This is key. I cannot count how many times I go to a client that is complaining that their software is slow and ends up that is because they don't have the right indexes
The difference between a poorly performing product that uses an RDBMS backend and a highly performant one is purely down to SQL query tuning and finding the right indexes.
It is the lowest hanging fruit to improve performance.
Even if you don't have the flexibility to change the SQL on a project, you may still have the ability to create/rebuild the indexes to make the query faster.
This will take you from the theory of indexes to actually seeing how it is used by the SQL engine.
This will go a long way to building the right indexes and SQL tuning in general.
I would suggest learning how to generate the query plans for your SQL engine of choice.