B odschelm, Flo ian; Ve e , Sebas ian; Hü l‐Maack, Ve ena; Laza o ici, Isabel‐
Sophie; Schumann, Jan Hend ik
A icle — Published Ve sion
The In e play o In la ed Expec a ions, (Dis)Con i ma ion,
and Emo ional Spillo e : Implica ions o Pos ‐Pu chase
Loyal y in Mys e y Deals
Psychology & Ma ke ing
P o ided in Coope a ion wi h:
John Wiley & Sons
Sugges ed Ci a ion: B odschelm, Flo ian; Ve e , Sebas ian; Hü l‐Maack, Ve ena; Laza o ici, Isabel‐
Sophie; Schumann, Jan Hend ik (2025) : The In e play o In la ed Expec a ions, (Dis)Con i ma ion,
and Emo ional Spillo e : Implica ions o Pos ‐Pu chase Loyal y in Mys e y Deals, Psychology &
Ma ke ing, ISSN 1520-6793, Wiley, Hoboken, NJ, Vol. 42, Iss. 11, pp. 2883-2901,
h ps://doi.o g/10.1002/ma .70017
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/330169
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h ps://c ea i ecommons.o g/licenses/by/4.0/
Psychology & Ma ke ing
RESEARCH ARTICLE
The In e play o In la ed Expec a ions,
(Dis)Con i ma ion, and Emo ional Spillo e : Implica ions
o Pos ‐Pu chase Loyal y in Mys e y Deals
Flo ian B odschelm
1
| Sebas ian Ve e
1
| Ve ena Hü l‐Maack
2
| Isabel‐Sophie Laza o ici
1
| Jan Hend ik Schumann
1
1
Chai o Ma ke ing and Inno a ion, Uni e si y o Passau, Passau, Ba a ia, Ge many |
2
Chai o Ma ke ing and Consume Beha io , Uni e si y o
Hohenheim, Hohenheim, Baden‐Wue embe g, Ge many
Co espondence: Isabel‐Sophie Laza o ici ([email p o ec ed])
Recei ed: 20 Decembe 2024 | Re ised: 12 July 2025 | Accep ed: 14 July 2025
Keywo ds: a ec i e eac ion | cus ome loyal y | discon i ma ion e ec | mys e y deals | posi i e a ec i e spillo e
ABSTRACT
Mys e y deals, a ma ke ing ac ic used o boos sales, in ol e wi hholding he exac e sion o he p oduc a cus ome will
ecei e un il a e pu chase. While p io esea ch has examined consume s' p e‐pu chase eac ions o mys e y deals, hei pos ‐
pu chase esponses emain unde explo ed. Add essing his gap, we con ibu e o he mys e y ma ke ing li e a u e by in es-
iga ing how consume s espond a e he pu chase when all de ails o he deal a e e ealed. Ac oss six empi ical s udies (1
ield, 3 online, 1 lab ha uses IA‐based acial exp ession ecogni ion so wa e, 1 supplemen a y), we disen angle he complex
in e play o cu iosi y‐induced in la ed expec a ions, (dis)con i ma ion, and posi i e a ec i e spillo e e ec s gene a ed du ing
he p e‐pu chase phase. The esul s highligh he c ucial ole o posi i e a ec in subs an ially enhancing cus ome loyal y
du ing he pos ‐pu chase phase—compa ed o adi ional deals whe e all de ails a e known in ad ance. These a ec i e
p ocesses e en coun e balance he nega i e e ec s on loyal y when consume s ecei e hei non‐p e e ed ou come. Fo
manage s, ou indings p o ide insigh s in o how mys e y deals impac cus ome e en ion, expanding beyond p io esea ch
ha has p ima ily ocused on cus ome acquisi ion. Thus, mys e y deals can se e as a iable s a egy o sus ain cus ome
loyal y, e en when indi idual p e e ences canno be ully me .
1 | In oduc ion
Many e aile s employ mys e y deals (MDs) as a ma ke ing
ac ic, whe e he key p oduc o se ice emains unknown
be o e pu chase (Guo e al. 2024; Hill e al. 2016; Ko ache a
e al. 2022; Ko ache a and Nikolo a 2023). MDs aim o boos
demand, op imize capaci y, o clea in en o y (e.g., Fay and
Xie 2008). Fo ins ance, Bu ge King o e ed a “mys e y box”
wi h one o eigh possible bu ge s a a discoun , wi hou
e ealing which one cus ome s will ecei e in ad ance (mein‐
deal.com 2019). Simila ly, Pu a Vida B acele s, an online
e aile , p omo es su p ise b acele s and ings online (Pu a
Vida 2024). Mo e examples appea in Appendix A, Table A.1.
Unlike o he mys e y ma ke ing ac ics—such as mys e y
discoun s, mys e y gi s, o mys e y ease ad e isemen s—
mys e y deals (MDs) inco po a e unce ain y and su p ise
di ec ly in o he co e o e ing i sel . In hese cases, he main
componen o he pu chase (i.e., he p oduc o se ice he
consume is ac ually buying) emains unknown a he poin
o pu chase, a he han being a supplemen a y o p omo-
ional elemen . This dis inguishes MDs om o he ac ics
whe e he mys e y is pe iphe al o addi i e, such as
ecei ing a su p ise gi alongside a known p oduc (Guo
e al. 2024). A clea classi ica ion and di e en ia ion o
mys e y deals om o he mys e y ma ke ing ac ics can be
seen in Figu e 1.
This is an open access a icle unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s use, dis ibu ion and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly
ci ed.
© 2025 The Au ho (s). Psychology & Ma ke ing published by Wiley Pe iodicals LLC.
2883Psychology & Ma ke ing, 2025; 42:2883–2901
h ps://doi.o g/10.1002/ma .70017
A de ining cha ac e is ic o MDs is he in en ional c ea ion o
an in o ma ion gap, which consume s a e mo i a ed o b idge
by comple ing he pu chase. Unlike adi ional deals (TDs),
whe e all de ails a e known up on , MDs in ol e unce ain y
and ele a e expec a ions while in oducing isk, as buye s
canno be ce ain hey will ecei e hei p e e ed op ion. In
Bu ge King's deal, o example, cus ome s had a one‐in‐eigh
p obabili y o ob aining hei p e e ed bu ge . Despi e hei
g owing p e alence, li le is known abou how consume s
e alua e MDs a e pu chase and wha psychological and
beha io al mechanisms shape hei inal judgmen .
Su p isingly, mos esea ch on MDs ocuses on p e‐pu chase
beha io , lea ing pos ‐pu chase eac ions la gely unexplo ed.
S udies sugges ha mys e y pu chases a e especially appealing
when po en ial ou comes a y in a ibu es such as as e o
s yle, a he han objec i e alue (Buechel and Li 2023), and in
expe ien ial con ex s (U umu a Hewage and He 2022).
