Published in the Journal of the Operational Research Society, the research with Xiaoying Cheng of Zhejiang University of Finance and Economics, Yanhong Sun and Xin Yan of Shanghai University, finds that platforms face three critical choices: whether to set prices themselves or leave them to suppliers, whether to share demand forecasts, and how much to invest in services that boost demand.
Using publicly available information, the researchers built a model to explain behaviours already seen in practice. Platforms usually hold better information on demand, while suppliers know more about the quality of what they offer. This asymmetry affects whether centralised or competitive pricing works best, and whether platforms should share demand data.
Quality can differ far more in accommodation than in app-based taxi services. A taxi journey must meet basic safety and licensing standards, so the ride experience varies only marginally. Airbnb options for a single traveller could range from a room in a shared flat, to a professionally managed, high-spec studio apartment. That wider spread in quality makes competitive pricing more effective for accommodation platforms, while a single, centrally set fare often suits taxi apps.
Professor Archibald said: “We noticed that some platforms set all the prices, while others give that responsibility to their suppliers. Rather than assuming one group had it wrong, we asked whether there might be rational explanations for both approaches. Our model shows that these decisions often depend on the level of variation in quality and the type of information each side holds.”
“By understanding these trade-offs, platforms can design strategies that make economic sense, even if they appear surprising. This research can help managers and policymakers think more clearly about how pricing, information sharing and service investment interact in digital markets.”
The findings show that the degree of quality variation strongly influences outcomes. Centralised pricing works best when quality differences are slight, since a single price smooths competition and raises returns. Where variation is greater, competitive pricing enables suppliers to segment the market more effectively. The research also finds that under competitive pricing, platforms may choose not to share demand forecasts if service investment costs are high and demand variability is low.
Thomas Archibald is our Professor of Business Modelling and Director of Research.