Ship the Smallest Honest Version
The MVP has been misunderstood into uselessness. Most teams hear "minimum viable product" and translate it as "the cheapest thing we can get away with shipping." That is wrong, and the error is expensive. The minimum viable product is not the cheapest build. It is the smallest thing that tells you the truth about whether the idea works.
Those two definitions point in opposite directions. The cheapest build optimizes for low cost. The smallest honest version optimizes for a real answer. Sometimes they coincide, and you get lucky. Often they do not, and chasing cheap gives you a product that ships fast, costs little, and teaches you nothing, because it was never a fair test of the actual idea.
Honesty is the word that matters. An honest MVP puts the core bet in front of real users under real conditions and lets reality vote. If your idea is that people will pay for faster onboarding, the honest version has to actually make onboarding faster and actually ask for money. Strip either of those and you have not tested your idea. You have tested a different, easier idea that happens to be cheaper to build.
This is where cutting corners sneaks in wearing the costume of being scrappy. Cutting a corner that is not part of the bet is good discipline. You do not need billing infrastructure to test demand. You do not need the full settings panel to learn if the main flow lands. But cutting a corner that is part of the bet quietly poisons the experiment, and you will draw a confident conclusion from a rigged test.
So the real question for any MVP is not "what is the least we can build" but "what is the least we can build that still gives a verdict we would trust." Those produce very different scopes. The first leads you to fake the hard part, the second forces you to keep the hard part and fake everything around it. Only one of them generates a result you can bet the next quarter on.
A useful gut check: when this version succeeds or fails, will you believe the result? If a failure would leave you muttering "well, it might have worked if we had built it properly," you did not build an MVP. You built an alibi. The honest version is the one whose failure you would actually accept as a failure of the idea, not the implementation.
There is a quiet courage in this. The smallest honest version exposes your idea to a real verdict sooner, and the verdict is sometimes no. Cheap-but-dishonest MVPs are popular precisely because they delay that verdict while feeling like progress. But a fast, fair no is worth more than a slow, comfortable maybe. It frees the team to go find the idea that actually works.
Build small. Build honest. Then believe what it tells you.