Innovation Diffusion and Facebook

15
INNOVATION DIFFUSION and FACEBOOK criteria to the successful introduction of innovations - by Werner Iucksch October.2008

description

A study of how facebook fits in existing innovation diffusion theory.

Transcript of Innovation Diffusion and Facebook

Page 1: Innovation Diffusion and Facebook

INNOVATIONDIFFUSIONandFACEBOOK‐ criteriatothesuccessfulintroductionofinnovations­

byWernerIucksch

October.2008

Page 2: Innovation Diffusion and Facebook

INTRODUCTION

Networkstudyisanewfieldforscience(Castells,2000)(Watts,2003),but

it’snowwidelyacceptedthatwhennodes(i.e.anindividual,inthecase

discussedhere)areconnectedinnetworks(suchassocialnetworks),theycan

behaveverydifferentlythaniftheyaresetapart(e.g.peoplerioting).By

organizingnodesinnetworks,thesystemasawholeisveryresilientandrobust

(Dijk,2006),butitalsomakesoutcomesofeventsveryhardtopredictorto

mould.(Watts,2003)

However,networksdochange.Theyadoptnewinformationanddiscard

uselessinformation.Thewaythisprocessofintroduction,adoptionanddiffusion

happensacrosssocialnetworkshasbeenstudiedforalongtime,buthow

individuals(nodes)caninfluencesociety(network)onlystartedtobecomeclear

intheearly70s.Muchprogresshasbeendonesincethen.Astheinternet

developed,someweb‐basedinnovationswereintroducedsuccessfully,

spreadingunderthemeasurableenvironmentofdatabanks,thusenablingsocial

networkscientiststoworktogetherwithinnovationresearcherstocreatefine

tunedtheoriesonhowchangecanbeintroducedinanetwork.Thatis,thestudy

ofhowtoprogramthenetworktoadoptagiveninnovationisadvancing.

Thispaperintendstorelyonsocialnetworktheoryandinnovationdiffusion

modelstooutlineandtest,againstarealinnovation,criteriathatarepredictedto

beassociatedwithlargescalesuccessofanetwork‐basedinnovation.Itisnotthe

objectivetopresentanexhaustivelistofwhatittakestohaveasuccessful

innovation,butitisexpectedthatsuccessfulinnovationsshowallofthe

characteristics.Totestit,thepaperwilldiscusswhetherFacebook,arguablythe

Page 3: Innovation Diffusion and Facebook

mostsuccessfulonlineinnovationinthelast5years,showssignsofstrategic

choicesandproductdevelopmentsthatsuitthecriteria.

NETWORKTHEORY,INNOVATIONDIFFUSIONandSUCCESSPREDICTION

Innovationcanappearanywherewithinanetwork;however,itismore

likelytoappearinthepartsthatarelessstronglyconnectedtothe“centre”or

wherethereislessaffinitywithestablishedculture.Granovetter(1973)argued

thatwhenlookingforajob,itisfarmorelikelytogetinformationabout

opportunitiesbyusing“weak”tiesthanwith“strong”ones,becausethe

informationsomeonehasaccesstowhenusingweaktiesaremorelikelytobe

differentfromthattheindividualalreadyhas,insuchcircumstance,thisweaktie

wouldbea“bridge”.Inhiswords:

“Thefewerindirectcontactsonehasthemoreencapsulatedhewillbein

termsofknowledgeoftheworldbeyondhisownfriendshipcircle”

(Granovetter,1973,p.1371)

Similarly,Rogers(2003)writesthatalthoughthelevelofhomogeneity

withinagroupmakesthecommunication(andthereforediffusion)more

effective,“theverynatureofdiffusiondemandsatleastsomedegreeof

‘heterophily’’(levelofheterogeneity).Ideally,accordingtoRogers(2003,p.19),

toincreasetheprobabilityofhavingasuccessfulinnovation,theinnovatorand

therestofthenetworkwouldbehomophilousineverything,exceptforthearea

oftheinnovation.Thisisabalancethatisdifficulttoachieve.Toomuch

homophilymayresultintolessradical,uninsterestinginnovationandtoomuch

Page 4: Innovation Diffusion and Facebook

heterophilymayresultintohighlyinnovativeideasthatalargeportionofthe

audiencesimplydon’tunderstand/havehighrejectionlevels.

