Tracking Consumer Behavior through Information Organization & Data Analytics
Sat, 2012-09-22 11:02 — divya
http://www.forbes.com/sites/kashmirhill/2012/02/16/how-target-figured-out-a-teen-girl-was-pregnant-before-her-father-did/
This Forbes article outlining the data mining systems used by Target has been much talked about in the marketing world. It is closely linked with the NYT article (http://www.nytimes.com/2012/02/19/magazine/shopping-habits.html?pagewant...) that outlined the entire process (and has additional insights into analyzing behavior and cognitive science). Both these articles give us a look into how retailers can study consumption patterns and use that information to pitch products to people before they even know that they need them.
Target has a huge amount of data about each of its customers. The key to pitching their products to customers who need them lies in <b>how the data scientists at Target have organized the information in a way that allows them to study their user behavior more effectively</b>. The system used by Target assigns every customer a Guest ID number and this is tied to their personal information like name, credit card details or email address. Every action a consumer takes is linked with this unique ID including survey responses, coupons, visiting the Target website or responding to their emails. The demographic information also encompasses things like how long it takes to drive to a store, estimated salary, the whether you've moved recently etc. Target can also buy further demographic information about customers ranging from their preferred coffee brands to political leanings to where they went to school and much more. This helps them create a rich database of information about each customer and use this data to analyze and predict future purchases.
The Target data analytics team studied the information they collected about their customers and found specific insights that enabled them to calculate a 'pregnancy prediction score' for their customers. The information also allowed them to predict with 87% accuracy the due date of their customers, and they used this to send them targeted coupons.
While this entire study and prediction system is fascinating from an information science point of view, it raises many ethical questions. The Target example led to some uncomfortable situations. Target was able to predict a teenagers pregnancy before she told her parents about it, and ended up sending her coupons in advance. Similarly, the science could predict impending pregnancy but there was no way of knowing if a person ended up miscarrying. This led to Target hiding their strategy by incorporating targeted coupons along with generic ones so people could no longer spot the patterns when they received coupons in the mail.
The Forbes article is worth a read not just from an information organization perspective, but also from an information policy perspective. They outline an efficient system of data collection and organization that in some ways could infringe upon privacy.
- divya's blog
- Login to post comments
