The Real Lessonof New Coke:

The Value of Focus Groups for Predicting the Effects of Social InfluenceThe embarrassing failure of Coca-Cola's attempt to change the flavor of its flagship brand has hecome a textbook case of how mar-ket research can fail. The lesson usually drawn is that Coke's researchers asked respondents the wrong questions. However, a morecareful examination of the events surrounding the reformulation attempt suggests an alternative explanation: that the error result-ed from the standard market research practice of considering focus groups to be only a form of preliminary research and not appre-ciating their unique ability to predict the effects of social influence.

By Robert M. Schindler

n April 1985, the man-agement of Coca-ColaCo. announced its de-cision to change the fla-vor of the cotnpany'sflagship brand. The

events that followed from this decision, aswell as the faetors which led up to it, havebeen reviewed, discussed, and extensivelyanalyzed in the popular press, the tradepress, and in marketing textbooks.

Two books and at least two marketingcases have been written on the events sur-rounding the flavor change decision. Also,a well-known, but somewhat older Har-vard Business School marketing case dealswith some of the key events which led upto the decision. Despite the extent of thisattention, more can be learned from thisdramatic pieee of marketing history.

The Attempt to ReformulateCoca-Cola

The 87-year old rivalry between Coca-Cola, the traditional market leader,and Pepsi Cola, the perennial runner

up, took an unexpected turn in the mid-1970s. Pepsi's consumer research discov-ered in blind taste tests that a majority ofconsumers preferred the taste of Pepsiover thai of Coke. In fact, even a majorityof loyal Coke drinkers reported preferringPepsi in the tests.

Pepsi began communicating these find-ings to consumers through "Pepsi Chal-lenge" television ads .showing taste testswhere Coca-Cola drinkers expressed pref-erences for a cola which was then revealedto be Pepsi, This campaign contributed toCoca-Cola's slow, but steady decline ofmarket share in the soft-drink category.This erosion was most apparent in food-store sales, which reflect consumer prefer-ences more directly than do vending-tiiachine or fountain sales. By 1977, Pepsihad actually pulled ahead of Coke in food-store market share.

Although publicly expressing a lack ofconcern about the Pepsi Challenge adver-tising, Coca-Cola's managetnent privatelywas quite worried because blind taste testsby the company's own market researchdepartment had confirmed Pepsi's claims.

Secretly, Coke's management beganresearching the possibility of reformulat-ing Coca-Cola to respond to the apparentchanges that had occurred in consumertastes. By 1984, researchers had arrived ata new formula for Coke whieh, in blindtaste tests, beat Pepsi by as much as six toeight percentage points. In addition tobeating Pepsi, cola drinkers chose this newformula over the old Coke formula by55% to 45% in blind taste tests and loyalCoke drinkers chose it over the old Cokeformula by 53% to 47%, In taste testswhere the drinks were identified as "newCoke" and "old Coke," cola drinkers pre-

ferred the new formula over the old for-mula hy 61 % to 39%,

Well aware of the importance of thereformulation decision. Coke's manage-ment made sure that the taste test resultswere checked and coiToborated in everymajor market in the country. Overall,Coke's market researeh on the reformula-tion was one of the most exhaustive mar-ket research projects in history; It cost $4million and included interviews withalmost 200,000 consumers. After the deci-sion to reformulate Coke was made, Coca-Cola chairman, Roberto Goizueta, termedthe decision, "one of the easiest we haveever made," according to Hale N, Tongrenin his book. Cases in Consumer Behavior.

On April 23, 1985, Coke announced thereformulation with a grand flourish, slag-ing a multicity satellite press conference inNew York, Atlanta, Chicago, Houston,Los Angeles, and Toronto. The next day, afront-page article in The New York Timesreported: "The Coca-Cola Company saidyesterday that il had scrapped the formulafor the world's best-selling soft drink. Therecipe, concocted 99 years ago, has beenpiaeed in the vault at the Trust Companyof Georgia Bank, never to be used again,said Roberto C. Goizueta, chairman ofCoca-Cola. "We have a new formula forCoke," he added."

