T.Dahlbergetal./ElectronicCommerceResearchandApplications14(2015)265–284Table5
Themesoftechnologyandtechnologicalenvironmentresearch.CategoryReferences#ofDescriptionsincludingspeculations6#ofPro-posedconstructions1#ofEmpirical,prototype,?eld,design,simulation2273EvaluationofAlternatives32Proposalsofm-paymentsystemsProposalsoftoolsormechanismsform-paymenttransactionsProposalsofprotocolsform-paymenttransactionsProposalsoftoolsormechanismsforsecurityandtrustTechnologydescriptionswithafocusonsecurityandtrustTechnologydescriptionsofm-paymentsSemiconductorelements,SIMcards,antennasTotalnumberofpapersAlshareandMousa(2014),deReuveretal.(2015),Gannamanenietal.(2015),Ghezzietal.(2010),HedmanandHenningsson(2012),Lim(2008),Liuetal.(2015),MisraandWickamasinghe(2004),Ondrusetal.(2009)Dahlbergetal.(2003),Dahlbergand??rni(2007),deAlbuquerqueetal.(2014),DennehyandSammon(2015),GaurandOndrus(2012),Hassinenetal.(2008),Hwangetal.(2007),KazanandDamsgaard(2014),Ondrusetal.(2015)AuandKauffman(2008),Dahlbergetal.(2008a,b),GoekeandPousttchi(2010),IsaacandZeadally(2014),KazanandDamsgaard(2013)Apanasevicetal.(2013),Kousaridasetal.(2008),Kungpisdanetal.(2004),Mallat(2007),Olsenetal.(2012),Ondrusetal.(2005),OndrusandLyytinen(2011)ChengandHuang(2013),Lametal.(2003)3337,8,14155,6,24432,27,431130DamsgaardandGao(2005),HedmanandHenningsson(2015),IsaacandZeadally(2013),Konidalaetal.(2012),Lietal.(2012),Linetal.(2008),Maetal.(2013),Apanasevicetal.(2013)JiaandHall(2014),KshetriandAcharya(2012),Magnier-Watanabe(2014)Ondrusetal.(2009)712632321131013Theclock-speedoftechnologiesusedinmobilepaymentser-vicesvariesalso.Forexample,theclock-speedofunderlyingISandnetworkinfrastructuretechnologies,suchastheshiftfrom3Gto4G,arecloseto10years–orevenlonger.MerchantsrenewtheirPOSterminalsevery3to7years,whereasthelifecycleofmobileandsmartphonesistypicallybetween6monthsand2yearsonaverage.Thus,thereshouldbeabetterrecognitionoftheroleofclock-speedintechnicalresearch.Forexample,develop-ingasecurityandtrustprotocolrelatedtoapaymentscenarioforaspeci?cmobiledevicemakeslittlesenseifthatdeviceisnolongeravailableafteroneyear–unlesstheprotocolcanbeeasilyportedtootherdevices.Those25%ofthearticlesthatcomparealternativetechnicalsolutionswouldbene?tgreatlyfromstandardizedteststoevaluateprotocolperformance,softwarequality,datatransmis-sionandhandlingspeed.Wealsopointoutthatmanytechnologi-calarticleslackasolidtheoreticalbasis,whichisrequiredforrepeatedstudiesandtheaccumulationofknowledgeovertime.TheliteraturereviewofDahlbergetal.(2008b)proposedthreeresearchquestionsforfutureresearch.Oneofthemencouragedresearcherstoinvestigatewhatsecurityandtrustmechanisms?tthevarioustypesofmobilepaymentservices(RQ6).Asdiscussedabove,themajorityofthetechnologicalresearchaddressesthisissue.Yet,theliteraturedoesnotreallyprovideanswerstothequestion,asnoarticleaddressesresearchquestionsfromthisper-spective.Thequestionappearstohavebecometoowideforanysinglearticlebutwewouldbene?tfromstudies,whichcomparesecurityandtrustmechanismsinspeci?cpaymentscenariosoraspeci?csecurityissueacrossseveralpaymentscenarios.Fortunately,afewstudieshavetackledthisissuethough.OuandOu(2009)comparedtheirownandsixothernon-repudiationprotocolsintermsof?vecriteria.Konidalaetal.