Ask HN: When to start building the MVP?

A part of me wants to talk to more users, before I build anything, with the belief that once I understand the problem well enough, the right product “seed” to experiment with, will be very clear.

Another part of me wants to quickly get an MVP out there, based on my current understanding of the problem, and utilize the feedback cycle with the target audience, to iterate towards a product that they actually want.

At what point did you say “fuck it, I’m going to build this product, and see if these people will actually use it”?

Would you have done anything differently?

context: B2B, Saas