Demog aphic ac o s, such as gende , also in luence consume s'
openness o mys e y o ma s (Ko ache a e al. 2022).
S udies ha e also iden i ied se e al mechanisms by which MDs
enhance pu chase mo i a ion, including s imula ing cu iosi y
(Hill e al. 2016), os e ing inna e op imism (Goldsmi h and
Ami 2010), igge ing a desi e o su p ise, and elici ing posi-
i e a ec (Buechel and Li 2023; La an and Tsi os 2013).
Resea ch in ope a ional and s a egic managemen u he
highligh s how e aile s bene i om MDs in he sho e m,
pa icula ly h ough p ice disc imina ion and in en o y man-
agemen (e.g., Fay and Xie 2008). Howe e , li le is known
abou how consume s eac in he pos ‐pu chase phase, pa -
icula ly when hey ecei e an ou come ha does no align wi h
hei p e e ences ( o a summa y o exis ing esea ch, see
Table A.2 in Appendix A).
This s udy seeks o add ess his gap by examining he impac o
MDs on cus ome loyal y, de ined as a consume 's a i ude and
beha io al in en ion o main ain a ela ionship wi h a i m by
making epea pu chases, ei he o he same o di e en p od-
uc s (Si deshmukh e al. 2002). We p opose wo key e ec s ha
in luence cus ome loyal y in he pos ‐pu chase phase. Fi s , a
(dis)con i ma ion e ec ha a ises as consume s compa e he
ac ual ou come o he MD o hei in la ed expec a ions and
p e e ed ou come. Second, we in oduce he concep o posi i e
a ec i e spillo e , whe ein he posi i e emo ions elici ed du ing
he p e‐pu chase phase ex end in o he pos ‐pu chase phase,
in luencing consume s' loyal y (B own and Ki mani 1999;
Zillmann 1971). These wo e ec s likely in e ac , and he esul
o hei combina ion has no been s udied, aising he ques ion
o hei o e all impac on cus ome loyal y.
Ou esea ch pu sues h ee p ima y objec i es: (1) o examine
consume eac ions in he pos ‐pu chase phase when he mys-
e y is e ealed; (2) o es he in e play o he (dis)con i ma ion
e ec and posi i e a ec i e spillo e e ec ; and (3) o assess he
o e all impac o MDs on cus ome loyal y, especially when
consume s do no ecei e hei p e e ed ou come. To achie e
hese goals, we p esen six s udies: h ee online expe imen s,
one labo a o y expe imen , one ield s udy, and one supple-
men a y s udy.
To ou knowledge, his is he i s s udy o sys ema ically
in es iga e how mys e y deals—whe e he unce ain y pe ains
o he co e componen o he pu chase a he han an ancilla y
bene i —a ec cus ome loyal y in he pos ‐pu chase phase. By
mo ing beyond he commonly examined domain o p e‐
pu chase exci emen and cu iosi y, ou esea ch shi s he
analy ical ocus o he long‐ e m consequences o mys e y deals,
speci ically hei impac on sus ained cus ome engagemen and
b and loyal y. We es ablish ha pos ‐pu chase eac ions a e
shaped by (1) he (dis)con i ma ion e ec , whe e loyal y
s eng hens when expec a ions a e me and emains s able e en
when hey a e no (and su p isingly, discon i ma ion e ec s
ha e no been s udied in he con ex o MDs), (2) posi i e
FIGURE 1 | Classi ica ion o and di e en ia ion om a ying mys e y ma ke ing ac ics. The aim o his igu e is no o p o ide a comp ehensi e
e iew o he li e a u e, bu a he o o e an o e iew o di e en b anches and ypes o mys e y ma ke ing ac ics and o posi ion he p esen
esea ch wi hin his con ex . (Ala i e al. 2015; Dha e al. 1999; Elsen e al. 2016; Gup a e al. 2020; Maza e al. 2017; Sch oll and G ohs 2019; Se illa
and Meye 2020).
2884 Psychology & Ma ke ing, 2025
a ec i e spillo e , which enhances emo ional esponses and
mi iga es nega i e e ec s o discon i ma ion, and (3) in la ed
expec a ions, which make discon i ma ion likely bu do no
necessa ily ha m loyal y. This ad ances MD esea ch by dem-
ons a ing how esol ing unce ain y impac s e en ion—an
aspec o e looked in p io esea ch ocused on p e‐pu chase
esponses (Buechel and Li 2023; Goldsmi h and Ami 2010; Hill
e al. 2016). Fu he mo e, ou indings also con ibu e o
esea ch on a ec by showing ha p e‐pu chase emo ions
pe sis in o he pos ‐pu chase phase, shaping sa is ac ion and
loyal y (e.g., Lee and Qiu 2009). Beyond heo e ical con ibu-
ions, ou indings p o ide ac ionable manage ial insigh s: MDs
no only d i e a en ion and sales bu also os e loyal y—
especially when designed o maximize posi i e p e‐pu chase
a ec . Re aile s can boos engagemen h ough isually
appealing deals, cu iosi y‐d i en messaging, and exclusi e
mys e y incen i es, making MDs a s a egic al e na i e o a-
di ional o e s ac oss a ious indus ies and e ail con ex s.
2 | Theo e ical Backg ound
2.1 | How Mys e y Deals Wo k—C ea ing and
Resol ing In o ma ion Gaps
In gene al, MDs consis o a c ea ion phase be o e pu chase and a
esolu ion phase a e wa ds. In he mys e y c ea ion phase, he
consume is con on ed wi h a deal ha in ol es delibe a ely c e-
a ed unce ain y o an in o ma ion gap abou he inal ou come. In
he case o he MD ype we in es iga e, he exac p oduc a ian
emains unknown a he ime o pu chase. Such in o ma ion gaps,
de ined as he di e ence be ween wha one knows and wha one
wan s o know, induce he pe cep ion o in o ma ion dep i a ion
and cu iosi y (Loewens ein 1994). Consequen ly, his mo i a es
consume s o close he in o ma ion gap and pu sue explo a o y
beha io o acqui e he missing in o ma ion (Hsee and Ruan 2016;
Loewens ein 1994).The eisconsensusin heli e a u e ha such
o ms o posi i e unce ain y (i.e., which in ol es unce ain y abou
non‐nega i e ou comes only; Ruan e al. 2018)cans imula epu -
chases and imp o e consume s' a i udes and a ec i e esponses
(Lee and Qiu 2009;Ruane al.2018).
