Hi Jack, thanks. You can view the full queries we run against your data from the three dot menu on the right. It is possible that the conversion windows, or the join to the experiment exposure reduces the values by a lot. You can run the query we run manually and adjust the date windows or other aspects to see whats happening