RxforMarketing Research

A diagnosis of and prescriptions for the recoveryof an ailing discipline in the business world

By Vijay Mahajan and Jerry Wind

It's easy to see why some people thinkmarketing research is dead. Considercompanies such as Sony, Chrysler, andCompaq, wbicb successfully "ignored tbecustomer" to create tbe Walkman''̂ '̂ ',minivan, and PC network servers. Others

listened to customers and created such flops asNew Coke (Coca-Cola Co.) and McLean burgers(McDonald's Corp.). Radical innovations such asinteractive television seem to defy all marketingtools to truly understand what services customerswant. Sony Corp.'s Akio Morita has commented:"The public does not know what is possible to do.but we do. So instead of doing a lot of marketresearch, we refine our thinking on a product andits use and try to create a market for it." Or, in thewords of minivan creator Hal Sperlich, as report-ed by Fortune magazine, "In 10 years of develop-ing the minivan, we never once got a letter froma housewife asking us to invent one."

Indeed, a study we conducted earlier thisdecade found a very low use of quantitativemarketing research techniques in new productdevelopment. The Marketing Science Institute-.sponsored study, which sought to assess the useof marketing research methods and models byFortune 500 companies, also found tbat many

E x e c u t i v e S u m m a r yAt a time when market research is ailing—yet more needed thanever—a prescription for its recovery is just what the doctor ordered.The authors examine some of the discipline's recurring problems,such as companies' failure to use market research to understandcustomers' unarticulated needs and to apply it strategically, and offersome ideas to help nurse the discipline back to health.

managers, especially top executives, gave market-ing re.search a low rating. In recent years, manycompanies have cut back on marketing research,whether eliminating the function altogether or rel-egating it to a marginal staff function with littleimportance and dwindling budgets. Managers rou-tinely outsource many of their marketing researchprojects and have tbe marketing researcbers serveas "researcb brokers."

Nevertbeless. these findings are especially sur-prising at a time of increased recognition of quali-ty as defined by the customer, rising attention tocustomer satisfaction, and increased focus on cre-ating knowledge-based organizations. But market-ing researcb, as it's currently viewed and prac-ticed, bas failed to rise to these opportunities. Forthis discipline to begin to address the currentneeds of our organizations, market researcbersneed to stop conducting researcb as usual andbegin to develop new and richer models for thecreation of marketing knowledge. We've pro-posed some prescriptions to help market researchto a full recovery.

MARKETING RESEARCH 7

FULL RECOVERY OPPORTUNITY

Market research is not dead; in fact, it's notfundamentally flawed. The discipline has beenmisapplied, but most of all, its potential has notbeen fulfilled. In rapidly changing global markets,its role is more important tban ever. In an agewhen knowledge is critical to success, marketingresearch can be a knowledge-creation engine.Marketing research can provide key insights onhow to achieve profitable growth. But to realizethe potential of marketing research, marketresearchers need to change the way they use it inthe following ways:

• Focus on diagnosing problems.

• Use new information tecbnology (IT) toincrease speed and efficiency.

• Take an integrative approach.

• Expand its strategic impact.

By following these guidelines, marketresearchers will change the content and nature ofmarketing research—so much so, that the next-generation marketing research perhaps should noteven be called "marketing research." Marketingresearchers have to reinvent the whole architec-ture for effective marketing research by reexamin-ing its culture, values, principles strategies, archi-tecture, and processes. In a knowledge organiza-tion, this function should be the eyes, ears, and akey part of the intelligence of tbe organization, hsbould belp make sense of the avalanche ofimpressions pouring in through the many humanand electronic portals of tbe firm.

Technology specialists have done a great jobexpanding the quantity and quality of tbe dataavailable, but even the best data warehouse falls farshort of providing true insights into the market andthe effective marketing strategies needed to suc-ceed. The strength of marketing research is makingsense of quantitative and qualitative inputs fromcustomers. With the right approach, it could play acentral role in tuming the overflowing informationof our organizations into marketing knowledge.

