I have my own set of product development stages that go before and beyond the MVP. They are:
Before worrying about building your MVP, you need to build your MUP.
There's nothing "viable" about an MUP. But there is the first possible increment of something useful in it. It's "usable" to solve a problem your product is meant to solve.
It might be manual. It might be horrible. It might be buggy. It might be 10x worse than solving the problem in any another way, but it's at least usable. No one else might ever use it, but you use it.
By using your MUP, you experience its shortcomings and understand how it needs to be improved. You know why it's not yet viable. With an MUP you get to point where you are reacting to a real thing, not vague ideas, documents, sketches and diagrams. It's a critical step.
Today I defined what my MUP will do (but not yet how it will do it).