Brand Loyalties

8/12/2019 Brand Loyalties

1/8

Journal of the Academy of M a r k e t i n g S c i e n c eall 1974 Vol. 2 No. 4 651-658

B r a n d L o y a l t i e s :

Q u a l i t a t i v e

Q u a n t i ta t iv e o r B o th ?

D a v i d R . W h e e le r D . B . A .

B a r u c h C o ll eg e T h e C i t y U n i v e r s it y o f N e w Yo r k

I N T R O D U C T I O N

M arke te rs wish th a t the i r f i rm 's o ffe rings wil l be perce ived byconsumers , no t as p roduc ts , bu t as b rands . Marke te rs hope tha t they cancause consum ers to discr iminate among the var ious brands that areava ilable and to purchase the co r re c t b rand . M arke ting abounds wi th

examples o f consum ers l ea rn ing to choose one p rodu c t o r s to re ra ther thananother. These consum er pre fe rences have been te rmed loya l t i es .Brand loy a l ty ( in var ious semant ic fo rm s) is a par t o f any mark e te r ' s

vocabula ry. B rand loya l ty has b een def ined (Engle , e t a l ., 1 968 , p . 609) asa sequence o f b rand choices o r expressed as p robabi l it i es o f repea tpurchases over t ime. Researchers (Brown, 1952) have found s ignif icantconsis tencies in consu m ers ' purchase beha vior concern ing var ious prod-uc t s. Th e consum er 's t endency t o deve lop b rand l oya l t y is an imp or t an tadvantage to the m arkete r of an es tablished br an d (Sturd ivan t , e t a l. ,

1970, p . 174) .Al though brand loya l t i es seem to ex is t , a cons i s ten t opera t iona l

def ini t ion is lacking. Defini t ions of brand loyal ty are not only e lusive buta lso num erous . Brand loya l ty has been def ined (Cunn ingham, 1956) inte rms of the pro po r t ion of purchases o f the mos t p opula r b rand . JagdishN. Sh e th (196 8) d ef ined brand lo ya l ty as the num ber of cho icesnsequence dur ing a s ta ted leng th o f t ime . Gues t (19 64) d ef ined brandloy a l ty as the cons i s tency of b rand purchas ing over time per iods up totw en ty years. There a re several behaviora l def in i tions of b rand lo ya l ty

which rep or t ac tua l purchas ing behavior by consum ers .In this paper, the problem s associated with qua nt i ta t ive and qual i ta t ive

approaches to b rand loya l ty a re d i scussed . The pos i t ion taken by th i sauth or is that the exclusive use of on ly one of these app roaches wil l lead

651


2/8

652 WHEELER

to a def ic ien t m odel o f b rand loya l ty because m arke te rs need to

unders tand and pred ic t consumer behavior.

Q U L I T T I V E M O D E L S

A num ber o f hyp o these s have been advanced t o exp l a in why consumer schange brand s. T here is usual ly som e em pir ical evidence for any par t iculartheo ry. Theor ies o f b rand swi tch ing can be ca tegor ized as be ing behaviora ldescr ipt ions of internal psychological s ta tes . Qual i ta t ive models viewconsumer purchas ing as be ing caused by some fac tor as opposed to

o c cu r ri ng b y c h a n c e. T h e t e r m l o y a l t y . . , h a s b e e n u s ed to . n a m e t h ecommonly obse rved phenomenon tha t consumer s do no t d i s t r i bu t e t he i rchoc ies rando m ly wi th in a g iven pro du c t a rea (Hansen , 1972 , p . 322) .

Behavioral descr ip tions of b rand swi tch ing a t tem pt to ana lyze consum erbehavior at a m icro leve l . Var ious in te rn a l fac tors have beenconceptua l ized in o rder to exp la in brand swi tch ing behavior. Thesedescr ipt ions of cogni t ive processes have included cur iosi ty, d isappoint-m en t , reassurance, avai labi li ty of a l ternat ives , and decis ions .

Cogni t ive explana t ions o f b rand loya l ty have been offe red in some

recen t exp l ana ti ons o f consum er cho i ce behav io r. Tucke r (196 4 ) p r e sen t edexper im enta l f ind ings on the cogni tive s t rength o f loya l t i es .