In he pos ‐pu chase phase, he in o ma ion gap induced by he MD
is esol ed by e ealing he speci ic ou come. Unlike mys e y ad-
e isemen s, MDs equi e consume s o make a pu chase—and
in es money— o close he in o ma ion gap. A e he p oduc is
e ealed, consume s will e alua e hei indi idual u ili y o he
ou come and will de elop some le el o sa is ac ion wi h he pu -
chase and u u e in en ions. In addi ion, he esolu ion o cu iosi y
and unce ain y elici s a ec i e esponses ha will also in luence
he ou come e alua ions. In sum, we a gue ha c ea ing and
esol ing a mys e y o he consume , h ough he MDs as de ined
ea lie , leads o a complex in e play o e ec s ha de e mine he
ou come e alua ion. The indi idual componen s o his o e all
e ec a e discussed in he ollowing pa ag aphs.
2.2 | The Role o In la ed Expec a ions
In he p e‐pu chase phase, consume s will de elop expec a ions
abou he ou come. Howe e , hese expec a ions will be in la ed in
heway ha heya emo eposi i e hani nomys e ywouldha e
been in ol ed, which can be a ibu ed o wo aspec s. Fi s , cu i-
osi y enhances expec a ions. I induces he an icipa ion o ewa ds
in indi iduals, which leads hem o expec o ecei e some hing
aluable o hei e o s unde aken o esol e i (e.g., Wiggin
e al. 2019). In line wi h his, i has been epo ed ha when con-
sume s a e in a s a e o cu iosi y, hey end o de elop o e ly pos-
i i e expec a ions abou he ou come (e.g., Daume and Hü l‐
Maack 2020). The second aspec is known unde he e m inna e
op imism (Ba ‐hillel and Budescu 1995;Goldsmi handAmi 2010).
When consume s a e con on ed wi h unce ain al e na i es, hey
o e es ima e he p obabili y o posi i e as opposed o nega i e
ou comes. In he con ex o mys e y gi s, Goldsmi h and Ami
(2010) ha e e en shown ha consume s e alua e an unce ain
ou come as aluable as he bes possible ou come unde ce ain y.
In he same ein, we expec consume s o be o e ly op imis ic when
con on ed wi h a mys e y deal p o iding a se o equally likely
(unce ain) p oduc a ian s. Because o he unce ain ou come o
he deal, consume s will o m high expec a ions and expec hei
mos p e e ed op ion. They will expec his bes possible ou come
e eni heac ualchanceo ob aining his ou come is as high as he
chance o ge one o he o he op ions. We he e o e assume:
H1. Mys e y deals will gene a e in la ed expec a ions in
consume s, such ha he ac ual sha e o consume s expec ing
o ecei e he bes possible ou come (i.e., hei p e e ed op ion)
will be highe han he p edic ed sha e o consume s ha
should ecei e i .
2.3 | Expec a ion‐(Dis)Con i ma ion E ec s
In he con ex o MDs, he expec a ion‐discon i ma ion pa adigm
plays a c ucial ole due o he unce ain ou come. Acco ding o his
pa adigm, consume s e alua e he ou come o a pu chase based on
a compa ison o hei p e‐pu chase expec a ions wi h he pu chase
ou come (Oli e 1980). The mo e a p oduc mee s consume s' ini ial
expec a ions, he highe he cus ome sa is ac ion and loyal y
(con i ma ion e ec ); he less i mee s hese expec a ions, he lowe
he cus ome sa is ac ion and loyal y (discon i ma ion e ec ; e.g.,
Habel e al. 2016;Oli e 1980). Gi en ha in he case o MDs,
expec a ions a e in la ed, as explained be o e, he isk o c ea ing
nega i e discon i ma ion is ex ao dina ily high. Fu he mo e, he
a ec i e eac ionscausedby hedeg eeo discon i ma ionplaya
c ucial ole; i dec eases he pleasu e o a consump ion si ua ion,
whe eas con i ma ion inc eases i (e.g., Oli e 1980), which u he
in luences consume s' sa is ac ion and loyal y in en ions (e.g.,
Wes b ook 1987).
Immedia ely a e he pu chase, cus ome s compa e he ou -
come o he MD wi h hei p e e ed al e na i e (Shen
e al. 2019). We expec ha consume s who ecei e a non‐
p e e ed ou come will expe ience a discon i ma ion such ha
hei posi i e a ec i e eac ions and loyal y in en ions a e lowe
han hose o consume s ge ing a TD, in which hey know om
he beginning wha hey will ecei e. Fu he mo e, consume s
who ecei e hei p e e ed op ion when he MD is e ealed will
expe ience a con i ma ion ha should be e en highe han o
consume s pu chasing a TD. Recei ing he desi ed, ye
unce ain ou come in an MD will lead o s onge posi i e
a ec i e esponses, such as posi i e su p ise (Oli e e al. 1997)
2885
and elie (Ruan e al. 2018; Yang e al. 2016), compa ed o a
TD. Based on hese conside a ions, we hypo hesize:
H2. The e will be a (dis)con i ma ion e ec in he pos ‐
pu chase phase such ha MDs (dis)con i ming consume s'
p e e ences will dec ease ( s. inc ease) loyal y h ough lowe
( s. highe ) le els o posi i e pos ‐pu chase a ec i e eac ions
compa ed o TDs.
2.4 | P e‐Pu chase Posi i e A ec and Posi i e
A ec i e Spillo e E ec s
An addi ional aspec , unique o MDs, ha in luences ou come
e alua ion is he posi i e a ec i e esponse gene a ed du ing he
p e‐pu chase phase, which can pe sis and con inue o in luence
heconsume expe iencewellin o hepos ‐pu chase phase.
S udies on posi i e unce ain y and cu iosi y in con ex s such as
easing and ad e ising ha e shown ha s ong posi i e eelings,
c ea ed in he p e‐pu chase phase, ul ima ely lead o posi i e
consume esponses (e.g., Ba ‐Anan e al. 2009;Hü l‐Maack
e al. 2024;LeeandQiu2009). Al hough nega i e a ec i e
esponses o cu iosi y a e also possible, such as eelings o us-
a ion and discom o associa ed wi h he expe ience o no
knowing (Be lyne 1954;Li manandJime son2004), esea ch has
demons a ed ha using mys e y deals in ma ke ing o e all elici s
mo e posi i e han nega i e a ec (Hill e al. 2016). The esul is a
ne posi i e hedonic expe ience (Ruan e al. 2018).