Focus ON DIAGNOSIS

To start moving the discipline forward, marketresearchers first need to stop using it only to testsolutions (specific products or services), andinstead should use it to diagnose the market.Consider this scenario using the doctor-patientanalogy: Patients don't typically have a specificsolution to their problem in mind. Instead, thepatients might tell the doctor they have an upset

stomach, trouble sleeping, and aches. (Witbincreased use of direct-to-consumer dmg advertis-ing, the patient may ask for a specific medicine tosolve that problem, yet it is up to the expert physi-cian to diagnose the problem and assure that theselected treatment is most appropriate.) Similarly,customers didn't ask for a Walkman or minivanbefore these products' invention, but they did helpidentify the needs for pttrtable entertaintnent and ofadditional space in family vehicles.

No doctor would propose identifying diagno-sis as worthless because the patients didn't comeup with the cure. But in marketing research, thattype of scenario is occurring. Although marketingresearchers cannot expect customers to identifyproduct or service solutions, they can use marketresearch as an interactive process of diagnosisand experimentation.

In some situations, particularly those in whichinnovations are continuous and uncertainty is mini-mal, companies can ask customers about the incre-mental changes they want in their product offer-ings. Such situations include those in which onemight use over-the-counter drugs. A patient typi-cally wouldn't visit a doctor for treatment of a mildheadache. Tbat's because solutions to tbe problemare fairly straightforward, and such a scenario doesnot warrant the added expense of a diagnostician.

On the other hand, companies need to lead themarket in situations in which there are discontinu-ous innovations and market uncertainty, lliis issimilar to the way doctors or architects "lead'"their clients through a process of understandingtheir needs by asking questions about their experi-ences. This is particularly important when:

1. Customer needs are not articulated

2. Customer preferences are not developed

3. Customers need to leam about unfamiliarproducts

4. The technology potential is unknown or evolv-ing over time

5. Regulations and other factors are uncertain orevolving

In tbis environment, customer responses toproducts and services may not be very useful. Ifcustomers are asked whether they would useinteractive television for education or to watchmovies, they might say education. But when theyactually have the boxes in their homes, they migbtbe just as likely to choose a movie over educationas they are to choose a Big Mac over a McLeansandwich. (For other examples, see the adjacent

8 Fall 1999

sidebar entitled "Understanding DiscontinuousOpportunities.") But marketing research can helpdiscover unarticulated needs and find ways tomeet them. For example, eustomer surveys maynot reveal what the demand tor interactive televi-sion will he. but they can identify unmet needs ofcustomers (such as the desire to find more infor-mation or purchase products).

As pointed out in Fortune, marketing researchfocuses (in current markets and articulated needs.(See Additional Readings on page 13.) But thereare ways to extend marketing research intounserved markets and unarticulated needs withoutasking customers lo come up with the nextWalkman. In fact. Morita's keen informal obser-vation of American youth led to the developmentof the Walkman, and Chrysler's understanding ofbaby boomers' unartieulated need for more spaceled to the ereation of the minivan.

The basis of an effective marketing diagnosis isan ongoing, iterative process, similar to the way adoctor makes a diagnosis. The tlrst diagnosis,made in real time, leads to a series of experiments.The doctor prescribes a treatment and, after tryingone treatment for a few weeks, the patient reportsbaek. The doctor assesses the effectiveness of thetreatment. In contrast, many marketing researchprojects are viewed as a one-shol event. At thesame time the experimental diagnoses are pro-gressing, the doetor is educating the patient aboutthe symptoms to pay attention to and the possibletreatments. This interactive process leads to thedevelopment of the best treatment.

Using a similar approach, marketing researcheould help customers articulate their needs and prob-lems, or understand and experience a new productand service solution. Effective marketing researchuses sampling, seiirehing, and experience to engagein this type of iterative, experimental priKess.

IT HELPS SPEED, EFFICIENCY

Speed is vital, particularly in new productlaunches. A study of new product development inthe Japanese automobile industry by HarvardBusiness School's Kim Clark found that for a earthat sells for $10,000, eaeh day of delay in marketintroduction costs a carmaker more than $1 mil-lion in lost profits.