While learning theories f rom psychological l i tera ture have been used ina t t emp t s t o exp l a in t he p rocess o f b r and l oy a l t y, the d i cho to m y be tweenconnec t ion is t and cogni t ive theor ies has no t g iven marke te rs a un i f iedexp lana t ion o f b r and l oy a l t y. J acoby and Ky ne r (1973 ) have foundempir ica l sup por t fo r a concep tua l def in i t ion of b rand loya l ty. The i rdef in i t ion i s based on the idea tha t b rand loya l ty is a fo rm o f repea tpurchas ing . Jacoby and Kyner ' sconceptu ldef in i t ion inc ludes thefo l lowing condi t ions : (1 ) purchas ing tha t is no nran dom , (2) behaviora lresponses , (3) temporal responses , (4) avai labi l i ty of a l ternat ives , and (5)beha vior tha t is a resul t o f a decis ion-making process . The m ajorcon t r ibu t ion of Jac oby and K yn er ' s a r ti c le is in the i r dev e lopm ent o f alogica l f rame wo rk tha t can be used to t i e toge th er two v iews of b randloy al ty : repeat purchases (behav ioral) to un derlying processes (cogni t ive) .

Marke te rs have a t tempted to f ind ev idence tha t a pos t ive re la t ionsh ipexis ts betw een users (classi fied e i ther as hea vy or light) and bran d lo yal ty.I f such a re la t ionship could be fo un d, a logical basis for segment ing

ma rke ts would ex is t . M arke t ing l i t e ra ture abounds wi th re fe rences tocont ras ts involv ing each o f the d imens ions [ to ta l purchases and brandloy a l ty ] . I t is o f ten a rgued tha t two of the m os t va luab le m arke t segments


3/8

BRAND LOYALTIES 653

to pene t ra te a re the 'heavy ha l f and those househo lds tha t exh ib i t a h igh

prope ns i ty to be brand - loya l (Frank , 1968 , p . 53) .In s tud ies o f heavy and l igh t p roduc t users , no d i scern ib le pa t te rns o f

loya l ty, person a l i ty, o r o ther charac te r i st i cs have been show n. I t w ouldappear tha t a b rand- loya l cu s tomer could be pos i t ive ly cor re la ted wi thheav y users of that pro du ct . R esearch has fa i led to sho w tha t th is is t rue.Rona ld E . F rank wr i te s : . . . t he 'h igh b r and - loya l ' househo ld appa ren t l yhas a p rof i l e o f perso na l i ty and soc ioeconom ic charac te r i st i cs tha t i sv i r tua l ly iden t ica l to tha t o f households exhib i t ing a lower degree ofloy a l ty (Frank , 1968 , p . 61) . Dik W. Tw edt (196 4) be lieves tha t a heavy

user household i s no t read i ly iden t i f i ab le except th rough ana lyses o fpurchase behavior. O nly very smal l d i ffe rences were foun d (G ot t l i eb ,1958 ) in the perso nal i ty character is tics betw een heav y and l ight users o f ab rand m ed ica ti on .

The conc lus ions f rom the preced ing d i scuss ion a re tha t marke te rs haveclass if ied pro du ct users into l ight and heav y categories , w hich are basicallynot d i s tinguishable f rom one ano ther exce pt by an a rb i t ra ry c lass if ica t ionscheme o f the par t i cu la r researchers .

Q U N T I T T I V E M O D EL S

Phys ica l p roduc t d i ffe ren t ia t ion i s becoming less p ronounced . Con-sumers ' ab i l it ies to d i scern phys ica l d i ffe rences among brands o f p rodu c tsin the m arke t p lace a re be ing ham pered b y f i rms produc ing near lyident ical prod ucts . Th e abi l i ty to dis t inguish am ong the var iousbrands . . , may technica l ly ex is t , bu t the magni tude i s qu i te smal l and i sun l ike ly to be o f g rea t va lue in the marke t p lace (Myers and Reyn olds ,1967, p. 18).

As phys ica l d i ffe rences among p rodu c ts have dec l ined , m arke te rs haveadvanced theor ies tha t b rand swi tch ing is due to chance . W henindividuals have diff icul ty discr iminat ing between s t imulus s i tuat ions ,behav ior becom es ran do m (Engel , e t al ., p . 125 ) . Purchases o f coffe e (aproduc t where brand d i ffe ren t ia t ion i s l imi ted) a re repor ted to be c lose lyappro xim ated by a rando m Poisson process (Day , 1970 , p . 68) . In an ear lyexper im ent (Husband and G od frey, 1934) involving sub jec t s' ab i li ti es tod isc r imina te be tween c igare t tes , cor rec t recogni t ion of the unmarkedc igare t tes was no b e t te r than could be expe c ted due to chance .