Soasinitialthreecriteriaofasuccessfulweb‐basedinnovation:

1) Innovationislikelytoappearawayfromthecentreofthenetwork

2) Theinnovatorsandtherestofthepublicareexpectedtohaveadegreeof

heterophilybetweenthem,butideallyonlyinthekeyareaofthe

innovation.

3) Thenetworkinwhichtheinnovationappearsisexpectedtohavea

relativelylargenumberofsocialclustersweaklyconnected.

Thesefirstfewcriteriapointtowheretheinnovationisexpectedtobe

bornandalsogivessomecharacteristicsofthenodesandactorsthatarethought

tobepresent.However,thelivesofordinarypeoplearenotmajorlyconstituted

ofweakconnectionsandheterophilouscommunication,thusitisimportantto

understandhowtobreakwiththe“pro‐homophily”environmentthatstrong

bondscreate.

TheworkofWatts(2003)aboutsocialcontagionshedssomelightover

thisissue.Accordingtotheauthor,socialcontagionofideasoccursundervery

specificcircumstances.Beforeaninnovativeideaisadoptedwidely,ithasto

percolatesuccessfullyintosocialclusters.Dependingontheconnectivityofthe

firstpercolatingcluster,itmaycascadethroughtopartsofthenetworkoreven

thewholenetwork,inwhatisusuallycalledaglobalcascade.

Thecreationofthecrucialpercolatingclusterisaprocessthatis

explainedbytheconceptofcriticalthresholds.Theinnovatorneedstobe

Page 5: Innovation Diffusion and Facebook

connectedtopeoplewithlowadoptionthresholdstothatkindofinnovation,

allowingfortheexistenceofearlyadopters(Valente,1996).Thesefirstfew

peoplecanformtheinitialpercolatingclusterthatwillbeconnectedtoanother

individuals,thatatthispointstilldidn’tadopttheinnovation(nodesthatare

“switchedoff”).

Thewaytheinitialclusterisconnectedtonodesthatare“off”andtheir

levelofconnectivitytoother“off”arealsocrucial.Theyareasimportantasthe

thresholdofeachnodeinenablingtheirconversionandallowforthecascadeto

happen,inaprocesscalledphasetransition.Aswecanseeinfigure1,thehigher

thenumberofneighboursan“off”nodehas,themoredifficultitistoadoptan

innovation,thus,thesenodeswouldneedtohaveaverylowthresholdtoadopta

giveninnovation.Anotheroptionwouldbetotryto“infect”nodesthathave

fewerconnections;inthiscaseahigherthresholdwouldbetolerable.However,

globalcontagiononlyhappenswhenconnectionlevelisratherhigh(Watts,

2003).

Theseviewsfurtherdevelopsourcriteria:

4) Peoplearoundtheinnovatormusthavelowresistancetowardsthe

innovativeidea/product.

5) Theinnovationmustbeabletoinfectawell‐connectedclusterinorderto

spreadglobally

Page 6: Innovation Diffusion and Facebook

Figure1–PhaseTransition

(basedonWatts,2003,p.238)

Iftheinnovationisgoingtobesuccessful,anotherphenomenashouldbe

observedjustafteritbeginstogetadoptedoutsidetheimmediateboundariesof

itslaunchingcluster.Innovationsthatareinteractiveand/orarebasedon

networkstendtobecomemoreusefulasthenumberofusersincreases,inwhat

isknownasnetworkeffect(Farrel&Klemperer,2007).Thenumberofuserswill

thenincreaseuntilapointwhenthepointatwhichreciprocalbehaviourgets

self‐sustainable,thatis,gainscriticalmass(Markus,1987,p.496).Onceit

happens,theinnovationwillcascadeactivatinghighnumber“offnodes”very

quickly,soonthethresholdsofthewholenetworkwillbeachieved(figure2).

Page 7: Innovation Diffusion and Facebook

Figure2–Criticalmass

(basedonRogers,2003,p.314)

Achievingcriticalmassearlyonisimportantinacompetitivemarket.Ifa

giveninnovationcanachievethisstagebeforeanalternativeconcept/innovation

forthesameproblemestablishesitself,itwillgrowinsizeandpeoplebeginto

gravitatetowardstheinnovation,inwhatisknowasaPowerLaw,thatis“The

richgetricher”.(Barabàsi,2003).Thisdiscouragescompetitorsfromentering

themarket.