A weakness of traditional marketing research isthe trade-off between moving quickly and con-ducting thorough research. To move quickly,companies often cut comers on their marketingresearch or abandon it completely.

Advances in IT offer unique opportunities toincrease the speed of marketing research and alsoputs an unprecedented amount of data in thehands of researchers. But they need to be prepared

i

UNDERSTANDING DISCONTINUOUSOPPORTUNITIES

AT&T videophones failed; Domino's delivered

Discontinuous innovations require a major change in consumer behavior. And

while AT&T Corp. failed to meet this challenge with its videophone products,

Domino's Pizza Inc. succeeded, with its pizza-delivery service in Japan.

AT&T's videophone has been on hold for

decades, despite high hopes for a consumer

market. In 1970 when the New York-based

company launched the Picturephone—an

early version of the modern videophone—it

predicted it would generate $1 billion in busi-

ness in a decade, which never materialized.

The videophone returned in 1992, with the

company predicting that millions of videophones would be in use by 1994.

However, by 1994 only 30.000 of them were sold worldwide, mostly outside the

U.S., according to Across the Board magazine.

The questions AT&T should have asked are; What market need does this prod-

uct fulfill? Do people at home really want to be seen? Will people pay extra to have

the whole world see them come to the phone in their bathrobes? The videophone

product required a discontinuous change in consumers' behavior—without a mean-

ingful customer benefit and at an increased cost. The product was a wonderful

solution to a problem that didn't exist.

When Ann Arbor, Mich.-based Domino's Pizza introduced its pizza delivery in

Japan, a major change in Japanese consumers' behavior was needed as well. Yet

Domino's managed to rise to the challenge successfully, unlike AT&T with its video-

phone, if Domino's had merely tested the acceptability of the service it delivered in

other parts of the world, it never would have entered Japan. Japanese consumers

typically don't eat tomato-based food, and Asians tend to have allergies to milk

products. Home delivery was not widely accepted, housewives were reluctant to

give up cooking, houses were small, and finding customers in the labyrinthine

streets of Tokyo seemed impossible. A market for pizza didn't exist, nor did any

sign of hope in creating one.

Instead of trying to sell its existing product and service to the market, Domino's

used its information about customers to design a new product and service offering

for Japan. It introduced toppings such as fish and sushi. To sustain its 30-mlnute

delivery. Domino's developed a complex address database and small scooters to

navigate the narrow streets in Tokyo. Through this interactive process of diagnosis,

this pizza-delivery service that no one asked for became a big hit In Japan.

MARKETING RESEARCH 9

to deal with the technology and information effec-tively. Such companies as Dallas-based TexasInstruments Inc. are experimenting wilh Internet-based marketing research, which allows almostinstantaneous feedback from actual consumers.White Plains. N.Y.-based Moskowitz Jacobs Inc.can provide a report based on information fromconsumer taste panels around the world within 72hours. Future innovations could further acceleratethis process, including the development of a'Vending machine" approach that allows man-agers to develop their own research projectsonline and send them out electronically to panelsaround the world.

Technology also can be used to give subjects adirect experience of a virtual product, avoidingthe costly, time-consuming, and competitivelyrevealing process of using real-life test markets.The "Information Acceleration" process, devel-oped by Massachusetts Institute of Technology'sGlen Urban and his colleagues, uses multimediastimuli to forecast consumer response to new-to-the-world products. These multimedia representa-tions of the product^—and the environment inwhich it is introduced to consumers—can bedeveloped earlier, quicker, and more efficientlythan physical prototypes and studies of actual testmarkets. Their tests over a five-year period—withproducts such as automobiles, medical equipment,and new cameras—showed the approach offeredinsights into the market that were borne from sub-sequent experience.