Par t ly due to the qu ant i t a t ive or ien ta t ion of the researchers , a t t en t ionhas been focused on var ious probabi l is t ic mode ls to explain bran dswi tch ing . These models a t t em pt to represen t behavior as some manner o f


4/8

654 WHEELER

stochas t ic p rocess 9 The par t i cu la r s tochas tic m odel depends on the

assumpt ions made regard ing the e ffec t s o f exper ience o n behavior 9 Brandswitching gen eral ly is prese nted s tochast ical ly in on e o f three models :Berno ul l i , M arkov, o r a l inear learning m odel 9

A Bernoul l i mode l o f b rand swi tch ing beh avior assumes tha t p rev iouspurchases have no e ffec t on presen t purchas ing . Each brand swi tch ingprobabi l i ty remains cons tan t f rom one purchas ing per iod to the nex t .Massy, M ontgo m ery, and Morr i son (197 0 , p . 18) assume:

9 tha t households reeva lua te the wor th of the var ious brands a t

discrete po ints in t im e, that the o utco m es of the successiveeva lua tions (drawn f rom the d i s t r ibu t ion o f p robabi l i ti es ) a reindependent o f one another, and tha t the purchase probabi l i t i es dono t chang e ( i .e . , the p rocess is Berno ul l i ) be tw een reevaluat ions .

F rank (1962 ) found ev idence t ha t p robab i l i ty o f buy ing was cons t an t ove ra per iod o f time for each buy er in h is sample . I f behavior has a cons tan tproba bi l i ty o f occur r ing i t is indepen dent o f purchase h i s to ry.

Most researchers have presen ted brand swi tch ing in t e rms o f a Markov

mo del . Markov models assume tha t the ana lys is o f p resen t purchasebehavior dep ends on ly on the imm edia te ly preced ing purchase event andany ear l i e r h i s to ry than tha t , i s i r re levant . Ehrenberg (1965 , p . 347)explains this mo del as fol lows:

Markov brand-swi tch ing models a im in genera l to dea l wi threpea t -buying and brand-swi tch ing behavior, p r imar i ly fo r f requen t lyb rough t nondurab l e consumer goods . Consumer pu rchas ing o fdiffe ren t brands w ithin a s ingle p ro du ct f ie ld is usual ly analyze d forsuccessive equal per io ds of t ime , e .g ., m on ths or quar ters 9

A th i rd s tochas t ic m odel tha t has been used to expla in brand swi tch ingis the l inear l ea rn ing m odel pa t te rn ed a f te r the genera l ized form presen tedby Bush and M oste lle r (1955) . The assum pt ion i f the l inear l earn ing mo delis tha t " . . . the p roba bi l i ty tha t a consum er will purchase a par t i cu la rbrand i s a func t ion of what she has l ea rned f rom pas t favorab leexper iences wi th tha t b ra nd " (Carman, 1966 , p . 23) . In the l inear l earn ingm odel , a consu me r ' s behavior is a ffec ted by h i s p rev ious brand choices ." . . . The ac t o f purchas ing and us ing a par t i cu la r b rand i s assumed toaffec t the pro babi l i ty tha t th is b rand will be se lec ted the nex t t ime thepro du ct c lass is to be pu rch ase d" (Massy, e t a l. , p . 141) .

A l inear l ea rn ing model tha t has a t t rac ted cons iderab le a t t en t ion in


5/8

BRAND LOYALTIES 6

marke t ing is the o ne deve loped by A l f red A. Kuehn . K uehn ' s 1962 s tudy

o f b r and l oya l t y a t t emp ted t o show tha t consum er 's b rand cho i ce behav io rcan b e repres ented as a l inear s tochast ic process . Ku ehn 's learning m ode l ispa t te rne d a f te r the Bush-Moste lle r m odel . K uehn def ined h i s l inear m ode lin term s o f the s lopes and in tercep ts o f tw o s t ra ight lines . Th e lines areca lled puchase op era tor and re jec t ion op era to r. These a re expla inedas:

I f the brand in ques t ion is purchased b y the consum er on a g ivenoccas ion , the consum er ' s p rob abi l i ty o f again bu ying the same brand

the nex t t ime th a t ty pe o f p rodu c t is purchased i s read f rom thepurchase opera tor. I f the brand i s re jec ted by the consumer on ag iven buying occas ion , the consu m er ' s p roba bi l i ty o f buying tha tb r and when he nex t buy s t ha t t ype o f p roduc t f rom the re j ec ti onopera tor (Kuehn , p . 104) .

