Assuming you have a tight budget (no budget for marketing or user testing) pick someone, yourself, a friend, or relative, then make some concept arts, and show it to them, if they like it, continue iterating, if they are not interested you have to pivot. Note that the user can also be yourself. It's unlikely that you would make a game that you yourself love to play but no other one does. It's however likely that you love developing it much more then playing it, then you are not the user! And you have to find an actual user. Test the game/product on your target user(s) throughout the development process to see what works and what doesn't. If the "users" are engaged they will have suggestions, actually you will get flooded with suggestions, so it helps if you also is a user yourself so you can prioritize features according to your vision and not just the most popular.

Website with preorder/newsletter sign up (maybe stretching it a bit to interpret "user" as "interested potential customer"?

but even if the answer is "for this kind of product it's impossible to get a user before writing code" it's not necessarily a bad heuristic -- maybe the conclusion could be "don't build this kind of product, pick something else"