Buying cars: Do product features matter more than brands?

It is a popular belief among marketers that online searches for brand names are good indicators of which products consumers plan to buy. But a new study in the Journal of Marketing suggests that searches for specific features may be far more telling.

"Consumers search online for much more than brand names," write authors Rex Yuxing Du and Ye Hu (University of Houston) and Sina Damangir (San Francisco State University). "Marketers should also tap into search trends for product features, which can reveal important changes in consumer preference. This opens the door to a whole new field in consumer interest tracking."

Researchers examined car sales statistics for a six-year period and correlated them to online searches for specific features such as body type, fuel economy, and acceleration. They found that consumer searches for specific features of a product, regardless of brand name, made it possible to predict with more accuracy what product that consumer would buy in the future.

The authors believe that the value of this new study goes beyond helping companies predict future sales. Because the results of consumer searches are available almost as they happen, companies can use this up-to-date knowledge to make sure they are developing products with features consumers want. Companies can also rapidly adapt their advertisements to emphasize features as they start to trend. Finally, managers may also consider spending less on overall advertising for products that are rich in features consumers already want, because these products are likely to sell themselves.

"This is a call to action for marketers to become more skillful in using big data sources such as Google Trends to follow changing consumer tastes. This is an area of research that is bound to grow as marketers continue to explore this promising new source of marketing intelligence," the authors conclude.

###

Rex Yuxing Du, Ye Hu, and Sina Damangir. "Leveraging Trends in Online Searches for Product Features in Market Response Modeling." Forthcoming in the Journal of Marketing. For more information, contact Rex Yuxing Du at (rexdu@bauer.uh.edu).