Have an omnichannel seamless interaction experience! Dimensions and effect on consumer satisfaction

1 “This is an Accep ed Manusc ip o an a icle published by Taylo & F ancis in Jou nal o Ma ke ing Managemen on 21s Aug 2020, a ailable online: h ps://www. and online.com/doi/ ull/10.1080/0267257X.2020.1801798.” Ha e an Omnichannel Seamless In e ac ion Expe ience! Dimensions and E ec on Consume Sa is ac ion. Paula Rod íguez-To icoa*, Lau en T abold Apadulab, Sonia San Ma ín Gu ié eza and Rebeca San José Cabezudoc aUni e sidad de Bu gos, Bu gos, Spain; bManha an College, New Yo k, USA; cUni e sidad de Valladolid, Valladolid, Spain *Co esponding au o : Paula Rod íguez-To ico, Uni e sidad de Bu gos, C/Pa alillos, s/n, 09005, Bu gos, Spain, [email p o ec ed], +34 947 258 973. The challenge o omnichannel e aile s is o o e a seamless expe ience ac oss all ouchpoin s. Howe e , he e is a lack o esea ch ha p o ides heo e ical and empi ical e idence abou how i ms can c ea e such expe iences. The aim o he cu en esea ch is o analyze: (1) he concep o omnichannel seamless in e ac ion expe ience (OSIE) and (2) i s e ec on cus ome sa is ac ion wi h he in e ac ion. Based on a sys ema ic li e a u e e iew and unning a con en analysis, consis ency, eedom in channel selec ion, and synch oniza ion ac oss channels we e iden i ied as OSIE dimensions. In wo s udies and using wo me hods, a su ey and a con olled expe imen , hese OSIE dimensions and downs eam e ec s we e es ed. The indings con i m he mul idimensionali y o OSIE –composed o consis ency, synch oniza ion, and eedom in channel selec ion– and i s posi i e e ec on cus ome sa is ac ion wi h he in e ac ion. Keywo ds: Omnichannel; Seamless; In e ac ion; Expe ience; Sa is ac ion; Consume Summa y S a emen o Con ibu ion: This esea ch is he i s o concep ually p opose and empi ically es he dimensions o he omnichannel seamless in e ac ion expe ience (OSIE) pu o h in he li e a u e: consis ency, eedom in channel selec ion, and synch oniza ion. Fu he , his esea ch con i ms he e ec o 2 he OSIE on cus ome sa is ac ion wi h he in e ac ion. We build upon esea ch on omnichannel e ail managemen and inc emen ally ex end he li e a u e by syn hesizing and empi ically es ing he p oposi ions pu o h hus a . In oduc ion The e ailing indus y has expe ienced a subs an ial change as a esul o he e olu ion o echnology o e he las h ee decades. Wi h he in oduc ion o new e ail channels, such as online and mobile, consume beha io has e ol ed d ama ically o inco po a e he use o se e al channels h oughou he decision-making p ocess. Consume s in e ac wi h companies ac oss mul iple e ail channels and a e seeking seamlessness as hey mo e be ween ouchpoin s (Laza is, V echopoulos, Doukidis, & F aidaki, 2015). This new omnichannel beha io has been ecognized by bo h academics and p ac i ione s as a cen al issue in e ail s a egy (Ewe ha d, Siso sky, & Johansson, 2019). Speci ically, omnichannel esea ch has emphasized he impo ance o seamlessness ac oss channels o acili a e mo e posi i e consume -b and in e ac ion expe iences (Pico -Coupey, Hu é, & Pi e eau, 2016; Ve hoe , Kannan, & Inman, 2015). Howe e , despi e acknowledgemen o i s impo ance, esea ch has no ye del ed in o how e aile s can c ea e a seamless in e ac ion expe ience, ei he heo e ically o empi ically. The seamlessness o he cus ome expe ience, such ha hey can in e ac wi h he b and by na iga ing be ween channels wi h con inui y and ease, is key in omnichannel managemen . Though “seamless” in e ac ion has been de ined as when “ he dis inc ions be ween physical and online will anish, u ning he wo ld in o a show oom wi hou walls” (B ynjol sson, Hu, & Rahman, 2013, p. 23), he e is unce ain y as o how i can be implemen ed by i ms. Howe e , i has only been supe icially desc ibed in he li e a u e and many e aile s a e s ill s uggling o deli e such an expe ience (Pio owicz and Cu hbe son, 2014). 3 As a esul , he e a e s ill many unanswe ed ques ions ega ding how i ms can c ea e a seamless in e ac ion expe ience (Mosque a, Pascual, & Juaneda Ayensa, 2017; Ve hoe , e al., 2015). The concep ualiza ion o he omnichannel seamless in e ac ion expe ience (OSIE) and he dimensions ha a e pa o his cons uc need o be iden i ied. D awing om insigh s in he li e a u e (Ba wi z and Maas, 2018; Haide , Zhuang, Hashmi, & Ali, 2020; Shen, Li, Sun, & Wang, 2018), his pape e e s o he OSIE as he o e all esul o consume s’ accumula ed in e ac ions, in which hey had he oppo uni y o eely and e o lessly swi ch be ween channels and ouchpoin s du ing he di e en phases o he cus ome jou ney, wi hou any in o ma ion loss o ei e a ion. Addi ionally, he downs eam e ec s, such as he impac o he OSIE on cus ome sa is ac ion, can be explo ed u he (Ewe ha d, e al., 2019; Lemon and Ve hoe , 2016). Li e a u e has emphasized i s c i ical ole in sa is ying cus ome s’ needs in he omnichannel en i onmen (Pico -Coupey, e al., 2016; Rigby, 2011; Saghi i, Wilding, Mena, & Bou lakis, 2017; Ve hoe , e al., 2015). The e o e, he p esen s udy ills his gap in omnichannel li e a u e by (1) p oposing and empi ically es ing he dimensions o he OSIE cons uc ha ha e been conjec u ed in he li e a u e and (2) analyzing i s e ec on cus ome sa is ac ion wi h he omnichannel in e ac ion. Mo e speci ically, wo esea ch ques ions a e examined: RQ1. Wha a e he dimensions o he omnichannel seamless in e ac ion expe ience and how do hey di e in impo ance? RQ2. How does he omnichannel seamless in e ac ion expe ience in luence cus ome sa is ac ion wi h he in e ac ion? These esea ch ques ions a e add essed in wo s udies. See Figu e 1 o an o e iew o he esea ch p ocess. Ex an li e a u e in he omnichannel domain was ho oughly examined o iden i y o e a ching hemes ega ding he impo ance o seamlessness. The p esen esea ch uses guidance om he li e a u e o ou line a clea , mul idimensional cons uc ha 4 de ines he dis inc dimensions ha con ibu e o he seamlessness o he OSIE. Th ough a sys ema ic e iew o he li e a u e, h ee impo an dimensions o he OSIE cons uc we e iden i ied: consis ency, eedom in channel selec ion, and synch oniza ion ac oss channels. In s udy 1, hese dimensions we e examined using a su ey ha measu ed cus ome pe cep ions o hei eal p io omnichannel shopping episodes and es ed he e ec o he h ee OSIE dimensions on cus ome sa is ac ion. In s udy 2, he dimensions o he OSIE we e manipula ed in a con olled expe imen o con i m he dimensionali y o he OSIE cons uc and i s in luence on cus ome sa is ac ion. The esul s a e eplica ed ac oss wo s udies using wo dis inc me hods o o e s onge suppo o he p oposed mul idimensional cons uc and i s esul an impac on cus ome sa is ac ion. Figu e 1. Resea ch p ocess o e iew. This pape con ibu es o omnichannel li e a u e wo old. Fi s , his esea ch iden i ies and empi ically measu es h ee dis inc dimensions o he OSIE. Second, his esea ch p o ides an inc emen al heo e ical con ibu ion by es ing he impo ance o he OSIE on cus ome sa is ac ion wi h he omnichannel in e ac ion. P io li e a u e has explo ed a ious aspec s o consume beha io in he omnichannel con ex , such as channel choice, shopping alue, omnichannel se ices usage, o consume pe sonali y (Hu é, Pico -Coupey, & Acke mann, 2017; Pa k and Lee, 2017; Rod íguez-To ico, San-Jose Cabezudo, & San- Ma ín, 2017; Shen, e al., 2018), all o which highligh he impo ance o p o iding RQ1 The omnichannel seamless in e ac ion expe ience: Dimensions