149
4, XIX, 2016
Ma ke ing and T ade
DOI: 10.15240/ ul/001/2016-4-011
In oduc ion
B ands, b anding, b and managemen , b and
o ien a ion and hei ela ed e ms ha e
ecei ed subs an ial a en ion om ma ke ing
schola s and p ac i ione s (Peng, Chen, & Wen,
2014). This could p obably be due o hei
s a egic impo ance o o ganiza ions (U de,
Baumga h, & Me ilees, 2013). Consume s on
he o he hand ha e also come o emb ace he
concep , as i helps hem in hei pu chasing
decisions by o e ing hem signals o imp o ed
e i ciency in in o ma ion p ocessing and o
selec ing p oduc s (Ko le & P oe sch, 2007;
Shi & Chow, 2015; Zablah, B own, & Don hu,
2010). As a esul , o ganiza ions and consume s
alike end o become b and o ien ed (Kelle ,
2009). U de (1999) de i ned b and o ien a ion as
a p ocess by which a i m de elops, builds and
p o ec s a b and and i s iden i y as i in e ac s
wi h ac ual and po en ial cus ome s wi h he
in en ion o ob aining sus ainable compe i i e
ad an age(s). This means ha b and o ien a ion
is mean o achie e a s a egic pu pose and he
cen al aim o his is o ea n a good epu a ion
in he ma ke place and, impo an ly, do so by
building good ela ionships wi h consume s
ia communica ion pla o ms (Xin, Ramayah,
So o-Acos a, Popa, & Ping, 2014; So o-Acos a,
Popa, & Palacios-Ma qués, 2016). Gi en he
ole o social media si es in ou con empo a y
socie y, i makes a lo o sense o businesses,
and in pa icula online e aile s, o build
a ‘ wo-way’ symbio ic ela ionship wi h hei
consume s h ough some o hese popula
social media pla o ms (Cu as-Pe ez, Ruiz-
Ma e, & Sanz-Blas, 2014; So o-Acos a,
Molina-Cas illo, Lopez-Nicolas, & Colomo-
Palacios, 2014). Again, i can be in e ed ha
engaging cus ome s on a social media si e is
c i ical o b and o ien a ion as social media
p omo es cus ome - i m (b and) in e ac ion
(Boa eng, 2014). Howe e , his is ye o be
p o en empi ically and his s udy seeks o
add ess hese gaps. I also seeks o asce ain
i consis en engagemen wi h endo s’ social
media si e by consume s will lead o an
elec onic wo d-o -mou h – (e)WoM e ec .
Again, since b and o ien a ion is a esou ce
(U de e al., 2013), his s udy aims o i nd i
he e is a co ela ion be ween b and o ien a ion
and epu a ion o he en e p ise (e.g. online
e ail endo ), which is o g ea impo ance o
ma ke ing p ac i ione s (Capozzi, 2005).
Some s udies (Fomb un, Ga dbe g,
& Ba ne , 2000; Sung & Yang, 2008) ha e
poin ed o he se e al ma ke ing ou comes o he
epu a ion o he business en e p ise, including
mone a y and non-mone a y ou comes.
Howe e , i is no clea i hese bene i s a e
a ainable in an online con ex . Ba ikowski
and Walsh (2011) no ed ha while co po a e
epu a ion has a ac ed conside able a en ion
in ma ke ing discipline and p ac ice, he e is
a dea h o li e a u e on he e ec s o co po a e
epu a ion on mos ac ions o consume s.
Fo example, can a endo ’s epu a ion lead
o (e)WoM e ec and d aw cus ome s o i s
social media si es? Does a endo ’s epu a ion
in l uence consume s’ epu chase in en ions
in an online en i onmen ? These ques ions
UNDERSTANDING COSMOPOLITAN
CONSUMERS’ REPEAT PURCHASING
IN THE eMARKETPLACE: CONTRIBUTION
FROM A BRAND ORIENTATION
THEORETICAL PERSPECTIVE
Ch is ian Nedu Osakwe, Hen y Boa eng, Simona Popa,
Milosla a Cho anco á, Ped o So o-Acos a
EM_4_2016.indd 149EM_4_2016.indd 149 30.11.2016 16:31:5230.11.2016 16:31:52
150 2016, XIX, 4
Ma ke ing a obchod
equi e answe s, which his s udy seeks o
add ess. Al hough, some s udies (e.g. Jalkala
& Salminen, 2009) ha e been conduc ed
in a business- o-business con ex , Aa ikka-
S en oos and Makkonen (2014) a gue ha he
ele ance o he i ndings migh no be same in
consume ma ke and he e o e call o mo e
s udies o be conduc ed in a business- o-
consume ela ionship.
In sum, in his pape , we add ess esea ch
issues connec ed o he ole ha b and o ien a ion
(po en ially) plays in c i ically in l uencing he
cons uc s social media si e engagemen and
(online) endo epu a ion; he assump ion ha
endo epu a ion may also s ongly in l uence
he cons uc s social media si e engagemen ,
(e)WoM e ec , and epu chase in en ion is
also in e oga ed. Addi ionally, his scien i i c
pape es s he assump ion abou he di ec
e ec o (e)WoM on cus ome s epu chase
in en ion wi hin he con ex o he elec onic
ma ke place (eMa ke place). Acco dingly,
he s udy’s majo objec i e is o acili a e he
be e unde s anding o he in e ela ionships
amongs he phenomena o b and o ien a ion,
(e)WoM e ec , social media engagemen ,
endo epu a ion, and epu chase in en ion in
he eMa ke place con ex .
