When Old Navy launched Bodequality, an effort to introduce more-inclusive sizing, a 2021 press release trumpeted the move as “revolutionizing the shopping experience” by offering every size of every women’s style at consistent prices. But shortly after introducing extended sizing, the chain was trying to unload much of the unsold merchandise at steep discounts—while shuttering the campaign, the Wall Street Journal reported.

Chicago Booth’s John R. Birge says Old Navy’s experience demonstrates how challenging it can be for retailers to set prices optimally and avoid costly markdowns. Birge, Chinese University of Hong Kong’s Hongfan Chen (a graduate of Booth’s PhD Program), and Duke’s N. Bora Keskin offer a two-step pricing process that takes into account customer behavior, which they contend will work better.

Half of all fashion items get marked down because vendors are notoriously inaccurate at forecasting demand, according to Birge, who says that the tech industry faces the same problem. When launching new products, fashion and tech companies often use historical sales data and patterns as guides because they don’t have good information about the potential market. Then they attempt to strategically lower prices to attract different consumers, for example requiring early tech adopters to pay top dollar to have the latest gadget, then dropping the price to lure in more price-sensitive buyers.

Because “sales data are extremely limited for a new product,” the researchers write, a seller “runs the risk of marking down the product price too early in the data-driven learning process.” When that happens, the company makes less money than it could have and hurts the market value of the product—and potentially the value of similar products coming out in the future.

Markdown strategies pose a big and important enough challenge that many companies turn to business analytics software and services for help and guidance—but still come up short.

The researchers approached this problem by considering the effects of forward-thinking shoppers who wait patiently for markdowns before buying. With modeling, they studied two forms of such behavior: one where customers are essentially watching both a product’s price and taking note of its past pricing, and one where customers focus on past markdown patterns.

For the first group, where forward-looking shoppers can influence the decisions of others, a company could help itself by announcing and committing up front to a pricing policy, the researchers find. And for the second group, they find that customer memory is crucial, and bargain-hunters with long memories can be costly.

The researchers’ model prompted them to propose a “learn-and-then-earn policy” to help guide companies. The strategy features two periods for pricing. During an initial learning period—the first two or three weeks of a merchandising season—a seller can set prices that are high enough to learn consumers’ purchasing behavior while also generating revenue. Then the seller can use the observations gleaned to assess consumer demand, comparing current sales patterns with those surrounding past markdowns.

“The difference would allow them to estimate the fraction of customers who are strategically waiting for the markdown before buying,” Birge says.

He suggests that such a policy might have helped Old Navy avoid its extended-sizing debacle. By analyzing purchasing data from the first few weeks after launching the new offerings, the company might have realized more quickly that there was limited demand for the new sizes. Marking down products earlier in the season might have helped them unload the remaining stock.

Birge offers some additional advice: start small. As he notes, “It would have been better for Old Navy to place a smaller initial order and then contract for secondary suppliers in the event that the new product, or products, were a big hit.”

More from Chicago Booth Review

More from Chicago Booth

Your Privacy
We want to demonstrate our commitment to your privacy. Please review Chicago Booth's privacy notice, which provides information explaining how and why we collect particular information when you visit our website.