In addition to the extensive publicity.Coke announced that the new Coke wouldcome in a new ean, with updated red and

2 2 Decetnber1992

silver graphics replacing the traditional redand white look. Clearly. Coke had decidedto make sure that consumers would beaware that Coca-Cola's flavor was beingchanged.

The initial reactions of most consumersappeared to be positive. Many bottlersreported that sales of new Coke weregreater than expected and. during the firstfew weeks after the new Coke introduc-tion, the company's weekly survey of 900respondents showed consumers preferringnew Coke over old Coke by a margin of53% to 47%.

However, during this period, there wasalso intense media coverage of those con-sumers who did not like the new Coke andwere angry about the change. In a numberof cities, old Coke loyalists sponsoredprotest rallies and boycotts and receivedwidespread media attention.

By June, it was becoming apparent toCoke's management that consumer dissat-isfaction with the reformulation wasincreasing. The stream of angry letters andphone calls was becoming a flood, andweekly tracking surveys confirmed thatconsumers were becoming increasinglynegative about the change. In a surveyconducted during the first week of Julyonly 30% of consumers interviewedreported preferring the new Coke to theold.

On July 10. the company announced itsdecision to respond to public pressure andbring back the old Coke formula. It wouldbe available in the form of a product withthe name "Coca-Cola Classic," and wasintended as a flanker brand to satisfy thoseconsumers who wanted the original tasteas an option. The reformulated soft drinkwas to be known simply as "Coke" andwould remain the flagship brand.

However, sales of new Coke erodedrapidly. In August, sales of Classic beganto exceed those of new Coke, and by theend of September, Classic had a 70%share of the combined volume of the twoproducts. Over the next few months, largefountain accounts, such as McDonald's,began switching back to the old formula.

In 1986. Classic outsold new Coke bymore than 8 to 1. despite the promotion ofnew Coke with over $48 million of top-rated television advertising. AlthoughCoke's advantage over Pepsi in the sug-ared cola category had decreased slightlycompared to 1984. Coke's advantage overPepsi in the overall market had increased,mostly due to the continued success ofDiet Coke and tbe introduction of CherryCoke, In response to questions aboutwhich was now the flagship brand,

Goizueta explained that the company hadadopted a "megabrand strategy" in whichsome promotion expenditures were direct-ed at enhancing the image of the compa-ny's entire line of cola drinks.

During the following years, the marketshare of new Coke continued to decline.By 1990. new Coke's share had droppedto .6%, In April 1990, the company begantest marketing new Coke under the name•'Coke II." It was advertised as having"real cola taste" with "the sweetness ofPepsi," according to the Wall Street Jour-nal. Although Coke's management hasclaimed that these tests were successful,they appeared to have developed a consid-erable degree of caution on the question ofsupporting Coke TT with a national rollout.

marketers have usedthese events to disparage

the discipline of ,marketing research I

The Wrong-QuestionExplanation

The widespread awareness of theseevents has led many observers tosearch for an explanation. The par-

ticular question which most frequently hasarisen is why Coke's extensive marketresearch was unable to provide manage-ment with better guidance in the reformu-lation decision.

Anyone who is or has been a marketingresearch practitioner will not find it sur-prising that some marketers have usedthese events to disparage the discipline ofmarketing research. For example, in Mar-keting News, one ad-agency managerattributed Coke's mistake to "executiveswho solely rely on statistics."

When announcing the re introduction ofold Coke in July 1985, the top people atCoke suggested that research is not capa-ble of measuring the types of consumerfeelings that resulted from the attemptedreformulation. When describing the emo-tional outpouring which led them to bring

back the old formula. Coke's president.Donald Keough told Time magazine,"...you cannot measure it any more thanyou can measure love, pride, or patrio-tism,"

However, most observers did notattribute the failure of Coke's research inthis instance to an intrinsic limitation ofthe capabilities of marketing research.Rather, they judged that the research wasconducted or interpreted incorrectly.Although some have argued that Coke'sresearch error was to overgeneralize frominexact taste test results, the vast majorityof people who have publicly voiced anopinion concerning where Coke's researchefforts went wrong espouse what could becalled the "wrong-question explanation."