As can be seen in these examples, new tech-nology offers ways to reduce the trade-offsbetween the speed and rigor of marketingresearch. Despite the obvious advantages, a recentstudy of marketing researchers at leading U.S,product and service companies found that only31% had conducted research using the Internet.More than two-thirds never conducted an Internetstudy, and these were among the heaviest andmost sophisticated users of marketing research.The encouraging news is that 100% of those com-panies that tried an Internet study said theyplanned to conduct future projects on the Internet,according to a Data Development Corp, survey of75 heads of marketing research at top U.S. firmsin May 1998. The remaining two-thirds need toget onto the Net, and other companies that feelthey don't have time for traditional marketingresearch should carefully examine the potential ofthese new approaches to deliver meaningfulresults more quickly.

TAKING AN INTEGRATIVE APPROACH

Marketing research, like many other manage-ment disciplines, has created its own silos.

Marketing researchers, by artificially distinguisb-ing among types of marketing research approach-es (such as qualitative vs. quantitative) and in sep-arating themselves from other information, miss ahuge opportunity to produce more effectiveresults and have a bigger impact on the firm.Among these divided processes that need to bebrought together are:

1. Qualitative and quantitative: Many firmsmake a distinction between qualitativeresearch (such as focus groups) or quantita-tive research (large-scale empirical studies).Some companies use qualitative studies ini-tially to help form questions and structurefollow-up quantitative studies, but manyfirms use either qualitative research alone orrely only on overly structured quantitativestudies. Market researchers need to breakthese silos to help integrate quantitative andqualitative research throughout ihe process(i.e,, incorporate open-ended, qualitativequestions into large-scale studies and analyzethem in a quantitative fashion). Many mar-keting research firms hesitate to include qual-itative questions in large-scale studiesbecause they find it more difficult to analyzethem. Firms usually use such questions inonly qualitative studies. With a little moreeffort, however, these open-ended questionscan be used in quantitative studies.

2. Marketing research and modeling: Modelingoften is conducted separately from marketingresearch. But it would be far more effective todesign integrated projects based on both appro-priate research and modeling. For example,advanced work in conjoint analysis integratesthe results of the marketing research phasewith simulations and optimization models.

3. Marketing research and decision supportsystems: Decision support systems aren'toften linked to marketing research activities.Marketing research should not be seen as adiscrete evcnl but rather a continuous processthat provides important marketing informationto decision support systems. The research hasto be updated constantly and immediately inte-grated so it can inform company decisions.

4. Marketing research and adaptive experi-mentation: Companies typically use market-ing research as part of discrete studies ratherthan as an integral part of a strategy of adap-tive experimentation. Aside from a few excep-tions, such as companies in direct marketingor those such as Anheuser-Busch, companies

10 Fall 1999

don't typically conduct adaptive experiments.But these experiments often provide moremarketing knowledge, insights into marketresponse to strategies and initiatives, and morestrategic tlexibility at a much lower cost than aseries of one-shot marketing research projects.Integrating the firm's adaptive experimenta-tion efforts with its research efforts bringsmuch value.

Market research and datahases: Marketingresearch usually is conducted on a sample sep-arate from the company's customer database—an approach that misses a huge opportunity touse the company's existing data on individualcustomers. Marketing researchers havedropped the ball in developing these customerdatabases, leaving data mining to the IT staff.It's time for marketing researchers to reclaimthis territory by better integrating their researchactivities with the company's database. Thiswill make the research more valuable and alsohelp to better shape the database by focusingon issues of importance to customers.

Marketing research and other informationsources: Companies have many sources ofinformation outside marketing research thatmanagement can use to improve their decisions.This infonnation that is collected separatelyfrom the marketing research function mayinclude customer complaints, customer satisfac-tion studies, product and service quality studies,financial data, customer transaction data, infor-mation on competitive behavior, security ana-lysts studies, and extemal databases. Marketing

research can help integrate these diversesources of data into coherent strategic insights.

