This is a summary of Courtland's mental model about validating a product
When validating your product, try different channels. The best channel should be the one you can use indefinitely.
For example: validating through ads can be quick and simple, but does your budget allow you to run in the long term?
Besides, Ad is one of the possible channels. If it doesn't work there, it doesn't mean your idea doesn't work. It can be a few different reasons (bad message, wrong audience)
See The Bullseye Framework Gabriel Weinberg - founder of DuckDuckGo
To have a good product, it needs 4 things:
Those are more like 4 wheels on a car. To get the car moving, it needs 4 wheels. Even if those are bad wheels.
Try to figure out the hardest one, then work backward to the easier ones. For example: if you're a tech person, try to figure out the market, then build the product around it.