First-Time Purchases Can Predict Loyalty

When prospecting for new customers, retail managers obviously want people who will keep coming back. It’s a gamble, but retailers can increase their odds by looking more at what first-time customers buy than at how many items they buy or how much they spend, according to research by Duncan Simester, a professor at MIT’s Sloan School of Management.

Simester’s research found that first-time customers who buy what he calls “harbinger products” are much less likely to return to the store in the future. For example, if a new customer visits a bookstore for the first time and only buys a calendar, this person is less likely to become a long-term customer compared to a shopper who buys a novel. Similarly, a customer who visits a home improvement warehouse and only buys a diet soda is unlikely to become a repeat customer.

One possible explanation for the harbinger product phenomenon might be customers’ underlying preferences. “Some people have preferences that are aligned with a store, and others do not. An initial purchase of a non-core item, such as a calendar from a bookstore or a can of soda from a hardware store, may signal that a customer has a weak preference for the retailer’s core offering,” Simester says.

Additionally, if harbinger products lead to negative experiences, this may reduce the likelihood that a first-time customer will come back to the store, he posits.

Based on his findings, Simester advises retailers to target first-time customers “with programs designed to reduce the probability of attrition. Most obviously, do not highlight or promote these [harbinger] products when targeting prospective customers.”

Instead, “focus on the products that loyal customers purchased on their first visits and avoid harbinger products,” he adds.

Retailers also often rent mailing lists to build up their databases of prospective customers, and when doing so, they “should avoid customers who purchased harbinger products at other retailers,” Simester also suggests.

Simester’s study, conducted with fellow MIT researchers Eric Anderson, Chaoqun Chen, and Caiyun Liu, looked at randomized customer data from a U.S. mass merchandise retailer, a U.S. private-label clothing retailer, a, U.S. convenience store chain, and a European retailer of books and music.