Title: Social Influence in Online Reviews: Evidence from the Steam Store

Abstract: How good are reviews as signals of product quality for consumers? Using a data-set derived from the popular Steam gaming platform I investigate the ‘priming’ of quality judgements as based on pre-existing consumer assessments. A policy reform on Steam in 2019 changed the average level of exposure to previous consumer quality ratings, with this randomly occurring within a game and reviewer’s life cycle. I find that removing the exposure of a reviewer to a product’s average rating leads to a 35% drop in the dependency of their review on such a rating. This is not driven by selection effects, and is robust to a wide range of alternate specifications and measures. The effect is heavily asymmetric: negativity compounds to inflate the gap between poorly-rated and well rated games. This is driven by users who are less experienced both within and across games. Finally, using estimates of owner data, I run a simple structural model of game choice based on rating. A 1% increase to product rating is equivalent to a 2.5 dollar sale price reduction, suggesting this effect has large implications for buyers and sellers.