Thus,asfinalcriteriatothepurposesofthispaper,wehavethefollowing

points:

6) Theinnovationshouldbeabletobenefitfromnetworkeffects,becoming

morerelevantasitgrow.

Page 8: Innovation Diffusion and Facebook

7) Successfulinnovationsarelikelytobefastintoachievingcriticalmass,

pointwhichit’sgrowthbecomeexponential.

Inthenextsectionofthispaper,allthesecriteriawillbeputtothetest

againstwhatFacebookdidtoachievethedominantpositionithastoday.

FACEBOOK

Facebook(www.facebook.com)isasocialnetworksite(SNS).Accordingto

boyd&Ellison(2008),SNS’sare“web­basedservicesthatallowindividualsto(1)

constructapublicorsemi­publicprofilewithinaboundedsystem,(2)articulatea

listofotheruserswithwhomtheyshareaconnectionand(3)viewandtraverse

theirlistofconnectionsandthosemadebyotherswithinthesystem.”

Itwasalreadyexposedthatitisaverysuccessfulinnovation,withmorethan

130millionuniquevisitors/month(comScoreWorldMetrix,2008),butthe

choiceforFacebookgobeyondnumbers.AsEllison,Steinfield,&Lampe(2007,p.

1144)putit:

“Facebookconstitutesarichsiteforresearchersinterestedinthe

affordancesofsocialnetworksduetoitsheavyusagepatternsand

technologicalcapacitiesthatbridgeonlineandofflineconnections.”

Previoussocialnetworkstudiesandhardstatistics,therefore,suggest

thatFacebook’sstructureisidealtoastudythatintendstobringtogether

NetworkTheoryandInnovationDiffusion.

Page 9: Innovation Diffusion and Facebook

FACEBOOKANDTHECRITERIA

Thefirstcriterionthatwasoutlinedisaboutthelocationofwhere

innovationappears.Itwaspredictedthatitwouldappearawayfromthecentre

ofthethenestablishedculture.Facebookwaslaunchedfromastudentroom

insideauniversity.Notanyuniversity,butoneofthemostinnovative

universitiesintheUnitedStates(Oldach,2008).Althoughitcanhardlybeargued

thatHarvardisatthemarginoftheeducationalsystem,universitiesareonthe

fringeofsociety.Theinnovationwasborninanappropriateplaceifweconsider

Facebookwaslaunched7yearsafterthefirstrelevantSNS(figure3).Soitwas

alreadycompetingwiththecorporateworld.Also,it’swasnotaformalproject

fundedbytheuniversity,sothemarginalityoftheinnovatorcouldbeobserved.

Therearesignsthatthecompanyactivelyfoughttoachieveabalance

betweenheterophilyandhomophily.Jones&Soltren(2005)documentedthat

Facebook,atthattimereachingabout2,000collegesintheUSA,wasn’tone

singularwebsite.Accordingtothem,“’Facebook’[was]acollectionofsites,each

focusedononeofthe2,000individualcolleges.”,withpermissionofaccess

restrictedtothecollege/universitytheuserbelongedto.Thisguaranteed

homophilousconditionsinmanyaspects,suchasagegroups,educationlevel,

eveneconomicbackground,toacertainextent.TheSNS’sfounder,Mark

Zuckerberg,wasastudentatthetimeofitslaunch,sohewasabletodevelopthe

innovationwithanhighlyintuitiveinterfacetoitstarget,aswellasrelevant

structure.Bydoingso,Facebookdidn’tfacemajor“noise”initsdiffusion.This

was,perhaps,amajorearlyadvantagethatthecompanyhadwhencomparedto

otherSNS’savailableatthattime.Theheterophilylevelsbetweeninnovatorand

Page 10: Innovation Diffusion and Facebook

theinnovation’saudienceresidedbasicallyinthecapacitytoprogramasocial

networksite.However,asSNS’swerealreadyavailable,heterophilywasnota

distinctivefeatureofthewebsiteinit’searlystages.Onceitwasopenedtothe

generalpublic,inSeptember2006,ashockofculturesbetweenstudentsand

worker/familiesmightbepresent.Asthisisthecase,thesecondpointoutlined

inourcriteriacouldbeobservedpartially.

