The term has been diluted until it means 'version one, but rushed'. That reading produces the two failure modes we are most often asked to rescue: a nine-month build of unvalidated features, or something assembled so carelessly that traction forces an immediate rewrite.
It is an experiment with a hypothesis
An MVP tests one belief your business depends on. Write the belief down as a sentence — 'clinics will pay ₹8,000 a month to cut appointment no-shows' — and the scope follows from it. Without a stated hypothesis there is no principle for deciding what to leave out, and everything ends up in.
If you cannot state the hypothesis in one sentence, you cannot scope the MVP.
Viable means someone would actually use it
The minimum is real. A signup form measures curiosity, not demand. To learn whether people will pay, the thing has to do the job well enough that using it is a rational choice, which means one workflow done properly rather than five done partially.
Cut breadth, not quality
The right reduction is scope: fewer features, fewer user types, one platform, one segment. The wrong reduction is care: no error handling, no tests, no thought about data structure. Breadth is recoverable later; a foundation you have to replace at the moment of traction is the expensive kind of shortcut.
Instrument it, or you learn nothing
An MVP without analytics is a launch, not an experiment. Decide before building what signal would confirm or refute the hypothesis, and instrument for exactly that. Teams that skip this end up interpreting the result through whatever they already believed.
Six to ten weeks, or rescope
If a proposed MVP will take six months, it is not one. That is a product build wearing the label to avoid a difficult prioritisation conversation. When we see a six-month scope, we work backwards to the smaller thing that would answer the question first.
Be willing to accept the answer
The uncomfortable outcome is the valuable one. If the evidence says no, that is a year and a great deal of money saved, and it is worth more than another phase of building in hope. Agreeing in advance what result would make you stop is what makes the experiment real.
