Sigma has a beta feature for Metrics, allowing the predefinition of formulas for commonly-calculated metrics. dbt has developed their own Metrics package, and is pushing hard in this direction following the acquisition of Transform/MetricFlow.
The proposed feature is to create an integration based on dbt job metadata such that when browsing a connection in Sigma, if there are dbt metrics defined on top of a table in the dbt project, they appear in the “Metrics” tab in Sigma and can be used without re-creating the metrics in Sigma.
What is the use case?
The use case is to be able to flexibly analyze metrics in Sigma without duplicating work that data teams may have already done in dbt. Further, many data teams will prefer to leverage dbt for metric definitions because they are more readily governed, version controlled, and documented than those defined in the Sigma UI.
How often would this feature be used?
For teams who use dbt and are building a semantic layer using its Metrics package, use would be widespread and virtually constant.
What is the impact of this feature on your organization?
This would 10x the ability of our non-data team users to create their own flexible analyses with no SQL skills. It would also 0.1x the time the data team spends calculating things for people, explaining metric caveats, and going through the same discussions over and over about data idiosyncrasies and why one persons calculation doesn’t match another.