EXPANDING STRATEGIC IMPACT

The biggest potential in the use of marketingresearch is not in the effective use of techniques,an integrative approach, or in the use of newinformation technology. Rather, it's in helpingmanagement ask the right strategic questions bygiving marketing research a more central place inshaping corporate strategy. Marketing researchneeds to be recognized as part of the organiza-tion's knowledge-creation process. In a dramati-cally changing global environment, business lead-ers need constant contact with the market to makethe best decisions. Marketing research finns oftencollect data in standardized or simplified formsrather than constructing more strategic measures.For example, companies that measure marketshare often neglect to ask the deeper questionssuch as; What is the market? Should the companylook at global market share, and if so, should itfocus on dollars or units, and at what exchangerates? Within that market, should the companylook at the share of the total market or the "shareof wallet" of its current customers? How can thecompany get more from its total spending? (Seethe example below.)

Coca-Cola, in its 1998 annual report, takes amore strategic view of its marketing measure-ments. Instead of merely reporting its marketshare of the global soft drink market, it looks atthe potential of the total global beverage market,including coffee, tea, and water, and asks how itcan increase its share of this broader market. This

The myth of market share

An example of marketing information's oversimplified use is the traditional approach to market share. Market share as a bald num-ber is almost meaningless. Consider four competitors, shown below. Each has 10% market share. What strategic information deesshare give you? Almost none. Who will control the market? To give the share any meaning, the intelligence should be added to theshares by looking at the factors that shape the future share of the market. These include numbers of first-time buyers, averagenumber of units purchased, number of repeat buyers, average number of units per repeat buyer, and fetal unit sales of eachbrand. These richer numbers are not so simple to deal vyith as basis market share, but they tell you much more. Instead of shovwinghow the market is divided up today (refiecting past performance), they provide insights into vwhere it's headed. More sophisticatedmeasures offer a deeper understanding of the market.

No. of First-Time Buyers1002050

50

No. of RepeatBuyers

2010

25

Av. No. of UnitsPer Repeat Buyer-45

2

Total UnitSales of Brand 1100100100

100

TotalMarket100010001000

1000

Market Shareof Brand 110%10%10%

10%

Major StrategicImplicationsGet repeat buyersIncrease triersIncrease repeatbuyersIncrease usageby repeaters

MARKETING RESEARCH / /

The biggest potential inthe use of marketingresearch is . . . in helpingmanagement ask theright strategic questions.

is a marketing research decision with tremendousstrategic implications. The shift in measuresimplies a very different focus for the company.Instead of looking at ways to compete with Pepsi,it now asks how it can replace British "tea time"with a "Coke time" or substitute a "Coke break"for a "coffee break." The strategic use of market-ing research—asking the right questions-—leads tofundamentally important insights and directionsfor the company.

Marketing research also can be used to informglobal strategy, but less than half of researchers atmajor U.S. corporations have conducted interna-tional studies. According to the 1998 DataDevelopment survey, 57% of researchers have con-ducted no international studies at all, and most ofthose who did such studies devoted less than 20%of their budgets to them. This is a shocking result.given that amount of resources and attention majorcorporations are devoting to global markets. Ifthese companies are not conducting internationalmarketing research studies, on what basis are theydeveloping strategies for entering global markets?

In a knowledge-based company, the role ofmarketing research is to develop knowledge aboutthe market and therefore generate value for share-holders. Market research also can help a companyunderstand customers, identify new potentialopportunities, and shape the strategic direction ofthe firm.

Marketing research also could be used toinform strategic decisions such as mergers andacquisitions strategy, in which it could be used toassess the value of the acquisition, potential syn-ergies, and the strength of the customer franchisebeing acquired. Some of the more sophisticatedinvestment bankers are including, as part of theirdue diligence, discussions with customers, formercustoniers, and occasionally prospects. Yet manyof these inquiries are informal discussions ratherthan systematic marketing research.

Many other management problems can benefitfrom quantitative insights into the needs of con-sumers and other stakeholders. Increasingly, com-panies are recognizing the importance of creating

value for these stakeholders but are slower to seehow marketing research can provide the informa-tion needed to act upon these intentions. In a fast-moving, global environment, careful studies areeven more important as companies fmd them-selves in foreign markets where they have littlehistory and where managers cannot rely on theirinstincts. The companies that commit to reinvent-ing their marketing research and developing capa-bilities to implement it well will have a significantadvantage over rivals.

