This is a common question — certainly, the standard advice is to pick a user, then pick a need. In my opinion, it’s very easy to box yourself in doing that: once you pick a user and then focus on narrow need for that specific user, it’s going to be difficult to come up with several big, ambitious solutions. Most people following that method generate very incremental solutions.
The best solutions solve multiple needs of a big user population.
Narrowing the search space too early and then having very incremental solutions is one of the most common mistakes I see with product sense interviews.
I wouldn’t categorically say that picking users/needs is bad, but if you do it you need to really make sure you’re getting something in return, i.e. that by picking a user segment, it might reduce the reach of your solution, but that these users have very differentiated needs and such burning needs that they’re going to use your solution so much more because of it (i.e. it’s worth it even if the target is smaller).
You need to believe that the decision maximizes the impact you can have and not just do it because it’s “part of the framework”.