User Experience (UX) Design and The Internet of Things

Download as pdf or txt
Download as pdf or txt
You are on page 1of 51

Cood evenlng. 1hanks for havlng me here.

1oday l wanL Lo glve you a blL of my


perspecuve on user experlence deslgn for Lhe lnLerneL of Lhlngs. 1he Lalk wlll be ln
Lwo maln parLs: a general overvlew of whaL l belleve Lhe lnLerneL of 1hlngs ls beyond
Lhe markeung hype and denluonal sllppage ln Lhe news, and Lo Lalk abouL some
speclc approaches l've Laken Lo developlng producLs ln Lhls space. l normally don'L
go deep lnLo Lhe nluy grluy of whaL l do, buL l LhoughL Lhls group may appreclaLe
knowlng aL leasL a blL abouL my deslgn pracuce.
Cverall, however, remember LhaL Lhls eld ls new. noL [usL Lhe lnLerneL of 1hlngs, buL
user experlence deslgn ln general, so one of Lhlngs l urge you Lo do ls Lo Lhlnk abouL
abouL all of Lhlngs l'm mlsslng, all of Lhe dlerenL ways Lo Lhlnk abouL Lhls space and
how Lo deslgn ln lL. We're [usL aL Lhe beglnnlng here, so Lhe eld ls wlde open for
lnnovauon aL all levels.
0
Let me begin by telling you a bit about my background. Im a user experience
designer. I was one of the first professional Web designers. This year I am celebrating
professionally designing Web sites for 20 years. This is the navigation for a hot sauce
shopping site I designed in the spring of 1994.
1
l've also worked on Lhe user experlence deslgn of a loL of consumer elecLronlcs
producLs from companles you've probably heard of.
2
l've wrluen a couple of books based on my experlence as a deslgner. Cne ls a
cookbook of user research meLhods, and Lhe second descrlbes whaL l Lhlnk are some
of Lhe core concerns when deslgnlng neLworked compuLauonal devlces.
3
l also sLarLed a couple of companles. 1he rsL, Adapuve aLh, was prlmarlly focused
on Lhe web, and wlLh Lhe second one, 1hlngM, l goL deep lnLo developlng hardware.
4
1oday l work for A8C, Lhe famous research lab, as a prlnclpal ln lLs lnnovauon
Servlces group.
3
A8C ls also ln many ways where Lhe lnLerneL of 1hlngs began. ln Lhe laLe 1980s Mark
Welser, who was Lhen Lhe C1C of A8C predlcLed LhaL [usL as malnframes-one
compuLer Lo many people-gave rlse Lo personal compuung-one compuLer per
person-Lhere would be a nexL perlod, whlch he called ublqulLous compuung, LhaL
would happen when Lhere was more Lhan one compuLer per person. 1haL's hls charL
from 20 years ago predlcung LhaL Lhe crossover would happen around 2003, whlch
was baslcally rlghL. l never knew hlm and he dled ln 1999, buL l wlsh he was around
Loday Lo glve hls LhoughLs abouL whaL's golng on Loday.
6
7
8
l wanL Lo sLarL by Lalklng abouL Moore's Law, slnce LhaL's where all conversauons
abouL Lhe lmpllcauon of dlglLal Lechnology sLarL. When people Lalk abouL Moore's
Law, lL's oen ln Lhe conLexL of maxlmum processlng power. 8uL lL's acLually
someLhlng dlerenL. lL's acLually a descrlpuon of Lhe cosL of processlng power. lL's a
model of how much more processlng power we can L lnLo a slngle chlp LhaL's prlced
aL a predlcLable prlclng polnL Lhls year Lhan we could lasL year. 1hls means LhaL lL's
noL [usL LhaL processors are gemng more powerful, lL's LhaL 8CCLSSlnC ls gemng
cheaper.

lor example, aL Lhe beglnnlng of Lhe lnLerneL era we had Lhe 486 as Lhe sLaLe of Lhe
arL and lL cosL $1300 ln Lodays dollars. lLs Lhe processor LhaL Lhe Web was bullL for
and wlLh. 1oday, you can buy LhaL same amounL of processlng power for 30 cenLs,
and lL uses only a fracuon of Lhe energy. 1haL decrease ln prlce ls Lhe same orders of
magnlLude drop as Lhe lncrease ln speed. 1hls ls noL a colncldence, because boLh are
Lhe producL of Lhe same underlylng Lechnologlcal changes.

