I think of it like listen to your users, but don’t do everything they ask. Use your judgement to make sure it aligns with your goals, and hopefully you have a goal beyond ‘make monies’.
Additionally, think about why they're asking for a certain thing, and whether their request is really the best way to solve the underlying problem they want solved
“Treat your users as you would a class of pre-school children. If they’re all screaming for a snack you don’t necessarily give them one, perhaps a healthy lunch would be better for the underlying need.”
I don't think pre-school children are the only people who are poor at recognising or expressing what they really want, I think that's a fairly universal thing. It's very easy to get an idea about what the solution should be & make a request based on that rather than basing the request on the problem you're facing.
A good analogy I heard recently (more to do with planning your own projects than requesting stuff in other people's, but I think it still works): imagine you have a large patch of grass and it is taking too long to cut. There are a couple of ways to define that problem:
* I need a better lawnmower
* I need a better way to cut grass
The first locks you in to a particular solution (lawnmower), but there may be a better possible solution out there, and the second statement opens it up a bit. So bringing that back to the user request thing, if your user requests a better lawnmower, make sure there isn't some better way to have the grass cut instead.
