Retail Network Analysis through the Branch

Online Access: www.absronline.org/journals International Journal of Operations and Logistics Management Volume 5, Issue 2 Pages: 83-97 June 2016 e-ISSN: 2309-8023 p-ISSN: 2310-4945 Retail Network Analysis through the Branch Ouma Denis 1, Mike Iravo 2, Agnes Njeru 3, Ismail Noor 4 1. 2. 3. 4. PhD scholar Jomo Kenyatta University of agriculture and technology, Nairobi. Kenya ([email protected]) PhD scholar Jomo Kenyatta University of agriculture and technology, Nairobi. Kenya PhD scholar Jomo Kenyatta University of agriculture and technology, Nairobi. Kenya PhD scholar Jomo Kenyatta University of agriculture and technology, Nairobi. Kenya Retail operations of supermarkets chains, hold a very important position in supply networks due to the dominant position that retailers hold in the downstream supply chains. Current literature agrees that there are less works on his area. The present study is a second flagship study, investigating branch network variables. Using a mixed research design (descriptive and exploratory) the study employed hoteling’s, retail location to establish the relationship between branch location and branch network expansion. The study used a sample size of 300 respondents in supermarket retail operations. With a response rate of 61%, the findings reveal that branch location is significantly related to branch network expansion and that supermarket retailers should ensure that good locations were identified through different search methods lowering distribution costs. Ideal locations were identified to be in malls and next to distribution centers. The study results propose that branch location is a significantly variable to be used in developing an ISM model for branch network expansion. Keywords: Branch network, Branch location, Supply chain management, Retailing INTRODUCTION Retail supply chain management is a contemporary and evolving field which is a culmination of two different areas of management, supply chain management and retailing. Even though there many refereed journals in the field of supply chain management and retailing, there are not many research papers in the area of retail supply chains especially supermarkets (Avirat, 2006).Due to the power that comes with the control over consumers, retailers are often dominant in a supply chain (Msimangira & Sitalakshmi, 2014).While providing their functions, retailers integrate customer demand and other channel member’s supply into the supply chain as well as managing own retail supply chains .Supermarkets like other retail members are affected by a number of issues that *Corresponding author: Ouma Denis PhD scholar Jomo Kenyatta University of agriculture and technology, Nairobi. kenya E-Mail: [email protected] 83 Retail Network Analysis through the Branch virtually concern all retail and service organizations reliant on branches. These include where best to site outlets; what size and format of stores to employ; what mix of products to incorporate; the area over which the outlets should be promoted and choice of the most efficient methods to solve logistical problems. These are generic problems, equally relevant to banks, grocery and superstores, and petrol stations. For banks, groceries and petrol stations, practical frameworks have been developed on branch network expansion modes (Sinha, & Uniyal, 2007: Srivastava, 2008) It is perhaps surprising that practical frameworks for helping retailers to plan their store own supply chains and networks expansion are all but absent from supermarket retail expansion literature. This has given selected supermarket retailers an advantage to expand their branch network creating oligopolies whose competitive edges cannot be explained. In Eastern Africa, the supermarket industry is dominated by few South African and Kenyan chains. Kenyan supermarkets are also present in Rwanda, and Burundi. Nakumatt has already entered the Burundian distribution sector which has a high concentration of operators from Belgium, China, India, the Netherlands and Pakistan. Surprisingly, foreign retailers such as the South African Metro Cash & Carry and Lucky 7 exited the market in 2005 after brief operations (Pan & Zinkhan, 2006). This was attributed to strong competition and expensive locations. Background of Supermarket retailing in Kenya Kenya is leading in Eastern Africa in terms of supermarket concentration. There is a growing demand for more outlets due to increased urbanization .It is estimated that the number of outlets will reach 129000 in 2017 from the current 112000.Supermarkets represents a third of the retail space and their annual growth is projected to increased at 18% yearly if it grow in tandem with self service demand (Neven & Reardon, 2010).According to their study ,the total sales by the top five leading supermarket chains amounted to $ 800 million in CY 2012 and are expected to keep increasing. These supermarkets include Nakumatt Denis et al. holdings, Tuskys, Naivas, Uchumi