Systems and methods of the present disclosure are directed generally to evaluating performance of a content group of a content provider. In some implementations, a system receives a request to display a performance score and accesses a data structure storing keywords, a quality metric for each keyword, and an impression count for each keyword. The quality metric can include a quality score or a first page bid of a keyword. The system can identify keywords of the data structure corresponding to the content group of the content provider. The system can obtain the quality metric and the number of impressions via the data structure for each keyword. The system can determine, for the content group, a performance score using an impression-based weighted average of the quality metric and impression count of each of the keywords.