In this view, the reason that Coke'smarketing research did not detect the con-sumer outcry which resulted from thereformulation was that they did not makeit clear to the taste-test respondents that ifmost people chose the new Coke flavor,then the traditional Coke flavor would nolonger be available. In other words, ratherthan ask. "Which flavor do you like bet-ter," consumers should have been asked amore relevant question. This questionmight have been something like. "Howwould you feel if we discarded Coca-Cola's current taste and replaced it withthis new taste?"

In the years since the introduction ofnew Coke and the reintroduction of Coca-Cola Classic, the wrong-question explana-tion is the one that has been most oftenrepeated. It appears to have become thestandard explanation for this highly publi-cized failure of market research.

An Alternative Explanation

S ince the intense publicity has dieddown, some further details of theresearch behind the new Coke deci-

sion have come to light. In particular, it isnow known that Coca-Cola's marketresearch department did indeed ask theright question. Considerable attention wasdevoted to testing consumer reactions tothe idea of changing Coke's flavor.

These tests and their results aredescribed in Thomas Oliver's book. TheReal Coke. The Real Story: "In 1982 theCoca-Cola Company conducted 2,000interviews in 10 major markets to investi-gate further the public's willingness toaccept a different Coke,

"Consumers were shown storyboards,comic-striplike mock commercials. Oneboard said Coke had added a new ingredi-ent and it tasted smoother, while another

December 1992 2 3

said the same about Pepsi. (To compensatefor the impact of first impressions, 1,000people were shown the Coke statementfirst and the other 1,000 saw the Pepsiannouncement first.) Then Coca-Colaasked consumers a long series of questionsabout what their reactions to such achange would be. Would you be upset?Would you try the new drink? Would youswitch brands immediately?

"'We estimated from the response that10%-12%i of exclusive Coke drinkerswould be upset, and that half of thosewould get over it, but half wouldn't,' said|Roy] Stout [director of marketingresearch I.

"...While the interviews pointed to peo-ple's willingness to try a new Coke,Zyman [vice-president of marketing] andStout discovered through other tests thatmany people just didn't believe anyonecould or should tamper with the king ofthe colas. To hear debate on the issue,Coca-Cola's research department usedfocus groups, a favorite marketing tool.

"...Some of Stout's focus groups wereshown a storyboard depleting a proposedcommercial. One said Brand X soft drinkwas going to be improved. 'Fine.' thegroup replied, and they were equally san-guine about a proposed improvement inPepsi, But when it came to changing Cokefor the better, the resounding response wasNO. 'It was like saying you were going tomake the flag prettier.' said Zyman. Asimilar response came in 1983 from agroup that included some exclusive Cokedrinkers. This group agreed that An-heuser-Busch could change Budweiser butin no way should the Coca-Cola Companytry to improve Coke."

Oliver goes on to describe how theseresults led Coke managers to continue tosupport the work of their technical peopleto come up with a new flavor for Cokethat would beat Pepsi. By September1984. product research had developedsueh a flavor, and taste-testing marketresearch began moving into high gear.

What Oliver describes sounds verymuch like what would be considered stan-dard market researeh proeedure for thedevelopment of a new product or the mod-ification of an existing one. The idealwould be to begin with focus-group test-ing of the product concept. Then, a surveywould be conducted, using individualinterviews with a large representative sam-ple of consumers, to verify and quantifythe results of the focus groups. If the newproduct survived these concept tests, thenthe testing of an actual product (or proto-type) would begin.

