A structure feels case-specific when each branch is tied to the client objective and the facts you were given.

Generic labels are not automatically bad, but they need a case-specific question underneath them.

For example, instead of just saying customers, state which customer behavior would explain the client's problem and what you would test.

Before speaking, force yourself to connect every bucket to one fact from the prompt. You can also use this CaseBasix guide and this related guide for a structured review.

I had this exact issue in mocks. Making the bucket names specific to the client changed the feedback pretty quickly.

I had a different experience: the interviewer did not seem to care about the label, only whether I could explain the logic.