Perhaps in the past. I think the approach now will be to vibe code multiple projects very quickly and see which one has traction even with a low quality product. You will get much better feedback than a discussion with a potential customer who may not even know what they want or have a false idea of what they want. You can always improve a product that has demand and abandon the ones that no one even downloads. Usage and payment are the real test if a product is worth doubling down on.

This might work to some degree if you can run your project by many eyeballs, but only if they aren't immediately made gun shy by interacting with a low quality product. A focus group environment would be good for this, but setting that up costs money.