Target's Pregnancy Prediction Algorithm Reveals Teen Pregnancy to Father Through Targeted Coupons

In early 2012, Target Corporation's sophisticated predictive analytics system, developed by statistician Andrew Pole, identified shopping patterns that could predict customer pregnancies with remarkable accuracy. The algorithm analyzed approximately 25 products that, when purchased together, indicated pregnancy likelihood, including unscented lotion, supplements, cotton balls, and hand sanitizers. The system assigned each customer a pregnancy prediction score and automatically triggered targeted marketing campaigns. The incident occurred when a teenage customer in Minneapolis began purchasing items that triggered the pregnancy prediction algorithm. Target's automated marketing system generated and mailed baby product coupons and advertisements to her family home address. The teenager's father received these materials and became angry, calling a Target store manager to complain about the company sending baby-related promotions to his high school daughter. The store manager apologized and promised to follow up. However, when the manager called the father back a few days later to apologize again, the father's tone had changed dramatically. He revealed that after confronting his daughter about the coupons, he learned that she was indeed pregnant and had not yet told the family. Target's algorithm had correctly identified her pregnancy before her own father knew. Charles Duhigg of The New York Times first reported this incident in February 2012 as part of a broader investigation into corporate data mining practices. The story highlighted how retailers use purchase data to identify major life events and target customers with relevant advertising. Target's system was so sophisticated that it could predict pregnancy during the second trimester with high accuracy, often before customers had made the news public or even confirmed it themselves. The incident exposed significant privacy concerns about predictive analytics and corporate data collection. While Target had not technically broken any laws, the case demonstrated how algorithmic predictions based on seemingly innocent purchase data could reveal highly sensitive personal information. Following the publicity, Target adjusted its marketing practices to make pregnancy-related advertising less obvious by mixing baby product promotions with other unrelated items to mask the targeting.