By and la ge, his scien i i c publica ion is
g ounded in he ela ionship ma ke ing li e a u e.
The emainde o he pape is o ganised as
ollows. Sec ion wo ocuses on he heo e ical
backg ound o he s udy, including he esea ch
hypo heses, while sec ion h ee p esen s he
me hodology employed. Sec ion ou o e s he
i ndings, which a e discussed in sec ion i e.
Conclusions, limi a ions and u u e esea ch
a e deal wi h in sec ion six.
1. Theo e ical Backg ound and
Hypo heses De elopmen
1.1 B and O ien a ion (BO) Social
Media Si e Engagemen (SME) and
Vendo Repu a ion (VRP)
B ands c ea e imp essions and emo ions and
elici beha iou al esponses h ough hei
unique iden i y (B akus, Schmi , & Za an onello,
2009). Whene e consume s come in o con ac
wi h a b and hey o m pe cep ions, and hese
in l uence hei decision o use he b and,
ecommend i o o he s and pay a en ion o
any p omo ion abou i (Amble e al., 2002).
Some s udies (e.g. Hu e , Hau z, Dennha d ,
& Fulle , 2013) a gue ha consume s con inued
in e ac ion wi h a b and con ibu es o hei
posi i e pe cep ions abou he b and. In o he
wo ds, b and o ien a ion can impac on
a endo ’s epu a ion since b and o ien a ion
adop s a con inued in e ac ion app oach wi h
cus ome s o build s ong b and iden i y (U de,
1999). As a esul , some o ganiza ions adop
a b and o ien a ion app oach whe e hey c ea e,
g ow and p o ec he iden i y o he b ands (U de,
1999). Sahin, Zehi and Ki apaci (2011) ound
ha , b and o ien a ion is posi i ely associa ed
wi h consume b and us . This p o ides
a basis o he e ec o b and o ien a ion on
a endo ’s epu a ion, since some o ganiza ions
gain hei epu a ion by being us wo hy
(Miyamo o & Rexha, 2004). A epu able endo
ins ils us and con i dence in i s cus ome s by
deli e ing se ices as p omised and so c ea es
alue o i s cus ome s (Agus in & Singh, 2005).
Algesheime , Dholakia and He mann (2005),
on he o he hand a gue ha , b and ela ionship
quali y in l uences consume s’ engagemen wi h
a b and communi y. I is sa e, he e o e, o a gue
ha b and o ien a ion con ibu es o consume s’
engagemen wi h he social media si e o
a endo . Mo eo e , in ano he ela ed s udy
by Jayawa dhena, W igh and Dennis (2007),
he au ho s p esume ha online shoppe s who
a e mo e b and mo i a ed (o o ien ed) a e
equally likely o b and-loyal shoppe s. Thus, i
is expec ed ha hese b and o ien ed (o sa y)
consume s would be mo e inclined o shop wi h
highly epu able online e ail endo s. Again,
since ce ain ac o s incline human beings o
bond wi h o he people o objec s, cus ome s
can bond wi h a endo ’s websi e; howe e ,
his will depend on he epu a ion o he endo
(Pa k & Kim, 2014). Fu he mo e, King, So and
G ace (2013) explo ed he e ec o se ice
b and o ien a ion on he a i ude o employees
in a ho el and ound ha se ice b and
o ien a ion posi i ely a ec s employee b and-
o ien ed beha iou . Again, applying he same
logic, i can be concluded ha b and o ien a ion
can con ibu e o cus ome s’ engagemen
wi h a endo ’s social media si e. Based on
he e idence abo e, he ollowing hypo heses
inco po a e ou expec a ions:
H1: B and o ien a ion is posi i ely associa ed
wi h endo ’s epu a ion.
H2: B and o ien a ion is posi i ely associa ed
wi h social media si e engagemen .
EM_4_2016.indd 150EM_4_2016.indd 150 30.11.2016 16:31:5230.11.2016 16:31:52
151
4, XIX, 2016
Ma ke ing and T ade
1.2 Social Media Si e Engagemen
(SME) and (e)WoM
The adical e olu ion o In e ne and i s
ela ed echnologies has c ea ed a new ool
o indi iduals and businesses o ne wo k,
engage and in e ac wi h each o he (Lei ne
& G echenig, 2007; Wang, Xu, & Chan,
2015; Zheng, Cheung, Lee, & Liang, 2015).
Social media echnologies especially p o ide
oppo uni ies o business o ganiza ions
o a ac consume s and engage hem in
con e sa ions on hei social media pla o ms.
Social media si es (SMSs) ha e been de i ned
as web-based applica ions ha help o gene a e
p o i les, upload pic u es, ideos and sha e wi h
o he people who a e connec ed in he ne wo k
(Boyd & Ellison, 2007; Wa en, Sulaiman,
& Jaa a , 2015). Social media pla o ms
ep esen powe ul ools o in e ac ion and
in o ma ion sha ing in gene al (Im an, 2014). On
he pa o endo s, hey p esen emendous
oppo uni ies o hese b ands o build an
engaging ela ionship wi h hei consume s ia
cus ome ( an)-b and ollowe ship. On he one
hand, cus ome s can use hese ools o pass
posi i e and nega i es commen s abou b ands.
