When consumers have access to a large number of reviews for a product on multiple e‐tailer websites, they may rely mainly on aggregated summaries of online reviews (SORs) (e.g., percentage of reviewers recommending a product) instead of on textual content of individual reviews. The SOR type may be “High” (when a majority of reviewers recommend a product on multiple merchant websites), “Low” (when only a minority of reviewers recommend a product on multiple merchant websites), or “Mixed” (when a majority and a minority of reviewers recommend the same product across different websites). Consumers’ prior product knowledge and regulatory focus moderate the impact of SOR type on consumer decision outcomes. Across three studies (representing low, moderate, and high levels of consumer knowledge), we find that negativity bias holds across both promotion and prevention foci in the Low‐SOR condition. In the High‐SOR condition, the level of prior knowledge moderates the relationship between SOR and purchase intention. Mixed‐SOR has no effect. We also propose and find support for a framework for SORs by extending Banduraʼs observational learning theory. Finally, we discuss the theoretical and managerial implications.