Tobacco Documents Online

o" ° I JJ tl I,~ Co~t:~.i~e= C:r¢=i=::.-."; "~F.S / RE~ i~=T No.: ~) , ' RE~'~:F. 1 t ITL¢_.: ~¢IJr::.~" C:lssiflc:~on: KE " ~,::.I, : I::I..= No.: ;; ", PR(:,E :" .E.ADER: 1..CFD ~-".1 : r.';H (1~ FS II) : CJP ~-t S (1) _Checx Cc, ~or : lli,~dinl~: r-¢... ;) ~GN (2l : .... ;"v - : File Figs. GraDh$ No.: P h ¢';O~i¢ill~l'lS No.: Re'Juctions: Total Pigls: DIS ~ ttu ION: R:,$ I2) : RL'G (1) ER (1) [~JK_[I)i : DGF ;1 L:s: T,Ikf .:; ~¢E F- Drll f. r t~ Di~ I1) (2) (3) 14) 15) I1) 12) ) , I I • . "t '. '.. ! Prl;,~.~,= ~ Bourx=l Col=Y: "'" " I m== .... " . ,-. j , ,, 0. . ¢ "- . ," , .; T :; Dig |lrllll il iltllnclId to t[ll¢iliUItl In¢l detail the I~rogrl$$ Of G,R. & D, C. fI.l~l~3 ~om tl~ tyl=l)lcriDt ~ ~mlil the t ~ finally ~ end iII~ml:l. The We h~rmi I~(~JId bI ¢ommentl¢t ill the time of the initial tylir~ ~lnd Ic¢omlIny the t'ypl~¢:~l~ lind I II"llf II~'ill~llr'ltJort) It Ill It,lbleo~Jlnt i'tl~ls. I1: mull be ill-it ~o u"l~lb'll filin| ,,-,,tith the ."~lltllr when ri~c "t km lUMl'~. NIt : ~Ilmr ~IIIIIle, Crlfll for Wl~ing I~oulc gl written in ink. ; &Pl:m" Wl~inIi, ft~e ryl=II~i~" li~ould normally be ~I~ur~ed to the Au.~. o,', .....ho w;;~ sul~e{:'~entl~" :t=l~Iin the II~ojeu'~ I.~I~er'l aM IIIMI~'I IIII~'O~IL ". :.~ :,ir~lltJlln llhotll= bl dlII,,~L'n=I~I~ I]Y ",,,he Illuer. i~lmte{= for ,-r'~tirlg unrii :Pe J$:;'Jpr ~2S :P.i::,li. il'~ .%"'! .~rO .~orm& 3..-3:nit "P.'~llrl .%~lllter". : The ty!DIII~-I~I rl'I~ll l~oI Ol ~, APtI~ I~Til'llil1(l. olle 130ur~ C:.'% m~j$: :! reI',.'nlC "c .'P, ~- !Ilu=-" ,=or =i~'~i 3=_-rovli Z.-,.~ IP.I )lllJef :. )i w~lllUrl trial one coo~ is s~e~ o-. the Man,,g:" ,E:¢ U'~ LY', C,C ~C ~J F ~, I C() LTD - MINNESOTA TOBACCO LITIGATION

THE INFLU~NCE OF BRAND IDENTIFICATION AND IMAGERY ON SUmJECTIVZ EVALUATION OF CIGARETTES REPORT NO. RD.1752-C RESTRICTED 18o7.1980 AU~IIOR: R.P. Ferrls GROUP LEADER: M. O1dman ISSUED BY: D.J. Wood DISTIIBUTION : Dro LoCoF. B1ackmmn Copy No. I, 2, 3, 4m 5, 6 Dr. ~.Wo Hughes " " 7 Dr. R.A, Sauford " " 8, 9 ~.M. Gibb, Esq. " " I0 I.S. Wade, Esq. " " 11, 12, 13 E.G. NichoiLs, Esq. " " 14, 15 Herr E. ~ctershaus" " 16 Dr. F. Seehofer " " 17 Dr. C.J.P, de Siqueira " " 18 Mr'. A.J. Kruegyusk/ " " 19 Dr. D.C. Fe~ton " " 20 L3brary " " 21, 22 t~ I Co LTD - MINNESOTA TOBACCO LITIGATION COPY .NO. wmmmmb CO C~D t~u CX~