Figure3–SNS’sTimeline

(boyd&Ellison,2008,p.212)

Page 11: Innovation Diffusion and Facebook

Atthesametime,byfocusinginacademicinstitutions,Facebookbecame

agreatalternativetomanypeoplewhoselifestagecarriesalotdisruptionwith

it.Universitylifemanytimesdemandmovingoutofhome,leavingfriends

behind,forinstance.Thiswayalmosteverystudenthastomakenewsocial

circles,whichtheydo;however,bondingtakeslongtoconsolidate.Thenumber

ofweaktiesofeachstudenttendstobeverybiginthesecircumstances,like

criterion3suggestsitwouldbe.Facebooktookthisintoconsiderationandbuilt

avarietyoftoolssuchas“Wall”,“NewsFeed”,“Photo”and“Events”tokeepusers

informedabouteachother,sharemoments,enrichsociallifewithouttheneedto

browseprofiles.Thiswaythelinksdonotdissolve,addingrelevancetotheSNS

asitgrew.

ZuckerbergcertainlyknewhowvaluableasystemsuchasFacebookcould

be,ashegrewupinalittletownaround300kmawayfromHarvard.Online

profiles,friendsandgroupsareconvenientwaysto(re)buildsomeone’ssocial

capital,thatis,theresourcesthat“allowapersontodrawonresourcesfrom

othermembersofthenetworkstowhichheorshebelongs”(Ellison,Steinfield,&

Lampe,2007).Additionally,Facebookhelpedtosolveabigprobleminacademic

life:itmadeeasytoorganizestudygroups,whichwassomethingmissingat

Harvard(TheCrimsonStaff,2005)andotheruniversities.Lateron,whenthe

websitewasalreadyopenforthegeneralpublic,thelaunchofafreeplatformto

developerstocreateapplicationsallowedforintroductionoftoolsthatcouldbe

relevanttoveryspecificregions/groups.Alltheseactionsshowthatlowering

thresholdofadoption(criterion5)wasoneoftheconcernsofthecompany.

Page 12: Innovation Diffusion and Facebook

Theinitialstructureofthewebsiteandhowitexpandedanddeveloped

aredeeplyrelatedtothelasttwocriteria.Bylimitingaccesstoprofilesofthe

schooltheuserbelongedto,thecompanychosetoacceleratetheachievementof

criticalmassineachoneofthoseschoolsindependentlytowhathappenedin

otherschools.Atthistime,althoughthetotalnumberofFacebookusersgrew

exponentially,networkeffectswerelimitedforitwasn’tpossibletocontact

usersfromotherschools.Thiswasachoiceofthecompany;itfeltthatitwas

moreimportanttoestablishlocalrelevancybeforegivingusersthechanceof

establishinglongdistanceconnections.Infact,thischaracteristicpersistsuntil

today,asitwasfoundthat‐contrarytootherSNS’s‐Facebookusersuseitasa

complementtoexistingoff‐lineconnections(Ellison,Steinfield,&Lampe,2007).

Oncethebaseincollegesandhigh‐schoolsweresolid,though,andthe

wallsaroundthemwerelifted,Facebookgrewasneverbefore.Familiesand

friendsthatwereleftbehindwhensomeonehadtouniversityandcolleagues

thatnolongerweretomakepartofdailylifeaftergraduationcouldnowbe

accessedanywhereintheworld.Thewebsitejumpedfrom5.5millionactive

users,inDec/05,to12.5million,inDec/06(closetothetimeitopened

registrationtoanyone).Thirteenmonthsaftertheopening,inOct/07,italready

hadmorethan50millionusers.(Facebook,2008).Suchstatisticsstrongly

suggestthatthewebsiteanditsactionsgeneratedvaluetousersasmorepeople

werejoining.

Page 13: Innovation Diffusion and Facebook

CONCLUSION

Basedintheliteratureavailableaboutsocialnetworksandinnovation

diffusionthetextoutlinedcriticalpointsthatarethoughttobepresentin

successfulinnovation.WhencomparingthesepointstoFacebook,asuccessful

recentinnovation,itwaspossibletoseeallpointsdescribed;however,they

didn’talwaysapplysmoothlyintothecasedescribed.

Inatleastoneoccasion,ourstudiedwebsiteneededtocompromise

betweentwooftheelementsthepaperwasobserving(“criticalmass”and

“networkeffects”).Facebookhadtochoseinitsearlystageswhichoneofthese

elementsitwantedtoprivilege,leavingtheothertoasecondstageof

development.