and ukwala supermarket. Together they have a five ratio concertration of 75%.This five have continued to flourish the harsh retail environment amidst the problems facing their chains and expanded their branch networks successfully to the extent of even threatening major south Africa giant that enjoy economies of scale in other Eastern Africa countries. Statement of the Problem The retail strategy index for the period 2009 – 2014 recognized branch network expansion as a valuable game plan that could be employed by major supply chain members at retail level. Highlighted in the index were location, branch numbers and use of skilled employees for knowledge sharing purposes .The retail study cited the northward and southern branch network expansion of Sainsbury and Asda. The study identified successful supermarkets as those having more than five branches regionally. The Nakumatt retail strategic plans for the period 2010 – 2014, corroborates these studies by highlighting supermarket moves closer to the customer. With all this reports and strategies, Supermarkets in Kenya still face branch network expansion challenges. Moreover, the network expansion reports for 2008/2009/2010/2011 and 2011/2012 describes theories explaining retail network expansion as descriptive to the extent that clear paths to branch network expansion cannot be extracted from different branch expansion variables. Additionally,information about supermarkets expansion in East Africa has traditionally been limited. In Kenya, focused research on branch network expansion and modeling is inadequate thus allowing five sister supermarkets to expand their supply chains monopoly powers in the retail industry with market concentration of 75% yet they only constitute 0.005 % of total supermarkets. The five supermarkets have owned the industry, moved into other Eastern Africa countries to outdo foreign giant supermarkets.Although the five have Kenyan roots, most other supermarkets are unable to benchmark themselves to the five .They have stagnated in a position of not opening more branches unlike the five although they harbor this 84 Int. j. oper. logist. manag. p-ISSN: 2310-4945; e-ISSN: 2309-8023 Volume: 5, Issue: 2, Pages: 83-97 ambition albeit studies which show that an increase in branch network by 0.26% increased the retail visibility by 6% and that 72% of channel expansion strategy used branches (Vida,Reardon & Fairhust 2007). n = N/{1 + N (e)2} Where n = Sample Size N = the total population I = constant Objective of the Study and Hypothesis The general objective of this study was to establish the reliability of branch location as a variable affecting supermarket branch network expansion and validate it for ISM supermarket branch network modeling.Specifically; the objective of the study was to determine the influence of branch location on supermarket branch network expansion in Kenya. The study was guided by the following hypothesis: H0: Branch location decisions do not influence supermarket branch supermarket network expansion. RESEARCH METHODOLOGY AND DATA COLLECTION This current study used a mixed research design (descriptive and exploratory) to describe practices of the five major supermarkets in Kenya and validate branch location with an aim of using the variable alongside other variables from literature review to formulate a working ISM model. The population for the study comprised of employees of five major supermarkets (Nakumatt, Tuskys, Uchumi, Ukwala and Naivas) working in operations and key decision areas. The supermarkets are characterized by having more than five branches across the country and with an annual turnover of 0.5 billion (Euromonitor international, 2014 Bryman and Bell (2010) define a sample as a subject of a specific population. The process of sampling involves the selection of a group of individuals or elements from a target population. The group sample can then stand for the whole population (Anderson, 2008). The sample of the researcher should select depends on the requirements of the products, its objectives and funds available. ).The sample selected for this study was selected using the slovin formulae as employed by Jankowicz (2011). E = limit of sampling error Assuming a sampling error of 0.05, this can be computed as shown below: n = 1200/{1+1200 (0.05)2} n = 1200/3 = 1200/(1+3) = 300 For structural interpretative modeling, sample population between 200-400 respondents is reliable and free from bias (Thakkar et al 2005: Kline 2011 & Tamorski, 2014).Purposive sampling was used to select the supermarkets. TABLE 1 HERE RESEARCH FINDINGS AND DISCUSSIONS Response rate A total of 300 questionnaires were distributed to the target population. Out of the 300 distributed, a total of 183 questionnaires were returned. This represents a response rate of 61%. The response rate was satisfactory to draw conclusion from for the study and was deemed representative. Moses and Karlton (1971) as cited by Ahmad, (2009) assert that a response rate