A 30% change in a marketplace’s suggested price produced an average 5% change in sellers’ listing prices in the same direction, according to a field experiment conducted with Mercari.
The research, published in the October issue of the Journal of Marketing Research, found that lower suggested prices increased the likelihood of a sale and seller revenue in the setting examined.
Jessica Fong and Puneet Manchanda of the University of Michigan and Yu Song of the Chinese University of Hong Kong studied a decision faced by many online sellers: what to charge when the platform offers a recommendation but the seller has the final say.
The results are specific to Mercari and the experiment. They do not establish that lowering prices increases revenue for every seller, category or platform.
How a suggestion affects a listing
Mercari is a marketplace where individuals can list goods for others to buy. The researchers varied whether sellers received a suggested price and, when a suggestion appeared, the amount shown.
A 30% change in the recommendation was associated with a 5% change in the chosen listing price in the same direction. Those figures describe percentage changes, not a five-percentage-point change in the share of items sold.
The recommendation was more influential when pricing was difficult, including for newer sellers and used goods. A used item can be harder to price because condition, age and available alternatives differ from one listing to another.
Sellers did not simply copy the suggested amount. The researchers describe the system as a way to offer guidance while allowing sellers to use information of their own, such as what they know about an item’s condition.
Revenue is not the same as price per sale
A higher asking price can increase the amount received on a completed sale, but it can also reduce the chance of finding a buyer. In this experiment, lower suggested prices improved both the chance of sale and the resulting seller revenue.
That result depends on the balance between price and demand in the tested marketplace. It does not mean sellers should always accept the lowest recommendation. Fees, shipping, acquisition costs and returns also affect what a seller keeps.
A later experiment addressed whether the results would extend when recommendations were applied more broadly across the marketplace. The authors report evidence consistent with their main findings, while the underlying company data are not public because of a data-use agreement.
Our earlier coverage of regional video game pricing examined another platform where sellers receive price guidance but make their own decisions. That research described listed prices across markets; it did not test the Mercari recommendation.
What sellers and platforms can measure
A seller evaluating a suggested price can compare similar completed sales, adjust for condition and calculate the amount left after platform fees and delivery costs. Listing prices alone show what sellers ask, not necessarily what buyers pay.
Platforms testing recommendations need to track sale rates, seller proceeds and whether sellers continue listing over time. A recommendation that produces a quick sale may be less attractive if it consistently leaves a seller with too little after costs.
The Mercari study provides experimental evidence that the number displayed beside a listing decision can influence the price a seller chooses. How much that influence helps either side depends on the quality of the recommendation and the market in which it is used.
