Part of the role of a product manager is to work with engineering and the data science team to define the set of metrics the product team should track. Setting the right metrics can be hard, and even the right metrics can sometimes drive the wrong behavior.

You need to consider metrics around a variety of categories.

PM doesn’t need to be an excel jockey or even know SQL (helps tho). But they need to be able to find patterns in data and draw conclusions that aren’t on the surface (common sense)