[~.PFI RA/46J Croup Research & Development Centre, British-American Tobacco Co. Ltd., SOUTKA~TON. 18th July 1980. THE INFLUENCE OF BRAND IDENTIFICATION AND IMAGERY ON SUBJECTIVE EVALUATION OF CIGARETTES (Report No. RD.1752-C Re|tr£cted) SUMMARY AND CONCLUSIONS Zn s future where increaslnsly the product may have to sell itself through the pack, • fuller understanding of the way in which perception ef such packs affects perceptlon of their contents is desirable. In this report a technique is demonstrated which allows controlled measurement of the extent and quality of influence which brand Ideutlf£catlon and pack image have on the individual's impression of a cIgarettets smokinE characteristics. It Is shown that, for some products, such imasery powerfully and measurably modifies che percelvad smoking c, haracterlstlcs of the ciSarattes assoclated srlth it. Further research is intended to determine both underlylns bases of pack /maKe Influence (e.K. colour, pattern, etc.) and levels of responsiveness within the consumer population to the influence of :lLmasery. Application of =he cechn£qua :o the screanlnE of pack dea1£n alCerua~Ives for produces Sn current develo~nc ~y nov be considered. CO CO ~ ~ I Cc LTD - MINNESOTA TOBACCO LITIGATION

w2e v INTRODUCTION The wa7 in which a product ts perceLved and ultimately received In the marketplace raZlms on two bas£c processes: (a) EVALUATION: the £nd/vidualts ab£1ity to obtain measures of that product's phTsicaZ or sensor,/ attributes (cost, size, weight, taste, mmeZ1, etc.). (b) IHpP, ZSS~ON: the individual's Interpretation of the "image" of the product (the term "/mage" usually beingrdeflned and discussed in the mmrkst£nJ; Ztterature as an abstract concept incorporating the influences of brand name, pack dsslSn, past prc~otlon and reputation). The tobacco industry has developed several means of apprals£ng and quant£fytns the avaluatlve component of both In-house and consumer product perceptions. These means are typified respectlvely by panel testing procedures of the WOODROSE and DELTA type, and by a range o£ consumer product tests. 1c is accepted practice co carry out product evaluations in "de- Identified" format; Chat is, using blank vh£te packs and cigarettes on • d~ich brand Identification markings have been obliterated or "m&sked" by a variety of techniques (usually overprinting or tapLn8) vhtch dl£fer in complexity accordtn8 to the type and distribution o£ ldenti£ters present on the ciBaretteo Such precautions are deemed nacessa~ In order that the potenti&1 effect o~ brand laage factors should not influence the results of the smokins evaluation. ~, l Co, LTD - MINNESOTA TOBACCO LITIGATION m O CO

-3- V¸ Brand de-ldentlflcaclon procedures art gensrally costly and tlme ~;:onsum~nS, confronting the product tester with three serious problems. ~£rst, for any siren set of cigarettes under test, since it is desirable :hac the maskln8 format 18 common to s11, the amount and complex£ty of ~ask£ng used is dlccated by the worst case (1,e. the c16arette wlth most Ldentlflcatlon mmrklnl). Second, the physical appearance of cigarettes ~hich have been masked is altered and it is posslble chac this sort ot man£pulatlon affects the taster's vlew of the product. Third, and most importantly, It Is likely that current mask£nK practice (p&rtlcularly =ape wh£ch can obllterate perforations end adversely affect paper porosity) alters the sensory smoking characteristics of c£8arettes. 1C should there£ore be understood thac the precautions which the Lndustry takes ~n order Co control the potentially confounding influences rOf brand tdentt~lcation/tmagery fro~ the sampling situation, themselves generate var$abtlttles which are uncontrolled. It t8 the purpose og this report to demonstrate that the variability ~£ch brand Identification/imagery contributes co the sensory evaluation situation may in fact be quantifiable, and the extent of this variable influence sensibly related to the sensory parameters on which a given product may be evaluated. lu short, the report will show how hitherto poorly controlled or $snored Influences bearing on product perception (imagery variables) ~ighc usefully be Incorporated into the integrated emplrlcal approach to praduct development, providing a ~urther stage of refinement in the development of s product's market aim. ~ r Co LTD - MINNESOTA TOBACCO LITIGATION O

