Unexpected metrics changes occur frequently for me. These changes can be attributed to factors like new feature launches, bugs introduced in new releases, or simply noise. Sometimes I leave the metrics as they are, hoping they will return to normal the next day or week. However, there are many times when I have to invest hours, if not days, in digging deeper into the issue.

While modern BI tools are good at showing what has changed and how a metric was changed, they fail to provide insight into the "why" behind the change. As a result, I have to run numerous ad-hoc queries to determine the root cause.

To address this challenge, we have developed DSensei, an open-source tool that quickly identifies root causes and helps organizations uncover critical drivers and segments that may otherwise be overlooked.

By conducting holistic analysis across all possible segments within one minute, DSensei saves the data team/product owner hours to days in performing root cause analysis. A similar tool is being used at Meta to diagnose key metrics changes, such as revenue and DAU. Now, DSensei is making it available for all businesses.

We would appreciate your thoughts on this!