WhaL Lhls means ln pracuce ls LhaL embeddlng powerful lnformauon processlng
Lechnology lnLo anyLhlng ls qulckly approachlng becomlng free.
1hls means LhaL because of compeuuve pressure, anyLhlng LhaL can have neLworked
lnformauon processlng embedded ln lL, wlll. Lveryday ob[ecLs wlll make sophlsucaLed
auLonomous declslons and acL uslng arblLrary lnformauon., whlch ls as deep an
lnfrasLrucLural change ln our world as elecLrlcauon, sLeam power, and mechanlcal
prlnung. Maybe lLs as blg of a deal as brlcks.
9
1he oLher domlnanL Lrend rlghL now ls of course pervaslve daLa communlcauon. 1hls
ls an lmage from 1lmo Arnall LhaL's envlslonlng how saLuraLed our envlronmenL ls
wlLh neLworks, and lL's noL even counung Lhe moblle phone neLwork. 1hls means
LhaL vlrLually any devlce, anywhere can share daLa wlLh Lhe cloud aL any ume. 1hls
means lL can oMoad funcuonallLy LhaL used Lo be local Lo Lhe cloud, whlch means
LhaL almosL all processlng, sLorage and mosL bauery llfe lssues dlsappear. lf your uny,
wlmpy devlce can Lalk Lo Lhe cloud, lL can be eecuvely as powerful as Lhe mosL
powerful compuLer on earLh.
10
11
1he lnLerneL of 1hlngs ls a really challenglng name Lo work wlLh, because Lhere are so
many denluons. 1hls ls 1lme Magazlne's lllusLrauon of Lhe lnLerneL of 1hlngs for
Lhelr 8esL lnvenuons of 2008" edluon. l love Lhls lllusLrauon ls because lL makes no
sense no mauer how you Lhlnk abouL lL, whlch ls acLually qulLe an accuraLe
represenLauon of how confuslng Lhe many denluons of Lhe lnLerneL of 1hlngs are
rlghL now.
12
1he rsL Lhlng Lo noLe ls LhaL Lhere are Lwo denluons of Lhe lnLerneL of 1hlngs,
whlch causes a blL of confuslon. ln Lhe orlglnal one kevln AshLon looked aL how 8llus
could aecL Lhe loglsucs lndusLry and saw an analogy Lo Lhe lnLerneL. 8llus could be
llke Lhe packeL header ln lnLerneL daLa packeLs, Lrucks were llke wlres, depoLs llke
rouLers, warehouses llke caches.

Somewhere around 2010 Lhe denluon sllpped and raLher Lhan belng an analogy
beLween shlpplng and Lhe lnLerneL, lL came Lo mean whaL Mark Welser had earller
descrlbed as ublqulLous compuung, whlch means LhaL everyday ob[ecLs as full-
edged cluzens of Lhe lnLerneL. lnsLead of belng LlkL packeLs, Lhe ob[ecLs now send
packeLs Lhemselves.

WhaL's especlally confuslng ls LhaL slnce so many of Lhe new ob[ecLs are moblle, Lhe
rsL denluon ALSC klnd of applles. When you're uslng Waze Lo gulde you around
Lramc, lL's Lreaung your car llke a daLa packeL because your phone has made your car
lnLo a neLworked compuung devlce.
13
1he lmpllcauons of Lhls Lechnologlcal change are broad, buL Loday l wanL Lo focus on
Lwo aspecLs: how consumers' relauonshlps Lo devlces change, and how companles
value chalns change.
14
1he rsL Lhlng l wanL Lo Lalk abouL ls how people's relauonshlp Lo Lhelr devlces ls
changlng.
13
The core of this change is a shift from generic devices and software to specialized
devices and software. When computing was expensive, you had general purpose
devices and general purpose software that had deal with almost every situation. This
necessitated design compromises that resulted in devices and software that could do
almost everything, but did none of it well.

