This paper studies a model of search engine competition with endogenous obfuscation. Platforms may differ in the quality of their search algorithms. I study the impact of this heterogeneity in consumer surplus, seller profits and platform revenue. I show that the dominant platform will typically induce higher prices but that consumers may benefit from asymmetries. I also show that enabling sellers to price-discriminate across platforms is pro-competitive. I then embed the static model in a dynamic setup, whereby past market shares lead to a better search algorithm. The dynamic consideration is pro-competitive but initial asymmetries are persistent.