Chu and Kim (2011) no ed ha , consume s
ha e employed SMSs such as Facebook,
Qzone, MySpace, Ins ag am, Twi e , and
LinkedIn o c ea e and sha e p oduc ela ed
in o ma ion wi h o he consume s, consequen ly
in l uencing hei pu chasing decision (Henning-
Thu au, Gwinne , Walsh, & G emle , 2004;
Wang & Doong, 2010). This means ha Social
media si es a e i ally impo an channels o
(e)WoM. Some consume s, especially hose
ha a e highly p ice conscious, spend ime
seeking opinions om o he s in a ious online
communi ies wi h ega d o bes p ices, and
subsequen ly sha e his in o ma ion wi h o he
cus ome s hey engage wi h online and e en
in some ins ances ia o l ine communica ion
(Kang, 2007). This acco ding o Ha is and
Dennis (2011) is one o he bene i s o social
websi es. This is an indica ion ha social media
engagemen can in l uence (e)WoM. Some
s udies (e.g. Walsh, Mi chell, Jackson, & Bea y,
2009) ha e ound ha cus ome s who a e loyal
and commi ed o some o ganiza ions suppo
hem h ough addi ional ole beha iou . In his
sense, hese consume s a e mo e likely no
only o engage (by means o in e ac ion) wi h
hei a ou i e endo ’s b and social media
ans’ pages, bu , mo e impo an ly, hey a e
also likely o pass on posi i e commen s abou
online e ail endo s. Agains his backd op, he
ollowing hypo hesis is o mula ed:
H3: Social media si e engagemen is
posi i ely ela ed o (e)WoM.
1.3 Vendo Repu a ion (VRP), (e)WoM
E ec and Repu chase In en ion
(RPI)
F om he esou ce-based iew o he i m
pe spec i e, co po a e epu a ion can be
ega ded as a aluable esou ce ha can enable
o ganiza ions o gain a las ing compe i i e edge
(Capozzi, 2005). Howe e , his will depend on
how well he i m uses his epu a ion. Repu able
endo s can in l uence hei consume s o pass
on a ou able commen s, which can lead o he s
o buy om hose endo s (Ah ens, Coyle,
& S ahile i z, 2013; Xun, 2014). Consume s
ha e posi i e a i udes owa ds epu able
endo s and his esul s in a posi i e b and
esponse such as passing on posi i e commen s
abou he endo , which in u n may lead o
epea pu chasing among he online shoppe s
as well as cus ome pa onage o some online
e ail endo s (Gup a, Melewa , & Bou lakis,
2010; Madi inos & Theodo idis, 2010; Huang,
2014; So o-Acos a, Pe ez-Gonzalez, & Popa,
2014). Addi ionally, Fomb un e al. (2000) and
Sung and Yang (2008) asse ha cus ome s
o well- epu ed endo s engage in suppo i e
beha iou s. This is an indica ion ha cus ome s
o epu able endo s will engage in (e)WoM o
he endo s and pa onise hei b ands. Lin, Lu
and Wu (2012) examined he e ec o co po a e
image and ela ionship ma ke ing on us and
consume pu chase in en ion and ound ha
co po a e image signi i can ly and posi i ely
impac s on consume s’ pu chase in en ions;
and wo d-o -mou h enhances his e ec . F om
hese i ndings, i is sa e o conclude ha endo s’
epu a ion and (e)WoM in l uence shoppe s’
epu chase in en ions online. B engman and
Ka imo (2012) a gue ha , since i is di i cul
o assess in insic ea u es o b ands online,
online shoppe s may seek unbiased opinions
in hei pu chasing decisions. Consequen ly,
online p oduc e alua ion has become common
among online consume s and a sou ce o
in o ma ion o pu chasing decision (Hu e al.,
2008). In ecen imes, some endo s ha e
employed and emb aced social media in o de
o ob ain consume insigh s (Hudson & Hudson,
2013), and use he oppo uni y o ini ia e
EM_4_2016.indd 151EM_4_2016.indd 151 30.11.2016 16:31:5230.11.2016 16:31:52
152 2016, XIX, 4
Ma ke ing a obchod
(e)WoM. They do his by including a “ ell-a-
iend” ea u e on hei an pages, o p omo ing
online p oduc a ings (Ah ens e al., 2013).
Based on he o egoing a gumen s, hese lead
us o o mula e he ollowing hypo heses:
H4: Vendo ’s epu a ion is posi i ely
associa ed wi h (e)WoM.
H5: (e)WoM is posi i ely associa ed wi h
cus ome s’ epu chase in en ion.
