Uncovering Consumer Preferences And Predicting Consumer Behavior With New Marketing Research Method

UNIVERSITY PARK, PA (June 14, 2004)—Researchers at Penn State's Smeal College of Business have designed a more effective method for uncovering consumer preferences that could dramatically improve a range of marketing activities including the success rate of new product development initiatives. Currently, new consumer products fail between 90 and 95 percent of the time, siphoning millions of dollars from companies in the process.

Traditional marketing research often relies on a process known as conjoint analysis in which participants are paid to answer questions about hypothetical purchase decisions. Such studies struggle to uncover true consumer preferences because participants have little stake in the answers they give. Building on the Noble Prize winning work of Vernon Smith, assistant professors Min Ding, Rajdeep Grewal, and John Liechty developed a new approach that provides study participants with incentives based on actual behaviors. Their paper, "Incentive-Aligned Conjoint Analysis," is forthcoming in the Journal of Marketing Research .

"The most relevant condition for conjoint analysis is salience, which requires that the reward be directly related to the decision the subject makes during a study," the authors write. "Paying a respondent a fixed amount is not salient, because there is no relationship between the respondent's performance/actions and the reward he or she receives. As a result, there is no reason to expect that the respondent's behavior during a study will be consistent with his or her behavior during a similar, real-world, economic activity."

Ding, Grewal, and Liechty conducted field experiments in a Chinese restaurant to test their incentive-aligned approach. Participants stated their meal preferences and eventually had to pay for and eat the preferred meal. Using the traditional, hypothetical conjoint approach they were able to predict consumer’s top choice only 26 percent of the time. In contrast, using the incentive-aligned approach, they were able to predict consumer’s top choice 48 percent of the time.

"We conjecture that a majority of new products fail because their sponsors do not understand the true preferences of their target market," the authors write in the executive summary to the paper. "In addition, pricing, placement, and advertising decisions that are based on a poor understanding of consumer preferences will be haphazard at best. We anticipate that incentive-aligned conjoint analysis will lead to significantly improved preference estimates and better marketing decisions."

About the Authors

Min Ding’s research interests are focused on marketing practices in life-science related industries and on the interaction of psychology and economics. His recent work has dealt with new drug development and portfolio management strategies, pre-launch forecasting models, hypothetical bias studies, and reverse auctions. He holds a Ph.D. in marketing from the University of Pennsylvania and a Ph.D. in molecular, cellular, and developmental biology from Ohio State.

Rajdeep Grewal’s research interests focus on modeling of strategic marketing issues. He has expertise in inter-firm relationships, product marketing strategies, and marketing management. Grewal holds a Ph.D. in marketing from the University of Cincinnati.

John Liechty’s research interests are focused on Bayesian statistical methods that can be used to gain new insights into problems in business and marketing. He is working on improving existing methods for estimating individual level demand/utility functions and is also interested in modeling involuntary eye-movement data and in data-mining problems. Liechty holds a Ph.D. in mathematical statistics from Cambridge University.