> Seems wasteful to build something only to scrap it once people say they want it.It can be. But you probably shouldn't build a product to test whether people in your chosen market say they want it, or to find out whether buyers or sellers are hardest to acquire to bootstrap a two-sided market, when you can probably validate that with some landing pages or user interviews.
If you need to test whether people will use it after buying (or perhaps which users will use it after buying, or how many users will keep using it and paying for it), you probably need to build something, but it may only need to be implemented using manual processes behind the scenes.
If you need to test whether you can serve the customer at scale for less than your cost to acquire them, that is a different sort of MVP.
In any case, there are definitely MVPs that can and should be scrapped when you are done using them to validate (or invalidate) whichever proposition is then your biggest source of risk.
Keep in mind that should you invalidate the hypothesis being tested, you may need to scrap the MVP anyway when you pivot away from that particular hypothesis. And, some of your hypotheses will be invalidated, so making sure you don't run out of resources for testing new ones is kind of important.