Unit 7 Case Study

THE SOCIETY FOR CASE RESEARCH

TARGET: THE CHALLENGE OF DATA MINING

Steven M. Cox, McColl School of Business, Queens University of Charlotte Melinda Harper, Queens University of Charlotte

This critical incident was prepared by the authors and is intended to be used as a basis for class discussion. The views presented here are those of the authors based on their professional judgment and do not necessarily reflect the views of the Society for Case Research. The names of individuals, the firm, and its location have not been disguised. Copyright © 2013 by the Society for Case Research and the authors. No part of this work may be reproduced or used in any form or by any means without the written permission of the Society for Case Research.

Andrew Pole began working for Target as a statistician in 2002. His job was to improve the effectiveness of Target’s promotions by statistically analyzing information on customer purchasing patterns and demographic characteristics. The technique was often referred to as data mining. One of his tasks was predicting whether a woman was pregnant through her purchasing patterns and demographic profile. Marketing could then target these women with information on products for prenatal care and infant needs. All was going well until an irate father asked to see the manager in a Target store outside of Minneapolis. Waiving a group of coupons for baby clothes and cribs that had been sent to his high school age daughter, the father fumed, “Are you trying to encourage her to get pregnant?”(Hill, 2012). Was Target’s utilization of customer data in this way ethical, were mistakes more damaging than the value of successes, and was this an invasion of privacy? These were questions that Target management had to resolve.

Target

From a single store in Roseville, MN, Target grew to over 1763 Target and Super Target stores by 2011. Target was the second largest discount retailer in the US and saw both sales and net earnings grow between 2006 and 2011. (Target, 2012). Target catered to a similar "money- saving" market as Wal-Mart, but offered a very different value proposition. Target focused on different capabilities and a different product portfolio, including:

· Target's “way to play” emphasized design-forward apparel and home decor for image-conscious consumers. Store layout and advertising focused on an eye for style.

· Its capabilities system supported this way to play with image advertising, "mass prestige" sourcing (with the use of private brand and exclusive offerings), pricing, and the management of urban locations.

· Target satisfied the needs of its younger, image-conscious shoppers by stocking more furniture, clothing and exclusive designer merchandise than Wal-Mart.

Gregg Steinhafel, Target’s president, boasted to investors that “heightened focus on items and categories that appeal to specific guest segments such as mom and baby” (Target, 2012) were responsible for these successes. The segment focus relied heavily upon the ability to determine

Journal of Critical Incidents, Volume 6

42

customer purchasing patterns through prior purchase behaviors and other demographic data. Target was one of the first major retailers to use predictive modeling (sophisticated data mining techniques) to identify customer segments and differentially market to those segments.

Consumer Behavior: Why Does Data Mining Work

It has long been a working assumption in psychology that one of the tendencies of human behavior was habituation (Crossley, 2001). One of the founding fathers of psychology, William James, described habit as “sequences of behaviors, usually simple….that have become virtually automatic” (James, 1890). With automaticity at its core, habituation was ideal in creating repetition of useful behaviors that ultimately required less mental exertion or effort to maintain. In fact, James suggested “the more of the details of our daily life we can hand over to….automatism, the more our mind will be set free…” (p.122). Habits were acquired through the gradual strengthening of a learned association between a situation (cue) and a routine action in a consistent context. In the formation of a habit, the control of the behavior transfer to cues in the environment. This transfer of behavioral control to the environmental cues increase the automaticity with which the behavior was performed when the situation was encountered again (Verplanken, 2006; Wood & Neal, 2007). From the behavioral perspective, habit strength was considered to be a function of repetition only when rewards were received for performing the behavior upon encountering a cue (Hull, 1943; 1951). Identifying the three-part process (cue, routine, reward) of the shopping habits of consumers allowed for retailers to market and exploit the habitual purchasing behavior of its consumers, all seemingly without the consumers’ knowledge. Data mining was ideal for retailers to utilize for identifying not only patterns of predicted behavior (based on data collected from purchasing history), but as the Target example illustrates, data mining was also useful for identifying potential disruptions in habitual behavior that would allow retailers the opportunity to possibly influence future purchasing habits.

Target and Data Mining

Target regularly collected data on every customer who made a purchase at a Target store. Each customer was assigned a unique identification number, or Guest ID number. Target also acquired various demographic and geographic data on each customer. With this information, Target was on the cutting edge of using the data to increase sales (Duhigg, 2012). Target then hired economists and statisticians like Andrew Pole to identify patterns in the data that were predictive of emerging purchasing opportunities. These analysts were part of Target’s Guest Marketing Analytics department. The theory behind Target’s Guest Marketing Analytics, often referred to as predictive modeling, was that humans are creatures of habit (Crossley, 2001). Our behaviors were learned and if satisfying, repeated. Similar to learning theory, Target’s Guest Marketing Analytics was based on the theory that most products purchased were routine rebuys and little or no individual analysis went into the brand choice once the product had satisfied the need. This habitual product purchase behavior often had other product correlates. An analysis of aspirin purchases might reveal that a person who purchased baby aspirin might also be likely to purchase children’s toys. If a Target Guest purchased baby aspirin from Target, but not toys, it might be advantageous to send that person a toy coupon since they may not have thought about purchasing toys at Target. A major problem in thinking about only the product to product correlations was that there were other factors such as purchase location habits that entered into purchase decisions. As a result, a customer might have habitually purchased baby aspirin at Target but purchased toys at Toys R Us even though Target sold toys. The theory of habitual

45

purchases suggested that if the customer was satisfied with the toy shopping experience at Toys R Us, then Target’s task of changing the toy shopping behavior would be very difficult.

According to Pole,

“There are, however, some brief periods in a person’s life when old routines fall apart and buying habits are suddenly in flux. One of those moments-the moment, really-is right around the birth of a child, when parents are exhausted and overwhelmed and their shopping patterns and brand loyalties are up for grabs”(Duhigg, 2012).

Since birth records were public, if Target was to maximize its profit potential then it needed an edge, knowledge of a pregnancy rather than a birth. With this information Target could begin the promotional process before other retailers. This was the task for which Target’s Guest Marketing Analytics team was hired. They developed predictive models that would identify pregnant women before other retailers by using data in the Guest ID profiles. With this information, product information could be sent to potential customers before competitors even knew that the individuals were pregnant.

Target’s Decision

The question for Target’s management was where should the line be drawn? Did Target have the right to use information collected during shopping visits to uncover personal information about customers? Was the collection, analysis, and use of customer data ethical when trying to increase sales? What are the ethical boundaries of privacy?

References

Crossley, N. (2001). The social body: Habit, identity, and desire. London, UK: Sage. Duhigg, C. (2012, February 16). How companies learn your secrets. New York Times, Retrieved