and ela i e impo ance RQ2 E ec s o he omnichannel seamless in e ac ion expe ience on cus ome sa is ac ion S udy 1 Su ey, N = 170 eal omnichannel consume s. S udy 2 Expe imen , N = 220 consume s. 5 seamlessness du ing omnichannel in e ac ions (e.g. Ie a and Ziliani, 2018; Ve hoe , e al., 2015). Howe e , he p esen esea ch is he i s o examine he p oposed heo e ical dimensions o OSIE h ough bo h su ey and expe imen al da a. This pape lays a ounda ion on which u u e omnichannel esea ch can build, by o e ing a comple e and o iginal amewo k o he OSIE. The es o he pape is o ganized as ollows. Fi s , he li e a u e e iew is p esen ed and he aims o his pape a e cap u ed in wo esea ch ques ions. Bo h he OSIE cons uc and i s e ec on cus ome sa is ac ion a e explained and a mul idimensional amewo k is laid ou . Nex , he me hodology and analysis o s udy 1 and s udy 2 a e desc ibed. Finally, he esul s a e desc ibed and he heo e ical and manage ial implica ions o he indings a e discussed. Omnichannel beha io : The new cus ome expe ience As he impo ance o omnichannel managemen has g own, i has been examined by an inc easing numbe o esea che s. A sys ema ic e iew o he li e a u e was unde aken o iden i y he ways in which omnichannel cons uc has been concep ualized. Table 1 illus a es he a ious ways in which omnichannel e ailing has been de ined in he li e a u e. Table 1. The omnichannel concep ualiza ion. Au ho /s De ini ions o omnichannel concep Rigby (2011, p. 4) Omnichannel e ailing: “an in eg a ed sales expe ience ha melds he ad an ages o physical s o es wi h he in o ma ion- ich expe ience o online shopping.” Le y, Wei z, & G ewal (2013, p. 67) Omnichannel e ailing: “a coo dina ed mul ichannel o e ing ha p o ides a seamless expe ience when using all o he e aile ´s shopping channels.” Laza is and V echopoulos (2014, p. 2) Omnichannel e ailing: “ he use o bo h physical and online channels combined wi h he deli e y o seamless shopping expe iences.” Ve hoe , e al. (2015, p. 176) Omnichannel managemen : “ he syne ge ic managemen o he nume ous a ailable channels and cus ome s ouchpoin s, in such a way ha he cus ome expe ience ac oss channels and he pe o mance o e channels is op imized.” 6 Cummins, Pel ie , & Dixon (2016, p. 5) Omnichannel ma ke ing (in a sales con ex ): “ he syne ge ic in eg a ion o cus ome ouchpoin s and communica ion oppo uni ies o he pu pose o c ea ing a uni ied b and expe ience ega dless o channel, pla o m o s age in he selling p ocess.” Blom, Lange, & Hess J (2017, p. 287). Omnichannel managemen : “is he way o c ea e an o e all e ailing expe ience ha is he same ac oss channels and ouchpoin s.” Hu é, e al. (2017, p. 315) Omnichannel shopping: “ he comple e alignmen o he di e en channels and ouchpoin s, esul ing in an op imal-b and cus ome expe ience.” Shen, e al. (2018, pp. 62, 63) Omnichannel se ice: “a kind o se ice ha allows cus ome s eely choose among all pa allel channels, and seamlessly swi ch among he di e en channels, wi hou any in o ma ion loss o ei e a ion.” “The concep o “omnichannel” e ol ed om mul ichannel, wi h a speci ic ocus on he in eg a ion and coo dina ion o de ached channels o mee consume s' needs o seamless channel ansi ions.” Despi e he a ie y in hese concep ualiza ions, a p e alen aspec o omnichannel e ailing eme ges; i p o ides an imp o ed consume expe ience by in eg a ing all e ail channels (Ba wi z and Maas, 2018; Laza is and V echopoulos, 2014). In line wi h hese de ini ions, p io esea ch has con i med ha omnichannel e ailing is cha ac e ized by o e ing a seamless in e ac ion ac oss channels wi hou in e up ions (Hu é, e al., 2017; Saghi i, e al., 2017; Shen, e al., 2018; Ve hoe , e al., 2015). The seamlessness o he ansi ion be ween e ail channels is he key di e en ia ing elemen in omnichannel e ailing. A mul ichannel s a egy in which channels a e managed sepa a ely is belie ed o be obsole e and should be eplaced wi h channel managemen ha ully in eg a es all ouchpoin s so ha consume s can in e ac wi h he b and using in e changeable channels (Beck and Rygl, 2015; Cao and Li, 2018; Ve hoe , e al., 2015). The dimensions o he omnichannel seamless in e ac ion expe ience In omnichannel e ailing, seamlessness means elimina ing he dis inc ions be ween all o he di e en a ailable channels, e ec i ely u ning he wo ld in o a “show oom wi hou walls” 7 (B ynjol sson, e al., 2013). This is a challenge o e aile s because i equi es changes o hei exis ing channel design o ensu e ha he e a e no up u es when na iga ing om one o ano he , e ec i ely blu ing he ba ie s be ween channels (Hu é, e al., 2017; Os om, Pa asu aman, Bowen, Pa icio, & Voss, 2015; Ve hoe , e al., 2015). Omnichannel li e a u e has clea ly heo ized he impo ance o seamlessness, bu no esea ch hus a has syn hesized wha i means o c ea e a seamless in e ac ion expe ience. As consume s na iga e ac oss he g owing numbe o b and ouchpoin s, including o line b ick and mo a s o es, online, mobile, and social media (Cao and Li, 2018), he aspec s o hese channels ha make he consume ’s in e ac ion seamless ha e no ye been de ined. The p esen esea ch akes he i s s eps owa d building a amewo k o his concep . In mul ichannel con ex , Wu and Chang (2016) iden i ied he dimensions o mul ichannel in eg a ion quali y based on p e ious esea ch. They concluded ha di e si ica ion, consis ency, and ecip oci y a e he h ee main cha ac e is ics o mul ichannel in eg a ion ha p o ide cus ome s an op imal expe ience ac oss channels. Omnichannel managemen has e ol ed om li e a u e on mul ichannel managemen and channel in eg a ion (Lee and Kim, 2010; Oh and Teo, 2010; Wu and Chang, 2016), so he p esen esea ch ies o iden i y how hese c ucial cha ac e is ics a e p esen ed in he omnichannel con ex . To his end, a li e a u e e iew was ca ied ou . S a emen s desc ibing wha i means o be “seamless” we e iden i ied in ex an esea ch o analyze he unde lying dimensions o he cons uc . Appendix 1 summa izes he ele an li e a u e and he associa ed dimension ha can be de i ed om he espec i e au ho s’ insigh s on OSIE. Applying he knowledge o mul ichannel li e a u e (Lee and Kim, 2010; Oh and Teo, 2010; Wu and Chang, 2016) combined wi h a ho ough li e a u e e iew o he omnichannel esea ch (Appendix 1), his pape p oposes h ee dimensions o an OSIE: consis ency, eedom in channel selec ion, and synch oniza ion ac oss channels. 