27% insurance drivers 4 0.15% asigurarea obligatorie 6 0.23% affordable auto 4 0.15% �ri auto 6 0.23% quote today 4 0.15% auto casco 6 0.23% insurance policy 4 0.15% auto bulgaria 6 0.23% cheap car 3 0.12% civil auto 6 0.23% liability insurance 3 0.12% mai ieftine 6 0.23% american family 3 0.12% asigurarea auto 5 0.19% cost auto 3 0.12% asigurare obligatorie 5 0.19% automobile insurance 3 0.12% urma unui 5 0.19% insurance company 3 0.12% mai ieftin 5 0.19% 311 save car 3 0.12% prin care 5 0.19% online auto 3 0.12% obligatorie raspundere 4 0.15% auto-insurance aspx 3 0.12% asigurare auto 4 0.15% purchase policy 3 0.12% preturi asigurari 4 0.15% enter search 3 0.12% despre asigurarile 4 0.15% press enter 3 0.12% societatile asigurare 4 0.15% quote minutes 3 0.12% cea mai 4 0.15% loading car 3 0.12% rca casco 4 0.15% insurance grouploading 3 0.12% unui accident 4 0.15% vehicle insurance 3 0.12% mici preturi 4 0.15% compare car 3 0.12% livrare gratuita 4 0.15% insurance companyloading 3 0.12% accident auto 4 0.15% insurance provides 3 0.12% mai mici 4 0.15% asigurari obligatorii 3 0.12% ieftine asigurari 3 0.12% incheierea unei 3 0.12% �sa enter 3 0.12% acestea sunt 3 0.12% inmatriculari bulgaria 3 0.12% polita asigurare 3 0.12% pune dispozi 3 0.12% ieftin rca 3 0.12% asigurari-auto asigurari 3 0.12% �in �tate 3 0.12% pentru asigurarea 3 0.12% str �in 3 0.12% totul despre 3 0.12% obligatorie �spundere 3 0.12% anun �www 3 0.12% Show/Hide more 312 mai mic 3 0.12% tertii prejudiciati 3 0.12% care sunt 3 0.12% mai buna 3 0.12% persoane-fizice asigurariauto 3 0.12% auto este 3 0.12% rca ieftina 3 0.12% anun �uri 3 0.12% asigurarea raspundere 3 0.12% auto care 3 0.12% enter pentru 3 0.12% pentru �uta 3 0.12% asigurari asigurari 3 0.12% aici gasesti 3 0.12% rca online 3 0.12% casco asigurarea 3 0.12% Show/Hide more https://www.google.com/search?num=100&clie nt=opera&q=car+insurance&oq=car+insu&gs_l =serp.1.0.0l10.21245.24967.0.28241.8.8.0.0.0.0. 110.754.5j3.8.0....0...1c.1.51.serp..0.8.752.gCCD aO5QVO8 https://www.google.ro/search?num=100&client= opera&q=asigurari+auto&oq=asigurari+auto&gs _l=serp.3..0l10.37706.40661.0.41168.14.8.0.6.6. 0.118.733.6j2.8.0....0...1c.1.51.serp..0.14.767.Ms KBe9ilVrw 3 word 3 word There are 2831 words on this page 156 of these words were filtered out as " stop words" There are 2666 words on this page 68 of these words were filtered out as " stop words" 313 So there are 2675 words that are being analyzed 3 Word Phrase Count So there are 2598 words that are being analyzed Density 3 Word Phrase Count Density car insurance quotes 11 0.42% �eaz acest anun 9 0.35% car insurance quote 9 0.35% afi �eaz acest 9 0.35% auto insurance quotes 9 0.35% raspundere civila auto 9 0.35% insurance quote online 7 0.27% asigur �ri auto 6 0.23% insurance quotes online 6 0.23% �spundere civil auto 5 0.19% auto insurance companies 5 0.19% asigurari auto rca 5 0.19% auto insurance quote 5 0.19% cel mai ieftin 4 0.15% auto insurance rates 5 0.19% cele mai ieftine 4 0.15% compare auto insurance 5 0.19% urma unui accident 4 0.15% insurance car insurance 4 0.15% unui accident auto 4 0.15% insurance auto insurance 3 0.12% cele mai mici 4 0.15% affordable auto insurance 3 0.12% mai mici preturi 4 0.15% auto insurance drivers 3 0.12% 4 0.15% low cost auto 3 0.12% obligatorie raspundere civila save car insurance 3 0.12% asigurari-auto asigurari auto 3 0.12% cost auto insurance 3 0.12% rca este asigurare 3 0.12% cheap car insurance 3 0.12% str �in �tate 3 0.12% insurance quote today 3 0.12% mai ieftin rca 3 0.12% press enter search 3 0.12% ieftine asigurari auto 3 0.12% car insurance online 3 0.12% auto rca este 3 0.12% rates auto insurance 3 0.12% este asigurare obligatorie 3 0.12% auto insurance coverage 3 0.12% asigurare rca ieftina 3 0.12% auto insurance car 3 0.12% 3 0.12% online auto insurance 3 0.12% asigurarea obligatorie raspundere compare car insurance 3 0.12% cea mai buna 3 0.12% civil auto rca 3 0.12% Show/Hide more 314 enter pentru �uta 3 0.12% obligatorie �spundere civil 3 0.12% cel mai mic 3 0.12% asigurari asigurari auto 3 0.12% despre asigurarile auto 3 0.12% totul despre asigurarile 3 0.12% �sa enter pentru 3 0.12% asigurarea raspundere civila 3 0.12% Show/Hide more Source: http://www.internetmarketingninjas.com/seo-tools/seo-compare/ 315 AppendixăG: Wine production in Europe, in tons from 2008 to 2012 (source: http://faostat.fao.org/site/636/DesktopDefault.aspx?PageID=636#ancor ) country 2008 2009 2010 2011 2012 Austria 299372 235188 173745 281476 215475 Belarus 25730 20990 25110 26790 27220 Belgium 3000 3000 3000 4700 2864 Bosnia and Herzegovina 4591 3655 4504 3354 3061 230046 164525 150083 122687 104000 Croatia 50869 55614 46346 48875 48000 Cyprus 15697 12238 10673 12000 12800 Czech Republic 49725 47995 46276 47000 47000 4268899 4679170 4531671 5106761 5286414 Germany 999100 922800 690600 913200 528515 Greece 400409 386910 336500 295000 295000 Hungary 344876 334368 181279 164634 187400 4609554 4624500 4580000 4673400 4089000 2500 1800 2000 2200 2300 66 67 60 60 60 8343 5400 7244 6123 6479 12967 13479 11025 13199 8504 2450 2450 2380 2450 2450 16980 17094 17986 18000 16000 Portugal 547962 571072 694612 546626 585700 Republic of Moldova 159700 125135 127040 124526 121358 Romania 554182 495732 328724 405817 123450 Russian Federation 503483 501000 760530 696260 700000 Serbia 154224 202368 238233 224431 217503 Slovakia 29441 30763 27785 31388 32500 Slovenia 25974 25750 22680 24000 25000 3736690 3250610 3610000 3339700 3150000 107447 111354 103094 101800 100387 Bulgaria France Italy Latvia Liechtenstein Lithuania Luxembourg Malta Montenegro Spain Switzerland 316 The former Yugoslav Republic of Macedonia 92435 95506 77209 66530 78077 Turkey 24531 23250 27950 29000 30000 250000 273000 300200 175400 185748 1348 1360 1400 1450 1500 Ukraine United Kingdom 317