Coke's only deviation from this standardsequence is that the quantitative survey ofindividuals appears to have been donebefore rather than after the focus groups.But this is a minor point. What we see hereare very normal market research proce-dures. Coke's researchers started out askingthe right questions, and in the right way.The results of the foeus-group phase andthe survey conflicted. The researcherstrusted the survey, whieh comprised a largenumber of interviews spread over a wide,and presumably representative, area.

- he focusgroup is a unique source of infor-mation about how the consumer

will respond in a situation wherethere will be an awareness of the

views of other consumers

A wave of focus groups usuallyinvolves less than 200 respondents and isconducted with minimal attention toobtaining a representative sample. It isstandard market research practice to trustsurvey research over focus groups, andthis praetiee appears, even to the nonre-searcher (i.e.. the executive decision-maker), to make good sense.

Moreover, Coke's research did a prettygood job of predicting consumer response,at least initially. Recall that when thereformulation was first introduced, theconsumer response was favorable. But bythe end of May 1985, it had begun tochange. It was this that Coke had notexpected.

The New York Times' report on Coke'sannouncement of the reintroduction of oldCoke began as follows: "When the Coca-Cola Company introduced a reformulatedversion of the world's best-selling softdrink on April 23, it was well aware that itmight alienate some faithful Cokedrinkers. The company, however, expect-ed that alienation to fade. It was complete-ly unprepared for how it would spread anddeepen in the two months following thedebut of the new Coke."

It is this change in eonsumer opinion,and only this change, that Coke's marketresearch had failed to predict.

With the benefit of hindsight, the locusof their research mistake becomes clear. Itwas to respond to the conflict between the

results of the focus groups and the surveyof individuals by trusting only the survey.As it turned out. one can see that both pro-cedures had provided important informa-tion. When new Coke was first introduced.people made individual decisions on it.and most at least acquieseed to the change.

This is what was predicted by the indi-vidual interviews which indicated thatonly \{)%-\2% of consumers would beupset. But over time, as the majority of thepopulation bad the opportunity to be stim-ulated by media reports and other socialinteractions with angry Coke loyalists.most changed their minds. This is whatwas predicted by the focus groups. Giventhe IO%-12% figure from the quantitativesurvey, a typical eight- to 12-memberfocus group is likely to have at least oneangry loyalist as a member. The focusgroup results showed that, in this situation,exposure to the views of angry Coke loy-alists is likely to sway the others in thegroup to their position.

By July 1985, Coke executives hadsensed that this social interaction was amajor factor in causing their problems; itwas reported in Advertising Age that Cokeofficials were blaming the press for "fan-ning public discontent." Of course, bythen it was too late. Coke had alreadyIgnored the research that told them howthe market would respond to a flavorchange carried out in a public context.

Moreover, by the summer of 1985,Pepsi had also come to appreciate the roleof publicity in causing Coke's problems,and then actually acted on that understand-ing. As David Gilman, manager of publicrelations for Pepsi Cola International,described it in Public Relations Journal:

"After new Coke was introduced herein the States, Coca-Cola said that it wasplanning to announce the new formulationoverseas. But months went by, and theynever made the announcement to the for-eign market. This didn't sit well with us,so as a public service we conducted pressconferences in 18 different countries onAug. 20. In effect, we made Coke'sannouncement for them."

In this view, the real lesson to belearned from the new Coke affair is thatthe focus group technique is more thanjust a means of getting a quick and vividlook at consumer opinion. It is a uniquesource of information about how the con-sumer will respond in a situation wherethere will be an awareness of the views ofother consumers.

Even if Coke's researchers had includ-ed in all of their individual interviews andtaste tests a question directly asking con-

2 4 December 1992 Marketing Research

sumers how they would respond if onlytbe new Coke flavor were available, it isvery unlikely that the results would havepredicted the shift in opinion which even-tually occurred. Without information onhow other consumers will respond, indi-viduals have no means of predicting howtheir own feelings will change after beingexposed to the responses of others.