above 30% is good and acceptable when the research uses survey questionnaires. According to Mugenda and Mugenda (2009) a response rate of above 50% is excellent. Other studies employing the interpretative structural modeling methodology and a response rate above 50% include studies by Thakkar et al (2006) and Sagheer et al (2009) with response rate of 52% and 67% respectively. TABLE 2 HERE 85 Retail Network Analysis through the Branch Designation of Respondents The researcher sought to get reliable information from the employees more conversant with supermarket operations and strategy as shown in table 3. TABLE 3 HERE Majority of the respondents were floor leaders whose total number was 56 (31%).This was closely followed by stores supervisors 38(21%) roving sales supervisors 32(17.5%) and Central Warehouse Supervisor 21(11.5%).According to Bowman and Ambrosini (1997) as cited by Kovil (2008) data collected from one class of top managers may not give a clear picture about a firms strategy. This clearly indicates that there was fair representation in the different levels of decision in supermarket operations. Duration of Branch Operation The study sought to establish how long branches had been in operation. This is shown in figure 1. FIGURE 1 HERE Supermarket branches with less than 1 year to more than 5 years were sampled. Sixty seven point two percent (67.2%) of the respondents rated their branches to have operated for a period more than 5 years. Twenty one point three (21.3%) percent between 2 to 5 years while 11.5% for less than 1 year. Experience of Respondents The study sought to establish how long the respondents had worked in the supermarket. This is shown in figure 2. FIGURE 2 HERE Sixty seven point two percent (67.2%) of the respondents indicated to have been working in the supermarket for a period of above 5 years. Eighteen percent indicated to have worked for a period between 2 to 5 years while 14.8% indicated having worked in the supermarket for a period of less than a year. The length of service could be used to infer the experience and knowledge of the supermarket culture .The long period of work in supermarket Denis et al. respond rate indicates that the data received for this study is reliable. Factor Analysis For Branch Location Items Branch location had a total of six (6) items .All the items were confirmed since their factor loads were more than 0.4.This information is presented in table 4. TABLE 4 HERE Location of Branch The study sought to establish where the supermarket branch was located. The findings are shown in table 5. TABLE 5 HERE Fifty four point one (54.1%) of the respondents indicated that their branches were located in the general business district. Thirty nine point three percent (39.3%) were located in the estates while 6.6% respondents indicated that most of their branches were located both in general business district and estates. The current studies corroborates studies on Kenyan supermarkets by Kamau ( 2008) .The study found out that most supermarket stores started opening in cities and then shifted focus to opening smaller stores next to bus stations in the central business districts. The study indicated that bus stations were targeted for convenience purposes of middle income groups without cars. Distance between the branch and the next bus stop The study sought to establish where the distance between the branch and the next bus stop. The findings are shown in table 6. TABLE 6 HERE On the distance between the branch and the bus station most branch respondents rated a distance less than 5 kilometers (83.1%).Six to ten (6-10) kilometres had a rating of 3.8% while 1115kilometres had 13.1%.On a study reviewing rural retailing by location, Addison and Calderwood (2007) found out that location decisions of most retail branches targeted the general central business district. Their study failed to qualify that stand 86 Int. j. oper. logist. manag. p-ISSN: 2310-4945; e-ISSN: 2309-8023 Volume: 5, Issue: 2, Pages: 83-97 alone retailers were located further away from bus stops since they targeted customers with cars. Tenant mix in location site The study sought to establish the tenant mix where the branches were located. The findings are shown in table 7. TABLE 7 HERE Ninety four point five percent (94.5%) of the respondents indicated that they were located adjacent to an assortment of retail providers while 5.5% indicated that their location had a combination of many tenants. Contradicting the current study findings Borgers at al (2010) citing Beyard and O’Mara (1999) argue that tenant groupings should follow mix and match principles in order to sustain shoppers’ interest and ensure that they are drawn throughout the entire centre. Although the studies propose the mix and match strategy, they comment that one type of location may be suitable for one business and bad for another. In this case, the retail location in relation to the composition is critical and the Times Model (time, information, money, energy