Empir ica l suppor t o f th i s l inear model came f rom a s tudy tha t Kuehncond uc ted f rom 1950-1952 on da ta f rom 600 Chicago fami l ies ' sequent ia lpurchase behavior. Brand swi tch ing pa t te rns fo r Snow Crop f rozen orangejuice was analyzed.

Even though Kuehn ' s l inear l ea rn ing model seems to p red ic t purchasebehavior qu ie t wel l f rom prev ious purchase h i s to ry, there a re some se r iousdefec t s p resen t in the m odel . A problem of cons iderab le im por tance i s tha tof es t imat ing the four param eters . I f th i s [de te rmin ing the param eters ]cou ld be done a p r i o r i the model might be a va lue to marke t ingman agem ent fo r use in fo recas t ing . Massy, M ontg om ery, and Morr ison(1970 , p . 157 ) wr i t e: . . .w e canno t t ru s t t he speci fi c p r ed i ct i ons heobta ins o r the impl ied es t imates o f param eters . Ph i lip Ko t le r (197 1 , p .506) po in t s ou t two o ther p rob lems wi th Kuehn ' s s tochas t ic l ea rn ingmode l :

The purchase op erato r is too r igid in implying inevi table sat isfact ionwi th u se . A l so, th i s m o d e l , . . , i gno re s t he e f fec t on b r and cho ice o fvar ia t ions in the m arke t ing mix . I t descr ibes the buy er ' s b randpurchase probabi l i t ies as be ing m odi f ied so lely th roug h pas t b randchoices .

S tochas ti c mode l s a re abs t r act i ons o f t he r ea l wor ld . Th ey m ay be t ooabs t ract to p resen t com ple te ly the m any face ted aspec ts o f b randswi tch ing . Probably no s imple s t ruc ture can d o an adequ a te job . In thetyp ica l s tochas t ic mo del , the inputs a re the sequence and the f requen cy of


6/8

656 WHEELER

b r a n d p u r c h a se s . H e r e in l ie s th e s t o ch a s ti c m o d e l s ' m a j o r w e a k n e s s - t h e

l a ck o f i n c o r p o r a t i o n o f s e v er al i m p o r t a n t m a r k e t i n g v a r ia b le s (M a c L a c h -lan , 1972 , p . 378) . The in f luences and re la t ionsh ips o f p r ice , p ro m ot io n ,p ro du c t , and d i s t r ibu t ion a re no t cons idered in the p robab i l is t i c mo de ls .There i s su ff i c ien t jus t i f i ca t ion to be l i eve tha t s tochas t i c mode ls a re no tab le to ad equa te ly exp la in b rand swi tch ing . Ehrenb erg (p . 361) w r i tes :

So fa r, it has never bee n show n th a t the re i s e ithe r any need o r anyv a lu e i n g o in g b e y o n d t h e n o n s t o c h a s ti c f lo w m o d e l t y p e o fin te rp re ta t ion . What i s more , the re i s a l ready ample ev idence tha t

the p robab i l i s t i c in te rp re ta t ion would be imposs ib le .

I t m a y b e t h a t s t o c h a st i c m o d e l s m a y e x p l a i n n o t h i n g a b o u t b r a n dswi tch ing . The va lue o f these p robab i l i s t i c mode ls may l i e in thed e v e l o p m e n t o f e x p e r i m e n t a l w o r k t o v e r i f y t h e b r a n d s w it ch i n gh y p o t h e s e s .

S U M M R Y

Researchers have p resen ted var ious exp lana t ions o f b rand lo ya l ty int e r m s o f e i th e r q u a n t i ta t i v e o r q u a l i ta t iv e m o d e l s . T h e d e v e l o p e r s o f b r a n dl o y a l t y m o d e l s h a v e a t t e m p t e d t o f o r m u l a t e t h e o r e t ic a l f r a m e w o r k s f r o mwhich b rand-swi tch ing behav ior cou ld be p red ic ted and exp la ined .Qua l i t a t ive exp lan a t ions o f consum ers beha v ior have no t b een success fu l inaccount ing fo r the va r iance in the d epe nde n t va r iab le. An ab i l ity to p red ic tbehav ior on the bas i s o f a ve rba l mode l has no t been a sa t i s fy ingexper ience fo r con sum er resea rchers . The var ious quan t i t a t ive mod e ls o fb r a n d l o y a l t y s e e m t o p r e d i c t c o n s u m e r b e h a v i o r m o r e a c c u r a t e l y t h a n

have the qua l i t a tive m ode ls . Q uan t i t a t ive m ode ls su ffe r f rom a l ack o f ac o n s i d e ra t io n o f w h y l o y a l t y e x is ts .