1.4 Vendo Repu a ion (VRP), Social
Media Si e Engagemen (SME)
and Repu chase In en ion (RPI)
Keh and Xie (2009) i nd ha co po a e epu a ion
posi i ely a ec s cus ome us and cus ome
iden i i ca ion, so o cus ome s o engage and
in e ac wi h a endo will depend on how
epu able he endo is. Vendo epu a ion can
ha e se e al ou comes. B engman and Ka imo
(2012) no e in hei s udy ha inco po a ing
social media in o co po a e communica ion can
in l uence cus ome s owa ds un amilia e- aile s
and “pu chase in en ions”, bu hey cau ion ha
a endo should only in eg a e app op ia e and
impo an ones. In an online en i onmen , cues
such as endo ’s image in o ma ion (Jin, Pa k,
& Kim, 2008) can ins il us and con i dence
in consume s and consequen ly hey us he
online e aile (Wu, Chen, & Chung, 2010; Wu
& Huang, 2015). Howe e , Wu e al. (2010)
no e ha his in o ma ion migh no be p esen
when dealing wi h an un amilia e- endo and
he e o e he ini ial us o med would be based
on he ea u es o he endo ’s si e. Benne and
Gab iel (2001) asse ha a posi i e co po a e
epu a ion gi es cus ome s consis en posi i e
ein o cemen , which commi s hem o he
o ganiza ion. Yoon, Choi and Sohn (2008) also
sha ed a simila opinion wi h he a o emen ioned
au ho s. Agains he iews ha ha e so a been
exp essed in he li e a u e, one may expec
ha he (pe cei ed) endo ’s epu a ion will
play a leading ole as o he online shoppe ’s
commi men o engage wi h he endo ’s
b and in social media pla o ms. Likewise, i
seems e y likely ha he endo ’s (pe cei ed)
epu a ion is e en mo e impo an han e e
be o e, as i is a c i ical de e minan o cus ome
pa onage. In line wi h he abo e a gumen a ion,
we make he ollowing hypo heses:
H6: Vendo ’s epu a ion is posi i ely ela ed
o social media si e engagemen .
H7: Vendo ’s epu a ion is posi i ely ela ed
o cus ome s’ epu chase in en ions.
The se o ela ions is illus a ed in Fig.1.
Fig. 1: Resea ch Model
Sou ce: own
EM_4_2016.indd 152EM_4_2016.indd 152 30.11.2016 16:31:5230.11.2016 16:31:52
153
4, XIX, 2016
Ma ke ing and T ade
2. Resea ch Design
2.1 Sample and Da a Collec ion
The a ge esponden s in ou s udy we e
‘cosmopoli an’ Slo ak ci izens. In his s udy, we
de i ned cosmopoli an Slo aks as hose who
could w i e and speak in English. Thus, ou
uni o analysis is a bilingual (Slo ak–English)
speake and a he same ime a Slo ak wi h
In e ne li e acy. Acco dingly, sample da a we e
collec ed using a s uc u ed ques ionnai e
and his was mainly despa ched h ough an
in i a ion o pa icipa e in a web-based su ey.
The su ey link was also pos ed on some
popula Facebook o ums ha a e cu en ly
been u ilized by Slo ak na ionals. To encou age
pa icipa ion, we assu ed he esponden s
o he con i den iali y o hei in o ma ion and
also made i explici ly clea ha he esea ch
se es o academic pu poses only. Da a
collec ion ook place du ing he i s and
second qua e s o 2014; in all, we we e able o
ga he 125 e ec i e esponses om he s udy
pa icipan s. In e es ingly, nea ly almos o he
pa icipan s epo ed o ha e online shopping
expe ience; and as migh be expec ed mos
o he esponden s we e Slo ak you hs (see
Tab. 1 o u he de ails abou he pa icipan s).
Impo an ly, ou sample size exceeds he “ en
imes ule” o e alua ing he sui abili y o
a sample size used in a mul i a ia e ( eg ession)
analysis (Cos a eda, 2009; De illeuxa, The on,
Beduin, & Hanzen, 2012; G een, 1991; Hill,
1998; VanVoo his & Mo gan, 2007). Mo e
speci i cally, he a io o sample size o he
numbe o pa ame e s o be es ima ed in ou
esea ch model is 25:1. F om he sample da a,
we also ound ha nea ly all he esponden s
use Facebook compa e o ew o he s who make
use o any one o Twi e , Ins ag am, Linkeldn
and Google+ (ou pu omi ed).
2.2 Measu es
A i e-poin Like scale, anging om comple ely
disag ee (1) o comple ely ag ee (5) was used
o elici esponses om he pa icipan s. All he
measu es used in he empi ical su ey we e
ob ained om ea lie s udies and ully modi i ed
o emo e any o m o ambigui y om he i ems
lis ed as pa o he i nal scale. The measu es
o endo epu a ion we e adap ed om Doney
and Cannon (1997); Kim, Yang and Kim (2013)
and Ja enpaa, T ac insky, and Vi ale (2000),
Sample cha ac e is ics
Gende : % In e ne Pu chase: %
Female 64.8 No 4.8
Male 35.2 Yes 95.2
Age G oup: F equency o In e ne Pu chase:
17-25 63.2 Daily 17.6
26-34 25.6 Weekly 42.9
35-43 7.2 Mon hly 29.4
44-52 3.2 Th ee o mo e imes a yea 10.1
53-61 0.8 Once o Twice yea ly –
61+ –
Educa ional S a us: Pa onage o Vendo s’ Websi e ou side
Slo akia:
High School 28 No 23.8
Unde g adua e/Bachelo ’s 29.6 Yes 76.2
Pos G adua e (Mas e ’s, PhD) 32.8
PhD/P o esso 8
O he s 1.6
Sou ce: own
Tab. 1: Demog aphic p o i le o s udy pa icipan s
EM_4_2016.indd 153EM_4_2016.indd 153 30.11.2016 16:31:5230.11.2016 16:31:52

154 2016, XIX, 4
Ma ke ing a obchod
while he i ems measu ing he cons uc o b and
o ien a ion we e adap ed om Ling, Chai, and
Piew (2010) and Seock (2003). The measu es
o (e)WoM e ec we e ob ained om Awad and
Ragowsky (2008), Mikale , Giannakos, and Pa eli
(2013) and Zei haml, Be y, and Pa asu aman
(1996). The i ems measu ing social media si e
engagemen wi h endo b ands we e based
on Ka akaya and Ba nes (2010), La oche,
Habibi, Richa d, & Sanka ana ayanan (2012),
and Ramna ainand and Go ende (2013). The
epu chase in en ion cons uc was adap ed
om Bha ache jee (2001), Ma hieson (1991),
and Thong, Hong, and Tam (2006). Cons uc s
and associa ed indica o s in he measu emen
model a e lis ed in he Appendix.
