What Target's Pregnancy Prediction Teaches Us About Health Advertising

In 2003, a man walked into a Target store in a Minneapolis suburb.

He was angry.

He had brought with him coupons for baby clothes and cribs. His high school–aged daughter had received them in the mail.

He demanded to speak to a manager. "Are you trying to encourage her to get pregnant?"

Target apologized. A few days later, the manager called again to apologize.

This time, it was the father who apologized.

It turned out his daughter was pregnant. And Target knew before he did.

The company wasn't using health records. It was using purchase data.

• Unscented lotion
• Vitamins
• Cotton balls
• Hand sanitizer

Patterns emerged. Target had built a model that assigned a "pregnancy score" based on shopping behavior. They sent her coupons based on the prediction.

The full story was reported in the New York Times.

This happened more than a decade ago, but the lesson is more relevant now than ever.

Because this is still how many health audiences are built.

• Predictive scores
• Lookalike models
• Inferences based on behavior

The problem is not the coupons. It is the assumption.

When advertisers rely on inferences about someone's health, they create risk. All it takes is one person to ask the question:

"Why was I targeted with this ad?"

The Target story is a reminder that data-driven predictions can feel invasive, even when technically accurate. For healthcare marketers, the risk is even greater. Regulations have evolved, and so have consumer expectations.

Predictive audiences may look efficient on a media plan. But they can also create compliance exposure and reputational damage.

The safer path is transparency: build audiences in ways that marketers can explain, and that consumers would reasonably expect.

Because if you can't answer the "why," the risk is already there.