and space) is proposed as the most generic. Supporting Borgers et al (2010), later studies by Chung et al (2012) seeking a shopping malls tenant mix model agreed that tenant mix was vital in relating the percentage of shop area occupied by different store in a shopping mall. The authors differed that there was a scientific model determining an optimal mix of tenants in a mall. TABLE 8 HERE The study sought to establish whether sales volume information was vital in branch network decisions. Sixty eight point seven percent (68.7%) were in agreement, 25.7% strongly agreed while 6.0% were ambivalent. Corroborating the findings Wood and Tusker (2008) study on retail location identified site visits as paramount in forecasting sales volumes of geographical areas and penetration of supply chains. In their study the authors illustrated that the measurement and analysis of logistical efficiency while establishing new networks, viability techniques addressing projected sales volume were the best guide to cost and benefit analysis. The authors propose the use of search techniques to discover areas of the country for new stores based on forecasted market share. Vias (2008) study on retail restructuring found that results of previous studies examining the relationship between sales volume and branch network expansion had been inconclusive. His study findings show some studies reporting positive relationship while others found no clear relationship. Using rural retailers he illustrated that they were disadvantaged due to geographical isolation and unfavorable cost structures and restricted population .Although the studies do not provide a solid solution to guide retailing market share, he illustrated that different retailers had a mixture of growth actions dependent on adaptation, diversification differentiation as controlled by market positions. On whether retail patronage assisted location decisions, 66.1% percent of the respondents were in agreement, 25.7% strongly agreed while in totality 5.1% disagreed. Corroborating the findings Alsultan and Al Fawzan (2009) stressed the importance of efficient and effective facility location. Their study however ranked competitors retail patronage information and information sharing vital particularly when locating in competitive environment. Contradicting the findings Penny and Broom (1988) as cited by Wood and Reynolds (2010) study of evolution of UK retailers found out that irrespective of the retail environment, the dominant factor in reaching decisions about new sites or in developing trade forecast was the experience of operational managers in the firm. The study also sought to establish the extent to which forecasted market share in a location could provide information on location decisions. Seventy four point three percent (74.3%) respondents indicated to be in agreement, 15.8% strongly disagreed while 5.5% disagreed .Corroborating the findings Daskin et al (2008) employed the fixed care facility problems in illustrating that any location model adopted needed vital market share information before models were validated. Based on 33 respondents from an exploratory survey, Wood and Tusker (2008) found out that while 100% of the affected firms used sales volume and market share, there was little evidence of database integration into strategic decision making and 87 Retail Network Analysis through the Branch Denis et al. therefore detailed exploration and the ‘search’ approaches were still vital. aimed at reducing transport and inventory costs of both new and established branches. On whether market size saturation information was vital in branch network decisions, sixty eight point three (68.3%) of the respondents agreed, nineteen point seven percent (19.7%) strongly agreed, 3.8% disagreed while 3.3% strongly disagreed. The study findings corroborate Mamoun and Akrous (2012) and Sandberg (2014) studies which established that market saturation was a good measure of over representation and could be employed to closure and assortment reduction of affected stores were flagship stores. Wood and McCarthy (2014) further concur with the above findings by using the UK food retailing industry retailers .The authors found out that the retailers controlled their expansion activities through new location space races and market saturation. On whether pedestrian flow in a branch was vital in network decisions, seventy five point four percent (75.4%) respondents’ agreed, 24.0% strongly agreed while 0.5% was in disagreement. Corroborating the findings Morscett et al (2005) and Chuan et al (2011) found out that retail inflow and outflow were vital elements of store success.Contraditing the findings Dass and Piyush (2012) study on category vulnerability a cross retailers, found out that pedestrian flow level mixes had no real bounds since they could be controlled by physical abilities of store checkout counters. Their study however proposed that what needed to be addressed was the speed of checkout as it