2.3 Common Me hod Bias (CMB)
Since we used a sel - epo ed ques ionnai e
s udy, we checked o CMB in he colla ed
da ase . Based on he sugges ion o Podsako ,
MacKenzie, Podsako , and Lee (2003), we an
a pos -hoc s a is ical analysis o he su eyed
da a by means o Ha man’s un o a ed single
ac o echnique, using he p incipal componen
analysis (PCA) oolbox in SPSS. Ou esul s
show ha he i s dominan dimension
accoun ed o a 21.6% a iance. Fu he mo e,
no single ac o eme ged om his un o a ed
ac o analysis; all i e dis inc componen s had
eigen alues o g ea e han one. This hin s a
he absence o CMB in he su eyed.
2.4 Psychome ic P ope ies
o Resea ch Cons uc s
Fi s , we checked he in e nal consis ency
eliabili y o ou measu emen ins umen s by
using C onbach’s alpha, mani es a iables’
loadings and composi e eliabili y. F om
ou ou pu (see Tab. 2), all he cons uc s’
Cons uc s αCR AVE Indica o s FL Boo s apped
T-S a is ics
BO 0.728 0.831 0.554
BO1 0.714 7.484
BO2 0.789 8.968
BO3 0.816 9.609
BO4 0.645 4.815
SME 0.778 0.852 0.591
SME2 0.711 4.084
SME3 0.693 3.429
SME4 0.785 4.547
SME5 0.875 5.089
VRP 0.725 0.828 0.547
VRP1 0.705 8.031
VRP2 0.812 12.965
VRP3 0.706 7.067
VRP4 0.731 10.359
(e)WoM 0.735 0.834 0.557
WoM1 0.686 11.479
WoM2 0.751 13.257
WoM3 0.771 13.262
WoM4 0.774 16.424
RPI 0.849 0.898 0.690
RPI2 0.844 22.863
RPI3 0.880 28.997
RPI4 0.901 43.600
RPI5 0.680 7.788
Sou ce: own
No e: Insigni i can i ems a e d opped (SME1 and RPI1)
Tab. 2: Reliabili y and con e gen alidi y o measu emen model
EM_4_2016.indd 154EM_4_2016.indd 154 30.11.2016 16:31:5330.11.2016 16:31:53
155
4, XIX, 2016
Ma ke ing and T ade
C onbach’s alpha alues a e abo e he
c i e ion alue o 0.7 (Nunnaly, 1978). The
mani es a iables’ loadings a e wi hin he
ange o 0.645 o 0.901 and a e all s a is ically
signi i can a he 0.01 le el; abou i e i ems
wi h insigni i can loadings we e pu ged ou om
he i nal analysis. Mo eo e , he composi e
eliabili y o each o he cons uc s is simila ly
abo e he sugges ed alue o 0.7 (c . Hai ,
Sa s ed , Ringle, & Mena, 2012; Shah ia ,
2014). In addi ion, he con e gen alidi y o all
he la en e l ec i e cons uc s was checked by
he a e age a iance ex ac ed (AVE), which
a e all abo e he commonly accep ed h eshold
o 0.5 (c . Rod iguez, Reina, & Ru i n, 2015).
The ea e , we checked o disc iminan
alidi y o he measu emen model using he
app oach sugges ed by Fo nell and Lacke
(1981). The esul s show ha he squa e oo s
o he AVEs, which a e in he diagonal cells (see
Tab. 3) exceed he in e -co ela ions be ween
he e l ec i e la en a iables. To u he con i m
he disc iminan alidi y o ou cons uc s,
we wen ahead by aking a look a he i ems’
c oss loadings. All he i ems we e ound o
load di e en ly on hei assigned cons uc s
and he a e age c oss loading was abou 0.2
(ou pu omi ed). To sum up, ou measu emen
model shows ha ou esea ch cons uc s
ha e sui able eliabili y and cons uc alidi y.
All he compu a ions we e ca ied ou using
he Sma PLS 2.0 (Ringle, Wende, & Will,
2005) s uc u al equa ion modelling so wa e
wi h a non-pa ame ic boo s ap o 5,000 sub-
samples used o gene a e he T-S a is ics.
In his a icle, we op ed o use PLS pa h
modelling due o he explo a o y na u e o ou
s udy. Besides, we a e mainly in e es ed in he
p edic i e alidi y o he esea ch model gi en
also ou limi ed sample size. All hese ac o s
and e en mo e wa an he use o PLS o e he
co a iance-based SEM me hod.