Based on p e ious e idence in o he domains (e.g., B own and
Ki mani 1999; Gee s and Lassi e 2002;Zillmann1971), we a gue
ha hese posi i e p e‐pu chase a ec i e eac ions will spill o e
o he pos ‐pu chase phase and posi i ely in luence he con-
sume 's inal esponse owa ds he deal. Fo ins ance, B own and
Ki mani (1999) and Zillmann (1971) p opose ha a ec i e
eac ions owa ds a s imulus in luence (1) a ec i e eac ions o a
subsequen s imulus and (2) consume s' cogni i e judgmen s o
his second s imulus. Simila ly, he a ec i e p iming pa adigm
sugges s ha he alence o an a ec i e s a e in luences he
alence expe ienced by a subsequen s imulus in a consis en
manne (Goldbe g and Go n 1987). In addi ion, he a ec i e
expec a ion model (e.g., Gee s and Lassi e 2002) indica es ha
people wi h posi i e a ec i e expec a ions o a si ua ion a e likely
o ac ually expe ience ha la e si ua ion mo e posi i ely. Based
on his e idence, we assume ha MDs ( s. TDs) will c ea e
highe posi i e p e‐pu chase a ec , which spills o e o he pos ‐
pu chase phase and posi i ely in luences consume s' posi i e
pos ‐pu chase a ec i e eac ions and loyal y by bu e ing any
po en ial nega i e e ec s. In ac , we posi ha posi i e a ec i e
spillo e e ec s will ou weigh he nega i e e ec s o dis-
con i ma ion and ein o ce he posi i e e ec s in case o a con-
i ma ion. We he e o e p opose:
H3. The e will be posi i e a ec i e spillo e e ec s such ha
MDs ( s. TDs) inc ease posi i e p e‐pu chase a ec ha
(a) inc eases cus ome loyal y di ec ly and (b) indi ec ly
h ough an inc ease in posi i e pos ‐pu chase a ec .
The in e play o he hypo hesized e ec s a e summa ized in
Figu e 2. I illus a es how he wo dis inc sou ces o a ec , he
p e‐pu chase s a e in ol ing cu iosi y and an icipa ion, and he
pos ‐pu chase si ua ion in which unce ain y is esol ed,
in e ac o shape he o e all consume esponse.
In he ollowing, we p esen a o al o six s udies, including
h ee online scena io‐based expe imen s, one labo a o y ex-
pe imen , one ield s udy, and one supplemen a y s udy. Fo an
o e iew o all s udies, see Table B.1 in Appendix B.
3 | S udies 1a and 1b: Examining he P oposed
E ec s
In he i s wo s udies, we in es iga e ou p oposed e ec s in a
p oduc ‐ ela ed (S udy 1a) and se ice‐ ela ed con ex (S udy 1b).
3.1 | S udy 1a: P oduc ‐Rela ed Con ex
(Chocola e)
3.1.1 | Design, Pa icipan s, and P ocedu e
We conduc ed a single‐ ac o (MD s. TD) be ween‐subjec s
online expe imen wi h p e‐and pos ‐pu chase phases. A o al
FIGURE 2 | Concep ual Model.
2886 Psychology & Ma ke ing, 2025
o 148 pa icipan s (50% women; M
age
= 46.32 yea s,
SD
age
= 14.63 yea s) om a Ge man p o essional panel, all in-
e es ed in chocola e, ook pa .
3.1.1.1 | P e‐Pu chase Phase. Pa icipan s iewed a Ge -
man ic i ious chocola e ad e isemen ea u ing ei he a TD o an
MD and we e andomly assigned o one condi ion. The TD o e ed
a10‐pack o one speci ic la o o €9(discoun ed om€10), while
he MD o e ed a 10‐pack o one o ou possible la o s o €7,
wi hou e ealing which. Fla o s—milk chocola e, hazelnu ,
s awbe y, and ca amel—we e chosen o hei popula i y
(Nie 2019; Splendid Resea ch 2017;seeAppendixC). Re lec ing
manage ial p ac ice and p io s udies (e.g., Fay and Xie 2008;
G anados e al. 2008), he MD was u he discoun ed (€7 s.€10).
Compa ed o eal‐wo ld cases (Table A.1 Appendix A), he MD
discoun was mode a e o conse a i e es ing.
Pa icipan s iewed he ad e isemen o a leas 15 s, a ed
hei posi i e a ec i e esponse using a 5‐poin smiley scale by
Shampanie e al. (2007), and indica ed hei expec ed la o .
They also a ed hei pu chase in en ion on a 7‐poin single‐
i em scale (e.g., La an and Tsi os 2013; see Appendix D, o
cons uc s and i ems). C onbach's alpha alues o mul i‐i em
cons uc s a e also epo ed in Appendix D. O he cons uc s,
mos ly single‐i em, we e adap ed om es ablished li e a u e o
ensu e eliabili y and heo e ical i .
3.1.1.2 | Pos ‐Pu chase PhasePa icipan s imagined ecei -
ing a 10‐pack o chocola e a e accep ing he deal. TD pa icipan s
go he ad e ised la o ; MD pa icipan s we e assigned ei he
hazelnu (mos p e e ed) o s awbe y (leas p e e ed) o con i m
o discon i m p e e ences. A e iewing he la o o 15 s, b and
loyal y in en ions we e measu ed using a 4‐i em scale (Chaudhu i
and Holb ook 2001) and epu chase in en ions wi h a single‐i em
scale (Habel e al. 2016;Oli e 1980). Pa icipan s also a ed pos ‐
pu chase a ec (Shampanie e al. 2007)and anked la o s o
assess p e e ence con i ma ion. All cons uc s and i ems a e de ailed
in Appendix D.
1
3.1.2 | Resul s o S udy 1a
3.1.2.1 | In la ed Expec a ions and P e‐Pu chase
A ec . We i s es ed whe he consume s had o e ly op i-
mis ic expec a ions in he MD condi ion. While a a ional ex-
pec a ion would be 25% (wi h all ou op ions equally likely),
86% o MD pa icipan s expec ed hei p e e ed la o . A chi‐
squa e es con i med a signi ican di e ence (i.e., 25%; χ
2
(1,
n= 75) = 149.15, p< 0.001), suppo ing H
1
ha MDs lead o
in la ed expec a ions.
3.1.2.2 | (Dis)Con i ma ion E ec s and Posi i e A ec-
i e Spillo e E ec s. We used Model 4 o he PROCESS
mac o (Hayes 2022) wi h 5000 boo s apped samples and 95%
bias‐co ec ed con idence in e als o es ou media ion
hypo heses. The MD sample was spli in o con i med (MD
con
,
n= 20) and discon i med (MD
dis
,n= 55) g oups based on
whe he pa icipan s ecei ed hei p e e ed la o . Then, wo
se ial media ion models we e es ima ed: one es ed (dis)con-
i ma ion e ec s by con as ing MD
con
and MD
dis
wi h he TD;
he o he examined posi i e a ec i e spillo e by compa ing
MD g oups wi h TD. We con olled o MD
con
e sus MD
dis
di e ences in all pos ‐pu chase measu es o align pa h coe i-
cien s ac oss models and included age and gende as co a ia es.
