Event based analytics database synchronization is described. In one example, an event tracking engine can sequence through batches of records in a transactional database that is being synchronized to an analytics database. The event tracking engine can determine a last version attribute associated with the batch, and query the transactional database to identify any records for the batch having a more recent version attribute than the last version attribute. When any records in the transactional database having a more recent version attribute are identified, the event tracking engine finds the data values in the record are different than those in the analytics database. The data values can be normalized, for example, before being forwarded to an analytics system for incorporation into the a analytics database. Over time, changes in the transactional database are identified and populated or mirrored in the analytics database.