Now that processing is so cheap, you can have a combination of 10, 20, or 30
computing devices and apps for the price of that one device, and you can acquire new
functionality as needed. This means that every device and software package can have a
narrower purpose.

Adobe brush
Haiku deck
16
1he second Lrend ls LhaL much of Lhe value of compuung ls now remoLe. 1oday, mosL
people undersLand LhaL Lhe experlence you see on one devlce ls oen a parL of
someLhlng LhaL's dlsLrlbuLed LhroughouL Lhe world. 1here's no longer a need Lo pack
everyLhlng lnLo a slngle plece of soware, and Lhere's no expecLauon LhaL everyLhlng
wlll be Lhere.

loursquare as a whole doesn'L llve on your phone ln any meanlngful way, and people
know LhaL.
17
If we chart these two tends, two broad classes of digital products emerge.

lf we follow Lhe local Lo remoLe axls, we nd general-purpose devlces LhaL do roughly
Lhe same seL of Lhlngs, and dler prlmarlly ln slze. They exist to provide access to
online services, in a form factor thats appropriate to the context in which theyre used.
l call Lhese devlces Lermlnals.

lf we follow Lhe general Lo speclc axls, we see a shl ls Lo more narrow-funcuon
devlces LhaL are deslgned Lo do a small seL of things with specialized hardware. A
parking meter can take quarters, which your phone cant do, and a digital SLR has a
giant lens on it, which you probably dont want on your phone. These devices differ in
their specialized hardware. l call Lhese devlces appllances.
18
The most interesting thing, however, to me is that these shifts are part of an even larger
transition, one where devices are simultaneously specic AND deeply tied to online
services. In this model, the service provides the majority of the value, and can be
represented either as a dedicated appliance, an app running on a terminal, or anything
in between.

I call these devices service avatars. And this is where much of the Internet of Things
currently lies, at least for consumer products.
19
As value shifts to services, the devices, software applications and websites used to
access itits avatarsbecome secondary. A camera becomes a really good appliance
for taking photos for Flickr, while a TV becomes a nice Flickr display that you dont
have to log into every time, and a phone becomes a convenient way to take your Flickr
pictures on the road.

Hardware becomes simultaneously more specialized and devalued as users see
through each device to the service it represents. The hardware exists to get better
value out of the service.
20
AnoLher example ls Lhe klndle. Peres a Lelllng older ad from Amazon for Lhe klndle.
lL's saylng Look, use whaLever devlce you wanL. We donL care, as long you sLay
loyal Lo our servlce. ?ou can buy our speclallzed devlces, buL you donL have Lo.
21
AL rsL WlLhlngs was klnd of a glmmlck. ?ou can LweeL your welghL Lo your frlends!"
was one of Lhe ways lL was orlglnally plLched. 1haL's of course noL parucularly
lnLeresung, buL LhaL was noL Lhe purpose of Lhe devlce. 1he scale ls Lhe avaLar Lo
WlLhlngs' healLh servlce LhaL helps you Lrack your welghL. 1he scale orlglnally
dlerenuaLed WlLhlngs from oLher welghL Lracklng apps, buL Lhe value ls noL ln Lhe
scale, buL ln Lhe servlce, whlch you can only fully experlenced uslng oLher avaLars,
such as Lhe web slLe and Lhe phone and LableL apps.