(2012)proposedanewanonymitysolutionandcomparedittocredit/debitcards,contactless(RFID)debit/creditcardssuchasmobileNFC,prepaidcontactlesscards,andanonymouselectroniccashsolutions.YangandWu(2013)developedthreealternativesformicro-paymentschemeswithanabilitytoreturnchangesbasedondifferenthashmodes(counter-modeencryptionhashing,aggressivemodehashingandbalancemodehashing).TheythencomparedthosetoschemeswithsingleandmultiplePayWordchainsintermsofabilitytoreturnchanges,computationalrequire-ments,storagecostsandkeyingofhashoperations.Theuseofhashinginmicro-paymentshasalsobeeninvestigatedbyNanetal.(2009).Theirsolutionisbasedonachaotichashfunction,andiscomparedtothePayWordprotocolandtothedivisionPayWordprotocoltechniques.TimeitselfhassolvedanotherresearchquestionnotedbyDahlbergetal.(2008b).Thestandardizationoftransactionproto-colsandinter-operabilitymechanismshelptosolvetheroamingproblembetweennetworksandfacilitatemobilecommerceandpaymenttransactions(RQ7).Standardizationhasnotprogressed,butthemobileInternethassolvedmostinter-operabilityissuesandprovidesameanstoreducethedependencyonmobilenet-workoperators.The?nalresearchquestionisaboutthetechnologicalandtechnology-relatedstrengthsandlimitationsofthemaintechnol-ogyarchitectures(RQ5).Thisresearchquestionhasnotbeeninves-tigated.Wefeelthatitisunrealistictoexpectthatresearcherscancontributesigni?cantlytothetechnologyofreal-worldmobilepaymentservicesandwriteaboutthat.Thisquestionneedsrefor-mulationaswedolaterinSection4.3.Inadditiontothearticlescitedabove,we?ndthefollowingarticlestoberepresentativeofgoodqualitytechnology-relatedstudiesfrom1999uptothepresent:MisraandWickamasinghe(2004),Valcourtetal.(2005),Hwangetal.(2007),Kousaridasetal.(2008),Linetal.(2008),andMaetal.(2013).3.3.MobilepaymentadoptionstudiesAsweexplainedearlier,allofthemobilepaymentadoptionstudiesthatwecoveredinvestigateconsumeradoption.Understandingconsumerpreferencesandthereasonstouseornotuseaspeci?ctechnology-enabledserviceisimportanttodesignviableservicesthatgeneratevaluetoconsumersandtheotherstakeholdersofanecosystem.Severalmodelsontheadop-tionofinformationtechnologies,suchasthetechnologyacceptancemodel(TAM),theuni?edtheoryofacceptanceanduseoftechnology274T.Dahlbergetal./ElectronicCommerceResearchandApplications14(2015)265–284(UTAUT),anddiffusionofinnovation(DOI)theoryhadbeenusedinconceptualandalsoinempiricalstudiesbefore2007,asdescribedbyDahlbergetal.(2008b).Inadditiontothefactorsinthosemod-els,trust,security,andcostwerereportedtostronglyin?uencetheadoptionofmobilepayments(Dahlbergetal.2003,Pousttchi2003,Zmijewskaetal.2004).Thesituationhasnotchangedafter2007.Onthecontrary,con-sumeradoptionhasbecomethelargestcategoryofmobilepay-mentresearchmeasuredbythenumberofarticles.Authorshavecontinuedtousewell-establishedadoptionanddiffusiontheories:TAM,UTAUT,DOI,butalsotask-technology?t(TTF)theory,thethe-oryofreasonedaction(TRA)andthetheoryofplannedbehavior(TPB).Itisprobablyjustamatteroftimebefore?rstUTAUT2arti-cleispublished.Ontheonehand,continuedresearchcontributionsonconsumeradoptionhavecon?rmedthe?ndingsofearlieradop-tionstudies,sometimeswithbetterempiricaldatacollectionandmorerigorousstatisticalanalyses.Cumulatively,weprobablyknowquitewelltherangeofmobilepaymentservicecharacteris-ticsthatarerelevanttoconsumers.Ontheotherhand,the34arti-clesincludedinthestatisticalanalysisinSection2relyonthesamesetoftraditionaladoptionfactorsasthestudiespriorto2007andadoptionstudiesingeneral.Inouropinion,consumeradoptionstudiesafter2007havethusfailedtointroduceinnova-tiveconstructsorapproachesintomobilepaymentresearch.Table6liststheadoptionfactorsbythenumberofarticlespub-lishedafter2007.ThetwofactorsofTAMtopthelistfollowedbytrustandrisk.Thesametwofactorstoppedthecomparablelistoffactorsfoundimportantinprior2007articles.Trustsharedthefourthpositioncost,whichwasinthirdplace,inarticlespriorto2007.Risksharedthesixthplaceinarticlespriorto2007togetherwithsecurity,socialin?uenceandconvenience.Theonlyfactornotreportedtobeimportantinarticlespriorto2007amongthetop-10factorsafter2007isdemographics.Inadditiontorelyingmainlyontraditionalconsumeradoptionfactors,theauthorsofarticlespublishedafter2007haveplayeditsafewiththeirmethodologicalchoicesbyavoidingnewerandrisk-ierapproaches.Oneofthe34articlesappliedthedesignsciencemethodology,twoarebasedoninterviews,andtheremaining31articleshavecollectedempiricaldatawithsurveyquestionnairestobeanalyzedstatistically,mainlywithalternativestructuredequationmodelmethods.Duetothesecharacteristicsofconsumeradoptionresearchafter2007,itappearsthatresearchershavenotrespondedtotherecommendationsoutlinedbyDahlbergetal.(2008b),accordingtowhichfutureresearchcouldtrytounder-standtheactualmeaningofeachadoptionfactorforconsumers,aswellastheirrelationshipstoeachother(RQ10).23articlesafter2007and12articlesbefore2007reportedthatperceivedeaseofuseisanimportantadoptionfactorformobilepaymentservices.Theproblemisthatwedonotknowwhateaseofusemeansandiscomparedto.Easeofuseandotheradoptionfactorshavebeeninvestigatedlargelyatagenericlevel,asopposedtothecontextsofactualreal-worldpaymentscenarios.Inreal-worldpaymentscenarios,mobilepaymentisonlyonealterna-tiveratherthantheonlyalternative.Anothermeanstosupportthedevelopmentofmobilepaymentknowledgecouldbetheuseofmixed-methodapproaches(e.g.,asurveyfollowedbyinterviews,ora?eldexperiment).Althoughwehopedthatadoptionresearchonmobilepaymentsasawholehadaddressedotherstakeholdersandhadbeenmoreinnovativeregardingconsumers,thereareafewpromisingopen-ings.TheuseofothertheoriesfromotherdisciplinesthanthosealreadyappliedinISresearch–especiallyinTAMandUTAUT–mightresultinadeeperunderstandingonthefactorsthatimpactconsumeracceptanceofmobilepaymentservices.AlshareandMousa(2014)investigatedtheeffectofculturaldimensions,Table6
Adoptionfactorsintheliterature.Construct#ofarticlesPerceivedeaseofuse23Perceivedusefulness22Trust22Risk21Demographic15Security15Compatibility10Socialin?uence10Cost10Mobility10Convenience7Subjectivenorm7Personalinnovativeness6Habit6Privacy5Self-ef?cacy5Quality5Experience4Paymentscenario4Income3Image3Knowledge3Satisfaction2Uncertaintyavoidance2Technologicalimpulse2Complementarity1Complexity1ChengandHuang(2013)incorporatedmentalaccountingtheoryintoTAM,andJiaandHall(2014)examinedthetransferoflearningtheoriesbyanalyzingthehabitualbehaviorofmobilepaymentusers.Otheralternativescouldachievericherconceptualizationsofthedependentvariable(s).Insteadofintentiontousemobilepay-ments,forexample,itmaybepossibletooffersuchalternativesaspostponementofadecision,intentiontouselater,intentiontouseinthenearfuture,andstepstakentostarttheuse.Newconceptu-alizationsarepossiblealsofortheindependentvariables.Forexample