Consumers are influenced by both online review valence and usefulness. Although some scholars have recognized the gatekeeping role of review usefulness, whereby consumers always follow “useful” reviews, this is the first study that attempts to provide a better understanding of the gatekeeping role of review valence. Drawing on loss aversion theory, we propose that review valence provides more diagnostic value than review usefulness. Its crucial role is demonstrated in that only when reviews are perceived as positive do highly useful (rather than lowly useful) reviews increase consumers' adoption of them, whereas the difference appears less evident when reviews are perceived as negative. We empirically test the model from two perspectives including product level and review level. The results of a mixed‐method approach – a large‐scale panel study (product level) followed by two experimental studies (review level) – provide empirical evidence for the above hypotheses.