WlLhlngs has slnce expanded Lhe servlce Lo lnclude a bunch of oLher hardware and
soware avaLars. Agaln, Lhe value ls noL ln Lhe devlces, buL ln Lhe knowledge LhaL
Lhey creaLe by collecung slmple pleces of lnformauon and Lhen provldlng users wlLh
Lhe full power of cloud-based servlces Lo make use of LhaL plece of lnformauon.
WlLhlngs can keep addlng avaLars, new sensors and new ways Lo dlsplay Lhe
lnformauon Lhe sensors collecL, wlLhouL fundamenLally changlng Lhe promlse of Lhe
servlce.

22
SmarL 1hlngs clearly sLaLes lLs servlce oerlng rlghL up fronL on Lhelr slLe. 1he rsL
Lhlng Lhey say abouL Lhelr producL llne ls noL whaL Lhe funcuonallLy ls, buL whaL eecL
Lhelr servlce wlll achleve for Lhelr cusLomers. 1helr hardware producLs' funcuonallLy,
how Lhey wlll Lechnlcally sausfy Lhe servlce promlse, ls almosL an aerLhoughL.
23
Compare LhaL Lo x10, Lhelr splrlLual predecessor LhaL's been ln Lhe buslness for more
Lhan 20 years. All LhaL x10 Lells ls you ls whaL Lhe devlces are, noL whaL Lhe servlce
wlll accompllsh for you. l don'L even know lf Lhere lS a servlce. Why should l care LhaL
Lhey have modules"? l shouldn'L, and l don'L.
24
llnally, l wanL Lo Lalk abouL Lhe change ln Lhe buslness sLraLegy LhaL all of Lhls enLalls.
23
Connecung everyday ob[ecLs Lo Lhe cloud lsn'L lnLeresung ln lLself. 1haL's why Lhe
naive value proposluon of many consumer elecLronlcs companles ls oen so
laughable. Any reason for you Lo be able Lo check wlLh your phone wheLher your
sLove ls on, or your garage door ls open, ls conLrlved and of Leruary value Lo
consumers, aL besL, buL LhaL's whaL we're gemng a loL of.
26
l Lhlnk LhaL Lhe blggesL change ls LhaL once you can unlquely ldenufy an ob[ecL and
auach lL Lo an onllne servlce, you can fundamenLally change Lhe ownershlp model.
?ou are no longer selllng ob[ecLs, generaung revenue ln slngle-unlL quanuues, you
are selllng servlces LhaL Lhe physlcal ob[ecLs are avaLars of. Companles ln Lhls new
model no longer sell producLs buL supply servlces, llke Lhe old phone company used
Lo, buL applled Lo everyLhlng.
27
uurable ob[ecLs are no longer owned, buL renLed. lor example, much heavy
consLrucuon equlpmenL ls noL owned by consLrucuon companles, buL renLed by Lhe
mlnuLe LhaL's lL's acLually runnlng, wlLh LhaL LelemeLry collecLed by a company LhaL
provldes Lhe consLrucuon equlpmenL servlce. lf a plece of equlpmenL breaks down,
Lhe company ls lmmedlaLely noued and Lhey send ouL a replacemenL equlpmenL.
1haL's parL of Lhe servlce, because Lhey're noL selllng a dlgger, buL Lhe servlce of
dlgglng. 1he same model ls rapldly Lrlckllng down Lo every form of heavy
LransporLauon from Zlpcar Lo ScooL Lo blke renLal.
28
Consumables become servlces. We already have Lhls ln some respecL wlLh Amazon
subscrlpuons and all of Lhe varlous 8lrchbox-llke subscrlpuon servlces, buL whaL
happens when processlng becomes cheap enough LhaL Lhese producLs can reorder
Lhemselves? Well LhaL already exlsLs for lnk[eL prlnLer lnk because lnk[eL prlnLers are
one of Lhe rsL common lnLerneL of 1hlngs devlce. lcLure Lhls servlce mapped Lo
everyLhlng LhaL you use up. 1haL's comlng.
29
Also leL me make lL clear LhaL Lhe servlce model--llke any Lechnology-has a posluve
