Advances in technology have allowed firms to assemble massive databases of customer behavior. This has led to a de-emphasis of “soft” attitudinal information — such as getting data on how a customer feels about a product. Instead, proponents of the “behavioral data approach” say that customer purchase behavior includes those underlying attitudes.
So there’s been a massive shift in marketing strategy toward focusing on customer buying behavior as opposed to figuring out customer attitudes and identifying a firm’s actions that drive those attitudes. In fact, the entire debate about marketing versus return on investment reflects the sentiment that insights into attitudes are insufficient and that behavioral insight is today’s benchmark.
In my current research, I question this conventional wisdom.
Several firms such as Siebel Systems, Dunnhumby USA, Yankelovich, ZS Associates and IBM have made large-scale investments in systems that track various customer attitudes, as well as information about the interactions between customers and businesses. Customers also talk about brands in blogs and in social media, such as Facebook and Twitter. This has led to the development of a new industry that infers consumer sentiments from mining chatter on the Internet. The technology used to capture customer attitudes, either directly or through inferences from online chatter, requires significant investments from a firm, typically on the order of tens of millions of dollars.
So is it worthwhile for firms to make such investments? In my research, we aim to answer the following questions:
Does including customer attitudes improve the prediction of customer lifetime value?
How do customer attitudes affect retention and spending?
Does inclusion of attitudes direct firm’s resources to customers that were previously ignored?
Are the returns from including attitudes worth the investment required to collect such information?
The results from a three-year study I conducted with a big pharmaceutical company pertaining to sales calls directed toward the physicians and survey data about the physician attitudes are telling.
Following the industry convention, I treated physicians as customers for this research. The number of new prescriptions from a physician, or sales, reflects the customer behavior of interest, and the sales calls directed toward the physician by the pharmaceutical firm constituted marketing actions. In this study, we define customer attitudes as physician perceptions of the efficacy of the firm’s drug and their perception of the performance, credibility and knowledge of the firm’s salespeople.
I found that customers (physicians) with better attitudes toward the pharmaceutical firm were more responsive to the firm’s sales calls. Inclusion of customer attitudes substantially improves the quality of customer lifetime value predictions. In other words, firms obtain better predictions of customer’s future profit potential if they include information on customer attitudes in addition to the firm actions and past customer behavior in their marketing mix models.
Inclusion of customer attitudes allows firms to identify mid- and lower-tier customers who have the potential to grow in the future. This is not possible in the absence of the knowledge of customer attitudes. Firms have therefore always focused their resources on their current top customers, not their future stars.
A forward-looking customer targeting strategy that is enabled by the knowledge of customer attitudes is 11 percent more profitable than an alternative that ignored customer attitudes. For the firm I study, this is equivalent to an annual whopping return of more than 300 percent on the investment for collecting customer attitudes.
The bottom line — pay attention to customer attitude as well as behavior.
Rajkumar Venkatesan is Bank of America Research Professor of Business Administration at the University of Virginia Darden School of Business. He co-teaches the Darden Executive Education program “Marketing Analytics: Effective Resource Allocation” Dec. 10-12 in the Washington area.