NCDM Speaker: Attitudinal Data Can Humanize Data MiningKISSIMMEE, FL -- Though the process can be difficult, incorporating attitudinal data into data mining can make the decision process more human-focused rather than action-focused, according to Jacob Zahavi, professor at the University of Pennsylvania's Wharton School.
Zahavi was yesterday's keynote speaker at the Winter 2004 National Center for Database Marketing Conference here.
Customers' purchase decisions are affected strongly by their attitudes and perceptions as well as by the way they feel and think, he said.
"Only the combination of all three data dimensions -- demographic data, behavioral data and attitudinal data -- provide a full picture of the customer," Zahavi said.
As an example, he said, "demographic data may indicate that a person can afford to buy a Porsche car, but only very few people own a Porsche car, and the reasons are due to differences in motivations and attitudes."
Combining attitudinal and demographic/behavioral data into a data warehouse can be challenging, Zahavi said, because traditionally, "hard data, or behavioral/demographic data, belong to the realm of data mining while soft data, or attitudinal data, sit in the realm of CRM and market research. They are not connected."
Zahavi said the most effective way to solicit attitudinal data is by getting it directly from customers through surveys and questionnaires. However, issues surround this method, such as:
· How does a marketer design and validate a questionnaire to get the best information?
· How can attitudinal data be used to make predictions?
· How can the economic value of attitudinal data be quantified?
"No universal procedure exists for humanizing data," he said. "It all depends on the application, objectives and availability of quality data." But if you do "attitudinal data right, there is a huge potential [for your company] because it all boils down to understanding the customer better, and if you understand the customer better, you can better communicate with the customer."