was a determinant of store selection. The study also sought to establish whether the number of shopping malls and shopping centers around an area influenced branch location decisions. Seventy point five percent (70.5%) of the respondents’ agreed, 14.8% strongly agreed while 6.6% were indifferent and 8.2% were in disagreement. The study findings corroborates ElAldly (2007) study which illustrated that shopping malls were an attractive location for retail outlets.El-aldly cited time,information,money, and space as efficiencies established in malls and could lower establishment costs of new retailers. The authors also illustrated that by locating in malls, retailers enjoyed low sunk costs such as advertising and tenant mix related problems which were cushioned by mall management and anchor stores. The study sought to establish the extent to which transport and inventory holding costs information was vital in branch network expansion. Sixty tree point four percent (63.4%) agreed and 30.6% strongly agreed. Corroborating the findings, Ernie and Rant (2008) reviewed the transport and inventory costs of Sainsbury’s and found out that the fulfillment factories established on 40acres and 650000 centres were targeted at lowering transport and inventory related costs. Other scholars such as Amrouce and Zaccour (2007) had earlier indicated that Sainsbury’s six dependency criteria that stressed the use of traded units’ bar codes (TUI) On whether the distance to the distribution centers was vital in branch network expansion decision making, 68.3% of the respondents were in agreement, 24.6% strongly agreed and 6.6% of the respondents disagreed.Corrroborating the study findings Wood and Browne (2007) study findings on convenience branch location in Europe, found that before branches are established site visits was rated 97% as the most important factor in making location decisions. Similar studies by Kan,Weinarter (2013) identified such information by illustrating how retailers were extending their control upstream of distribution centers(from DC to manufacturers) in an effort to improve utilization of branch and store logistical assets to reduce wastage and also improve efficiency. Contradicting the findings Calvo and Lang (2015) explain that the distant to distribution centers is not significant as a factor. To illustrate this they used Sainsbury’s new supply chain strategy of replacing existing networks of 25 regional distribution centers with automated distribution facilities known as fulfillment factories which have significantly increased efficiency in UK branches. Mapped with flagship fulfillment of 160 docks, supplier goods are received in one side while Sainsbury’s trucks are loaded for deliveries to the stores at the other side. Branch Location Computation Pearson Correlation 88 Int. j. oper. logist. manag. p-ISSN: 2310-4945; e-ISSN: 2309-8023 Volume: 5, Issue: 2, Pages: 83-97 The results of Pearson correlation between branch location and branch network expansion are represented in table 9. TABLE 9 HERE The correlation coefficients between branch location and branch network expansion were found to be .473** at P = 0.000 which is less compared to P benchmark value of .05.This therefore demonstrates some positive relationship between branch location and branch network expansion.The results support the argument by Holweg and Lorentz (2010) that good location decreases distribution costs of the retail supply chain making branch expansion cheaper. The authors analysis also affirm that location is the most optimal tool of quick analysis of stores traffic to existing, would be branches and competitor locations when opening new branches. Poor location increases distribution costs making branch network expansion hard. Employing location analytics approach Hillebrand and Bieman (20011) also argue that location is among the main factors positively influencing retail performance particularly using organic growth. Results of the regression analysis on branch location The Results of the regression analysis on branch location are presented in table 10. TABLE 10 HERE Predictors: (Constant), BRANCH LOCATION The model of y = β1 X1 + e, explained 21.9% of the variation in branch network expansion as shown by the adjusted r. This supports arguments advanced by Rigby (2007) that a significant level of the variations in branch network expansion can be explained by retail location decisions .Explaining the significance of branch location, the author cites Carrefour’s strategy of analyzing a city with potential, looking for suitable suppliers and income levels to sustain a network of stores before moves are made. Kwok (2012) confirms that there is an important and inextricable link between the network strategy and the location. The author argues that location decisions have positive relationship with the branch network decision and therefore location decision should be an integral part of retail strategy, not designed as an afterthought. Results of