Deliberately introducing obstacles to locating discounted items in online stores can improve retailers’ margins, according to the authors of a Journal of Marketing Research paper
In “The Impact of Increasing Search Frictions on Online Shopping Behavior: Evidence from a Field Experiment,” co-authors Donald Ngwe, Kris Johnson Ferreira and Thales Teixeira explore how adding search frictions to online stores can benefit sellers. For their study, the researchers partnered with an online retailer and introduced search frictions by removing navigation links to discounted sections of the website. Prices or product availability were left untouched.
“Our results challenge the conventional wisdom that online retailers should eliminate all frictions,” says Ngwe, an economist at Microsoft. “In our setting, the additional frictions require no additional cost to the firm and produce large gains in gross margins. Without these frictions, the firm is effectively granting discounts to high-value consumers who are willing to buy at full price.” The treatment is similar to placing sales racks at the back of the store, Ngwe says, which forces shoppers to walk past the rest of the store’s goods before reaching the discounted items.
The total number of online shoppers that fell into the main experiment within a two-week period.
The number of versions of the online store that shoppers were randomly assigned to as part of the experiment. Each version had an added search friction: fewer product sorting options, a missing link to the sale section or missing percentage discount information. Shoppers bought more items at higher selling prices or lower discounts in each.
Percentage point increase in full-priced items sold by the retailer in the treatment conditions. The researchers found that the increase in average selling prices came mostly from consumers who bought full-priced items rather than discounted items.
Percentage point improvement in gross margin for each product sold with increased search frictions.
The increase in conversion rates in the treatment arms. “This is the most surprising outcome of our experimentation,” Ngwe says. “We anticipated that, as a result of adding search frictions, per-unit margins may increase at the expense of lower conversion, or total number of items sold. Instead, we found that consumers bought more items in addition to buying them at higher prices.”
Highest increase in basket size in the treatment arms. Not only did the conversion rate and average selling price increase, but—conditional on purchase—shoppers also checked out with higher basket sizes.
Additional seconds spent shopping by consumers in the treatment conditions. Because consumers spent more time searching in the treatment arms, they also considered more items—which in turn resulted in higher conversion rates.
Effect of the treatment on the most price-sensitive consumers. “We found that the least price-sensitive consumers, who had the highest baseline likelihood of buying full-priced items, were the most impacted by our treatments,” Ngwe says. “On the other hand, the most price-sensitive consumers did not change their purchase behavior.”