8 The i s dimension o OSIE ha was highligh ed ac oss he li e a u e is consis ency. Consis ency is de ined as he consume s’ pe cei ed cohe ence o e ail ouchpoin s (Pico - Coupey, e al., 2016). Re aile s mus de elop consis ency ac oss all he ouchpoin s (Pio owicz and Cu hbe son, 2014; Saghi i, e al., 2017; Valos, Maples one, Polonsky, & Ewing, 2017), because ensu ing cohe ence and uni o mi y ac oss channels is a cen al pa o he omnichannel s a egy (Mosque a, e al., 2017; Shen, e al., 2018) and is expec ed by he majo i y o consume s (Hu é, e al., 2017; Ie a and Ziliani, 2018; Shanka , Inman, Man ala, Kelley, & Rizley, 2011). B and image, p oduc s, p omo ions, p ices, and a ailable se ices a e i e aspec s o he e ailing mix o which consis ency ac oss channels is undamen al (Beck and Rygl, 2015; Cao and Li, 2015; Hu é, e al., 2017; Laza is and V echopoulos, 2014; Neslin e al., 2006). The e o e, i ms mus ensu e ha all channels p esen hese pa icula a ibu es consis en ly, as his will con ibu e o he pe cei ed seamlessness o he omnichannel in e ac ion. In addi ion o consis ency, eedom in channel selec ion was also emphasized in he li e a u e, as shown in Appendix 1. I e e s o consume s’ pe cep ion abou he le el o eedom hey ha e o selec di e en channels o a ious ypes o in e ac ions wi h he b and, such as shopping, e u ning, deli e y, and sea ching o in o ma ion (Lee and Kim, 2010). A i m needs o conside consume s’ mo emen ac oss channels h oughou he shopping p ocess in o de o p o ide a seamless in e ac ion expe ience (Hu é, e al., 2017; Ve hoe , e al., 2015). Speci ically, p omo ing he use o mul iple channels is conside ed an inno a i e way o p o ide seamlessness (Shanka , e al., 2011). This in ol es gi ing consume s con ol and allowing hem o choose hei p e e ed channel a each s ep o he pu chase p ocess, including in o ma ion sea ch, pu chase, e u n, deli e y, and sel -se ice (Cha e jee, 2010; Os om, e al., 2015; Pio owicz and Cu hbe son, 2014; Shen, e al., 2018). Omni-consume s a e cha ac e ized by he capabili y o selec o use any o he a ailable channels du ing any 9 phase o hei decision-making p ocess depending on hei needs, and when channels a e cons uc ed o gi e consume s his eedom o do so, i enhances he seamlessness o he consume in e ac ion expe ience (Pel ola, Vainio, & Nieminen, 2015). Channel synch oniza ion is he hi d and inal dimension o he OSIE. I cap u es he idea ha consume s use all ouchpoin s in e changeably o sea ching, shopping, o de ing, pu chasing, pickup, deli e y, and e u ns (Sands, Fe a o, Campbell, & Pallan , 2016) and he e is no up u e when mo ing om one ouchpoin o ano he one (Hu é, e al., 2017; Laza is and V echopoulos, 2014). This dimension goes hand-in-hand wi h eedom in channel selec ion. No only is he abili y o choose be ween channels a di e en poin s in he decision p ocess impo an (Cha e jee, 2010; Juaneda-Ayensa, Mosque a, & Sie a Mu illo, 2016; Pio owicz and Cu hbe son, 2014; Saghi i, e al., 2017), bu , mo eo e , synch oniza ion be ween he channels such ha consume s may in e changeably use any channel o any pa o he decision p ocess is necessa y (Kim, Ahn, & Fo ney, 2014; Shen, e al., 2018; Ve hoe , e al., 2015; J. Zhang e al., 2010). Channels may be connec ed in a way ha consume s can comple e hei shopping p ocess whene e hey wan (Huang, Lu, & Ba, 2016; Ie a and Ziliani, 2018). Simila ly, Pico -Coupey, e al. (2016) ound ha in he omnichannel en i onmen , i is challenging o i ms o ensu e he synch oniza ion ac oss ouchpoin s, bu impo an o do so in o de o p o ide a be e expe ience. In he p esen esea ch, hese h ee unde lying dimensions a e examined, as well as he ela i e impo ance o each o he p oposed dimensions. Fo mally, RQ1. Wha a e he dimensions o he omnichannel seamless in e ac ion expe ience and how do hey di e in impo ance? 16 change in one o he componen s can cause a change in he cons uc , which is a key heo e ical conside a ion o assessing he cons uc s (Col man, De inney, Midgley, & Venaik, 2008). Fo ins ance, i one o he i ems o dimensions is emo ed, he a iable ha is measu ed adically changes because he expec ed seamlessness disappea s, and a mul i- o c oss-channel expe ience would be measu ed, a he han an OSIE. (3) The co a ia ion among he indica o s is assessed. In he case o e lec i e cons uc s, he indica o s mus co a y wi h each o he . Con e sely, he indica o s o a o ma i e cons uc migh no necessa ily co a y. In his case, he indica o s and dimensions o OSIE do no p esen high co ela ion alues, which indica es hei o ma i e cha ac e (see Appendix 3). (4) An eceden s and consequences o he indica o s a e examined. Ja is, e al. (2003) p opose o analyze he nomological ne o he cons uc indica o s, ha is, whe he all o he indica o s a e equi ed o ha e he same an eceden s and consequences o no . In he case o OSIE, his ule allows us o de e mine ha he cons uc and dimensions a e o ma i e because hey a e no equi ed o ha e he same an eceden s and consequences. Consequen ly, om he concep ualiza ions ha eme ged om he omnichannel li e a u e (Hu é, e al., 2017; Laza is and V echopoulos, 2014; Ve hoe , e al., 2015) and using he scales p e iously alida ed in he li e a u e o measu e hose dimensions, a scale was cons uc ed o cap u e h ee dimensions: consis ency, eedom in channel selec ion, and synch oniza ion. Rega ding consis ency, he i e i ems de eloped by Lee and Kim (2010) and Wu and Chang (2016) we e adap ed o he o ma i e scale. These measu es ela e o he main aspec s 17 o he e ailing mix o which pe cei ed consis ency is men ioned in he li e a u e and equi ed in omnichannel managemen (Beck and Rygl, 2015; Cao and Li, 2015; Hu é, e al., 2017; Laza is and V echopoulos, 2014; Neslin, e al., 2006): image, p oduc , p omo ions, p ice, and se ices. Simila ly, ou i ems adap ed om Lee and Kim (2010) we e used o measu e eedom in channel selec ion o all he s ages o decision-making. To measu e channel synch oniza ion, six i ems we e adap ed om Oh and Teo (2010) o cap u e he abili y o swi ch be ween all channels a he di e en s ages (sea ch o in o ma ion, pu chase, pick up, e u n, and pos -pu chase se ice) (Kim, e al., 2014). Finally, sa is ac ion wi h he omnichannel in e ac ion was measu ed using h ee i ems adap ed om Walsh, Shiu, & Hassan (2014) (“O e all, I am sa is ied wi h his las in e ac ion wi h he b and,” “I am pleased wi h his las in e ac ion wi h he b and,” and “I am deligh ed wi h his las in e ac ion wi h he b and”). Resul s o he Measu emen Model This esea ch akes he wo-s age app oach o app oxima e he second-o de cons uc . In his case, he h ee dimensions o he OSIE, consis ency, eedom in channel selec ion, and synch oniza ion, ha e an unequal numbe o indica o s: i e, ou and six, espec i ely. Consequen ly, he wo-s age app oach sol es he p oblem o unequal numbe o indica o s a he i s -o de le el and i has he ad an age o es ima ing a mo e pa simonious model on he second-le el analysis wi hou needing he i s -o de cons uc s (Becke , Klein, & We zels, 2012; Hai , e al., 2018; Ringle, Sa s ed , & S aub, 2012). Conside ing ha he elec ion o an app oach should be based on each esea ch pa icula i ies (Becke , e al., 2012), he assessmen o his measu emen model was unde aken a wo le els. Fi s , a he i s -o de le el, he mul icollinea i y o he dimensions was assessed (Hai , Ringle, & Sa s ed , 2011) (Table 2): a iance in la ion ac o (VIF) alues a e below 5; and he ole ance alues (IT) a e abo e .10, as li e a u e ecommends (Hai , Hul , Ringle, & 18 Sa s ed , 2017). In o de o analyze cons uc alidi y, i em weigh s we e examined. As some indica o weigh s a e no signi ican , loadings signi icance was obse ed. I em loadings p esen ed signi ican alues in all he cases, he e o e, he indica o s in he o ma i e cons uc s we e e ained (Hai , e al. (2017). Table 2. Fi s -o de measu emen model. Va iable Fo ma i e dimension I ems Weigh s ( -Value) Loadings ( -Value) OSIE Consis ency “The b and p o ided consis en … …s o e images be ween he all channels. (VIF=1.225, IT=.811) .396 (2.957) .704 (7.179) …p oduc in o ma ion be ween he all channels. (VIF=1.562, IT=.639) .258 (2.124) .741 (9.864) …p omo ional in o ma ion be ween he all channels. (VIF=1.569, IT=.644) .373 (2.899) .787 (10.960) …p icing policy be ween he all channels. (VIF=1.381, IT=.726) .067 (.690) .554 (6.047) …cus ome se ices be ween he all channels. (VIF=1.458, IT=.678) .288 (2.483) .691 (7.612) F eedom in channel selec ion “The b and allowed me… … o choose whe e o shop o me chandise. (VIF=1.198, IT=.835) .586 (5.042) .815 (10.828) … o choose a way o e u ning he me chandise. (VIF=1.364, IT=.733) .322 (2.433) .704 (7.493) … o a ange deli e y op ions. (VIF=1.432, IT=.698) .262 (1.693) .617 (5.439) … o a ange a ious se ice op ions. (VIF=1.378, IT=.726) .212 (1.445) .634 (6.295) Synch oniza ion “The b and allowed me… … o examine p oduc s physically once I ind hem in ano he online/mobile channel. (VIF=1.322, IT=.756) .111 (1.025) .451 (4.327) … o sea ch o p oduc in o ma ion in one channel and hen pu chase i in ano he channel. (VIF=1.344, IT=.744) .065 (.539) .469 (4.227) … o pick up p oduc s bough in one channel h ough ano he channel. (VIF=1.466, IT=.682) .218 (1.264) .540 (5.006) … o e u n p oduc s bough in one channel h ough ano he channel. (VIF=1.390, IT=.719) -.016 (.141) .370 (3.490) … o eques pos -pu chase se ices o any p oduc bough in one channel h ough ano he channel. (VIF=1.364, IT=.733) -.035 (.305) .411 (3.732) 19 … o choose he mos con enien way o in e ac ing wi h his endo (e.g., sea ch, pu chase, pick up, e u n, pos -pu chase…) h ough all he channels. (VIF=1.318, IT=.759) .853 (8.953) .964 (22.809) Thus, in o de o espond o RQ1, ollowing he li e a u e guidelines (Becke , e al., 2012; Hai , e al., 2018; Ringle, e al., 2012), he p e iously alida ed i s -o de cons uc can hen be inco pos o he second-o de measu emen model a e p esen ed in Table 3. Simila ly o he i s s age, mul icollinea i y was uled ou in he second le el. As can be seen, VIF alues a e below 5 and IT alues a e abo e .10. In addi ion, cons uc alidi y was assessed. As can be seen, all weigh coe icien s show signi ican alues a a con idence le el o 95% ( > 1.96) excep o eedom in channel selec ion. Al hough his dimension does no show a con idence le el o 95%, he cons uc alidi y can be con i med due o he ac ha i s loading alue is signi ican , as Hai , e al. (2017) ecommend. These indings add ess RQ1 and illus a e h ee dis inc dimensions o he OSIE cons uc and hei ela i e impo ance, as indica ed by hei weigh s (Table 3). Table 3. Second-o de measu emen model. Va iable Fo ma i e dimension Weigh s ( -Value) Loadings ( -Value) VIF IT OSIE Consis ency .550 (5.069) .873 (18.325) 1.477 .677 F eedom in channel selec ion .175 (1.361) .754 (10.252) 1.851 .540 Synch oniza ion .457 (4.541) 8.47 (17.027) 1.809 .553 Wi h ega ds o sa is ac ion, he e lec i e la en a iable, he eliabili y and alidi y o he scale was con i med. The con i ma o y ac o analysis (CFA) o he h ee sa is ac ion measu es yielded a C onbach’s alpha abo e .7 (α = .821), composi e eliabili y abo e .6 (CR 20 = .894), and a e age a iance ex ac ed abo e .5 (AVE = .738), as ecommended (Hai , e al., 2017). Linea eg ession model analysis A e alida ing he measu emen model, he p oposed RQ2 was es ed by es ima ing a mul iple linea eg ession model. To accomplish his, he la en a iables sco es ob ained om PLS algo i hm esul s we e used in SPSS. Gende , age and annual income we e included in he model as co a ia es o con ol o he demog aphics. Table 4 shows he model wi hou co a ia es (Model 1) and wi h co a ia es (Model 2). In bo h cases, he model is signi ican : Model 1: F(1, 168) = 206.695, p < .001 wi h a R-squa ed o .522 and Model 2: F(4, 158) = 52.172, p < .001 wi h a R-squa ed o .569. The e o e, he p oposed model conside ing he con ol o he co a ia es accoun s o 56.9% o a iance in ou da a. As can be obse ed in Table 4, he esul s show he signi ican posi i e in luence o OSIE on sa is ac ion wi h he omnichannel in e ac ion, suppo ing he p oposed esea ch ques ion, RQ2. As pe cei ed seamlessness o he in e ac ion expe ience inc eased, pa icipan s indica ed highe le els o sa is ac ion wi h he omnichannel in e ac ion. In addi ion, gende p esen s a ma ginal impac (p < .10) on cus ome sa is ac ion wi h he omnichannel in e ac ion, such ha emales p esen highe le els o sa is ac ion wi h he omnichannel in e ac ion han males. Table 4. Mul iple linea eg ession analysis esul s o sa is ac ion. Model 1 Model 2 β s anda dized -Value β s anda dized -Value VIF Omnichannel seamless in e ac ion expe ience .743 14.377*** .731 13.254*** 1.088 Gende - - .098 1.852┼ 1.026 Age - - .062 1.153n.s. 1.044 Income - - -.025 -.463n.s. 1.029 R2 .522 .569 F-s a is ic 206.695*** 52.172*** ***p <.001; **p < .01; *p < .05; ┼p < .10; n.s., no signi ican . 21 Discussion S udy 1 explo es he OSIE dimensionali y and i s e ec on cus ome sa is ac ion wi h he omnichannel in e ac ion wi h da a based on eal omnichannel in e ac ion expe iences. Using a sample o consume s wi h p io omnichannel expe ience, he h ee hypo hesized dimensions o he cons uc , as well as cus ome sa is ac ion wi h he in e ac ion, we e measu ed h ough an online su ey. Fi s , in o de o add ess RQ1, he esul s o his s udy o e e idence abou he mul idimensionali y o he OSIE cons uc composed o he h ee dimensions. Consis ency, eedom in channel selec ion, and synch oniza ion ha e been ound in he li e a u e and con i med in he con en analysis as OSIE dimensions. Speci ically, he esul s o S udy 1 u he suppo he mul idimensionali y and show ha consis ency is he mos impo an dimension, ollowed by synch oniza ion, and eedom in channel selec ion. This esul expands he indings o Hu é, e al. (2017), who concluded ha consis ency is a p e equisi e o seamlessness, bu no enough on i ’s own o conside he in e ac ion expe ience as such. This esea ch goes ano he s ep o wa d by including synch oniza ion and eedom in channel selec ion as o he equisi es o an OSIE. Second, he di ec posi i e e ec o he OSIE on cus ome sa is ac ion wi h he in e ac ion was con i med, add essing he p oposed second esea ch ques ion. The e o e, as consume s’ pe cep ions o seamlessness inc ease, hei sa is ac ion wi h he omnichannel in e ac ion is signi ican ly inc eased. In addi ion, a e including he demog aphics, a ma ginal e ec o gende on cus ome sa is ac ion wi h he omnichannel in e ac ion was e ealed, showing ha emales ha e highe le els o sa is ac ion o e all han males wi h he omnichannel in e ac ion. Limi a ions The e a e some aspec s o he s udy ha may limi he gene alizabili y o he indings. Pa icipan s we e asked o ecall hei mos ecen pu chase in which hey in e ac ed wi h a 22 b and on wo o mo e channels. Al hough his allows o he measu emen o consume pe cep ions in eal b and in e ac ions, he esul s o ou su ey ely on pa icipan s’ ecall o hei beha io du ing he pu chase scena io and he con en ha was p o ided by he b and ac oss channels. Addi ionally, he e could be a ia ion in he numbe o channels ha we e used be ween subjec s, as well as he pu pose o which each channel was used. Thus i is di icul o assess whe he his may impac cus ome pe cep ions o he dimensions o OSIE o hei esul an sa is ac ion wi h he in e ac ion. S udy 2 In his s udy, he esul s o s udy 1 a e eplica ed using a con olled expe imen al design (see Appendix 2). The p oposed dimensions o OSIE and i s in luence on cus ome sa is ac ion wi h he omnichannel in e ac ion is es ed using a hypo he ical shopping scena io. This s udy also add esses some o he limi a ions o s udy 1. A hypo he ical scena io allows o con ol o e bo h he ype and numbe o channels wi h which he pa icipan in e ac s and he b and- ela ed con en ha pa icipan s will be exposed o du ing he omnichannel in e ac ion. Addi ionally, he pa icipan s will espond o he su ey measu es immedia ely, so he ouchpoin s will be esh in hei mind. Las ly, his expe imen al design allows he p oposed dimensions