COMMENTS OF THE WARTON SCHOOL CONCERNING CONSUMER ON-LINE PRIVACY-P954807
Consumer Privacy 1997 -- Request to Participate, P954807
From: Eric J. Johnson
I would like to participate in Session Two of the Public Workshop on Consumer Information Privacy. I am Director of the Wharton Forum on Electronic Commerce, a partnership of the Wharton School of the University of Pennsylvania and a number of firms to conduct research in this emerging arena. My comments are attached, and I look forward to hearing from you.
2.1 The Wharton Forum on Electronic Commerce is currently conducting a study examining the usage of click stream data. While we are still gathering data, primarily through the use of expert interviews, our initial results suggest that click stream analysis, at this point in time, is not widespread. A more complete set of results and analysis will be available by the time of the workshop in June.
2.2 Despite its apparently infrequent use, I am convinced that individual level click stream data could be of great benefit to both firms and customers. I base this upon both the opinions of our experts, and my own experience in the analysis of click stream data in understanding the decisions of consumers, a research project that has been underway for the last 15 years. A couple of important facts: Preference and looking time are highly correlated: In computer mediated environments, we see that the options that receive the most attention are clearly the most likely to be chosen. This raises the possibility of customizing the virtual store to the needs of the consumer. On the positive side, a consumer might find a set of options which reflect the options they would most prefer, limiting search costs. Another application would be adapting the marketing mix to each consumer. Another fact: Information acquisition is highly related to one's knowledge of the product class: By click stream data alone, a firm could separate the 'expert' consumer from the less knowledgeable.
The risks of this collection are without the mechanisms that the commission has discussed are substantial, however. Based an research on societal risks, the collection of all personal information could be seen as a substantial problem for consumers. Much of the profile of the fear of this information maps into the literature on the psychology of risk perception. Risks which cause widespread societal concern are perceived as having several characteristics in common: Based on the work of Paul Slovic, Sarah Liechtenstein and Baruch Fischhoff the first group of these characteristics are labeled Risks that are dread, the second risks which are unknown. In the first group, the characteristics of involuntariness, uncontrolability, and personal exposure, seem quite relevant to assessing the public's' perception of such data. The second factor consists of perceptions of observability, knowledge to those exposed, delayed effects, and newness, all seem to have great relevance. In discussing this work, it seems clear that the perceived risk to customers, and the subsequent and real risk to firms caused by a suspicious public, is very real.
Further these principles suggest some additional considerations to consumer choice, the basis of prior commission efforts in this area. In addition to education, which is a major result of the last workshop, we will suggest that the identity of the default option might play a major role in consumer responsiveness to choice.