In he i s model, we ound signi ican discon i ma ion
(a×b=−0.16, 95% CI [−0.2830, −0.0485]) and con i ma ion e -
ec s (axb= 0.18, 95% CI [0.0479, 0.3439]) on b and loyal y
in en ions. Compa ed o a TD, MD
dis
lowe s and MD
con
aises
b and loyal y in en ions ia pos ‐pu chase a ec . Simila e ec s
we e ound o epu chase in en ions—discon i ma ion
(a×b=−0.17, 95% CI [−0.3361, −0.0492]) and con i ma ion (a
xb= 0.20, 95% CI [0.0648, 0.3663]). These esul s suppo H
2
.
The second model showed posi i e a ec i e spillo e : MDs ( s.
TDs) inc eased b and loyal y and epu chase in en ions h ough
highe p e‐pu chase a ec (MD s. TD →p e‐pu chase a ec →
b and loyal y, axb=0.12, 95% CI [0.0177, 0.2552]; MD s. TD →
p e‐pu chase a ec → epu chase in en ions, axb=0.14, 95% CI
[0.0339, 0.2847]) and h ough bo h p e‐and pos ‐pu chase a ec
(MD s. TD →p e‐pu chase a ec →pos ‐pu chase a ec →b and
loyal y in en ions, axb=0.13,95% CI [0.0361, 0.2371]; MD s. TD
→p e‐pu chase a ec →pos ‐pu chase a ec → epu chase in en-
ions, axb= 0.13, 95% CI [0.0409, 0.2639]). These esul s suppo
H
3a
and H
3b
. Key esul s a e shown in Figu e 3.
2
3.1.2.3 | O e all E ec s on Cus ome Loyal y. The
indi ec e ec s show ha posi i e a ec i e spillo e (1) ully
o se s he nega i e impac o discon i ma ion and (2) ampli ies
he posi i e impac o con i ma ion. As a esul , o al e ec s
e eal no signi ican dec ease in b and loyal y (b= 0.16,
p= 0.37) o epu chase in en ions (b= 0.20, p= 0.25) o MD
dis
e sus TD. In con as , MD
con
( s. TD) led o signi ican ly
highe b and loyal y (b= 0.51, p< 0.05) and epu chase in en-
ions (b= 0.59, p< 0.05).
3.1.2.4 | Supplemen a y S udy. In a supplemen a y
s udy (n= 213), we es ed whe he he e ec s held when MDs
and TDs we e p iced equally a €7, despi e pas esea ch sug-
ges ing MDs a e ypically cheape (e.g., G anados e al. 2008).
We ound simila esul s and pa e ns in he case o he same
p ices o b and loyal y in en ions and epu chase in en ions
owa d he deal (see Appendix E). In addi ion, we applied an
al e na i e 3‐i em scale o o e op imis ic expec a ions owa d
he MD ou come adap ed om Daume and Hü l‐Maack (2020)
o measu e in la ed expec a ions. A he low p ice (€7), MDs
igge ed signi ican ly highe expec a ions han TDs (M
MD, low
p ice (€7)
= 3.53 s. M
TD, low p ice (€7)
= 2.54; F(1, 222) = 14.69,
p< 0.01). A he high p ice (€9), no di e ence was ound (M
MD,
high p ice (€9)
= 3.08 s. M
TD, high p ice (€9)
= 2.81; F(1, 222) = 1.07,
p= 0.30). Thus, hese indings addi ionally suppo he a io-
nale ha no he p ice pe se, bu he mys e y deal (a a speci ic
discoun a e) d i es hese e ec s.
3.2 | S udy 1b: Se ice‐Rela ed Con ex (Pas a
Res au an )
3.2.1 | Design, Pa icipan s, and P ocedu e
S udy 1b used a simila design and p ocedu e o S udy 1a. We
ec ui ed 248 pa icipan s om a Ge man p o essional online
2887
panel (64% women; M
age
= 31.21 yea s, SD
age
= 13.56 yea s) o
an online scena io expe imen . In con as o S udy 1a, he deal
ea u ed a pas a meal in a es au an , a ypical se ice se ing
(see Appendix C o s imuli).
3.2.1.1 | P e‐Pu chase Phase. Pa icipan s we e asked o
imagine encoun e ing a deal a a pas a es au an . In he TD
condi ion, a speci ic pas a meal was o e ed o €9 (discoun ed
om €10). In he MD condi ion, a “su p ise pas a”was o e ed
o €7. Fou possible meals we e used: mea sauce (mos p e-
e ed), ish (leas p e e ed), ca bona a, and go gonzola,
selec ed based on Ge man consume as e p e e ences
(YouGo 2019).
3.2.1.2 | Pos ‐Pu chase Phase. Pa icipan s we e hen
asked o imagine accep ing he deal. Those in he TD condi ion
ecei ed he ad e ised pas a meal, while hose in he MD
condi ion we e assigned ei he he mos p e e ed meal (mea
sauce) o he leas p e e ed ( ish), o ei he con i m o dis-
con i m hei p e e ences. The same scales and measu es om
S udy 1a we e used in bo h phases. Addi ionally, pa icipan s
we e asked abou hei pe cei ed discon i ma ion o alida e
whe he hei ou comes we e con i med o discon i med
(Oli e 1980). All measu es a e depic ed in Appendix D.
3.2.2 | Resul s o S udy 1b
3.2.2.1 | (Dis)Con i ma ion E ec s and Posi i e A ec-
i e Spillo e E ec s. As in S udy 1a, we spli MD pa ici-
pan s in o con i med (MD
con
,n= 32) and discon i med (MD
dis
,
n= 97) g oups. A one‐sample ‐ es showed ha pe cei ed dis-
con i ma ion was below he scale midpoin o MD
con
(M
MDcon
= 3.13; (31) = −3.22, p< 0.01) and abo e he mid-
poin o MD
dis
(M
MDdis
= 4.96; (96) = 6.39, p< 0.001), con-
i ming ha MD
con
pa icipan s pe cei ed he ou come as be e
and MD
dis
pa icipan s as wo se han expec ed.
We ollowed he same p ocedu e as in S udy 1a, es ima ing wo
se ial media ion models using PROCESS Model 4 (5000 boo -
s apping samples, 95% bias‐co ec ed con idence in e al;
Hayes 2022). The i s model e ealed signi ican dis-
con i ma ion (axb=−0.45, 95% CI [−0.6055, −0.3009]) and
con i ma ion e ec s (axb= 0.38, 95% CI [0.2372, 0.5486]) o
b and loyal y, as well as o epu chase in en ions
FIGURE 3 | Resul s o Media ion Analyses o S udy 1a. (a) Dependen a iable: B and loyal y in en ions, (b) Dependen a iable: Repu chase
in en ions o he deal. No e: *p< 0.05, **p< 0.01. AFF ( 1) = posi i e p e‐pu chase a ec i e eac ions, AFF ( 2) = posi i e pos ‐pu chase a ec i e
eac ions, BLI = b and loyal y in en ions, RI = epu chase in en ions, X1 = MD
con
e sus TD; X2 = MD
dis
e sus TD.