3. Hypo heses Tes ing
Acco ding o ele an s udies (e.g. see Hai e
al., 2012; Tenenhaus, Vinzi, Cha elin, & Lau o,
2005), au ho s a e expec ed o e alua e he
inne model’s coe i cien o de e mina ion (R2),
pa h coe i cien s, p edic i e ele ance (S one-
Geisse ’s Q2) o he model and i s global
Goodness-o -Fi (GoF) in o de o assess he
o e all quali y o he model. Based on he
esul s, ou esea ch model is able o cap u e
abou 26% o he a iance in (e)WoM e ec and
21.2% o he a iance in epu chase in en ion
is cap u ed in he esea ch model. Please e e
o Tab. 4 o mo e ela ed in o ma ion. In e ms
o he p edic i e ele ance o he exogenous
cons uc s, we used he blind olding p ocedu e
(wi h omission dis ance, d = 7). Mo e speci i cally,
we used he c oss- alida ed edundancy
measu e (S one-Geisse Q2 es ) and all Q2
alues o each o he endogenous cons uc s
a e posi i e alues. The e o e, we can conclude
ha he p edic i e abili y o he exogenous
cons uc s is ela i ely s able and high. To
assess he e ec size o he o e all global
model, we used he GoF index (Tenenhaus e
al., 2005). A GoF index alue o 0.1 s ands o
a small e ec size, 0.25 (GoFmedium) and 0.36
(GoFla ge) (c . Osakwe & Cho anco á, 2015).
We ob ained a GoF index alue o abou 0.3,
which is conside ed as a sligh ly la ge e ec
size o R2, hus, indica ing an adequa e global
alida ion o he o e all PLS model (We zels e
al., 2009).
Al hough he hypo heses we e s a ed in
a one-di ec ional o ma , ne e heless, we used
a wo- ailed es (wi h a cu -o p obabili y alue,
La en Va iables Mean SD 12345
1. SME 2.949 0.818 0.769
2. BO 3.298 0.668 0.263 0.744
3. RPI 3.483 0.726 0.167 0.106 0.831
4. VRP 3.933 0.661 0.093 0.256 0.266 0.740
5. (e)WoM 3.830 0.348 0.192 0.161 0.457 0.484 0.746
Sou ce: own
Tab. 3: Disc iminan alidi y o measu emen model
EM_4_2016.indd 155EM_4_2016.indd 155 30.11.2016 16:31:5330.11.2016 16:31:53
156 2016, XIX, 4
Ma ke ing a obchod
p < 0.05) o de e mine he s a is ical signi i cance
o ou hypo heses. Mo e impo an ly, he esul s
(see Fig. 2 and Tab. 4) show ha ou ou o
he se en s a ed hypo heses we e suppo ed.
B and o ien a ion is posi i ely associa ed wi h
a endency o use social media si es o engage
wi h endo s’ b ands (β = 0.286, p < 0.01).
Also, consume s ha a e mo e b and o ien ed
a e mo e likely o ha e a highe pe cep ion o
a endo ’s epu a ion (β = 0.283, p < 0.01).
Mo eo e , a endo ’s epu a ion was ound
o posi i ely and o signi i can ly con ibu e o
IV -> DV Pa h Coe i cien T-S a is ics
(Boo s apped)
BO -> SME 0.286 2.498
BO -> VRP 0.283 2.979
VRP -> (e)WoM 0.471 5.518
(e)WoM -> RPI 0.429 5.088
SME-> (e)WoM 0.148 1.623
VRP -> RPI 0.059 0.857
VRP ->SME 0.028 0.291
Coe i cien o De e mina ion C oss- alida ed edundancy (Q2) Model’s GoF
R2 o (e)WoM 0.256 Q2 o (e)WoM 0.13
0.298
R2 o RPI 0.212 Q2 o RPI 0.127
R2 o SME 0.070 Q2 o SME 0.040
R2 o VRP 0.065 Q2 o VRP 0.037
Sou ce: own
Tab. 4: Resul s o he Pa h Modelling
Fig. 2: Empi ical esul s
Sou ce: own
EM_4_2016.indd 156EM_4_2016.indd 156 30.11.2016 16:31:5330.11.2016 16:31:53
157
4, XIX, 2016
Ma ke ing and T ade
(e)WoM e ec (β = 0.471, p < 0.01). Simila ly,
(e)WoM e ec is posi i ely associa ed wi h
online cus ome s epu chase in en ions
(β = 0.429, p < 0.01). On he o he hand, social
media engagemen wi h a endo ’s b and is
posi i ely ela ed o (e)WoM e ec , bu i is no
s a is ically signi i can (β = 0.148, p > 0.05).
Simila ly, we ound ha VRP does no ha e
any signi i can in l uence on SME (β = 0.028,
p > 0.05). This was equally he case wi h endo
epu a ion and epu chase in en ion (β = 0.059,
p > 0.05). Jus o b ie l y men ion ha a pos -hoc
analysis, hough esul s no epo ed he e due
o space limi a ions, is indica i e ha (e)WoM is
a media o be ween VRP and RPI.
4. Discussion o Findings
The esea ch o e s empi ical on he s ong
ela ionship be ween b and o ien a ion and
consume s engagemen wi h endo s’ social
media si e. This s udy con ibu es o he b and
o ien a ion li e a u e while i also b oadens
ou unde s anding o he b oade subjec a ea
o ela ionship ma ke ing a he same ime. I
i nds ha b and o ien a ion leads o a highe
pe cep ion o endo s’ epu a ion online.