slde and a negauve slde. Cn Lhe posluve slde are all of Lhlngs l [usL descrlbed:
predlcuve behavlors, access Lhrough many dlerenL LouchpolnLs, a hollsuc vlew of
someone's llfe, a wln-wln buslness model LhaL Lrades people's cognluve overhead for
a recurrlng revenue sLream for Lhe servlce provlder. ?ou never have Lo Lhlnk abouL
runnlng ouL of cereal, or anyLhlng else, and Lhe producLs you use lmprove based on
how you use Lhem, eLc. Cn Lhe oLher hand, Lhe Lhlngs ln your llfe are lncreaslngly
embedded wlLh a slngle corporaLe enuLy. 1oday door locks don'L sLop worklng when
Lhe company LhaL made Lhem goes ouL of buslness, or geLs hacked, buL ln Lhe servlce
model Lhey mlghL, and a Sony radlo doesn'L only geL sLauons LhaL Sony has approved,
buL ln Lhls model lL mlghL. ?ou already experlence some of LhaL ln Coogle or Apple's
ecosysLem and lL's only golng Lo become more pervaslve, so LhaL's someLhlng we
need Lo keep ln mlnd as deslgners of such sysLems.
30
now l would llke Lo Lalk abouL how Lhls ls changlng deslgn pracuce. When l sLarLed
20 years ago, lnLeracuon deslgn was a comblnauon of graphlc deslgn and cognluve
psychology. 1here was parL of lL LhaL was abouL guldlng people vlsually around a aL
surface and anoLher parL LhaL was abouL undersLandlng Lhelr lnLenL so LhaL you could
gulde Lhem beuer. ln Lhe lo1 ux space, graphlc deslgn has been largely replaced by
lndusLrlal deslgn, whlch guldes people's use of an ob[ecL, and cognluve psychology
has been replaced wlLh servlce deslgn, where you pro[ecL speclc buslness models
Lhrough a range of cusLomer lnLeracuons. Powever, Lhls ls noL a slmple one-Lo-one
Lradeo, slnce lo1 ux deslgn sull lncludes all of Lhose vlsual deslgn and cognluve
psychology quesuons, plus a whole new seL of challenges because Lhe experlences
span muluple LouchpolnLs, muluple avaLars.
31
Cne Lhlng LhaL's for sure ls LhaL Lhe era of a hero deslgn ob[ecL-LhaL slngle Lhlng,
wheLher ls a devlce, a posLer, or a Web page, LhaL ls Lhe prlmary focus of a deslgn
process-LhaL era ls over. Lvery servlce avaLar, whaLever lL ls, ls merely one faceL of a
larger whole, and deslgn as a pracuce ls no longer Lhe creauon of ob[ecLs, buL a
pracuce for solvlng problems.
32
8ecause deslgn LhaL focuses only on Lhe experlence of a slngle LouchpolnL produces
gadgeLs, noL soluuons Lo problems, and gadgeLs are Lhe klnd of Lhlng LhaL lluer your
closeL, noL Lhlngs you use every day and LhaL you love and assoclaLe wlLh your own
happlness.
33
1he deslgn arufacLs you make are all over Lhe board. l'm golng Lo go Lhrough a
handful LhaL l've found Lo be useful, buL llke l sald aL Lhe beglnnlng, lL's a new
dlsclpllne and one of Lhe mosL lnLeresung deslgn challenges ls acLually deslgnlng Lhe
arufacLs LhaL use Lo undersLand Lhe problems you're Lrylng Lo solve and how Lo
communlcaLe your ldeas Lo Lhe approprlaLe audlences ln Lhe approprlaLe way.
34
Many of my conversauons wlLh cllenLs sLarL our wlLh a loL of Lhe Lyplcal whlLeboard
and posL-lL noLe dlscusslon where we descrlbe whaL Lhey belleve Lhe value of Lhelr
servlce wlll be Lo poLenual cusLomers. 1yplcally no one has very good answers aL Lhls
polnL, so we Lake guesses.

