Creating Questions-as-Tables using SQL

In the last section, we learned that while SQL can often be an easier option, it has a major drawback: SQL questions are hard for non-SQL users to explore. This presents a clear problem for a product such as Metabase, since Metabase aims to democratize data and allow anyone, no matter how technical, to explore it and find answers to their questions. Their solution to this is the notebook editor, but we've already seen that the notebook editor either can't answer a question, such as Items Ordered, which we learned about back in Chapter 6, Creating Questions, or is too unwieldly, as we saw in the last section with the review rate.

Unfortunately, having messy or unwieldy data in your application database is more of a rule than an exception. It's often quoted that 80% of a data scientist's job is cleaning messy data. In my experience, that sounds about right, and a lot of that cleaning is done with SQL. Remember that application...