We use aggregated, non-personal search data to improve search algorithms, for spell correction or search suggestions purposes, and for aggregated statistical analysis. In other words, we do save certain data on searches, but never in a personally identifiable way.