Abstract
This paper develops a dynamic model of price competition in online marketplaces
with technological asymmetry and platform-controlled product visibility. The key fea-
ture is a platform-designed ranking algorithm that determines which product appears
at the top of the page. I show that when ranking depends partly on past sales, which
the platform views as a signal of strong past demand and continued consumer appeal,
and when there exists a large enough share of attention-limited consumers, the strategic
seller adopts an invest–then–harvest strategy: it cuts price initially to build sales and
secure a past-sales-dependent ranking advantage, then raises price in the subsequent
period to extract higher margins from attention-limited consumers. As a result, equi-
librium prices exceed the benchmark Bertrand level on average. An infinite-horizon
extension shows that the same mechanism can recur over time, generating intertempo-
ral price fluctuations. These findings imply that platform design has important welfare
consequences: interface designs that make cross-product comparison easier, together
with ranking rules that place greater weight on price, can reduce supracompetitive
pricing and price fluctuations, thereby improving consumer welfare.