Addi ionally, i i nds ha endo s’ epu a ion has
a signi i can e ec on (e)WoM. Fu he mo e,
he s udy con i ms he signi i can e ec o
(e)WoM on epu chase in en ions. Al hough he
s udy ound a ela ionship be ween consume s
engagemen wi h endo s’ social media si e,
his was no subs an ial. Simila ly, he e was
a ela ionship be ween endo s’ epu a ion and
consume s’ engagemen wi h endo s’ social
media si e bu i was no signi i can . Al hough,
endo s’ epu a ion was ound o posi i ely
in l uence online shoppe s’ epu chase
in en ions, we could no es ablish any s a is ical
signi i cance o he esul .
As mos shoppe s (especially, b and
o ien ed consume s) p e e o buy amous
b ands online (Hu e e al., 2013), being
b and o ien ed is one way o d i ing a i c and
in l uencing consume s o engage wi h endo s’
social media si es. As no ed by Algesheime
e al. (2005) mos consume s ( e y likely o
be b and-o ien ed shoppe s) p e e isi ing
and shopping on websi es o well-known
b ands. Since mos consume s a e isk a e ,
especially when online, hey will usually wan
o choose popula b ands and he e o e will
shop mos o he ime on websi es ha a e
less isky. Consequen ly, once consume s i nd
hei p e e ed b and on a pa icula si e, hey
s ick o i (Hu e e al., 2013). In line wi h some
exis ing s udies (e.g. Sahin e al., 2011), his
s udy ound ha b and o ien a ion (on he pa
o online shoppe s) c ea es a highe pe cep ion
o endo s’ epu a ion. In o he wo ds, b and
o ien a ion c ea es an imp ession in consume s’
minds ha a endo is eliable and is us wo hy.
The esul sugges s ha his g oup o shoppe s
is mo e pa icula abou he epu a ion o he
endo hey buy om online because some
online endo s do no keep o hei p omise
and sell in e io b ands. Consequen ly, some
endo s spend mo e esou ces on building
s ong b ands o dis inguish hemsel es om
he compe i ion (U de, 1999). The esul s in
se e al ways align wi h he adi ional hinking
abou he ole ha a b and plays in consume
decision-making, as a shoppe ’s b and
o ien a ion has been es ablished in he s udy o
c i ically in l uence he shoppe ’s in e ac ion wi h
he endo ’s social pla o ms and his is also
likely o make he consume o m an imp ession
o he epu a ion o an online e ail endo .
Vendo s’ epu a ion as no ed by Yoon e
al. (2008) in l uences consume beha iou . I
is he e o e no su p ising ha his s udy ound
a signi i can e ec o endo s’ epu a ion on
(e)WoM. Consume s who buy om epu able
endo s no mally engage in p oduc e iew
online and suppo endo s by ecommending
hem o o he consume s online. Consume s
who encoun e ed and enjoy he se ices o
epu able endo s alk o hei iends and
amily abou hem. I is impo an o no e ha
consume s do no only pass on good commen s
abou endo s. Dissa is i ed online shoppe s a e
mo e likely o sp ead nega i e commen s abou
endo s. As a esul , ha ing a good epu a ion
online is highly essen ial o posi i e (e)WoM.
In a simila ein, since ‘mos ’ consume s migh
no us some o he online endo s; hey a e
highly likely o ely on o he consume s when
making pu chasing decisions. Consis en wi h
Lin e al. (2012), his s udy ound ha (e)WoM
a ec s epu chase in en ions. This is p obably
because hi d pa y in o ma ion sou ce is seen
o be mo e c edible han in o ma ion coming
om an online e ail endo .
On he o he hand, al hough his s udy
ound a posi i e ela ionship be ween cus ome s
engagemen wi h endo s’ social media si e
and (e)WoM, his ela ionship was s a is ically
insigni i can . This means ha cus ome s
EM_4_2016.indd 157EM_4_2016.indd 157 30.11.2016 16:31:5330.11.2016 16:31:53
164 2016, XIX, 4
Ma ke ing a obchod
In e na ional Jou nal o Resea ch in Ma ke ing,
27(3), 248-260.
Zei haml, V. A., Be y, L. L., Pa asu aman,
A. (1996). The Beha io al Consequences o
Se ice Quali y. Jou nal o Ma ke ing, 60(2),
31-46. doi:10.2307/1251929.
Zheng, X., Cheung, C. M. K., Lee, M.
K. O., & Liang, L. (2015). Building b and
loyal y h ough use engagemen in online
b and communi ies in social ne wo king si es.
In o ma ion Technology and People, 28(1),
90-106. doi:10.1108/ITP-08-2013-0144.
Ing. Ch is ian Nedu Osakwe
Tomas Ba a Uni e si y in Zlin
Facul y o Managemen and Economics
Depa men o Managemen and Ma ke ing
[email p o ec ed]
Hen y Boa eng, Ph.D. Candida e
Uni e si y o Technology, Sydney
School o Communica ion
Hen y[email p o ec ed]
Simona Popa, Ph.D.
Uni e si y o Mu cia
Depa men o Managemen & Finance
[email p o ec ed]
doc. Ing. Milosla a Cho anco á, CSc.
Tomas Ba a Uni e si y in Zlin
Facul y o Managemen and Economics
Depa men o Managemen and Ma ke ing
[email p o ec ed]
P o . Ped o So o-Acos a, Ph.D.
