All over the Web, ads are getting more personal. They follow you from one site to the next and know your browsing history. But are such ads really effective? The answer may not be as obvious as digital marketers assume.

“There’s been a focus on trying to identify the customers likely to purchase a product. But that’s distinct from which customers will be influenced by advertising,” says Catherine Tucker, an MIT Sloan School of Management professor who gave a keynote address at an international data mining conference in February.

In the rush to use ever-more data about people, Tucker says, ads are too often shown to those who have already decided whether to buy or not buy the product, or who have bought it already. “What we may be doing is wasting a lot of money.”

Tucker has focused her studies on the growing number of personalized ads. In an experiment with an online travel firm, for example, she saw that ads tailored to a specific browsing history were, on average, less effective than generic ads for the site when shown to people who had recently visited. “You’ve been to the website and looked at the products. There’s probably a good reason why you didn’t buy it,” she says. This is akin to an age-old marketing maxim—“timing is everything”—that Tucker says is being lost in the digital age.