Kerby Anderson

Yesterday I talked about the privacy concern that results from you being required to share lots of personal information, including a photo and a video of yourself. Today I wanted to illustrate how you can lose your privacy even when you don’t share information. Twenty years ago, the Target Corporation was able to conclude that a shopper was pregnant and even estimate her due date. The story was written up in the New York Times and has become a classic example of targeted marketing. The article had the arresting title, “How Companies Learn Your Secrets.” A data scientist at Target began to analyze the massive database Target stores were amassing from the purchases of their customers. He began to mine the data and discovered a few interesting things about Target customers. He discovered, for example, that pregnant women are more likely to buy unscented lotion, and that they start doing this at the beginning of their second trimester. This correlation between pregnancy and changes in shopping behavior was one of about two-dozen data points he and the other analysts were able to identify. When they combined all these correlations together, they were able to establish a “pregnancy prediction” score. This score told them two things. First, that the female consumer was pregnant. Second, it also gave them a good estimate of her due date. The value to Target was considerable. They were able to send coupons to the woman during different times in her pregnancy so that they would arrive when she was most likely to need them. The timing of the coupons brought more pregnant women into Target stores to use them. This marketing success story illustrates how big data and sophisticated data analysis can invade your privacy without you providing any information. Whether you click on a link online or buy a product in the store, someone is watching.