Uni e si y o Mu cia
Depa men o Managemen & Finance
[email p o ec ed]
EM_4_2016.indd 164EM_4_2016.indd 164 30.11.2016 16:31:5430.11.2016 16:31:54

165
4, XIX, 2016
Ma ke ing and T ade
Cons uc s & indica o s Li e a u e suppo
B and o ien a ion
BO2 – Mos imes, I p e e o buy a well-known b and om a web e aile
BO2 – The mos popula b ands a e usually my i s pick/choice
BO3 – I am pa icula abou a good b and name while shopping online
BO4 – Fo me, buying a popula b and online is less isky
Ling e al. (2010); Seock
(2003)
Social media si e engagemen
SME1 – I would p e e o buy om a websi e ha I am connec ed o on
a social ne wo k si e- acebook, wi e , ins ag am, l icke , e c…
SME2 – I hink i is good o an online selle o ha e a Facebook / Twi e /
Linkeldn / Google+ Page
SME3 – I would like my online shopping websi es o keep me upda ed wi h
la es sales/p oduc s on my social ne wo k si es
SME4 – To me, i ’s impo an ha companies engage wi h hei cus ome s
h ough social ne wo k si es
SME5 – On he whole, I ollow he ac i i ies o he companies I like on
social ne wo k si es
Ka akaya and Ba nes
(2010);
La oche e al. (2012);
Ramna ain and
Go ende (2013)
Vendo Repu a ion
VRP1 – I can only buy om an online endo ha is eliable
VRP2 – I p e e o deal wi h a us wo hy web me chan
VRP3 – I am mo e pa icula abou he epu a ion o any In e ne selle
VRP4 – I like a websi e which is u h ul abou i s o e s
Doney and Cannon
(1997); Kim e al. (2013);
Ja enpaa e al. (2000)
(e)WoM
WoM1 – Fo me, online p oduc e iews will in l uence my pu chasing
decision
WoM2 – My iends and I some imes alk abou ou shopping expe iences
WoM3 – I I like/dislike a p oduc , I ell my iends and amily membe s
WoM4 – I will easily ecommend a good online shopping websi e o o he s
Awad and Ragowsky
(2008); Mikale e al.
(2013); Zei haml e al.
(1996)
Repu chase in en ion
RPI1 – Gi en he con enience o Shopping on he Web, I will always use i
RPI2 – I will con inually use he Web o my shopping needs
RPI3 – I ha e made some ecen pu chases online and would mos likely
buy mo e i ems online
RPI4 – I will con inually use he In e ne o my shopping
RPI5 – I see my use o In e ne shopping inc easing in he nea by u u e
Bha ache jee (2001);
Ma hieson (1991); Thong
e al. (2006)
Sou ce: own
No e: i e-poin Like - ype scales
Appendix 1: Measu es
EM_4_2016.indd 165EM_4_2016.indd 165 30.11.2016 16:31:5530.11.2016 16:31:55
166 2016, XIX, 4
Ma ke ing a obchod
Abs ac
UNDERSTANDING COSMOPOLITAN CONSUMERS’ REPEAT PURCHASING
IN THE eMARKETPLACE: CONTRIBUTION FROM A BRAND ORIENTATION
THEORETICAL PERSPECTIVE
Ch is ian Nedu Osakwe, Hen y Boa eng, Simona Popa, Milosla a
Cho anco á, Ped o So o-Acos a
As his scien i i c pape is posi ioned unde he ela i ely big umb ella o ela ionship ma ke ing;
i hus makes a ui ul a emp o b idge he gap be ween schola ship and p ac ice. Ou o e iding
objec i e o his s udy was o explo e c i ically he con ibu ion o cus ome s’ b and o ien a ion as
well as o he i al cons uc s such as social media engagemen , (e) endo epu a ion and (e)WoM
on epea pu chasing in en ion amongs cosmopoli an consume s in eMa ke place con ex . Da a
we e collec ed h ough a non-p obabilis ic sampling echnique om cosmopoli an consume s in
one o he EU-27 coun ies, Slo akia. Da a was analysed using he Pa ial Leas Squa es s uc u al
equa ion modelling. This s udy modelled online consume s’ epea pu chasing decision using
cons uc s such as b and o ien a ion, endo epu a ion, endo s’ social media si e engagemen .
The s udy ound ha b and o ien a ion leads o a highe pe cep ion o endo s’ epu a ion online.
The i ndings showed ha a posi i e and signi i can ela ionship exis s be ween b and o ien a ion
and consume s engagemen wi h endo s’ social media si e. Fu he mo e, his s udy ound ha
endo s’ epu a ion has a signi i can e ec on (e)WoM. Impo an ly, his s udy con i med he
subs an ial e ec o (e)WoM on epu chase in en ions. These i ndings imply ha online e ail b ands
should ‘push’ hei p oduc o e ings ia some o he popula social ne wo king si es (SNSs) and,
impo an ly, endea ou o keep ack o hose online use s in he social media communi y who mos
likely ha e a i ni y owa ds hei b ands and/o o he simila b ands. By and la ge, he pape has
demons a ed ha he s udied cons uc s a e key in consume s’ decision making online. Hope ully,
he i ndings o he esea ch will assis he online e ail endo in i s execu ion o ( obus ) cus ome
iendly policies.
Key Wo ds: Social media engagemen , (e) endo epu a ion, b and o ien a ion, (e)WoM,
epu chase in en ion.
JEL Classi i ca ion: L81, M30, M31, O33.
DOI: 10.15240/ ul/001/2016-4-011
EM_4_2016.indd 166EM_4_2016.indd 166 30.11.2016 16:31:5530.11.2016 16:31:55