Viewing the events surrounding Coke'sattempted reformulation in this light, oneis less likely to fault Coca-Cola's manage-ment. After reintroducing the old Coke,these executives were subjected to unflat-tering suggestions that they had failed toappreciate the importance of Coca-Cola inthe American psyehe or that they hadabsentmindedly neglected to ask con-sumers an obviously relevant question.

Rather, with the perspective of somedistance, it appears that they indeed appre-ciated the importance of consumerresponse to the idea of reformulating Cokeand used very acceptable market researcbprocedures to try to predict consumerreaction to this idea. Moreover, in contrastto many executives who devote resoureesto market research studies and then ignorethe results of these studies in favor of theirown personal intuition, it appears thatGoizueta and other top Coke executivesactually trusted and used their marketresearch results.

It is a flaw in accepted market researcbpractice, and in the understanding of con-sumer behavior that supports this practice,which must bear at least some of theresponsibility for the embarrassing failureof the Coca-Cola refonnulation.

Predicting the Effects ofSocial Influence

The failure of market researchers toappreciate the important and uniqueinformation which focus groups pro-

vide may be one reason behind whatEdward Tauber has argued is the relativelypoor validity of eoneept-testing research.However, the recent rise in theoreticalwork on qualitative methods has set thestage for a rethinking of the traditionalrole of focus group methods.

Consumer behaviorists have pointedout that focus groups and individual inter-views each have distinctive strengths andweaknesses, and that each is appropriatefor different researeh tasks. The opportu-nity for respondents to interact with othersmakes focus groups the best method formeasuring the effects of interpersonalinfluence.

Although the interpersonal interactionsin a focus group setting cannot be expect-ed to be identical to those that will occurunder natural conditions, they can beexpected to yield a far better indication ofthe effects of group interaction than meth-ods where consumers are interviewed indi-vidually. In individual interviews, whetherqualitative or questionnaire-based, there isno opportunity at all for individual opin-ions to be influenced by the views of othereonsumers.

This distinctive strength of the focus

-1 is theactive awareness

of the opinionof influential others

which is the key factor

group method is not important when indi-viduals can accurately anticipate the viewsof influential others and are actively con-scious of these views when they are beinginterviewed. For example, suburbanhomeowners are likely to be highly andaccurately aware of how their neighborswould react to a new line of exterior housepaints featuring iridescent colors. In sucheases, individual and group methods willyield the same results and the distinctivestrength of focus groups is not important.

However, when individually inter-viewed consumers either cannot or do notaccurately anticipate the views of others,then the ability of a focus group to bringtbese views to awareness becomes animportant advantage.

It is the active awareness of the opinionof influential others which is the key fac-tor here. Consider the case of medical doc-tors being interviewed by a drug companyconcerning a new product that dispensedtwo types of birth-control pills whichneeded to be taken together.- When inter-viewed individually, physicians indicateda generally positive reaction to the newproduet. When interviewed in focu,sgroups, physicians initially expressed pos-itive reactions.

However, at some point in the typicalgroup, one physician would express a verynegative view to the effect that he or sheperceived this new product as nothingmore than a repackaging of two existing

products and felt that it was devious andinsulting to try to pass it off as a newproduct. This invariably led to a change inthe views of the other group members, andthe consensus of most of the focus groupswas negative. Note that the differencebetween the individual and group resultsin this ease was due to the ability of thefocus group to bring to awareness a view-point which was probably not even con-sidered during most of the individualinterviews.

Apparently, such also was the case withnew Coke. When interviewed individuallyabout the proposed refonnulation, the vastmajority of the respondents were not eveneon.scious that some others would expressfeelings of loss and anger. However, theability of the group setting to make themconscious of such views as they were for-mulating their opinions made it possiblefor them to be influenced by this opinion.During the introduction of new Coke, itwas largely the media rather than personalinteractions which caused this conscious-ness, but the effect of this awareness on theopinions of most consumers was the same.

