I exploit a platform policy change at Amazon that removed the requirement to write a text review in order to rate a product. This effectively created a substitute feedback system to the previous regime of only allowing text reviews. To study this, I formulate a simple theory to explain consumer behavior in selecting between two alternative feedback systems. My complimentary empirical analysis, using a number of methods, finds that the policy change caused a 7% reduction in review volume and increased the objectivity of comments by 4%.