Marketing research can play a broader rolewithin and throughout organizations. New rela-tionships between manufacturers, suppliers, andcustomers create the need for the broader applica-tion of marketing research. With direct informa-tion about customers purchasing its products inWal-Mart stores, for instance. Procter & GambleCo. may understand the retailer's customers bet-ter than it does. This gives P&G a unique oppor-tunity to use that marketing information, but alsomeans it needs to do so to remain competitive.Similarly, Intel Corp. needs to focus not only onits direct customers such as Dell Computer Corp.but also on its customers' customers—the endusers of its computer products. More types ofcompanies need to use marketing research effec-tively to diagnose problems and needs of cus-tomers, prospects, and other participants in theglobal supply chain and to test the viability ofalternative solutions for them.

THE CHALLENGE: REBUILDING TRUST

Instead of stripping down the marketingresearch function, market researchers need to giveit a more central role by connecting it more close-ly to strategy processes and information technolo-gy initiatives. If marketing researchers and theorganizations in which they work begin to makethese changes, they can transform marketresearch's role.

If marketing researchers expect to be given amore strategic role in organizations, however,practitioners need to do more to prove themselvesworthy of that trust. Marketing researchers, muchlike physicians, need to develop an independentsystem of certification to assure clients they'requalified to do the research effectively. The cur-rent lack of certification makes it harder for thebest marketing research firms to differentiatethemselves, and it perpetuates mediocre or poorresearch, eroding the entire profession.

As part of this certification, some type of con-tinuing marketing research education is needed.With the development of new tools and technolo-gies, a person trained five or ten years ago inresearch techniques may be missing important

12 Fali 1999

innovations or insights in new tools or practice.An ongoing certificalion process wotild assureclients that marketing research firms are workingwith Ihe most current and effective approaches.

Such a certitlcation process will lead toincreased accountability for marketingresearchers. If marketing research professionalsdon't take up the certification issue on their own,the courts will. Already, a company tiled a courtcase against a marketing research firm that gross-ly overestimated the potential market tor a newproduct. If the industry does not set its own stan-dards, companies will enforce their own de factostandards throtigh the legal system.

Trust is at the heart of the doctor-patient rela-tionship, and it's also critical to successful mar-keting research. Certification, continuing devel-opment, and other initiatives on behalf of themarketing research community can rebuild trustamong customers. For marketing researchers towork with clients in developing accurate diag-noses to their challenges, the profession needs totake steps to build this trust. And with this trust,a focus on diagnosis, the use of the latest infor-mation technology, increased integration, and amore strategic focus, marketing research can rise

to its full potential in contributing to buildingvalue for a firm. •

ADDITIONAL READINGS

Hamel. Gary and C.K. Prahlad (1994), "Seeingthe Future First," Fortune, (September 5). 64-7.

Mahajan. Vijay and Jerry Wind (1992). "NewProduct Models: Practice. Shortcomings, andDesired Improvements." Journal of Producthtnovation Monagement, 9 (June), 128.

"Videophone 2" (1994), Across the Board,(October). 26.

ABOUT THE AUTHORS

Vijay Mahajan is a John P. Harbin CentennialChair in Business at the department of marketingat the University of Texas at Austin.

Jerry Wind is the Lauder Professor andprofessor of marketing at the Wharton Schoolof the University of Pennsylvania.

Strong healthy marketing efforts start with pure,wholesome data. That's why L&S TeleServices has beendelivering clean, accurate, high-integrity data for morethan a decade. Our professional research team followsstrict qtiality and data collection protocols to ensure thehigh standards you expect on projects like...

• Attitude and usage • Image and positioning• Benchmark studies • New product research• Market share • Customer satisfaction

and loyalty

TELESERVICESAmerica's Most Responsive Call Center"

MARKETING RESEARCH 13