8ased on Lhose guesses l puL LogeLher a hlgh-level descrlpuon of whaL pleces Lhe
servlce ls golng Lo conLaln. 1he dlagram separaLes daLa sources, analyucs, dlsplays,
conLrollers, sensors and acLuaLors so l can sLarL a dlscusslon abouL how many avaLars
we're acLually golng Lo deslgn and Lhe value each one adds Lo cusLomers and Lo Lhe
buslness. 1he dlagram ls super slmple, buL one of lLs goals ls Lo relnforce LhaL all of
Lhe pleces are connecLed and LhaL nelLher Lhe app or Lhe waLch ls more lmporLanL.
1haL helps me Lalk Lo sLakeholders abouL Lhe scope of whaL we're golng Lo bulld.
33
nexL, someone on Lhe producL Leam needs Lo creaLe one of Lhese. lL's a 8uslness
Model Canvas, from Alex CsLerwalder and ?ves lgneur's 8uslness Model Cenerauon
book, whlch ls a slngle-page checkllsL LhaL descrlbes whaL Lhe servlce ls golng Lo do
for people, who's lL's for, and why Lhey'll pay for lL.

1he user experlence of mosL mulu-LouchpolnL lnLerneL of 1hlngs servlces ls much
more closely ued Lo Lhelr buslness model Lhan ln a Lradluonal Lradluonal consumer
elecLronlcs company. lor example, say you're maklng a smarL lock. Pow much
funcuonallLy should you puL ln Lhe lock mechanlsm? Well, LhaL depends on whaL
cusLomer problem Lhe servlce ls Lrylng Lo address. lf you prlmary audlence ls Alr8n8
properLy owners who wanL Lo creaLe a unlque code for each LenanL, LhaL's a dlerenL
seL of needs Lhan for a securlLy-concerned homeowner. Same lock, perhaps even
same locklng hardware, buL a dlerenL emphasls based on Lhe key buslness model.
36
Aer Lhe 8uslness Model Canvas, we're ln sLralghuorward ux deslgn for a whlle. l llke
personas, so l creaLe 4-3 dlerenL users of Lhe producL based on faceLs of Lhe
buslness model-here's one we dld for a SLanford pro[ecL. 1hese can be bullL wlLh
some up-fronL eLhnography, whlch ls whaL l prefer, or one of Lhe oLher Lechnlques
LhaL geLs you a reasonable persona.
37
1hen you geL Lo use scenarlos and sLoryboards. 1hese help Lo Lell Lhe sLory of how
people lnLeracL wlLh varlous avaLars, whaL role each does and does noL play ln
people's llves. lL's useful for boLh communlcaung a vlslon among Leam members and
sLakeholders, and you oen learn lmporLanL ux lessons [usL by Lalklng Lhrough a
couple of sLorles. 1hls ls one we dld wlLh SmarL ueslgn. My approach Lo wrlung
scenarlos ls Lo sLarL by assumlng LhaL people don'L wanL Lo use your producL, buL
your producL helps Lhem accompllsh a goal fasLer, easler or more eleganLly Lhan Lhey
would have oLherwlse, so my scenarlos and sLoryboards oen feaLure slLuauons
where personas don'L use Lhe producL much, or only use one aspecL of lL, buL lL plays
a key role ln Lhelr llves.

SLoryboard by SmarL ueslgn
38
nexL, you sLarL brlnglng Lhlngs LogeLher by pumng varlous LouchpolnLs agalnsL
lmporLanL goals your personas' and your sLakeholders' goals. 1hese LouchpolnLs can
be full-edged servlce avaLars or [usL lnLeracuons wlLh Lhe servlce. Cne way Lo do
LhaL ls by maklng a swlm lane dlagram LhaL puLs Lhe LouchpolnLs across Lhe Lop and
Lhen maps lnLeracuons across.