The Validity of FocusGroup Results

This new appreciation of the distinc-tive capability of the focus groupmethod leads to the following practi-

cal advice. In new-products research,rather than have focus groups be the pre-liminary research and the individual inter-views of a quantitative survey be the con-firming research, conduct preliminaryresearch using both individual and groupmethods. If the two forms of qualitativeresearch agree, then the confirmatoryresearch ean proceed using the traditionalsurvey methods.

However, if the results of the prelimi-nary individual interviews and focusgroups diverge, then the researcher needsto consider how much awareness of theviews of influential others will exist in themarketing situation the research is beingused to predict.

The researcher should carry out confir-matory research using a procedure thatprovides tbe respondent with a level ofawareness of tbe views of others compara-ble to that which is expected to oecur dur-ing tbe rollout of the marketing program inquestion. If the researcher expects thatvery few consumers will be aware of thereactions of other consumers to the intro-duction of the new or altered product, thenconfirmatory research using individual

December 1992 2 5

interviews should yield the most accurateprediction.

But if the researcher expects that manyconsumers will be aware of the reactionsof others, then the focus group results aremore likely to be correct and should beconfirmed with a research technique thatgives respondents a realistic awareness ofthe views of other consumers.

At least three factors can be used toestimate the level of avi'areness of theopinions of others:

•The visibility of product. Productsthat are highly visible, such as automo-biles, beer, clothing, and magazines dis-played on the living-room coffee table, arevery likely to be associated with highawareness of the views of other con-sumers.

•The importance of the product.Consumers will tend to seek out tbe opin-ions of influential others when makingdecisions about risky and otherwiseimportant products such as medical ser-vices, day-care centers, and banking ser-vices.

• T h e difficulty of the decisionregarding the product. Even if the prod-uct is not especially important to con-sumers, they will tend to seek out theopinions of others if they regard the prod-uct as complex (e.g., personal computers,audio equipment) or if it is a service that isdifficult to evaluate before purchase.

For products deemed high in any ofthese three factors, the results of groupinterview methods are more likely to beaccurate than those of individual interviewmethods in cases when the results of thetwo research methods conflict.

For the introduction of new or alteredproducts that are not highly visible orimportant and do not involve a difficultdecision, the marketer can often exertsome control over the degree of the con-sumer's awareness of the opinions of oth-ers. This makes it possible for the mar-keter to resolve conflicting results of indi-vidual and group research by designingtbe marketing program to create the condi-tions which produced the more favorableresult.

For example, the drug company thattested the new birth-control product wasadvised to inform physicians about theproduct through its detailers so that eachphysician could form an opinion on anindividual basis, as opposed to launchingthe produet at a iiu-ge conference or othergroup setting.

Further, if the marketing program

involves the alteration of an existing prod-uct, then even if the product is associatedwith high visibility, importance, or diffi-culty of decision, there are many caseswhere management can choose to roll outthe program with little or no consumerawareness.

Coke's reformulation was almost cer-tainly one of those eases. A gradual phase-in of the new flavor with no announce-ment or other changes would very likelyhave gone unnoticed. Pepsi had been qui-etly reformulated from time to time, andsome believe that Coke had. in f'act,secretly modified its formula .several timesduring its first 99 years of existence. Con-sider also that researcher David Piercefound that, even after all the flavor-change

onsiderationof the effects of socialinfluence must becomea standard part of the

new-productdevelopment process

publicity, consumers rated the new Cokeflavor labeled as old Coke to be moredesirable than the old Coke flavor labeledas new Coke,

Even if it was later revealed—by Pepsi.for example—that Coke had secretlyphased in a reformulation, it is unlikelythat there would have been any seriousconsumer reaction. Consumers' awarenessthat they had already been drinking thenew formula with no decrease in theirenjoyment would probably have made itsomewhat difficult for them to summonthe indignation necessary to voice publiccomplaints.

Conclusions

Tbe Coke reformulation attempt was adramatic example bow consumerawareness of the reactions of other

consumers can play a critical role in thesuccess or failure of a new or altered prod-uct. It highlights the necessity for someexplicit investigation of social-interactioneffects during concept-testing research.

