Abstract: Systems and methods are provided that presents interesting local content to users at times when the content is most actionable and engaging without the user explicitly entering a search term or expressing intent. In one implementation, the system uses contextual signals about a user to rank local content. Signals may include, for example, who they (one or more users) are, what kinds of places they like, where they are, how familiar they are with the area, the time of day, where the user's friends have been nearby, among other information related to the one or more users. A system may be provided that uses contextual rules and machine learning to target content to users. The system learns which of these contextual signals are most important and alters its ranking function to optimize user engagement in terms of conversions.