1hls ls a chunk of a [ourney dlagram l puL LogeLher for Sleo, Lhe game company, a
couple of years ago. We developed a seL of goals for Lhe ux of Lhe devlces, and Lhose
goals spanned Lhe documenLauon, Lhe lnLerface on Lhe devlces, and Lhe Web slLe.
Across Lhe Lop ls a seL of LouchpolnLs, whlch range from Lhe LexL on Lhe box Lo Lhe
lnsLrucuons LhaL people won'L read, Lo Lhe rmware and Lhe slLe. PorlzonLally l
creaLed swlm lanes for people's mouvauon, whaL lnformauon we gave Lhem, whaL
Lhey can do aL LhaL polnL, eLc. And Lhen across Lhls grld l mapped ouL a sLraLegy for
lnLeracuons LhaL we hoped would lead people Lo do Lhlngs such as downloadlng Lhe
soware and buylng new games.
39
nexL comes Lhe deslgn for all of Lhe avaLars. l dlvlde Lhls process lnLo 2u, whlch ls
essenually screen-based lnLeracuon deslgn llke how you would deslgn an app or a
Web page.
40
.and 3u. Pere we Lake our process cues from lndusLrlal deslgn pracuce and sLarL
wlLh skeLches Lhen move Lo models and renderlngs, wlLh looks-llke proLoLypes.

1hls ls a renderlng from a pro[ecL we dld wlLh SmarL ueslgn.
41
.and works-llke proLoLypes. 1hls ls by ?olchl nagashlma for a pro[ecL he dld, buL l
Lhlnk lL's a good represenLauon of whaL our sLu looks llke aL Lhe crazler end.

1hls process ls essenually Lhe Lradluonal consumer elecLronlc producL developmenL
process, don'L Lhlnk lL's easy. lL's noL and lL's where much of your ume ls golng Lo go,
buL lL's falrly well esLabllshed, so l'm glosslng over lL here.
42
ln facL, we oen make Lhe renderlngs less reallsuc so we can focus on Lhe concepL,
raLher Lhan Lhe acLual execuuon. 1oday's soware make lL really easy Lo produce
lmages or 3u models LhaL are compleLely nonvlable as producLs and aL Lhls sLage
we're much more lnLeresLed ln undersLandlng whaL's valuable Lhan maklng Lhlngs
LhaL are preuy. 1oo much dellLy leads Lo sLakeholders and poLenual cusLomers
focuslng on Lhe wrong aspecLs of Lhe devlce.

