In early 2012, a story broke about a man who had come into a Target store in Minneapolis to complain about a book of coupons sent to his home. He was in fact quite incensed about these coupons, which had been addressed to his daughter, a high school student at the time. And although it might seem like an odd reaction to a potential money-saving opportunity, learning that the coupons were exclusively for products such as prenatal vitamins, diapers, baby formula, cribs, and so on might change your view.

The manager, upon hearing the complaint, apologized profusely. He felt bad enough, in fact, that he called several days later to follow up and explain how this could have happened. But before the manager was able to even begin his apology, the father began to apologize to the manager. As it turned out, his daughter was in fact pregnant and her shopping habits had given her away.

The algorithm that gave her away was likely based—at least in part—on one type of algorithm used in recommendation engines called collaborative filtering.