2888 Psychology & Ma ke ing, 2025
(discon i ma ion e ec : axb=−0.40, 95% CI [−0.5659,
−0.2610]); con i ma ion e ec : axb= 0.35, 95% CI [0.2149,
0.5002]). These esul s suppo H
2
.
The second media ion model showed signi ican posi i e
a ec i e spillo e e ec s. MDs ( s. TDs) posi i ely in luenced
b and loyal y and epu chase in en ions h ough highe p e‐
pu chase a ec (MD s. TD →p e‐pu chase a ec →b and
loyal y, axb= 0.09, 95% CI [0.0245, 0.1694]; MD s. TD →p e‐
pu chase a ec → epu chase in en ions, axb= 0.15, 95% CI
[0.0693, 0.2550]) o h ough bo h p e‐and pos ‐pu chase a ec
(MD s. TD →p e‐pu chase a ec →pos ‐pu chase a ec →
b and loyal y, axb= 0.12, 95% CI [0.0618, 0.1978]; MD s. TD →
p e‐pu chase a ec →pos ‐pu chase a ec → epu chase
in en ions, axb= 0.11, 95% CI [0.0532, 0.1808]).
4
Resul s a e
shown in Appendix F.
3.2.2.2 | O e all E ec on Cus ome Loyal y. To al e -
ec s showed ha MD
dis
( s. TD) did no signi ican ly educe
b and loyal y in en ions (b=−0.07, p= 0.59) o epu chase
in en ions (b=−0.12, p= 0.36), e en hough he dis-
con i ma ion e ec was la ge han he wo a ec i e spillo e
e ec s. This indica es ha he spillo e e ec s we e s ong en-
ough o coun e balance he nega i e impac o discon i ma ion.
Mo eo e , b and loyal y and epu chase in en ions we e highe
o MD
con
pa icipan s han o TD pa icipan s (b and loyal y:
b= 0.62, p< 0.01; epu chase in en ions: b= 0.50, p< 0.05).
3.3 | Discussion o S udies 1a and 1b
The esul s o S udies 1a and 1b e ealed wo e ec s ha
in luence he e alua ion o MDs a e he pu chase. Fi s , a (dis)
con i ma ion e ec in which cus ome loyal y depends on
whe he he ou come o he MD ma ches he cus ome s' p e -
e ences (con i ma ion) o no (discon i ma ion). In he case o
(dis)con i ma ion, MDs lead o mo e (less) cus ome loyal y
because o highe (lowe ) le els o posi i e pos ‐pu chase
a ec . Second, posi i e a ec i e spillo e e ec s in which con-
sume s' posi i e a ec i e eac ions o MDs in he p e‐pu chase
phase ca y o e o he pos ‐pu chase phase and posi i ely
impac cus ome loyal y. Mo eo e , hese spillo e e ec s
coun e ac nega i e e ec s o he discon i ma ion and comple-
men posi i e e ec s o he con i ma ion. Consequen ly, cus-
ome loyal y is a equal le els o discon i med MD and TD
cus ome s—which is impo an gi en he high likelihood o
ha ing discon i med cus ome s h ough MDs due o in la ed
expec a ions. The e ec s hold in a se ice con ex whe e
po en ial nega i e e ec s o a discon i ma ion a e pa icula ly
s ong due o he need o immedia e consump ion. Impo -
an ly, hese esul s emain consis en e en when he p ices o
MDs and TDs a e iden ical, as demons a ed in he supple-
men a y s udy.
4 | S udy 2: Elabo a ing on A ec i e Reac ions
wi h Facial Exp ession Da a
In S udies 1a and 1b, we demons a ed ha consume s' a ec i e
eac ions in bo h he p e‐and he pos ‐pu chase phases play an
impo an ole in unde s anding cus ome loyal y in MDs.
Howe e , we ocused on sel ‐ epo ed a ec i e eac ions and
did no cap u e consume s' spon aneous and unconscious
esponses. In S udy 2, we used AI‐based acial ecogni ion
so wa e o di ec ly ack consume s' a ec i e esponses o MDs
in he p e‐pu chase phase be o e and a e he deal and in he
pos ‐pu chase phase immedia ely be o e and a e esol ing he
mys e y.
4.1 | Design, Pa icipan s, and P ocedu e
We ec ui ed 285 s uden s om a Ge man uni e si y o a
labo a o y expe imen conduc ed in a specialized lab equipped
wi h acial emo ion ecogni ion so wa e. Pa icipan s ecei ed
mone a y compensa ion. A s uden sample was chosen due o
he on‐campus na u e and echnical complexi y o he lab se up,
which equi ed s anda dized, in‐pe son p ocedu es. While s u-
den samples ha e known limi a ions, hey a e commonly used
in heo y‐d i en expe imen s in es iga ing psychological p o-
cesses such as a ec i e esponses. The design mi o ed S udy 1a
(Appendix C), wi h ideo eco ding added o cap u e acial
exp essions. A e in o med consen , we ob ained usable da a
om 264 pa icipan s M
age
= 21.64 yea s, SD
age
= 2.50 yea s),
analyzed ia he TAWNY Emo ions Analy ics Pla o m
(TAWNY n.d). This so wa e is based on he es ablished acial
ac ion coding sys em (FACS) by Ekman and F iesen (1978)
used o iden i y he se en basic emo ions. I epo s second‐by‐
second index alues anging om ze o o 100 o each emo ion,
indica ing he p obabili y ha a pe son is exp essing ha
emo ion.
The p ocedu e ollowed S udy 1a wi h h ee adjus men s:
(1) andom assignmen o pa icipan s o 15 wo ks a ions,
(2) pa icipan s spen 60 s showing a neu al exp ession be o e
s a ing, and (3) in he pos ‐pu chase phase, pa icipan s had a
chance o ecei e any la o , wi h hazelnu o s awbe y
assigned 40% o he ime and milk chocola e o ca amel 10%.