V BACKGROUND The research literature dealing with the Influence of brand imagery ~)n product parcapt$on £s domLnated by hypotheses generated from psychology theory, particularly those dealing with the concept of "self" (1). In terms of this theory, an Individual's behav£our is largely a function of his self image, such that he wtll tend to behave in such a ~mnner as to maintain and enhance his view o£ h£msel£. As a resultp ~:onsumer preferences should tend to be determined, noc so much by the l)roduct's objective properties, huc by the relevance o£ these propercles ~or the indlvldual's attempt to ma£ncaln and enhance self image. Thls should mean thac products which complement the £ndlvidual's self image should be preferred to Chose whlch are seen Co fall Co complement :~t. Hence IndlvlduaLs should prefer those brands In a glven product :etegory which have images stmliar to thelr self image r~ther than those ,4hlch have d~ssln~LLar Images. (Th£s predlct£on would be equally supported ~y cognltlve d£ssonance theory which suggests that when an indlvldual ~econ~s aware that hls expectations differ from hls percepC£on of reallty, d£ssonance occurs and the indlvldual will seek to reduce his d£ssonance ~)y allgnlng his perceptions to his current choice or by seeking a new .~iternatlve.) These theoretical predictions have been tested, notably by Marclneau ;~2), Wescfall (3), end Evans (4), who each conflrm chic, faced vlCh brand ~iternaclves in a 81van product caceKory, consmners tend co prefer the )ne whose lmaKe £s vest similar to their self £mage. Evans, for example, ~ound thac he could dlscrlminate owners of Ford and Chevrolet cars to a • taCist£cal~y stgntf£cant degree with the use of personality variables, ,~I ~,, Co LTD - MINNESOTA TOBACCO LITIGATION

-5- demographic varlables, or both in comblnatlon. Similariy, Reid end Buchanan (5), and Woodslde (6), for cat foods and beer brands respectively, find that consumers attribute different persomalicy characteristics to r.~urchasers of different brands. ~nterestlnSly, cigarettes are a favoured subject Eor the study of brand imagery, largely because the user is free to purchase in response to .Lndividual needs without serious complications due to different prices, availability, or trade-offs among Joint users or u~as. Fry (7), for exmnple, finds brand preference for cigarettes on the Canadlan market eo be related eo chelr perceived Image (deeermlned by non- :~ecric scaling to fall into two groups, one of which yes seen to be ~Llder, more feminine and more elegant timn the other). Again, the Lnfluence of the Inmsery was laportantly related ~o a number of personallty ~=.e~sures including social confidence, mascullnlty/femininlCy and need for .change, these being further moderated by socioeconomic varlables. In the same vein, Vitz and Johnston (8), investigating the reXatlonshlp between mascullnity of smokers as measured by personality inventory and rated masculinity of brand images, found for both male and female groups a low buc etatlstlcally signlflcant positive correlation between their masculinity and the rated masculinity of their regular brand cigarette. Feiedman and Dipple (9) have tested Chic relationship empirically in a ~udy designed to determine the influence of brand name on consumer ~rod~ct evaluation. In this case ali subjects smoked identical cigarettes. ~owever, one group of subjects knew the cigarettes by the name of +'Frontiersman", a masculine name, while the other group knew the cigarette as "April", 8 feminine name. Hale and fenmle subjects were asked to rate ~ ~ ICe LTD - MINNESOTA TOBACCO LITIGATION O C~

-6- the cigarette on seven sensory attribute measures and the results showed that women gave a more positive evaluation to the cigarette purportedly r.~amed "April") vhile men gave s more positive evaluation to the identical ~isarette when they thought it was named "Frontiersman". In addition, women reacted more stronsly than d£d men to brand name Ln£1uence. Orpen and Low (I0) have united the type of research conducted into car and cigarette imagery by Evans (4) and Vltz and Johnston (8) ~eapectively, in a s£nllm experiment vh£ch euKKests that the e££ects o£ brand imaKar7, and in particular, the way in which the consmnmr's psychological dlspos£clon relates to that imagery, may be generallsed across differing product categories. In this case, correlations were computed between subject's self rating of masculinity, an inventory measure o£ mascullnlty and the rated • asculintty of image of the s~lhJect's most liked and disliked cigarettes and motor cars. The correlation between the self rating and the inventory =ensure of masculinity (r - 0.73) was slgnlficant at the 0.01 level, end as predicted the correlat£ons between both measures and the rated ~asculinity of their most liked cigarette were also significant at the 0.0i level. The am relatlonshlp held for most favoured motor cars. Zn addition, the mean mascullnlty rating of the most llked motor cars was significantly higher (p < 0.01) than Chat of the most disliked motor cars, and the name relationship held for the cigarettes. Thls t adicates that the cars and cigarettes with more masculine images were preferred to those with more femtnlne images despite the fact that the sample consisted of an equal number of males and females Cthls flnd£ng relates yell to the fact that sales of Harlboro, a clsarette vith a ~Lstin~t£vely masc,,T£ne £mage, are supported by a large female contingent).._. ~ I Ce LTD - MINNESOTA TOBACCO LITIGATION

-7- Within CR&DC two reported studies (11, 12) have examined discriminations mad• by tan •ub~ects of fifty cigarette packs from ehs Malaysian market, iollowing a pilot study of the discriminations of one indtvidal on (;~stemalan packs. The data in chls came was analysed by ~ltl-d£menslorml st•ling from whlch • fro-dimensional model of discrimination was suggested, each axis being t•nc•cively labelled on the basis of subjective end c,bJective characteristics. Axis £ appears to refer to "low colour contrast high colour contrast" and "old fashioned - modern" and axis II might be labelled "small pack slze - large pack slze", "cheap - expensive" and "low class - hlgh class". With th• introduction of repertory grld research to GR&DC. imagery investigation was extended to three areas: (a) BRAND I}LqGE: the mental associations stlmul•ted solely by the pack appearance. (b) INFERRED SMOKING CHARACTER~STZCS: the type of smoking experienc• which m/ght be anticipated (in the absence of sampling) from • cigarette contained in such a pack, (c) SHOKZR ATTRIBUTION: the anc£cipated personality/character type of the sort of person who mlght be expected to regularly smoke cls•rettes contained In such a pack. This report (13) demonstrated that a pack not only Eeneraces powerful tndepmndent ImaBes, but also provides important and predlctable cues or s~sestlons about the type of smoke whlchmay be expected from a cigarette contained In such • peck, and even the type of person ~,n might typically s~oke such cigarettes. The power o£ brand image in inferring smoking ~ar•cteristlcs is clnarly demonstrated in the case of Marlboro, a brand .alcha strong prominent image which almost invariably suggests, even to S t-~ ~ 1 Ce LTD - MINNESOTA TOBACCO LITIGATION O C~ C~O

-8- those vho aeeually smoke she brand, chat the clgaretee is fully mlddle tar in strensth, hence belying the face that the product is lo~a~Lddle tar. From this position it is a short seep co the understanding that brand imaSe factors will sIKnifieantly modify the indlvldual's perception of the actual smokinS attributes o£ his cigarette, supporting tentative research by DeHare (14) indlutinE that differences in objective parameter :atluas of "mildness". "dlstlnct taste", etc. depended upon whether consumers were JudglnK on the basis of their ~mprese~on of the clsarette (a) unsmoked but with brand lease prominent, (b) as smoked and accompa~Led by brand 1maKe, or (c) as smoked £n the absence of any cues reZatlng to brand. Tl~e possibility of measuring the influence which brand image has in modlfyinE the Indlvidua1's perception of the "objective" sensory attributes of his ciEsrette is of sreat importance, particularly if the influence can be quantified with sufflclent precision co allo~ prediction of the extent and type of modification from particular aspects of the brand image. For example, it is already anecdotal knowledse that bright red on a pack connoees a stroker product, but It would be considerably more useful it if were possible to empirically relate such a thing as area of red on the pack to the amount by ~h£ch the consumer's evaluation of Chat c£Karette's H£mpact", say, is £ncrmased, If the InfXuenee of brand image could be rendered quantlflable and predlctably z~slated to modi£1catlons of the perceived sensory attributes of clgarettes, wm m48ht profitably add "image enalneering" to the series of processes by whleh a cigarette is designed and developed as a total produce. Imagery actshe then be Mde~mystlfled" and made to work in the ~ a, ICc, LTD - MINNESOTA TOBACCO LITIGATION