Let’s say you begin with some idea in mind. You mull over it until you are convinced about its success theoretically. Then you go about building a business or a product or both around it. Then you figure out it falls short in attracting the customers you originally thought it would.
This seem to be the case when you are doing a Startup. You are all geared up to build stuff, slog, read things you have never come across before, get excited about technologies which you have barely heard of. And then you find there are no takers.
It appears like the act of building precedes the act of finding takers. But to build something with such intensity you need to have identified the takers. If you are a build enthusiast you ignore a large part of the takers and end up building some large piece and end up finding there are no one enthused as much as you about it which later comes to include you as well.
If you are the one who finds takers before you build, you are definitely in a better position than the build enthusiast, but still you fail as you misjudged the takers.
And it comes back to saying that you misjudged the takers either by interviewing a low number or by asking the wrong people or by extrapolating the numbers in your zeal to see things scale or by lack of knowledge of the domain or by bad co-founders or bad execution or lacking the technical edge or your compromising attitude or your hurry to bring it out and so on and on.
Plenty of reasons to end up somewhere else than where you thought you would be when you began the whole thing.
So it is not about trying to find fault with building something or finding takers before build. Both are needed and most of us do that in our own way. No one has that uncanny eye to see the right from the wrong. There seem to be plenty of paths from where you begin and where you end up.
Finding what to build that will have takers is THE problem you signed up to solve as part of a Startup and unless you entered here with this problem in mind, you will find every problem you work on leading back to the above problem.