Minimum viable test
Why it matters
Building comes before evidence by habit, because building feels like progress. A test reverses that order. It asks the cheapest possible question first, so a wrong idea costs days, not months.
A week of testing can save a quarter of work on something nobody wanted. It also makes stopping easy. When the pass mark was set in advance and the result misses it, nobody has a half-built product to defend.
The test is also the first real look at price. "Would you pay for this?" gets polite answers. "Here is the price, reserve a place" gets honest ones. That is why a minimum viable test belongs in pricing work as much as in product work.
How to apply it
- Write the assumption as one sentence: "Agencies with five to twenty staff will pay a monthly fee to automate client reporting."
- Choose a result that could change the decision. A test that always says yes is not a test.
- Set the pass mark before starting, such as "at least ten of the first hundred visitors join the waiting list".
- Test with real behaviour and real money where possible. A stated intention is weak evidence.
- Stick to the pass mark. A miss means change or stop, not "run it again until it passes".
What it is
Every new idea rests on assumptions: that people have the problem, that they will pay, that they will pay this much. A minimum viable test picks the riskiest assumption and checks it with the least work that can still give a real answer. It is smaller than an MVP, because nothing needs to be built that works.
Common forms are a landing page that describes the offer with a button to pre-order or join a waiting list, a manual "concierge" version where the owner delivers the service by hand, and a direct ask where ten likely customers are offered a price and a start date.
The abbreviation MVT also means multivariate testing in website optimisation, which compares several page elements at once. That is a different thing.
Common mistakes
- Testing the wrong assumption. Pick the one that would end the idea if false, not the easiest one to test.
- Setting no pass mark. Without one, any result can be read as success.
- Moving the goalposts. After a miss, lowering the bar until it passes defeats the purpose.
- Counting compliments. "Sounds great" costs nothing. A reservation, a deposit or a signed letter of intent does.
- Testing on friends. Use people who resemble the real customer.
- Confusing it with an MVP. The test comes first and needs nothing that works. The MVP is what you build if it passes.