At the very least, exploratory researchshould routinely be carried out using indi-vidual as well as group methods. If theresults of the two methods disagree, then

careful attention should be given to thesocial context of the consumer decisionsthat will determine the product's successor failure.

Although focus groups are the bestresearch method for revealing criticalsocial interaction effects, there is a needfor methods which make possible the con-firmation of the insights of focus groupresearch. One approach would be thedevelopment of procedures for conductinggroup research on a large representativesample of consumers.

For example, such "confirmatory focusgroups" might be larger and less homoge-neous than traditional exploratory focusgroups, and recruitment could be guidedby expectations concerning the types ofopinions consumers are likely to becomeaware of after the product is introduced.

A second approach to the developmentof research techniques to verify focusgroup results would be to experiment withindividual interview procedures that makeconsumers aware of the full range of view-points expressed during the exploratoryfocus group pha.se and which are likely toeome into eonsumer awareness after theproduct's appearance in the market.

The real lesson of the failed new Cokeintroduction is that consideration of theeffects of social influence must become astandard part of the new-product develop-ment process. This can best be done byappreciating that focus groups can revealimportant insights about social interactioneffects that other qualitative methods arelikely to miss, and by developing newresearch procedures that make possible thereliable and representative quantificationof these insights. | ^

Footnotes'Atihough both the locus groups atid the survey

provided indications that there would be consumerdissatisfaction, the survey results indicated that thisdissatisfaction would be limited to a smalt segmentof the market; the Twus groups suggested the dissat-isfaction would be widespread.

-Case conlributed by Edith Krieger of Psychono-metrics Inc., Southampton. Pa.

About the AuthorRobert M. Schindler isAssociate Professor of Mar-keting a( Rutgers University,Camden, N.J.

2 6 December 1992

ReferencesAllvine. Fred C. (1987), Marketing: Principles and Practices. New York:

Harcourt Brace Jovanovich.

Alsop, Ron (1985), "Coke's Flip-Flop Underscores Risks of Consumer TasteTests," Wall Street JournaHiuly 18), 23.

(1987), "New Coke Is a Smash Success With Consumers in ThisPoll," Wall Street Journal {Feb. 26), 23.

Alter, Stewart (1985) "Coke Hurting Research Image," Advertising Age (July22), 58.

Anderson, W. Thomas Jr. and Linda L. Golden (1984), "Bank PromotionStrategy," .lournal of Advertising Research, 24 (April/May) 53-65.

Bearden, William O. and Michael J. Etzel (1982), "Reference Group Inllu-ence on Product and Brand Purchase Decisions," Journal of ConsumerResearch, 9 (September), 183-194.

Boure, Hrancis S. (1957). "Group Influence in Marketing and Public Rela-tions," in Some Applications of Behavioral Research, Rensis Likeri andSamuel P. Hayes Jr., eds. Paris: The UNESCO Press.

Clark, John B. (1987), Marketing Today: Successes. Failures, and Turn-arounds. Englewood Cliffs, NJ: Prentice-Hall.

Crawford. C. Merle (1987), New Products Management, 2nd ed. Homewood,IL: lrwin.

Engel, James F., Roger D. Black well, and Paul W. Miniard (1990), ConsumerBehavior, 6th ed. Hinsdale, IL: The Dryden Press.

. . and Robert J. Kegerreis (1969), "How Information isUsed to Adopt an Innovation." Journal of Advertising Research, 9(December), 3-8.

Enrico. Roger and Jesse Kornbluth (1986). The Other Guy Blinked. NewYork: Bantam Books.

Feldman, Sidney P. and Merlin C. Spencer (1965), "Ihe Effect of PersonalInfluence in the Selection of Consumer Services," in Proceedings of iheFall Conference of the Ameriean Marketing Association, Peter D. Ben-nett, ed. Chicago: American Marketing Association.

Fisher, Anne B. (1985), "Coke's Brand-Loyalty Lesson," Fortune, i 12 (Aug.5). 44-46.