,itmaybepossibletoevaluateeachadoptionfactoronascalefromnicetohavetoavitalnecessity,ortoenquireaboutactiv-itiestakentoadoptamobilepaymentservice.Anotherpossibledirectiontoincreaseourunderstandingistocomparethesigni?canceofadoptionfactorsinvariouspaymentscenarios.GoekeandPousttchi(2010)investigatedtheuseofmobilepaymentstopayparkingfees,tobuyfareticketsandtomakemoneytransfers.Therearealsoarticlesthataddresstheadoptionfactorsfordifferenttechnologies,mostnotablyNFCbutalsomobilewallets,indifferentpaymentscenarios,suchasmobileticketingandpaymentsforgovernmentalservices.Suchstudiescouldbemademorerealisticbyinvolvingchoicesamongcompet-ingpaymentinstruments.Dahlbergand??rni(2007)didsoandintroducedtheconceptofpaymenthabit,whichmapspaymentinstrumentsandpaymentscenarios.Theycomparedthechoiceof11paymenthabitsinpaymentsatthepointofsale(e.g.,theuseofcash,creditcard/debitcards,amobiledeviceetc.)andremotely(e.g.,theuseofelectronicbanking,credit/debitcards,amobiledeviceetc.).Itisalsopossibletoinvestigatetheantece-dentsofadoptionfactors.Forexample,Xinetal.(2013)investi-gatedtrustfromfourperspectives:trustinamobileserviceprovider,mobilepaymentvendor,institutionalenvironment,andthetechnologycomponent.Dahlbergetal.(2008b)putforwardanotherrecommendationforfutureresearchregardingconsumerstudies.Intheonlyquali-tativestudyafter2007,Mallat(2007)discoveredwithfocusgroupinterviewsthattheadvantagesofmobilepaymentservicesaredif-ferentthanthoseinestablishedISadoptionmodels.Amobilepay-mentserviceisseenasameanstobehave(=payforpurchases),T.Dahlbergetal./ElectronicCommerceResearchandApplications14(2015)265–284275notasanIStobetakenintouse.Accordingtoher,timeandplaceindependence,queueavoidanceandpossibilitiesforremotepay-mentsareperceivedtobevaluable.ThesecondrecommendationofDahlbergetal.(2008b)(RQ11)encouragedresearcherstoinves-tigatetheinvolvementofconsumersinthedevelopmentofmobilepaymentservices.Rapidlydevelopingdesignsciencemethodspro-videtoolsforthatbutwefoundonly1designsciencearticlethataddressedthisissue.Olsenetal.(2012)involvedintendedusersinthedesignoffunctionalitiesandinteractionwiththeservicesofamobilewallet.Consumer-centricmobilepaymentadoptionresearchhastriedtoextendTAM,UTAUTandotherestablishedadoptionmodelswithnumerousextensionsduringthelast15years.Thescienti?csophisticationandrigorhasincreasedduringthisperiod.Itwasperhapsviabletoinvestigatemobilepaymentadoptioningeneral15yearsago.Itwasperhapsalsoacceptabletouseconveniencesamplessuchasstudents,andtoconductdescriptivestatisticalanalysisofdatacollectedwithsurveys,whichappliedgenericcon-structsofadoptionmodels.Currently,however,weneedhypothesis-basedstudies,whichrelyonstrongtheoreticalback-ground,andareconductedinreal-worldpaymentscenariocon-texts.Thereisenoughreal-worlddataaboutactualmobilepaymentusageandthevolumeofthisdataisincreasing.Thishasmadeconvenientsampleslessjusti?ed.Relianceonawidervarietyofresearchmethods,includingmixed-methodstudies,wouldalsodeepenourunderstanding.Currently,thechallengeofmobilepaymentadoptionresearchisnotinrigorbutinrelevance.Mobilepaymentsshouldbeinvesti-gatedasdynamicallyevolvingservices–ratherthanITorIS–intheirreal-worldcontexts,wheretheycompetewithotherpaymentinstrumentsandmethodsforpopularityinpaymentscenariosapplicabletotheseservices.Similartomulti-sidedplatformandecosystemstudies,thereisaneedformulti-sidedadoptionstudieswithinthesameecosystems.Suchstudies,forexample,couldinvestigatewith?eldresearchwhatpaymentinstrumentsthevar-iousstakeholdersofanecosystem(consumers,merchants,issuers,acquirers,traf?ccarriers)preferandwhy.Italsomaybepossibletoexaminetheimpactof?nancialandotherincentivesandpaymentfeesonthesedecisions.Classicadoptionfactorsareprobablyusefulwhencontextualizedforsuchmobilepaymentstudies.3.4.AuthorshipanalysisOverall,ourreviewofliteraturefrom2007to2014showedfairprogressinthequantityandqualityoftheresearch.Itispossibletoarguethatmobilepaymentresearchisgettingmoremature,sincetherehavebeenmorejournalarticlesandempiricalstudiesinrecentyears.Wealsodiscoveredthatnewermobilepaymentresearchersdonotsystematicallyrefertopriorworkthough.Hence,wehavetriedtounderstandthelackofcumulativeresearchbyinvestigatingthelongevityofmobilepaymentresearchers.Themajorityofauthorshavepublishedlessthan3articlesintheselectedoutletsofthisdescriptiveliteraturereview.Weidenti?ed23authorswhopublishedmorethan2articlesbetween1999and2014.InFig.5,welistedtheauthorsinchronologicalorderaccordingtotheir?rstarticlepublished.Onecanseethatthereareseveralgenerationsofauthors.Thepioneerspublishedtheir?rstpapersin2002–2003.Anothergroupstartedin2004whenthetopicseemstohavebeenthehottest.Morerecently,otherdedicatedresearchershavejoinedthemobilepaymentresearchcommunity.Wefoundthatonlyafewresearcherswhostartedearlyandkeptaconsistentproductionofpapersovertheyearsthough.Thelackofcommittedresearchersmayexplainwhynewerresearchtendstobe‘‘re-inventingthewheel’’.Wetookacloserlookatthe23authorsinourlistandexam-inedtheiraf?liations.2ApioneeringteamfromAaltoUniversitySchoolofBusiness(formerlyHelsinkiSchoolofEconomics)inFinlandpublishedseveralarticles,whichbelongedtotheearliestacademicworkonmobilepayments.TheiractivitystartedduringtheriseofNokiatothenumberonepositionasamobilehandsetmanufacturer.Atthetime,NokiatogetherwithFinnishtelecomoperatorsandbankschairedmostmobilepaymentstandardizationandindustryconsortia(Dahlbergetal.2008a).TheyfundedDahlberg,Mallat,??rni,TuunainenandRossi’sresearchprojects.TheEUprojectSEMOPS(SecureMobilePaymentService)projectwasledbyKarnouskosfromtheFraunhoferInstituteinGermany,andincludedanumberofindustryresearchersfromHungaryandGreece,suchasVilmos.Thisprojectwastechnically-orientedandtheresearcherssoughttodevelopanarchitectureformobilepayments.Thisteamwroteearlyarticlesonmobilepayments.AlsoinvolvedinSEMOPS,Pousttchiestab-lishedaresearchteamattheUniversityofAugsburginGermany.Histeam,called‘‘Wi-mobile’’involvedanumberofPh.D.studentsworkingonmobilepayments(e.g.,Wiedemann,Goeke).Wi-mobilehasremainedactiveinmobilepaymentresearchforalongtime.In2004,OndrusandPigneur,basedattheUniversityofLausanneinSwitzerland,publishedtheir?rstarticleonmobilepayments.OndruswrotehisPh.D.thesisunderthesupervisionofPigneur.In2007,LyytinenjoinedtheteamwhilehewasonsabbaticalinLausanne.Morerecently,OndrusstartedcollaborationsattheESSECBusinessSchoolwithhisstu-dents,inparticularGaurandGannamaneni.In2004,anotherresearchteamwasorganizedaroundthelateLawrenceinAustralia.HerteamatUniversityofTechnologySydneyincludedSteeleandherPh.D.studentZmijewska,whoceasedtoworkonmobilepaymentsafterhergraduation.Also,ZeadallyfromtheUniversityofKentuckywroteanumberoftechnicalpapersonmobilepayments.Somewereco-authoredwithIsaacfromtheUniversidaddeCaraboboinVenezuela.