Also by SmarL ueslgn
43
And of course ln beLween many of Lhese sLeps ls a loL of user research. Pere's a
phoLo of one of my colleagues dolng prellmlnary eLhnography for a shopplng
producL. l'm a rm bellever LhaL no ldea ls worLhwhlle unless lL's been valldaLed by
people who are noL your Leam. CusLomer LesL and valldauon ls a crlucal parL, and we
Lry Lo lnclude lL as oen as we can, whlch ls never as much as we'd llke.
44
ln pracuce Lhls means LhaL Lhe range of Lhlngs LhaL need Lo be deslgned ls enormous.
Cne mlnuLe you're debaung Lhe buslness model, Lhe nexL you're slmng nexL Lo an
englneer plcklng ouL sensors. 1here's a loL of uncerLalnLy.
43
8roadly speaklng, our process ls lLerauve and we Lry Lo use aglle and lean besL
pracuces, buL as any of you who have ever done LhaL wlLh a cllenL know, lL's Lough Lo
be docLrlnalre. We Lry Lo be as llghL on our feeL on Lo commlL Lo plasuc or code as
laLe as posslble, whlle Lesung for cusLomer value early and oen.
46
Cne of Lhe blggesL quesuons LhaL our cllenLs have, and frankly we have, ls how Lo know lf a
largely new Lechnology or servlce ls golng Lo be successful wlLh people wlLhouL acLually golng
Lhrough Lhe expense and process of developlng a compleLe producL. no one ln an lnLervlew
or a focus group ls golng Lo Lell you LhaL Lhey'd pay $230 for a wl-connecLed LhermosLaL
wlLh predlcuve behavlor almed aL reduclng energy usage when Lhey've never seen such a
Lhlng before. nesL was able Lo bulld one because 1ony laddell had Lhe bankroll and
connecuons of Lhe guy who had creaLed Lhe lod, buL mosL oLher pro[ecLs have lndlegogo
and hunches aL besL.
Pere's Lhe process l'm currenLly uslng Lo help me push Lhe bounds of exlsung producLs and
servlces, and Lo do lL wlLh some ldea of whaL people value and whaL's Lechnologlcally
feaslble. 1he Lop layer descrlbes a hlghly lLerauve, hlghly explorauve phase LhaL exlsLs solely
Lo help us undersLand whaL experlences provlde value Lo people and whaL Lechnology our
Leam can reasonably dellver. Aer Lhls phase ls done, we Lhrow everyLhlng we dld away and
sLarL deslgnlng Lhe real producL from scraLch, and LhaL's where Lhe proLoLyplng process
sLarLs.
robes, whlch l based on Lhe Lechnology probes ldea descrlbed by Plllary PuLchlnson and
Wendy Mackay, are Lhlngs LhaL resemble nlshed producLs-perhaps Lhey're nonfuncuonal
apps or boxes wlLh llghLs-buL exlsL Lo explore hypoLheses abouL whaL klnds of avaLars
people wlll respond well Lo. We glve Lhem Lo people for a couple of days or a week and Lhen
we lnLervlew Lhem when Lhey've had a chance Lo llve wlLh Lhem. roofs of concepL exlsL Lo
LesL Lhe core Lechnologlcal ldeas. l don'L care lf we have Lo Lhrow a $3000 server aL Lhe
problem when Lhe end producL has Lo cosL $10, ls lL posslble for us Lo do LhaL Lhlng A1 ALL?
Aer we've done Lhese Lwo Lhlngs lLerauvely, we have a preuy ldea of whaL goes lnLo Lhe
47
So, sLepplng back, why do Lhls? Why go Lhrough Lhls process where l as a deslgner
don'L end up dolng LhaL much deslgn? Well, lL's because l Lhlnk LhaL Lhere's
enormous posslblllLy ln Lhe lnLerneL of 1hlngs, buL lL's noL ln Lhe Lhlngs.
48
1o me why we make lnLerneL of 1hlngs ls noL abouL creaung new gadgeLs, buL
developlng servlces LhaL exLend ourselves from our bodles ouL lnLo Lhe world. Cur
blggesL challenge as lnnovaLors, deslgners, developers and enLrepreneurs wlll noL be
addlng cheap elecLronlcs Lo everyday ob[ecLs. 1haL's a solved problem. Cur challenge
wlll be Lo undersLand how Lo use Lhose Lhlngs Lo creaLe value, and Lo creaLe
susLalned and susLalnable change, ln a world where Lhere ls an enormous amounL of
poLenual for embedded, neLworked Lechnology, and an enormous number of paper
Lhln, one noLe gadgeLs. l once heard an lndusLrlal deslgner say LhaL everyLhlng you
deslgn ls garbage from Lhe mlnuLe lL comes o Lhe assembly llne, and LhaL lL's your
responslblllLy Lo keep lL ouL of Lhe Lrash bln for as long as posslble. l Lhlnk LhaL ln Lhe
lnLerneL of Lhlngs LhaL ls especlally Lrue and Loday l Lhlnk LhaL our blggesL challenge
as deslgners wlll be Lo Lhlnk abouL why we wanL Lo make Lhlngs aL all, and Lhen how
we can make Lhem provlde Lhe mosL value, for Lhe leasL cosL, for Lhe mosL people, for
Lhe longesL ume.
49
1hank you.
30

You might also like