analysis of variance on branch location The analysis of variance (ANOVA) indicated that the model of branch network expansion with branch location at F value of 52.113, p > 0.05 indicate that there was a highly significant relationship between branch location and branch network expansion in Kenyan supermarkets. The results are presented in Table 11. TABLE 11 HERE The table shows that branch location play a crucial role in branch network expansion of Kenyan supermarkets. This supports Cao and Dupuis (2009) who argue that the success of retailing significantly depended on lean retailing, a practice synonymous with location standardization, location based on cost-effective relationships with suppliers as well as distribution which reduces retail chains minimization of distribution and selling labor costs. Results Of The Coefficients For Regression Between Branch Location And Branch Network Expansion Branch location was found to have a positive influence on branch network expansion. This is illustrated by the regression results at 5% level of Significant and unstandardized beta coefficient of 0.257 and t-value of 7.219 at P=0.000. TABLE 12 HERE The significance of branch location on branch network has also been supported by Schiele, (2008) who argue that the location of retail activities in relation to each other as well as buyers and suppliers often contribute to logistics efficiency, supplier access and branch network strategy success. The author argues that firms located within clusters have been found to enjoy productivity, innovation and profitability advantages compared to their isolated competitors and that branch location correlated with branch network between 6 and 7. Branch Location Hypothesis Results There is no significant relationship between branch location and branch network expansion:This hypothesis was stated as: 89 Retail Network Analysis through the Branch Ho β = 0 HA β ≠ 0 and tested using a two tailed . TABLE 13 HERE The calculated t value of 7.219 is greater than the tcritical (1.96) at (183-1)(0.005) and therefore the study rejected that null hypothesis that there is no significant linear relationship between branch location and branch network expansion in Kenyan supermarkets. Studies conforming to the current study are Aoyama (2007) and Gereffi and Ong (2007) who employing DEA models for analysis of intra-chain comparative store efficiency, significantly related the value of branch location to branch network expansion in examining the competitiveness of the chain as a whole. The authors argue that branch expansion competitiveness should be based on benchmarking the retail outlets which compose the chain for retail success. CONCLUSION AND RECOMMENDATIONS Branch Location The study established that most supermarket retailers located their branches in the general business district. The study found out that most supermarket stores started opening in cities and then shifted focus to opening smaller stores next to bus stations in the central business districts and sought shopping malls tenant mix model. The study indicated that bus stations were targeted for convenience purposes of middle income groups without cars. The study also established that transport and inventory holding costs information was vital in branch network expansion and the distance to the distribution centers was vital in branch network expansion decision. Branch location was established to belong to the key drivers of retail chain branch network expansion implementation gave quick results. Recommendations The study established that most supermarket retailers located their branches in the general business district. Good location was adjacent to distribution centers, bus stops and in shopping Denis et al. malls. The study recommends that flagship branches needed to be started in urban centres before extending to other areas. The study proposes that more efforts be channeled towards Branch location as key drivers of retail chain branch network expansion and its implementation gave quick results. Areas For Further Research Despite the agreed importance attached on branch location, the brick and mortar model is embracing on line retailing. Future researchers are encouraged to account for the impact of online retailing on branch expansion efforts. Secondly, the data are from one country yet the successful retailers have extended to other East African countries and caution should be exercised when generalizing findings to other geographic regions. REFERENCES Al-Sultan K.S & Al-Fawzan, M.A (2009). A Tabu Search Approach to the incapacitated facility location problems. An Operational Research 86:91-103 Amrouce, M & Zaccour, H, (2008),"“Supply chains in the era of turbulence",International Journal of Physical Distribution & Logistics Management, 4 (1) 3-12 Aoyama, Y. (2007), “Oligopoly and the structural paradox of retail TNCs: an assessment of Carrefour and Wal-Mart in Japan”, Journal of Economic Geography, 7 (4),471-90. Avirat., S (2006), “A Valuable System – Different aspects of managing the supply chain in retail”, Retail Biz, 12(2)37-38. Borgers, A., Brouwer, M., Kunen, T., Jessurun, J. & Janssen, I. (2010). A virtual reality tool to measure shoppers’ tenant mix preferences. Computers, Environment and Urban Systems, 34(5), 377-388. 