Inrecentyears,otherteamsinthenorthernpartofEuropebecameactivelyinvolvedinmobilepaymentresearch.InDenmark,Damsgaard,HedmanandHenningssonattheCopenhagenBusinessSchoolrecentlyformedaproductiveresearchgroupwithseveralPh.D.students(e.g.,Kazan,Olsen,Staykova).TheirresearchissupportedbytheCopenhagenFinanceITRegion,whichhostedthe?rstInternationalCashlessSocietyRoundtable.TeamsfromIreland(withCarton)andSweden(withMarkendahl,Arvidsson,andtheirstudentApanasevic)organizedfollow-upeditionsoftheRoundtable.Bothgroupshavebeenactivelyinvolvedinmobilepaymentsresearchduringthepastyears.AnothergroupofresearchersledbyBouwmananddeReuveratTUDelftintheNetherlandshasinves-tigatedmobilepaymentslately.InSpain,Sanchez-Fernandez,Munoz-LeivaandLiebana-CabanillazfromtheUniversityofGranadaalsocontributedtomobilepaymentresearchwithpapersonacceptanceandadoption,mostlypublishedin2014.ZhoufromtheHangzhouDianziUniversitypublishedseveralpapersbyhim-self.AtSingaporeManagementUniversity,KauffmanandMahavebeenworkingonmobilepaymentswiththeirPh.D.studentLiuforthelastseveralyears.4.CriticalcommentsandrecommendationsforfuturemobilepaymentresearchIntheprevioussection,wereviewedandcommentedstudiesinthethreedominantcategoriesofmobilepaymentresearch:mobilepaymentstrategyandecosystems,technologyandtechnological2Thelistofauthorsandteamsisbasedonourliteraturereviewandthereforemaynotbeexhaustive.Therearemostlikelysomeomissions,andweapologizeforthem.
276T.Dahlbergetal./ElectronicCommerceResearchandApplications14(2015)265–284Fig.5.Mobilepaymentsresearcherswithmorethan2papersinourliteraturereview.environment,andmobilepaymentadoption.Nowwewilldiscussissuesthathaveimpededtheharmoniousdevelopmentofmobilepaymentresearch.Wealsowillproposewhatmobilepaymentresearchshouldbeaimingat.Wehopethatresearcherswillmain-tainrigorwhereithasbeenreached,striveforitelsewhere,and?nallyengageinmorerelevantresearchduringthecomingyearsespeciallyinadoptionresearch.Webelievethattheconditionstoperformsuchresearchexistbutwillrequireconsistenteffortandnewideas.4.1.OverallcommentsonmobilepaymentresearchOverall,mobilepaymentresearchershaverarelyengagedinempiricallongitudinalstudies.Dahlbergetal.(2008a)isoneofthefewexceptions.Itdrewonrepeatedinterviewswiththesame12expertsover6years.Mobilepaymentshaveturnedouttobeahighlycomplextopictostudy.Amobilepaymentserviceormarketevolvescontinually.Evenfailedservicesmayevolveforseveralyearspriortheirdiscontinuation.Consequently,one-timestudieswithoneorafewarticleshavebeenandareprobablylikelytopro-videlimitedcontributions.Wehopethattheauthorshipdescrip-tionabovehelpsnewresearcherstojoinproductiveandinsightfulteams.Evenacademicandindustryexpertsinvolvedwithmobilepaymentsformorethanadecadestillstruggletocom-prehendthephenomenon.Despitehundredsofscienti?carticlesandprobablyevenmorepilotsandservicelaunches,westilldonotknowwhenandhowmobilepaymentservicesreachmassmarketsindevelopedeconomies.Althoughanythingispossible,itisunlikelythatopportunisticresearcherswillbeabletomakebreakthroughswithashort-termvisionandlimitedknowledgeaboutthetopic.Manypioneershaveleftthe?eldandevenacademia.Newcomersareencouragedtoreviewpriorworksincetheywillbene?tfromittostrengthentheirnewideas.Wediscoveredthatnumerousarticlespublishedinrecentyearsignoredthe?ndingsandcontributionsofpastwork–orsimplydidnotcitethem–andreachedthesameresultsastheuncitedarticlesdid.Oursecondcommentisaboutthequalityofdata.Duringtheearlydays,empiricaldataweredif?culttocollectasthephe-nomenonwasonlyemerging.Therewereonlyafewindustryexpertsthatcouldbeinterviewedalso.Mobilepaymentswerealmostthestuffofscience?ctiontomostconsumersandmer-chants.Moreover,secondarydatawerenon-existentintheearly2000s,withtheexceptionofsomeconsultingreports.Butlater,theavailabilityofexperts,experiencesandsourcesincreasedsig-ni?cantly.Soweexpectresearcherstousebetterqualitydatatovalidatetheirresearch.Theuseofstudentandotherconveniencesamplestoassessmobilepaymentadoptionwillbemoredif?culttojustify.Similarly,technologyarticlesthatdonotincludeproperevaluationsofthetechnologyconstructorcomparisonswithalter-nativeconstructswillbemuchlessviable.Toconclude,westronglyencourageresearcherstocollectdatafromtherealworld.Forexample,?eldtests,experimentsandexperiencesaboutrealservices,ordataabouttheactualusageofmobilepaymentserviceswillimprovetherelevanceandimpactofacademicresearch.Earlyconceptualarticlesseldomreliedontheoriestoproperlyframetheirresearchproblems.Whenaphenomenonisemerg-ing,researchoftenlacksanappropriateandsoundtheoreticalbasis.Asaresultofthis,alimitednumberofarticleswerepub-lishedinmajorjournals.Overtheyears,usingexploratoryapproachesandbyimportingtheoriesfromotherdisciplines,researchershaveattemptedtocreateandadapttheoriesthatweresuitedtomobilepaymentsresearch.Ourthirdcommentisthattheorydevelopmentandthestrengtheningoftherelevanttheorybasewerestillessentialtoimprovetherigorandrele-vanceofmobilepaymentresearch.Thefourthcommentregardsthefragmentationofresearchandthelackofaholisticviewaboutmobilepaymentsandtherelatedservices.Thiscommentdoesnotmeanthatweshouldinvestigatemobilepaymentsasoneissuebutthatecosystemcharacteristics,technologyfeaturesandadoptionfactorsimpacteachother.Thusstudiesonecosystem,technologyoradoptioninisolationonlyareabletocontributetoapartialunderstandingofmobilepayments.After15yearsofresearch,itistroublingthatweknownexttonothingaboutmerchantadoption,compe-titionbetweenmobileandotherpaymentinstruments,andtheimpactsofchangesincommercial,legal,regulatory,socialandculturalenvironments.The?fthcommentaddressesthegeneralizationoftheresearch?ndings.Mobilepaymentshavecreatedbothglobalandlocalissues.Forexample,GoogleWalletandApplePayareavailableonlyinNorthAmerica,AliPayisavailableonlyinChina,andinEuropemostservicesareavailablelocallyinoneorafewcountries.Themostpopularcreditcardsaretheonlyglobalpaymentinstruments.Crossingbordershasprovedtobedif?cultindevelopedcountries,sincelocalbanks,merchants,telecomcompaniesandregulatorsfeelthattheirvestedinterestsarethreatened.Evenifthiswerenotthecase,eachmarketisdifferent.Telecom,banking,andretail,aswellasothereconomicinfrastructures,regulation,culture,commercehabitsandpay-mentinstrumentsused,differsigni?cantlyacrossmarkets.Therefore,manymobilepaymentresearch?ndingsarenotnec-essarilygeneralizable.Weencouragemulti-countryandmulti-marketstudieswheneverpossible.InTable7,welistthe15most-citedarticlesfromourliteraturereview.Wemadethecutat100citationsfromGoogleScholar(asofApril20,2015).The‘‘overview’’articlesinjournalswithhigher