Gelb, Betsy D. and Gabriel M. Gelb (1986), "New Coke's Fizzle—Lessonsfor the Rest of Us," Sloan Management Review (Fall), 71-76.

Giges, Nancy (1985), "Coke's Switch a Classic," Advertising Age, 56 (July15), 1,82.

___^__ (1986), "Adman of ihc Year: Coca-Cola's Roberto Goizueta Engi-neers Startling Comeback," Advertising Age, 64 (Dec. 29), 1,26-27.

Greenwald, John (1985). "Coca-Cola's Big Fizzle," Time (July 22), 48-52.

Harvard Business School (1978), "Pepsi-Cola (A)," Case # 9-579-108.

Hendon, Donald W. (1989), Cta.ssic Failures in Product Marketing. NewYork: Quorum Books.

Hollie, Pamela G. (1985), "Coca-Cola Changes Its Secret Formula, In Use far99 Years," New York Times (April 24), 1.

HonomichI, Jack (1985), "Missing Ingredients in "New' Coke's Research."AdvfiUsing Age (July 22), 1,58.

Kotler, Philip (1988), Marketing Management: Analysis, Planning, imple-mentation, and Control, 6th cd. Englewood Cliffs, NJ: Prentice-Hall.

Louis, J. C. and Harvey Z. Yazijian (1980), The Cola Wars. New York: Ever-est House.

Martin, Stephen H. (1987), "Marketing Can ^Make A Good Life Better' If ItIs Properly Practiced," Marketing News, 21 (April 10), 2.

McCarthy. Michael J. (1990), "Coke II Survives One Test City, Heads to Sec-ond," Wall Street Journal {Oct. 5) BLIO.

__̂ (1991), "Coca Cola is Facing New Pepsi Challenge: AvoidingSigns of Age," Wall Street Journal <Ocl. 1), Al,14.

McQuarrie, Edward F. and Shelby H. Mclntyre (1990), "What the GroupInterview Can Contribute to Researeh on Consumer Phenomenology,"Research in Consumer Behavior, 4, 165-94.

Oliver, Thomas (1986), The Real Coke. The Real Story. New York: RandomHouse.

Pierce, W. David (1987), "Which Coke Is It? Social Influence in the Market-place," Psychological Reports, 60, 279-86.

Policano, Christopher (1985), -Dueling Colas," Public Relations Journal(November), 16-17.

Piiltz, Kim (1985), "Wi/.ards of Marketing," Newsweek (July 22), 42-44.

Quickel, Stephen W. (1989), "Hard Lessons of the Eighties," Business Month(June), 62-68.

Ringold, Debra J. (1988), "Consumer Response to Product Withdrawal: TheReformulation of Coca-Cola." Psychology and Marketing, 5 (Fall), 189-210.

Robertson, Thomas S., Joati Zielinski. and Scott Ward (1984), ConsumerBehavior. Glenview, IL: Scolt, Foresman and Company.

Schiffmati, Leon G., Joseph F. Dash, and William R. Dillon (1975), "Interper-sonal Communication: An Opinion Leadership/Opinion Seeking Compos-ite Approach," in Combined Proceedings, Edward M. Mazze, ed. Chica-go: American Marketing Association.

Scredon, Scott and Marc Frons (1985), "Coke's Man on the Spot,'" BusinessWeek (My 29). 5(}-6].

Stevenson, Richard W. (1985). "The Revival of the 'Old' Coke," New YorkTimes (July 12), Section 4, 1,39.

Tauber, Edward M. (1975), "V '̂hy Concept and Product Tests Fail to PredictNew Product Results." ./wwrau/ of Marketing, 39 (October), 69-71.

Tongren, Hale N. (1987), Ca.ses in Consumer Behavior. Englewood Cliffs,NJ: Prentice-Hall.

Urban, Glen L. and John R. Hauser (1980), Design and Marketing of NewProducts. Englewood Cliffs. NJ: Prentice-Hall.

December 1992 2 7