o consis ency, eedom in channel selec ion, and synch oniza ion o be manipula ed, while holding he es o he s imuli cons an be ween condi ions, o isola e he impac o hese speci ic a iables on pe cei ed seamlessness and, subsequen ly, on sa is ac ion. Me hodology and P ocedu e A new sample o MTu k wo ke s (N = 220) we e ec ui ed o pa icipa e in a 2 cell (omnichannel in e ac ion expe ience: seamless s. non-seamless) be ween-subjec s expe imen . MTu k esponden s ecei ed $1.25 upon comple ion o he su ey. Pa icipan s 23 we e 55% male and a mean age o 38.2 yea s. 94.5% had an annual income below $100.000. To ensu e he quali y o he esponden s, MTu k Mas e Wo ke s wi h a pas su ey app o al a ing o a leas 95% we e selec ed (Sheehan, 2018). The scena ios we e p e es ed (N = 29, Mage = 36.8, 69% male) using he 5-poin Like scales in s udy 1. The scena ios we e pe cei ed o be seamless and non-seamless, espec i ely. The scena ios measu ed signi ican ly di e en ac oss he dimensions o consis ency (Mseamless = 4.2, Mnonseamless = 2.3, F(1, 27) = 30.762, p < .001), eedom in channel selec ion (Mseamless = 4.3, Mnonseamless = 2.6, F(1, 27) = 20.190, p < .001), and synch oniza ion (Mseamless . = 4.2, Mnonseamless = 2.1, F(1, 27) = 50.417, p < .001). As clo hing has been one o he indus ies ha has success ully implemen ed omnichannel managemen (Gao and Yang, 2016), his p oduc ca ego y was selec ed o use in his s udy. Pa icipan s we e shown s imuli depic ing consume ouchpoin s o a aux ashion b and, XBRAND, including a mock websi e, social media page, and imagined in-s o e shopping scena io. Two e sions o he shopping scena io and isual s imuli we e c ea ed o ep esen a seamless and a non-seamless in e ac ion by a ying he consis ency, eedom in channel selec ion, and synch oniza ion be ween channels. These dimensions we e con i med in s udy 1 o be componen s o he OSIE. Pa icipan s we e andomly assigned o ei he he seamless o non-seamless condi ion and p esen ed wi h a co esponding omnichannel shopping scena io. All pa icipan s we e asked o imagine ha hey we e shopping o a black -shi , because i is a s anda d, unisex p oduc . They we e guided h ough a hypo he ical omnichannel in e ac ion, in which hey we e o p e end ha hey we e in a s o e speaking o a sales associa e, examining he XBRAND websi e, and social media page (see Appendix 2) o in o ma ion abou he -shi . 24 In he seamless condi ion, he p oduc in o ma ion, p ice, a ailabili y, b and logo, and sales p omo ions o he -shi we e consis en ac oss channels, as hey a e he aspec s o he e ailing mix o which consis ency ac oss channels is undamen al (Beck and Rygl, 2015; Cao and Li, 2015; Hu é, e al., 2017; Laza is and V echopoulos, 2014; Neslin, e al., 2006). In he non-seamless condi ion, -shi in o ma ion, p ice, a ailabili y, b and logo, and sales p omo ions we e inconsis en ac oss channels. Mo eo e , in he seamless condi ion he pa icipan s could eely choose hei desi ed channel o pu chase (e.g.“A ailable in s o es and online”). On he con a y, in he non-seamless condi ion, he elec ion o he channel was es ic ed (e.g.“Only a ailable on ou websi e”). Finally, synch oniza ion was p o ided in he seamless condi ion, whe e he esponden s could see he connec ions among channels (e.g. a bu on ha o e ed he op ion o “in s o e pickup”). This was no an op ion in he non- seamless condi ion. A e eading he shopping scena io and iewing he mock websi e and social media pos o XBRAND, pa icipan s esponded o a se ies o ques ions abou hei pe cep ions o he b and and in e ac ion expe ience and, inally, epo ed hei demog aphics. The same scales as we e used in s udy 1 we e included o measu e he pe cei ed consis ency, eedom in channel selec ion, synch oniza ion ac oss channels, and sa is ac ion wi h he omnichannel in e ac ion. Resul s Manipula ion checks. To ensu e ha he XBRAND in e ac ion di e ed in seamlessness be ween condi ions, pa icipan s esponded o scales assessing each o he h ee dimensions o an OSIE: consis ency, eedom in channel selec ion, and synch oniza ion. Consis ency. The 5-poin Like scale used in s udy 1 was used o measu e consis ency. Pa icipan s we e asked o indica e how consis en XBRAND was in b and image, p oduc , 25 p omo ions, p ice, and se ices be ween channels on a 5-poin scale om s ongly disag ee (1) o s ongly ag ee (5). Fo ease o analysis, he measu es we e a e aged in o an index (𝝰 = .927). An ANOVA was conduc ed o examine whe he his measu e di e ed signi ican ly be ween condi ions. Resul s sugges a signi ican di e ence in pe cei ed consis ency be ween condi ions such ha he seamless condi ion was pe cei ed o be mo e consis en han he non- seamless condi ion (Mseamless = 4.24, Mnonseamless = 2.30, F(1, 218) = 217.695, p < .001). F eedom in channel selec ion. F eedom in channel selec ion was measu ed using he ou i ems used in s udy 1 on a 5-poin Like scale om s ongly disag ee (1) o s ongly ag ee (5). Fo ease o analysis he esponses we e a e aged in o an index (𝝰 = .916). An ANOVA was conduc ed and e ealed ha pa icipan s in he seamless condi ion pe cei ed g ea e eedom in channel selec ion han hose in he non-seamless condi ion (Mseamless = 4.38, Mnonseamless = 2.48, F(1, 218) = 205.333, p < .001). Synch oniza ion. Simila ly, o measu e synch oniza ion ac oss channels, pa icipan s esponded o six i ems as in s udy 1. The ques ions assessed he ease o na iga ing be ween channels o sea ch o in o ma ion, pu chase, pick up, e u n, and pos -pu chase on a 5-poin scale om s ongly disag ee (1) o s ongly ag ee (5). The measu es we e a e aged o c ea e an index (𝝰 = .955). An ANOVA was pe o med and sugges s ha pa icipan s in he seamless condi ion pe cei ed g ea e synch oniza ion ac oss channels han hose in he non- seamless condi ion (Mseamless = 4.34, Mnonseamless = 1.98, F(1, 218) = 384.338, p < .001). OSIE. The seamless s imuli we e pe cei ed o be mo e consis en ac oss channels, o e ing g ea e eedom in channel selec ion, and p o iding g ea e channel synch oniza ion han he non-seamless s imuli. An ANOVA wi h an index o he h ee dimensions (𝝰 = .949) was pe o med and con i ms ha in he seamless condi ion he pa icipan s pe cei ed g ea e OSIE han hose in he non-seamless condi ion (Mseamless = 4.31, Mnonseamless = 2.25, F(1, 218) = 320.674, p < .001). A mul i a ia e analysis o a iance (MANOVA) was un o assess he 32 be ween male and emale sa is ac ion, such ha emales epo ed signi ican ly lowe sa is ac ion han males when a non-seamless, e sus seamless, scena io was p esen ed. The e o e, an inconsis en and in e up ed omnichannel in e ac ion will mo e nega i ely a ec women’s sa is ac ion han men’s. This is consis en wi h ma ke ing li e a u e ha has sugges ed ha emales eac mo e s ongly han males o some en i onmen s wi h which hey a e no com o able (Luo, McGold ick, Bea y, & Keeling, 2006; San-Ma ín, López-Ca alán, & Ramon-Je onimo, 2012). Theo e ical implica ions This pape makes h ee con ibu ions o omnichannel li e a u e. Despi e he in e es ha li e a u e has gi en o seamlessness (e.g. Pio owicz and Cu hbe son, 2014; Shen, e al., 2018), he concep , i s dimensionali y, and i s in luence on omni-consume beha io need mo e a en ion. The e is a lack o ex an esea ch ha explains he unde lying dimensions o he OSIE in a comp ehensi e way. Thus, he i s con ibu ion o his esea ch is o ill his knowledge gap, by p oposing ha he OSIE is a mul idimensional cons uc . The empi ical esul s o his esea ch indica e he mul