4.2 | Measu es
We ocused on he emo ion o happiness, as i bes ep esen s
posi i e a ec in his s udy a ea (Ruan e al. 2018). We analyzed
pa icipan s' happiness index in ou phases: (1) he calib a ion
phase, cap u ing na u al exp essions o happiness; (2) he c e-
a ion phase, whe e pa icipan s encoun e ed he TD o MD;
(3) he p e‐ esolu ion phase, measu ing happiness jus be o e
opening he i ual pa cel; and (4) he esolu ion phase, when
he speci ic TD o MD was e ealed. Fo he calib a ion phase,
we used he ull 60‐second ime ame. Fo he c ea ion and
esolu ion phases, we analyzed he i s 10 s o cap u e quick,
unconscious emo ional eac ions (Ekman 1992), a oiding cog-
ni i e in luences. Ou da a suppo ed his decision, wi h mean
imes o minimum and maximum happiness alues wi hin
hese phases being 2.22 (SD = 3.80) and 6.35 (SD = 7.69),
espec i ely. Simila ly, he mean ime (seconds) o each a
minimum alue in he esolu ion phase was 2.16 (SD = 3.41)
and o each a maximum alue was 6.42 (SD = 6.33). Fo he
p e‐ esolu ion phase, we used a 33‐second ime ame, as i was
2889
he median ime om when pa icipan s accep ed he deal o
when he p oduc was e ealed. We calcula ed mean happiness
sco es o each phase and deal ype. In bo h he p e‐ esolu ion
and esolu ion phases, we spli he MD g oup in o MD
con
(con i med) and MD
dis
(discon i med) subg oups, based on
whe he pa icipan s ecei ed hei p e e ed al e na i e.
5
Co a ia es included wo ks a ion, session, and gende .
4.3 | Resul s o S udy 2
A ec i e Reac ions in he P e‐Pu chase Phase.Weused
planned con as s o ou analysis and compa ed he happiness
sco es o each deal (n
TD
=132;n
MD
=132)in hec ea ionphase
o he sco es in he calib a ion phase o in es iga e whe he he
deal signi ican ly inc eased pa icipan s' happiness. The MD sig-
ni ican ly inc eased happiness sco es in he c ea ion phase com-
pa ed o he calib a ion phase (M
MD, calib a ion
=1.04 s.
M
MD, c ea ion
= 3.49; F(1, 262) = 8.07, p< 0.01). In con as , he
happiness sco es we e no signi ican ly a ec ed by he TD (M
TD,
calib a ion
=1.93 s.M
TD, c ea ion
=3.27;F(1, 262) = 2.42, p=0.12).
A ec i e Reac ions in he Pos ‐Pu chase Phase. We com-
pa ed happiness sco es ac oss TD, MD
con
(n= 28) and MD
dis
(n= 104) g oups be o e and a e mys e y esolu ion. MD
con
pa icipan s epo ed signi ican ly highe pos ‐ esolu ion hap-
piness (M
MDcon, p e‐ esolu ion
= 3.97 s. M
MDcon, esolu ion
= 9.06;
F(1, 261) = 4.72, p< 0.05). In con as , MD
dis
sco es we e
una ec ed by he un a o able ou come o he MD, (M
MDdis, p e‐
esolu ion
= 2.93 s. M
MDdis, esolu ion
= 4.04; F(1, 261) = 0.83,
p= 0.36). TD pa icipan s' sco es emained s able ac oss phases
(M
TD, p e‐ esolu ion
= 3.78 s. M
TD, esolu ion
= 5.39; F(1,
261) = 2.22, p= 0.14). Figu e 4summa izes ou indings.
4.4 | Discussion o S udy 2
The indings o S udy 2 elabo a ed on consume s' posi i e
a ec i e esponses o MDs in he c ea ion and esolu ion
phases. Using AI‐based emo ion acking, we obse ed how
consume s na u ally espond o MDs, con i ming ha posi i e
a ec inc eases du ing bo h phases. When an MD's ou come
ma ches cus ome p e e ences, cus ome s expe ience e en
highe posi i e a ec pos ‐ esolu ion. Mo eo e , he posi i e
a ec o hose discon i med by he MD emains simila o
(1) ha o TD consume s and (2) o hei p e‐ esolu ion le els,
sugges ing ha p e‐pu chase happiness coun e ac s he nega-
i e e ec s o discon i ma ion on pos ‐pu chase happiness.
5 | S udy 3: Robus ness Checks
S udy 3 compa ed he MD condi ion wi h a di e en TD base-
line e sion (TD
choice
) o ule ou al e na i e explana ions o
he highe posi i e a ec in he MD p e‐pu chase phase. I
con olled o (a) he g ea e numbe o op ions in he MD
condi ion and (b) he lack o choice in he o iginal TD condi-
ion. In TD
choice
, pa icipan s saw all ou la o s and chose one
eely. The s udy also educed demand e ec s by emo ing he
MD p omp asking pa icipan s o s a e hei expec ed la o .
5.1 | Design, Pa icipan s, and P ocedu e
We ec ui ed 299 pa icipan s om a p o essional online panel
ep esen ing he Ge man popula ion (50% women;
M
age
= 44.71 yea s, SD
age
= 13.77 yea s) o an online scena io
expe imen . We used a single‐ ac o (MD s. TD
choice
s. TD)
be ween‐subjec s design. Ou p ima y ocus was he compa i-
son be ween he MD and TD
choice
condi ions, wi h he o iginal
TD condi ion included o es obus ness.
P e‐Pu chase Phase. Pa icipan s we e andomly assigned o
one o h ee condi ions: MD, TD
choice
,o TD
non‐choice
. TD con-
di ions mi o ed hose om S udy 1a. In he TD
choice
condi ion,
pa icipan s saw all ou chocola e la o s (like he MD condi-
ion) and we e in o med hey could choose one la o (see
Appendix C). The same measu es om S udy 1a we e used, bu
pa icipan s we e no asked which la o hey expec ed o
ecei e. Addi ionally, TD
choice
pa icipan s also selec ed one o
he a ailable la o s.
Pos ‐Pu chase Phase. As in S udy 1a, pa icipan s we e asked
o imagine pu chasing he deal and ecei ing a pa cel wi h a
10‐pack o chocola e. In he MD condi ion, hey we e andomly
gi en one o ou la o s. In he TD condi ions, pa icipan s
ecei ed ei he he ad e ised la o (TD
non‐choice
condi ion)
o he la o hey had chosen (TD
choice
condi ion). The p oce-
du e and measu es we e iden ical o hose in S udy 1a (see
Appendix D).
FIGURE 4 | Resul s o he Happiness Index in he Di e en Phases o he Lab Expe imen .
2890 Psychology & Ma ke ing, 2025
TABLE B.1 | S udy o e iew.
N . Sample S udy pu pose Design
Me hodological
app oach
1a 148 In es iga ion o consume s’pos ‐pu chase eac ions owa d
he esolu ion o a mys e y deal in a p oduc con ex .
Single‐ ac o be ween
subjec s
Online‐scena io
expe imen
1a supplemen a y 213 Supplemen a y in es iga ion o consume s’pos ‐pu chase
eac ions when MD and TD a e p iced equally.
Single‐ ac o be ween
subjec s
Online‐scena io
expe imen
1b 248 In es iga ion o consume s’pos ‐pu chase eac ions owa d
he esolu ion o a mys e y deal in a se ice con ex .
Single‐ ac o be ween
subjec s
Online‐scena io
expe imen
2 264 In es iga ion o consume s’na u al and spon aneous
eac ions o mys e y deals.
Single‐ ac o be ween
subjec s using AI‐based
acial ecogni ion
so wa e
Labo a o y
expe imen
3 299 Elimina ion o he p esence o a highe numbe o
p esen ed op ion as a cause o highe posi i e a ec i e
eac ions o mys e y deals in he p e‐pu chase phase.
Single‐ ac o be ween
subjec s
Online‐scena io
expe imen
4 235 Addi ion o ex e nal alidi y o p e ious indings. Single‐ ac o be ween
subjec s
Field s udy wi h
online su ey
TABLE C.1 | S imuli in he P e‐Pu chase phase in s udies 1a, 1b, 2, and 3.
Scena ios o S udies 1a and 2 (P oduc Con ex )
TD MD
Scena ios o S udy 1b (Se ice Con ex )
TD MD
Scena ios o S udy 3
TD TD
choice
MD
Appendix B
Appendix C
(See Table C1)
2897
TABLE D.1 | Measu emen i ems and alidi y assessmen o cons uc s.
Cons uc sou ce αMeasu emen i ems
P e‐Pu chase Phase
Pu chase In en ion
e.g., Goldsmi h and Ami (2010); La an and Tsi os (2013)
single i em
(used in S udies 1a,
1b and 3)
How likely would you buy he deal you ha e
jus seen?
Posi i e P e‐Pu chase A ec i e Reac ions
Shampanie e al. (2007)
single i em
(used in S udy 1a,
1b and 3)
How do you eel wi h ega d o [ he ad]?
Posi i e Expec a ions Adap ed om Daume and Hü l‐
Maack (2020)
S udy 1a:no used
S udy 1b: α= 0.79
S udy 3: no used
The ad I jus saw…
…made me expec some hing in e es ing.
… igge s high expec a ions abou wha is
ad e ised he e.
… igge s my hope o ecei e my mos ly p e e ed
la o om [ he p omo ion].
Unce ain y
e.g., Lee and Qiu (2009)
single i em
(used in S udies 1a,
1b, 2 and 3)
I eel unce ain abou which [p oduc ] I would
ecei e wi hin he p e iously seen ad, when I
pa icipa e in he deal and buy he [p oduc ].
Pos ‐Pu chase Phase
B and Loyal y In en ions
Chaudhu i and Holb ook (2001)
S udy 1a: α= 0.88
S udy 1b: α= 0.87
S udy 3: α= 0.85
I will buy [ his b and] he nex ime I buy
[p oduc name].
I in end o keep pu chasing [ his b and].
I am commi ed o [ his b and].
I would be willing o pay a highe p ice o [ his
b and] o e o he b ands.
Repu chase In en ions o he P omo ion
Adap ed om Oli e (1980); Habel e al. (2016)
single i em (used in
S udies 1a, 1b
and 3)
How likely would you pu chase [such a p omo ion] o
[ he b and] again?
Posi i e Pos ‐Pu chase A ec i e Reac ions
Shampanie e al. (2007)
single i em (used in
S udy 1a, 1b and 3)
How do you eel wi h ega d o [ he ecei ed p oduc /
se ice]?
Sa is ac ion
e.g., Habel e al. (2016)
single i em (used in
S udies 1a, 1b
and 3)
Gene ally, I am sa is ied wi h [ he p oduc ] I ecei ed.
Pe cei ed Discon i ma ion
Oli e (1980)
single i em (used in
S udies 1b and 2)
Gene ally, [ his p oduc ] is…(“be e han expec ed”–
“as expec ed”–“wo se han expec ed”)
No e: Posi i e a ec i e eac ions we e measu ed wi h a 5‐poin scale; all o he cons uc s we e measu ed wi h 7‐poin scales.
Appendix D
(See Table D1)
2898 Psychology & Ma ke ing, 2025
Appendix E
(See Table E1)
TABLE E.1 | Resul s o Supplemen a y S udy wi h TD and MD a he Same P ice (€7).
MD (€7) e sus TD (€7) Resul s
E ec s on B and Loyal y In en ions
Discon i ma ion E ec
MD
dis
e sus TD →posi i e pos ‐pu chase a ec →b and loyal y in en ions
axb=−0.11, 95% CI [−0.2670, −0.0004]
Con i ma ion E ec
MD
con
e sus TD →posi i e pos ‐pu chase a ec →b and loyal y in en ions
axb= 0.26, 95% CI [0.0496, 0.4985]
A ec i e Spillo e E ec s
MD e sus TD →posi i e p e‐pu chase a ec →b and loyal y in en ions axb= 0.09, 90% CI [0.0036, 0.1857]
MD e sus TD →posi i e p e‐pu chase a ec →posi i e pos ‐pu chase a ec →b and loyal y
in en ions
axb= 0.02, 90% CI [0.0001, 0.0625]
To al E ec s MD
dis
e sus TD: b= 0.17, p= 0.38
MD
con
e sus TD: b= 0.74, p< 0.01
E ec s on Repu chase In en ions o P omo ion
Discon i ma ion E ec
MD
dis
e sus TD →posi i e pos ‐pu chase a ec → epu chase in en ions
axb=−0.21 95% CI [−0.4393, −0.0139]
Con i ma ion E ec
MD
con
e sus TD →posi i e pos ‐pu chase a ec → epu chase in en ions
axb= 0.52, 95% CI [0.2971, 0.7838]
A ec i e Spillo e E ec s
MD e sus TD →posi i e p e‐pu chase a ec → epu chase in en ions axb= 0.08, 90% CI [0.0049, 0.1919]
MD e sus TD →posi i e p e‐pu chase a ec →posi i e pos ‐pu chase a ec → epu chase
in en ions
axb= 0.05, 90% CI [0.0029, 0.1104]
To al E ec s MD
dis
e sus TD: b=−0.01, p= 0.95
MD
con
e sus TD: b= 0.81, p< 0.01
No e: We ollowed he same p ocedu e and conduc ed he same analyses (PROCESS models; Hayes 2022) as in S udy 1a; n= 110.
2899
FIGURE F.1 | In eg a ed Resul s om Media ion Analyses o S udy 1b and S udy 3. No e:
†
p< 0.10, *p<0.05, **p< 0.01, AFF ( 1) = posi i e p e‐
pu chase a ec i e eac ions; AFF ( 2) = posi i e pos ‐pu chase a ec i e eac ions, BLI = b and loyal y in en ions, RI = epu chase in en ions, X1 =
MD
con
e sus TD/TD
choice
,X
2
=MD
dis
e sus TD/TD
choice
.
Appendix F
(See Figu e F1)
2900 Psychology & Ma ke ing, 2025
FIGURE F.1 | (Con inued)
2901