Brand loya l ty needs to be de f ined in bo th qua l i t a t ive and quan t i t a t ivet e r m s . A b r a n d l o y a l c o n s u m e r n o t o n l y m u s t p u rc h a s e a p a r t ic u l a r b r a n dcons i s ten t ly over t ime , bu t a l so the loya l consumer mus t have pos i t ivepsycho log ica l p re fe rences fo r the b rand under cons idera t ion .

REFERENCES

Brown, George H. 19 52 . Brand Lo ya lty- Fa ct or Fiction? Advertising Age 23(June 9) 52-55.

Bush, Ro ber t R. and M osteller, Frederick. 19 55 . Stochastic Mod els for Learning.New York: John W iley and Son s, Inc.

Carm an, James M. 196 6. Brand S witching and Linear Learning M odels. Journal ofAdvertising Research 6 (June) 23-31.


7/8

BRAND LOYALTIES 657

Cunningham, Ross M. 1956. Brand Loyalty-What, Where, How Much? HarvardBusiness Review 34 (January-February) 116-128.

Day, George S. 1970. Buyer Attitude and Brand Choice Behavior. New York: TheFree Press.

Ehrenberg, A. S. C. 1965. An Appraisal of Markov Brand-Switching Models.Journal of Marketing Research 2 (November) 347-362.

Engel, James F., Kollat, David, and Blackwell, Roger D. 1968. Consumer Behavior.New York: Holt, Rinehart and Winston, Inc.

Frank, Ronald E. 1962. Brand Choice as a Probability Process. Journal of Business35 (January) 43-57.

Frank, Ronald E. 1968. Market Segmentation: Findings and Implications.Applications of the Sciences in Marketing Management. New York: John Wiley and

Sons, Inc.Gott lieb, M. J. 1958. Segmentation by Personality Types. Advancing Marketing

Efficiency. Chicago: American Marketing Association.Guest, Lester. 1964. Brand Loyalty Revisited: A Twenty-Year Repor t. Journal of

Applied Psychology 48 (April) 93-97.Hansen, Flemming. 1972. Consumer Choice Behavior. New York: The Free Press.Husband, R. W. and Godfrey, J. 1934. An Experimental Study of Cigarette

Identi fication. Journal of Applied Psychology 18 (April) 220-223.Jacoby, Jacob and Kyner, David B. 1973. Brand Loyalty Vs. Repeat Purchasing

Behavior. Journal of Marketing Research 10 (February) 1-9.Kotler, Philip. 1971. Marketing Decision Making: A Model Building Approach. New

York: Holt, Rinehart, and Winston.Kuehn, Alfred A. 19 68 . Consumer Brand Choice as a Learning Process.

Perspectives in Consumer Behavior. Glenview, Illinois: Scott Foresman andCompany.

MacLachlan, Douglas L. 1972. A Model of Intermediate Market Response. Journalof Marketing Research. 9 (November) 378-384.

Massy, William F., Montgomery, David B. and Morrison, Donald G, 1970. StochasticModels of Buying Behavior. Cambridge, Massachusetts: The M.I.T. Press.

Myers, James H. and Reynolds, William H. 1967. Consumer Behavior and MarketingManagement. Boston: Houghton Mifflin Company.

Sheth, Jagdish N. 1968. How Adults Learn Brand Preference. Journal ofAdvertising Research. 9 (September) 25-36.

Sturdivant, Frederick D. et al. 1970, Managerial Analysis in Marketing. Glenview,Illinois: Scott, Foresman and Company.

Tucker, William T. 1964. The Development of Brand Loyalty. Journal ofMarketing Research t (August) 32-35.

Twedt, Dik W. 1964. How Important to Marketing Strategy is the 'Heavy User'?Journal of Marketing 28 (January) 71-72.

BOUT THE UTHOR

DR. DAVID R, WHEELER received his doctorate in business administra-tion from Texas Te ch Univer sity in May, 1974. He is curre ntly t eaching atBaruch College of the City University of New York. He was an assistant


8/8

658 WHEELER

professor of ma rketi ng at the Universit y of North Dakota from 1972 to1974. His prima ry te aching and research activities are cons umer behaviorprinciples of market ing quan tit ati ve methods and marke t research. Dr.Wheeler has b een empl oyed as an engineer for an aerospace firm in Dallasas an au ditor for the state of Texas and as a meteorologist. Presently he isengaged in research c oncerning multi- dimens ional approaches to explainingand predicting consumer behavior.

Brand Loyalties

Documents