Self serve doesn't stop at providing data

You can’t do self-serve just by providing datasets that you understand, but your users don’t.

You need to consider how those datasets are provided, and how they are going to be used.

Are they designed so users cannot do it “wrong” when querying, joining and aggregating them?

Who is the dataset aimed at, and will it meet the requirements of those users?

Is it obvious which dataset to use, and when?

Do the users know which use cases they are suitable for, and (perhaps more importantly), what they are not suitable for?

Are they discoverable? Are they well documented?

This is just scratching the surface. There’s a lot you need to do if you want to truly allow self-serving of data that enables your users.

If you’re just throwing raw data at them, you’re probably doing more harm than good.