idimensionali y o his cons uc and con i m ha consis ency, eedom in channel selec ion, and synch oniza ion ac oss channels p o ide he ounda ion o a seamless in e ac ion expe ience. As a esul , his s udy o e s an o iginal amewo k o unde s anding he OSIE, con ibu ing, in his way, o cu en and u u e esea ch. The indings o his esea ch p o ide an ins umen ha in o ms schola s and p ac i ione s how o compose a seamless in e ac ion expe ience, one o he mos emphasized a iables in he omnichannel con ex . Mo eo e , his esea ch can be used as he basis o building u he esea ch. The mul idimensional OSIE amewo k con i med in his pape was es ed using b ick and mo a s o es, online, and mobile channels. Howe e , i also p o ides insigh in e ms o managing he e e -changing e ail en i onmen h ough i s applicabili y ac oss all channels. Changes in 33 echnology and consume beha io con inuously challenge i ms o s ay up- o-da e. Howe e , he cu en amewo k p o ides insigh in o omnichannel managemen by highligh ing he impo ance o consis ency, eedom in channel selec ion, and synch oniza ion ac oss all channels, bo h exis ing and new, as hey con inue o e ol e. The cu en esea ch acknowledges he e olu ion o channels and is one o he i s o explo e social media in he omnichannel domain. As esea ch con inues o explo e new channel inno a ions, such as he ad en o oice echnology in he mobile channel (Pagani, Raca , & Ho acke , 2019) o augmen ed eali y in he online channel (Fan, Chai, Deng, & Dong, 2020), he cu en esea ch p o ides a heo e ical and manage ial amewo k ha can be applied o in eg a e all channels in o a seamless expe ience. As a esul , his amewo k does no limi he seamless in e ac ion expe ience o he cu en channels, bu i is openly desc ibed o be adap ed o he e ol ing con ex . Acknowledging he as e olu ion o his con ex , he way he dimensions a e desc ibed enables u u e esea ch o adap he scales and scena ios including he new channels and ouchpoin s appea ed. The second con ibu ion o his wo k is o ad ance omnichannel esea ch by examining he subsequen impac ha pe cei ed seamlessness be ween ouchpoin s has on he cus ome - b and in e ac ion. Speci ically, he c ucial ole o a seamless in e ac ion expe ience on sa is ac ion wi h he omnichannel in e ac ion is empi ically es ed. Thus, his pape mo es his s ain o li e a u e o wa d by con i ming ha when cus ome s pe cei e he omnichannel managemen as seamless, hei sa is ac ion wi h he omnichannel in e ac ion is enhanced di ec ly and posi i ely. Al hough i was no p oposed, he hi d con ibu ion o his pape is he iden i ica ion o di e ences ha exis be ween men and women’s sa is ac ion wi h omnichannel in e ac ions. Speci ically, women in his s udy exhibi ed signi ican ly lowe sa is ac ion han men when aced wi h a non-seamless in e ac ion expe ience. This can o ien he esea ch in he 34 omnichannel ield o explo e addi ional nuances o he OSIE. Al hough gende has been s udied deeply, esea ch has gene ally ocused on examining i s e ec in posi i e si ua ions (e.g. A ulka and Kesa i, 2017). This esea ch shows di e en gende e ec s, speci ically in in e ac ion expe iences ha a e no seamless, and opens new ques ions abou i s ole in he omnichannel con ex . As a esul , he impo ance o ocusing no only on he eac ions in ideal scena ios, bu also in possible uncom o able o unpleasan scena ios, is e idenced in his pape . This con i ms he impo ance o con inuing o s udy gende di e ences in consume beha io , as A ulka and Kesa i (2017) s a e. Fou h, his s udy p esen s obus esul s ha a e eplica ed using wo me hods, a su ey o eal omni-consume s and a con olled expe imen , which allows o a conse a i e es o he p oposed esea ch ques ions. In s udy 1, pa icipan s e lec ed on hei own pas expe iences as omni-consume s and e alua ed he seamlessness o hei in e ac ion expe ience. In s udy 2, a con olled expe imen al se ing was used o c ea e an omnichannel en i onmen ha was ei he seamless o non-seamless. In bo h s udies, esul s obus ly suppo he p oposed dimensions o he OSIE and hei ela i e impo ance, and he ela ionship be ween he OSIE and highe cus ome sa is ac ion wi h hei omnichannel in e ac ion. Manage ial implica ions Omnichannel beha io has become a bu ning opic in ma ke ing oday. As a esul , academic esea che s ha e emphasized ha p ac i ione s need o manage mul iple channels in a seamless way. Howe e , a ecen epo edi ed by he Bos on Consul ing G oup (BCG) shows a su p ising eali y. 83% o he companies could no make connec ions ac oss consume ouchpoin s, and 80% su e ed om inadequa e channel coo dina ion (Field, Pa el, & Leon, 2019). As guidelines o achie e his c ucial aspec o e ail s a egy ha e no p e iously been speci ied, his esea ch o e s se e al impo an con ibu ions o p ac i ione s ega ding he 35 omnichannel en i onmen . Fi s , he esul s in o m omnichannel manage s abou how o design channels o o e he undamen al seamless in e ac ion expe ience o hei cus ome s. Conc e ely, he indings o his esea ch show ha consis ency is he mos impo an dimension o c ea e an OSIE, ollowed by synch oniza ion and, inally, eedom in channel selec ion. Thus, manage s may c i ically analyze hei s a egy and modi y i o p esen a uni ied ep esen a ion o he company ac oss all he channels (e.g. same aes he ics and b and image, same p ices and asso men o consis en p omo ions ac oss he channels). Then, channels should be synch onized and he men ali y o “wha you s a online (o line) has o inish online (o line)” mus disappea . Ins ead, o achie e synch oniza ion ac oss channels, manage s should adop he philosophy ha `channels ha e no ba ie s´. Fi ms should in eg a e he managemen o channels o o e lexible pu chase and e u n policies ha allow consume s o b idge mul iple channels in a single pu chase. Consequen ly, i ms may allow consume s o p oceed in each s age o he decision-making p ocess whe e e hey wan , in e changing and combining all he channels acco ding o hei needs wi hou es ic ions. Companies will a oid ee- iding beha io (Fla ián, Gu ea, & O ús, 2019), which is one o he mos c i ical challenges ha i ms ace. Second, he esul s o his pape con i m he impo ance o de eloping an imp o ed s a egy o omnichannel managemen o inc ease cus ome s’ sa is ac ion. Fi ms ha u ilize mul iple e ail channels should implemen he dimensions o OSIE o c ea e a seamless en i onmen in o de o inc ease cus ome sa is ac ion wi h he omnichannel in e ac ions. To o e a seamless in e ac ion expe ience and imp o e sa is ac ion i ms should, o example, allow cus ome s o buy a p oduc online and hen pick i up in he s o e, o p o ide he same in o ma ion abou p oduc s, p ices, and sales p omo ions ac oss all he channels. A uni o m ep esen a ion ac oss channels should make he cus ome eel like hey a e in e ac ing wi h a single uni ied b and ac oss all ouchpoin s. 36 The esul s o his esea ch also p esen insigh s o b ands ha speci ically a ge a emale ma ke . I is especially impo an o such b ands o manage omnichannel ouchpoin s seamlessly, because non-seamless in e ac ion expe iences mo e nega i ely impac emale cus ome s’ sa is ac ion wi h hei in e ac ion expe ience, ela i e o males. Consequen ly, i ms should be pa icula ly conscious o he impo ance o OSIE and i s impac on sa is ac ion o emale consume s. Fo example, i ms ha speci ically a ge emale consume s migh emphasize he seamlessness by including messages ha accen ua e he consis ency, eedom, and synch oniza ion ac oss channels (e.g. “buy