90 Int. j. oper. logist. manag. p-ISSN: 2310-4945; e-ISSN: 2309-8023 Volume: 5, Issue: 2, Pages: 83-97 Calvo M A, & Lang S (2015),"Corporate social responsibility in global supply chains", Operations and Supply chain management.4(3)4-17. Cao, L. & Dupuis, M. (2009), “Core competences, strategy and performance: the case of international retailers in China”, The International Review of Retail, Distribution & Consumer Research, 19 (4) 349-69. Chuang,M (2005), Business Research Methods, McGraw-Hill Chung-Yim Y., Sherry Y.S. & Xu,H,K (2012) "A tenant‐mix model for shopping malls", European Journal of Marketing operations , 46 (3/4),524 – 541. Dass, M., & Piyush, K (2012) "Assessing category vulnerability across retail product assortments", International Journal of Retail & Distribution Management, 40 (1),64 – 81 El Aldly ,M. (2007), “Shopping Malls: the case of international retailers in China”, The International Review of Retail, Distribution & Consumer Research, 9 (4) 49-69. Ernie, S & Rant, B (2008),"Managing sustainability performance of supply chains", Supply ChainManagement: An International Journal, 19 (3)32 – 41 Euromonitor International. (2014). Retail report –Kenya Retrieved from GMID: Global Market Information Database Gereffi, G. & Ong, R. (2007), “Walmart in China can the world’s largest retailer succeed in the world’s most populous market?” Harvard Asia Pacific Review, 9(1), 46-9 Hillebrand , R.D. & Bieman , S. (2011), “Collaboration in the supply chain”, Production and Operations Management, 12 (3), 36-52. Holweg, D. & Lorentz., M.N (2010), “Managing the interface between suppliers and organizations for environmental responsibility: an exploration of current practices in the UK”, Production and Operations Management, 11 (2), 71-84. Kamau, F, 2008. The growth of supermarkets in Kenya;opportunities and limitations for pineapple producers.Conference Presentation, ESAMI, Nairobi. Kan, G. (2008), “Customer relationship management: in B2C markets, often less is more” California Management Review, 44 (3) 87-104. Kovil.V.N.M (2008) Strategy formulation and implementation in manufacturing organizations:the impact on performance.international journal of retail and distribution management 28(4)25-32. Kwok. P. (2012), “Strategic alliance in a closedloop supply chain: a case of manufacturer and eco-nonprofit organization”, Technovation, 26 (10),127135. Mamoun, N., & Akrous , K (2012) "Assessing retail assortment across retail product assortments", International Journal of Retail & Distribution Management, 4 (1),6 –8 Msimangira,K. A., & Sitalakshmi, V (2014) Supply chain management intergration:acritical problem and solution.Operations and Supply chain management .7 (1) 23-31. Morsett,M,. Tachizawa, C & Yew W (2005)," multi-tier supply chains ", Supply Chain Management: An International Journal, 9 (6) 43-63 Mugenda O.M. & Mugenda A.G. (2006). Research Methods Quantitative and 91 Retail Network Analysis through the Branch qualitative approaches. Press. Nairobi: Acts Mugenda, A.G. (2009) Social science research. Nairobi: Acts Press. Newing, A., Graham P. Clarke, & Clarke, M (2015) “Developing and Applying a Disaggregated Retail Location Model with Extended Retail Demand Estimations,” Geographical Analysis, 47(3)219-239. Paddison, A & Calderwood, E (2007) "Rural retailing: a sector in decline?", International Journal of Retail & Distribution Management, 35 (2),136 – 155. Pan , G. & Zinkhan, M. (2006), “Do customer loyalty programs really work?”, Sloan Management Review, 38, 71-82. Rigby , C. (2007),“International codes of conduct and corporate social responsibility: can transnational corporations regulate themselves?”, Transnational Corporations, 8 ( 1), 143-181. Schiele,H (2008),"Location, location: the geography of industry clusters", Journal of Business Strategy, 29 (3) 29 – 36 Sagheer, S.,Yadav, S. & Deshmukh, S.G. (2009) "An application of interpretative structural modeling of the compliance to food standards", International Journal of Productivity and Performance Management, 58 (2), 136 – 159. Sekaran R.B.(2009), Introduction to Supply Chain Management, Prentice Hall, Upper Saddle River, NJ. Denis et al. using interpretive structural modeling(ISM) and analytic network process (ANP)”, Supply Chain: An International Journal, 6 (1)32-46. Vias, A.C (2006) Bigger stores, more stores or no stores: paths in retail restructuring in rural america,Journal of rural retailing studies.20(3)303-318. Vida, I., Reardon J & Fairhust, A. (2007): Determinants of International Retail Involvement: The Case of Large U.S. Retail Chains, Journal of International Marketing & operations , 8( 4 )37-60 . Wood.S & Browne .S (2006) Convenience store location planning and forecasting– a practical research agenda International Journal of Retail and Distribution Management, 35 (4)233-255. Wood .S .A & Mccathy .D (2014)The UK retail race for space and market saturation .International review of retail,distribution. 