Feedback loops are self-regulating mechanisms that include the practice of trying to anticipate—and outsmart—society's immediate rewards, a new study from the University of Michigan found. When data collection systems go too far, feedback loops can be self-sustaining, said senior author John McCrindle, assistant professor of marketing at the Ross School of Business.

The results of the study, which will be published in the March issue of the Journal of Marketing Research, were based on examining sales data in what was billed as the largest randomized impact analysis study of chain drugstores' loyalty programs. As part of the study, pharmacists read rewards email notifications so the participants could write down the responses they already expected. The reactions varied from cynicism and frustration to exaggerated enthusiasm and delight.

Participants were presented with a rotating bottle of a product named "Lil' Ramen" and told that after three months, it would disappear. It never did, but the participants came up with razors, shoes, clothes, and much more.

Although the findings were interesting, McCrindle cautioned that the effectiveness of feedback loops has been largely overlooked. He and his colleagues, Jennifer Tinnis and Paul Van Horn, used chain drugstores, such as Duane Reade, CVS and Rite Aid, to investigate the phenomenon.

Rewards, explained McCrindle, can be linked to safety, comfort, perceived utility, or knowledge. Information can be sought out and consumed. If the information is not sought out, it will never be consumed. The Internet is chock full of instant information, but so much of that instant information has the effect of not accessing the information.

"That's one of the reasons why people post often on Twitter, to get messages," said McCrindle. "But if the content or the channel is passive, then that does nothing." That's what happened when participants saw the email notifications about the disappearing bottles.

The response elicited from the participants was not instantaneous. The best survey responses came from those who anticipated immediate rewards and wrote their responses immediately.

The researchers suspect this is because constant interruption, constant feedback, and "steer" by the data system, blur information—or "focus"—the brain receives from receiving the messages. The brain in these situations starts to anticipate what the rewards will be, said McCrindle. The paradox is that one of the program's objectives is to give the product back to the consumer who didn't see them.

This can happen more quickly than users would expect. For example, the researchers in one study said, only six people in a large group responded positively to the invitation to write a reply to the emails: The other 180 felt that they'd received a message, but hadn't been able to write anything.

Another experiment found that, after being sent the emails, about 20 percent of participants who planned to give it a try went to the company's website. Only half completed the shopping experience.

Sociologist and former Cornell University researcher Robert Sutton has written frequently about feedback loops and how they can lead to overconsumption. He has said the feedback loop system is a product of ubiquitous sensing: the Internet, the mobile phone, the vehicle, the device, the accelerator, the television, and a variety of other physical or psychological ways that "zoom in and out on the same person."

In 2000, Sutton wrote in the Yale Law Review: "Efficient products can tell us just where to turn to get what we want. Overconsumption is the killer app."

Before real-world feedback loops can become self-sustaining, something has to happen on the other side, which could be social groups, said McCrindle. "This is all to do with influencing social groups" and how those groups interact.

But there may be other ways in which people can influence feedback loops. "Many of the approaches [in McCrindle's lab] take place within a context," said McCrindle. "'Push' approaches aren't that effective in a very social setting," said McCrindle.

That's how one of the experiments went. Spooky.

Read: A new supercomputer, The Bore, could revolutionize businesses.