he p oduc online and pick i up in he s o e” o “ o assis ance you can ask us online o isi ou s o e”). Mo eo e , i ms could use sa is ac ion su eys o ask cus ome s abou hei in e ac ion expe iences o de ec a eas o po en ial imp o emen . Limi a ions and u he esea ch This s udy is subjec o some limi a ions ha can be add essed in u he esea ch. Fi s , only he e ec o he OSIE on cus ome sa is ac ion is es ed. Fu u e esea ch can conside i s impac on addi ional a iables, such as b and p e e ence (J. Zhang, e al., 2010), b and lo e (Palusuk, Koles, & Hasan, 2019) o wo d-o -mou h beha io s (Manse Payne, Pel ie , & Ba ge , 2017). The impac o seamlessness in he omnichannel in e ac ion expe ience can also be examined on consume decision- ela ed a iables, o example, choice beha io , decision con idence, o decision com o . Simila ly, u he analysis could ex end he esea ch by including some mode a ing and/o media ing e ec s (Cummins, e al., 2016; Ve hoe , e al., 2015). As mode a o s, he indi idual di e ences among consume s, such as omnichannel endency, channel p e e ence, o channel use can be included. Sa is ac ion may wo k as media o be ween he pe cei ed OSIE and subsequen consume beha io . The examina ion o downs eam e ec s o cus ome sa is ac ion wi h he omnichannel in e ac ion expe ience 37 would add a g ea e unde s anding o how he seamless ansi ion be ween channels in luences consume beha io . Second, al hough s udy 1 does no ocus on one speci ic sec o , s udy 2 ocuses on a single p oduc ca ego y. Because o he echnical limi a ions and complexi y o an online expe imen , a black basic -shi was used as a s anda d p oduc ha p esen s unisex cha ac e is ics. To alle ia e his limi a ion, u u e esea ch should eplica e he s udy in o he ca ego ies. Fu u e esea ch may also examine he e ec s o consis ency, eedom in channel selec ion, and synch oniza ion by a ying di e en aspec s o hese h ee dimensions han we e used in he p esen s udy. The expe imen al s imuli used in s udy 2 we e c ea ed by a ying elemen s o he shopping scena io ha a e assessed in he scales in he li e a u e and used in his pape . Speci ically, p oduc a ailabili y, p ice, b and logo, and sales p omo ions we e he elemen s o he shopping scena io ha di e ed be ween seamless and non-seamless condi ions. Fu u e esea ch may examine each o hese a iables sepa a ely o disce n hei indi idual e ec s on cus ome sa is ac ion wi h he omnichannel in e ac ion. Mo eo e , u u e esea ch may manipula e consis ency using o he aspec s o he in e ac ion expe ience, such as cus ome se ice, o examine he e ec s o on cus ome sa is ac ion. Addi ionally, he p esen esea ch is one o he i s o inco po a e social media as an impo an ouchpoin in he omnichannel domain, and u u e s udies may expand on his channel. As consume beha io has e ol ed o inco po a e social media as a means o in e ac ing wi h b ands, sea ching o in o ma ion, and as a pla o m o bo h i m and consume -gene a ed ma ke ing con en , i is impo an o omnichannel esea ch o begin o inco po a e his channel (Cummins, e al., 2016; Sands, e al., 2016). Fu u e esea ch is needed o examine he ways in which i can be seamlessly inco po a ed in o i ms’ po olio o consume ouchpoin s (e.g. wea ables, digi al assis an s, oice…). 38 Finally, we iden i y he OSIE cons uc s om he o e a ching hemes ound in he li e a u e. Conside ing ha he e ail en i onmen cons an ly e ol es and ha sa is ac ion wi h he OSIE is based on each consume ’s subjec i e pe cep ion o he seamlessness o hei in e ac ion expe ience, quali a i e esea ch may be a ui ul a enue o u u e wo k in his domain. Re e ences A ulka , S., & Kesa i, B. (2017). Sa is ac ion, loyal y and epa onage in en ions: Role o hedonic shopping alues. Jou nal o Re ailing and Consume Se ices, 39, pp. 23-34. Ba wi z, N., & Maas, P. (2018). Unde s anding he Omnichannel Cus ome Jou ney: De e minan s o In e ac ion Choice. Jou nal o In e ac i e Ma ke ing, 43, pp. 116-133. Beck, N., & Rygl, D. (2015). Ca ego iza ion o mul iple channel e ailing in Mul i-, C oss-, and Omni‐Channel Re ailing o e aile s and e ailing. Jou nal o Re ailing and Consume Se ices, 27, pp. 170-178. Becke , J.-M., Klein, K., & We zels, M. (2012). Hie a chical la en a iable models in PLS- SEM: guidelines o using e lec i e- o ma i e ype models. Long Range Planning, 45(5- 6), pp. 359-394. Blom, A., Lange, F., & Hess J , R. L. (2017). Omnichannel-based p omo ions’ e ec s on pu chase beha io and b and image. Jou nal o Re ailing and Consume Se ices, 39, pp. 286-295. B ynjol sson, E., Hu, Y. J., & Rahman, M. S. (2013). Compe ing in he age o omnichannel e ailing. MIT Sloan Managemen Re iew, 54(4), p 23. Cao, L., & Li, L. (2015). The impac o c oss-channel in eg a ion on e aile s’ sales g ow h. Jou nal o Re ailing, 91(2), pp. 198-216. Cao, L., & Li, L. (2018). De e minan s o Re aile s' C oss-channel In eg a ion: An Inno a ion Di usion Pe spec i e on Omni-channel Re ailing. Jou nal o In e ac i e Ma ke ing, 44, pp. 1-16. Col man, T., De inney, T. M., Midgley, D. F., & Venaik, S. (2008). Fo ma i e e sus e lec i e measu emen models: Two applica ions o o ma i e measu emen . Jou nal o Business Resea ch, 61(12), pp. 1250-1262. 39 Cummins, S., Pel ie , J., & Dixon, A. (2016). Omni-channel esea ch amewo k in he con ex o pe sonal selling and sales managemen : A e iew and esea ch ex ensions. Jou nal o Resea ch in In e ac i e Ma ke ing, 10(1), pp. 2-16. Cha e jee, P. (2010). Causes and consequences o ‘o de online pick up in-s o e’ shopping beha io . The In e na ional Re iew o Re ail, Dis ibu ion and Consume Resea ch, 20(4), pp. 431-448. Chin, W. W., & News ed, P. R. (1999). S uc u al equa ion modeling analysis wi h small samples using pa ial leas squa es. In R. H. Hoyle (Ed.), S a is ical s a egies o small sample esea ch (pp. 307-342). Thousand Oaks, US: Sage. Ewe ha d, A.-C., Siso sky, K., & Johansson, U. (2019). Consume decision-making o slow mo ing consume goods in he age o mul i-channels The In e na ional Re iew o Re ail, Dis ibu ion and Consume Resea ch, 29(1), pp. 1-22. Fan, X., Chai, Z., Deng, N., & Dong, X. (2020). Adop ion o augmen ed eali y in online e ailing and consume s’ p oduc a i ude: A cogni i e pe spec i e. Jou nal o Re ailing and Consume Se ices, 53, pp. 1-10. Field, D., Pa el, S., & Leon, H. (2019). The di idends o digi al ma ke ing ma u i y. Fla ián, C., Gu ea, R., & O ús, C. (2019). Feeling Con iden and Sma wi h Web ooming: Unde s anding he Consume 's Pa h o Sa is ac ion. Jou nal o In e ac i e Ma ke ing, 47, pp. 1-15. F asque , M., & Miquel, M.-J. (2017). Do channel in eg a ion e o s pay-o in e ms o online and o line cus ome loyal y? In e na ional Jou nal o Re ail & Dis ibu ion Managemen , 45(7/8), pp. 859-873. Gao, R., & Yang, Y.-X. (2016). Consume s’ Decision: Fashion Omni-channel Re ailing. Jou nal o In o ma ion Hiding and Mul imedia Signal P ocessing, 7(2), pp. 325-342. Haide , S. W., Zhuang, G., Hashmi, H. b. A., & Ali, S. (2020). Ch ono ypes’ Task- Technology Fi o Sea ch and Pu chase in Omnichannel Con ex . Mobile In o ma ion Sys ems, 14(1), pp. 148-167. Hai , J. F., Hul , G. T. M., Ringle, C., & Sa s ed , M. (2017). A p ime on pa ial leas squa es s uc u al equa ion modeling (PLS-SEM) (2ª ed.) Los Angeles, US: Sage Publica ions. Hai , J. F., Ringle, C., & Sa s ed , M. (2011). PLS-SEM: Indeed a sil e bulle . Jou nal o Ma ke ing heo y and P ac ice, 19(2), pp. 139-152. Hai , J. F., Sa s ed , M., Ringle, C., & Gude gan, S. P. (2018). Ad anced issues in pa ial leas squa es s uc u al equa ion modeling Thousand Oaks, US: Sage Publica ions. 40 Hayes, A. F. (2013). In oduc ion o media ion, mode a ion, and condi ional p ocess analysis: A eg ession-based app oach New Yo k, US: Guil o d P ess. Huang, L., Lu, X., & Ba, S. (2016). An empi ical s udy o he c oss-channel e ec s be ween web and mobile shopping channels. In o ma ion & Managemen , 53(2), pp. 265-278. Hu é, E., Pico -Coupey, K., & Acke mann, C.