24(4)121-144. Wood .S & Tusker .A (2008) The importance of content in store forecasting,the site in retail location decision.Journal of targeting,measurement & analysis for operations,16(3)139-155. Wood.S & Reynolds .C (2010) store location planning and Retail agglomeration forecasting–International Journal of Retail and Distribution Managementvol 3 (4)2325. Sinha, P.K. & Uniyal, D.P. (2007), Managing Retailing, Oxford University Press, New Delhi.31. Thakkar, J., Deshmukh, S.G., Gupta, A.D. & Shankar, R. (2005), “Selection of thirdparty-logistics (3PL): a hybrid approach 92 Int. j. oper. logist. manag. p-ISSN: 2310-4945; e-ISSN: 2309-8023 Volume: 5, Issue: 2, Pages: 83-97 APPENDIX Table 1: Distribution of the selected supermarkets branches in Kenya and respondent distribution among them Supermarket Number of branches Respondents Nakumatt Holdings Ltd. 34 63 Tusker Mattresses Ltd. (Tuskys) 60 110 Uchumi Supermarkets 27 50 Ukwala Supermarket chains 11 20 Naivasha Self Service Stores Ltd 31 57 163 300 Source: (Euromonitor international, 2014). Secondary data was be collected using journal, academic documents and expert opinion. Table 2: Response Rate Supermarket Questionnaires Questionnaire distributed completed Nakumatt 63 50 Tuskys 110 68 Naivas 50 35 Ukwala 20 12 Uchumi 57 18 300 183 Table 3. Designation of Respondents Job Designation Team Leader/Branch Manager Floor Leaders Stores Supervisor Central Warehouse Supervisor Roving Sales supervisors Number of respondents 36 56 38 21 32 183 % of total respondents 19.8% 31% 21% 11.5% 17.5% 100% 93 Retail Network Analysis through the Branch Denis et al. Table 4: Branch Location component Matrix Item Extraction Retail area population growth .890 Retail patronage numbers .862 The projected sales volume of an area .798 Branch retail inflow/outflow .692 Transport and inventory holding costs .425 Distance to distribution centres .388 Table 5: Location of branch Branch Location Frequency Percent Valid Percent Cumulative Percent General Central 99 54.1 54.1 54.1 Estate 72 39.3 39.3 93.4 Mix 12 6.6 6.6 100.0 Total 183 100.0 100.0 business district Table 6 :Distance between your store and the next bus stop Frequency Percent Valid Cumulative Percent Percent Less than 5 kms 152 83.1 83.1 83.1 6-10 kms 7 3.8 3.8 86.9 11-15 kms 24 13.1 13.1 100.0 Total 183 100.0 100.0 94 Int. j. oper. logist. manag. p-ISSN: 2310-4945; e-ISSN: 2309-8023 Volume: 5, Issue: 2, Pages: 83-97 Table: 7. Tenant mix in location site Tenant mix Frequency Percent Valid Percent Cumulative Percent Assorted service 173 94.5 94.5 94.5 Mix and match 10 5.5 5.5 100.0 Total 183 100.0 100.0 providers Table 8: Respondents opinion on branch Location Item projected sales volume of an area Retail patronage numbers Forecasted market share Market saturation /market size(sales) Number of malls and shopping centres around the area Transport and inventory holding costs Branch retail inflow/outflow Retail area population growth Distance to distribution centres Strongly disagree .0% .0% .0% 3.3% Disagree Ambivalent Agree .0% 2.7% 5.5% 3.8% 6.0% 5.5% 4.4% 4.9% 68.3% 66.1% 74.3% 68.3% Strongly agree 25.7% 25.7% 15.8% 19.7% .0% 8.2% 6.6% 70.5% 14.8% .0% .0% 6.0% 63.4% 30.6% .0% .0% .0% .5% 3.3% 6.6% .0% 6.0% .5% 75.4% 61.2% 68.3% 24.0% 29.5% 24.6% Table 9: Branch location Pearson correlation computation BRANCH NETWORK EXPANSION BRANCH NETWORK EXPANSION 1 Pearson Correlation Sig. (2-tailed) N 183 BRANCH Pearson .473** LOCATION Correlation Sig. (2-tailed) .000 N 183 **. Correlation is significant at the 0.01 level (2-tailed). BRANCH LOCATION .473** .000 183 1 183 95 Retail Network Analysis through the Branch Denis et al. Table 10: Results of the regression analysis on branch location R R Square .473a .224 Adjusted R Square Std. Error of the Estimate .219 1.67652 Table 11: ANOVA results for branch location and branch network expansion Model Regression Residual Total Sum of Squares 146.473 508.739 655.212 df Mean Square F Sig. 1 181 182 146.473 2.811 52.113 .000a a. Predictors: (Constant), BRANCH LOCATION b. Dependent Variable: BRANCH NETWORK EXPANSION Table 12: coefficient for regression between Branch location and Branch network Expansion Model (Constant) BRANCH LOCATION Unstandardized Coefficients β 12.923 .257 Std. Error .507 .036 Standardized Coefficients Beta .473 t 25.510 7.219 Sig. .000 .000 a. Dependent Variable: BRANCH NETWORK EXPANSION Table 13: Hypothesis testing for Coefficients of Regression between Branch location and branch network expansion Model β t-cal t-critical (Constant) 12.923 25.510 BRANCH .257 7.219 1.96 LOCATION 96 Int. j. oper. logist. manag. p-ISSN: 2310-4945; e-ISSN: 2309-8023 Volume: 5, Issue: 2, Pages: 83-97 Figure 1: Duration of branch operation Figure 2: Length of Service 97