FindTheBest’s smart rating has always been one of the things that users always wonder as to how it was rated and what are it’s constituents. So, we worked on messaging this to the user in a more digestible manner. My work involved fully back end changes to consistently calculate/save every attribute used for rating calculation of each listing/comp. This involved quite a bit changes from saving them in Mongo Db and then retrieving it for easier access to the front end layer which creates the pie chart visible to the user.
User tests indicated that the feature helped understand ratings better and in a more detail.