-L. (2017). Unde s anding omni-channel shopping alue: A mixed-me hod s udy. Jou nal o Re ailing and Consume Se ices, 39, pp. 314-330. Ie a, M., & Ziliani, C. (2018). Mapping ouchpoin exposu e in e ailing: Implica ions o de eloping an omnichannel cus ome expe ience. In e na ional Jou nal o Re ail & Dis ibu ion Managemen , 46(3), pp. 304-322. Ja is, C. B., MacKenzie, S. B., & Podsako , P. M. (2003). A c i ical e iew o cons uc indica o s and measu emen model misspeci ica ion in ma ke ing and consume esea ch. Jou nal o consume esea ch, 30(2), pp. 199-218. Juaneda-Ayensa, E., Mosque a, A., & Sie a Mu illo, Y. (2016). Omnichannel Cus ome Beha io : Key D i e s o Technology Accep ance and Use and Thei E ec s on Pu chase In en ion. F on ie s in Psychology, 7, pp. 1-11. Kang, J.-Y. M. (2019). Wha d i es omnichannel shopping beha io s? Fashion li es yle o social-local-mobile consume s. Jou nal o Fashion Ma ke ing and Managemen , 23(2), pp. 224-238. Kassa jian, H. H. (1977). Con en analysis in consume esea ch. Jou nal o consume esea ch, 4(1), pp. 8-18. Kim, H., Ahn, S.-K., & Fo ney, J. A. (2014). Shi ing pa adigms o ashion: F om o al o global o sma consume expe ience. Fashion and Tex iles, 1(1), p 15. Kuma , V., & Reina z, W. (2016). C ea ing endu ing cus ome alue. Jou nal o Ma ke ing, 80(6), pp. 36-68. Law, K. S., Wong, C.-S., & Mobley, W. M. (1998). Towa d a axonomy o mul idimensional cons uc s. Academy o managemen e iew, 23(4), pp. 741-755. Laza is, C., & V echopoulos, A. (2014) F om Mul ichannel o "Omnichannel" Re ailing: Re iew o he Li e a u e and Calls o Resea ch. Pape p esen ed a he 2nd In e na ional Con e ence on Con empo a y Ma ke ing Issues,(ICCMI). Laza is, C., V echopoulos, A., Doukidis, G., & F aidaki, K. (2015) The In e play o Omni e ailing & S o e A mosphe e on Consume s’ Pu chase In en ion owa ds he Physical Re ail S o e. Pape p esen ed a he 12 h Eu opean, Medi e anean & Middle Eas e n Con e ence on In o ma ion Sys ems (EMCIS). 41 Lee, H.-H., & Kim, J. (2010). In es iga ing dimensionali y o mul ichannel e aile 's c oss- channel in eg a ion p ac ices and e ec i eness: shopping o ien a ion and loyal y in en ion. Jou nal o Ma ke ing Channels, 17(4), pp. 281-312. Lemon, K. N., & Ve hoe , P. C. (2016). Unde s anding cus ome expe ience h oughou he cus ome jou ney. Jou nal o Ma ke ing, 80(6), pp. 69-96. Le oi-We elds, S., S eukens, S., B ady, M. K., & Swinnen, G. (2014). Assessing he alue o commonly used me hods o measu ing cus ome alue: A mul i-se ing empi ical s udy. Jou nal o he Academy o Ma ke ing Science, 42(4), pp. 430-451. Le y, M., Wei z, B., & G ewal, D. (2013). Re ailing Managemen (9 h Edi ion ed.) New Yo k, US: I win/McG aw-Hill. Luo, J., McGold ick, P., Bea y, S., & Keeling, K. A. (2006). On-sc een cha ac e s: hei design and in luence on consume us . Jou nal o Se ices Ma ke ing, 20(2), pp. 112-124. Manse Payne, E., Pel ie , J., & Ba ge , V. A. (2017). Omni-channel ma ke ing, in eg a ed ma ke ing communica ions, and consume engagemen : a esea ch agenda. Jou nal o Resea ch in In e ac i e Ma ke ing, 11(2), pp. 185-197. Mi sch, T., Leh e , C., & Jung, R. (2016). Channel in eg a ion owa ds omnichannel managemen : a li e a u e e iew. 20 h Paci ic Asia Con e ence on In o ma ion Sys ems, Chiayi, Taiwan. Mosque a, A., Ola e-Pascual, C., Juaneda Ayensa, E., & Sie a Mu illo, Y. (2018). The ole o echnology in an omnichannel physical s o e: Assessing he mode a ing e ec o gende . Spanish Jou nal o Ma ke ing - ESIC, 22(1), pp. 63-82. Mosque a, A., Pascual, C. O., & Juaneda Ayensa, E. (2017). Unde s anding he cus ome expe ience in he age o omni-channel shopping. Icono14, 15(2), p 5. MSI, M. S. I. (2018). 2018-2020 Resea ch P io i ies. Bos on, US Neslin, S. A., G ewal, D., Legho n, R., Shanka , V., Tee ling, M. L., Thomas, J. S., & Ve hoe , P. C. (2006). Challenges and oppo uni ies in mul ichannel cus ome managemen . Jou nal o Se ice Resea ch, 9(2), pp. 95-112. Neslin, S. A., & Shanka , V. (2009). Key issues in mul ichannel cus ome managemen : cu en knowledge and u u e di ec ions. Jou nal o In e ac i e Ma ke ing, 23(1), pp. 70- 81. Oh, L.-B., & Teo, H.-H. (2010). Consume alue co-c ea ion in a hyb id comme ce se ice- deli e y sys em. In e na ional Jou nal o Elec onic Comme ce, 14(3), pp. 35-62. 48 Appendix 2 Manipula ion (2-cell be ween subjec s) “Imagine you a e in a s o e shopping o a black -shi (o some hing ha is ela i ely uni e sal/gende neu al/mos people would buy and wea ). While you a e in he s o e, you a e using you mobile de ice o look a he b and’s websi e and social media pages o in o ma ion. Below is some o he in o ma ion ha you see on hei websi e and social media page” ( he ollowing pic u es a e mockups o wha was p esen ed o pa icipan s): Figu e A.1. Condi ion 1: seamless Figu e A.2. Condi ion 2: non-seamless Scena ios (seamless condi ion in bold): You ask he sales associa e abou he -shi and a 20% o sales p omo ion ha is being ad e ised online. [The associa e ells you ha hey can hono he same sales p omo ion in- s o e] [The associa e ells you ha hey canno hono he same sales p omo ion in-s o e]. A e you’ e ound enough in o ma ion abou he -shi ha you wan o pu chase on he b and’s online pla o ms, you ask he sales associa e o y on he -shi . The associa e [ ells you ha hey do ca y he -shi in he s o e, bu ha you size is ou o s ock] OR [ ells you ha hey don’ ca y he -shi in he s o e, because he p oduc s a e di e en om he websi e]. A e ha you ask he sales associa e abou he possibili y o o de ing online o pick up in he s o e and whe he you can exchange o e u n he -shi in he s o e i you buy i online. [The associa e ells you ha you can pick up and exchange o e u n he -shi in he s o e ega dless o whe e you buy i ] [The associa e ells you ha you canno pick up, o exchange o e u n he -shi in he s o e, because he sales channels a e di e en ]. You buy he -shi on hei websi e and [decide o pick up i in he s o e] [wai o i o be deli e ed a home]. 49 Appendix 3 Co ela ion ma ix ( i s -o de model) C1 C2 C3 C4 C5 F1 F2 F3 F4 S1 S2 S3 S4 S5 S6 C1 1.000 C2 .346 1.000 C3 .332 .518 1.000 C4 .323 .316 .412 1.000 C5 .253 .454 .419 .430 1.000 F1 .315 .353 .325 .242 .389 1.000 F2 .131 .273 .304 .199 .399 .331 1.000 F3 .098 .154 .216 .223 .240 .197 .444 1.000 F4 .171 .183 .286 .198 .306 .332 .338 .453 1.000 S1 .149 .236 .241 .096 .314 .206 .344 .214 .314 1.000 S2 .189 .282 .119 .089 .198 .180 .332 .189 .309 .388 1.000 S3 .245 .144 .181 .177 .317 .296 .300 .220 .153 .359 .288 1.000 S4 .214 .116 .103 .156 .144 .158 .400 .311 .130 .279 .269 .476 1.000 S5 .083 .145 .210 .221 .405 .296 .357 .217 .399 .327 .353 .353 .356 1.000 S6 .241 .379 .371 .338 .405 .465 .480 .427 .326 .296 .369 .333 .290 .370 1.000 No e: C: Consis ency; F: F eedom in channel selec ion; S: Synch oniza ion Co ela ion ma ix (second-o de model) Consis ency F eedom in channel selec ion Synch oniza ion Consis ency 1.000 F eedom in channel selec ion .554 1.000 Synch oniza ion .455 .671 1.000