Althoughabalancebetweenhomophilyandheterophilycouldbeseen,it

isdifficulttoaccesswhetherFacebookactuallyisastrongexampleofsuch

delicatebalance.Itispossiblethatother,more(orless)innovativeSNS’sactually

hadabetterbalancethanFacebook,butwereeclipsedbyFacebook’ssuccessin

otherareas(speciallyinthe“speed‐to‐criticalmass”,whichcanbeverypowerful

ineliminatingcompetitors).Thehigherdegreeofheterophily,inthiscase,

actuallyresidedinthecontextinwhichFacebookwaslaunchedwhencompared

tothatofotherSNS’s,ratherthanthatbetweeninnovatorandadopters.

Theseobservationsanddifficulties,however,donotinvalidatethe

criteriathatwereselected.Theverynatureoftheanumberofthekeynetwork

theoryconceptsthatwereexposedinthispaperrequiresocialinteractions,

marketcontext,collectivedecisionmakingandadequatetimingtoanextentthat

Page 14: Innovation Diffusion and Facebook

todatenomodeliscapableofpredictingnetworkbehaviourwithaccuracy

(Watts,2003,p.29).

Overall,itispossibletosaythatthecriteriaselectedcouldbeseenas

importanttothesuccessoftheFacebook,buttheymaynotbenecessaryin

100%ofthecasesandsometimesitwillbenecessarytoprioritizeonecriterion

overanother.Suchconclusionindicatesthatcompaniescanprobablybenefitif

theyactivelyusethesecriteriainthedevelopmentofbusinessplansandproduct

development,butitisimportanttoallowroomforchangeofplans.Thisway

networktheoryandinnovationdiffusionresearchcanhelppavethewayfor

otherstoalsoprogramthenetwork.

Page 15: Innovation Diffusion and Facebook

BibliographyBarabàsi,A.‐L.(2003).Linked­Howeverythingisconnectedtoeverythingelseandwhatitmeansforbusiness,scienceandeverydaylife.Cambridge,Massachusetts,USA:Plume.

boyd,d.m.,&Ellison,N.B.(2008).SocialNetworkSite:Definition,HistoryandScholarship.JournalofComputer­MediatedCommunication(13),210‐230.

Castells,M.(2000).TheRiseoftheNetworkSociety(Vol.1).Oxford,UK:BlackwellPublishers.

comScoreWorldMetrix.(2008,August12).SocialNetworkExplodesWorldwideasSitesIncreasetheirFocusonCulturalRelevance.RetrievedOctober05,2008,fromcomScore.Inc:http://www.comscore.com/press/release.asp?press=2396

Dijk,J.v.(2006).NetworkSociety.ThousandOaks,California,USA:Sage.

Ellison,N.B.,Steinfield,C.,&Lampe,C.(2007).TheBenefitsofFacebook"Friends:"SocialCapitalandCollegeStudents'UseofOnlineSocialNetworkSites.JournalofComputer­MediatedCommunication(12),1143‐1168.

Facebook.(2008).Timeline.RetrievedOctober7,2008,fromFacebook:http://www.facebook.com/press/info.php?timeline

Farrel,J.,&Klemperer,P.(2007).CoordinationandLock‐In:Competitionwithswitchingcostsandnetworkeffect.HandbookofIndustrialOrganization,3,1967‐2072.

Granovetter,M.S.(1973).Thestrengthofweakties.TheAmericanJournalofSociology,78(6),1360‐1380.

Markus,M.L.(1987).Towarda"CriticalMass"TheoryofInteractiveMedia:UniversalAccess,InterdepenceandDiffusion.CommunicationResearch,14(5),491‐511.

Oldach,S.(2008,September).ThePatentScorecard2008­Universities.RetrievedOctober10,2008,fromIntellectualPropertyToday:http://www.iptoday.com/articles/2008‐9‐oldach2.asp

Rogers,E.M.(2003).DiffusionofInnovations(4ed.).NewYork,USA:TheFreePress.

TheCrimsonStaff.(2005,November13).FacingOffOverTheFacebook.RetrievedOctober13,2008,fromTheHarvardCrimson:http://www.thecrimson.com/article.aspx?ref=503343

Valente,T.W.(1996).Socialnetworkthresholdsinthediffusionofinnovations.SocialNetworks,69‐89.

Watts,D.J.(2003).SixDegrees­Thescienceofaconnectedage.NewYork,USA:Norton.