University of Free Knowledge
HD 62.5 · fol. 5

The Smallest Thing That Answers the Question

A minimum viable product is the least you can build to test the one assumption most likely to sink the idea. · 11 min

You do not have to build the whole idea to find out whether it will work. Building the whole thing first is the most expensive way to be wrong. Every new idea rests on a few things you are quietly hoping are true, and one of those hopes, if it turns out false, ends the idea no matter what else you get right. The job of a first build is not to impress anyone. It is to find out, as cheaply and quickly as it can, whether that one fragile hope holds.

Guess before you learn

You have an idea for an app that plans a week of meals and orders the groceries. What is the smallest first build that would best tell you whether people actually want it?

THE DEPTH DIAL — the same idea, younger or deeper
9–12

9–12

Write your idea as a set of claims that all must hold for it to work: people have this problem, they will change what they do, they will pay, and you can reach them. Some of these you already know; others are only guesses. The guess whose failure would be fatal is the leap-of-faith assumption. A minimum viable product is the cheapest experiment that puts exactly that assumption to a real test — and returns evidence you can act on, not opinions.

minimum viable product

The least you can build to get an honest answer about the one assumption most likely to sink the idea. Often rough; sometimes no software at all.

List everything that must be true for the idea to workMark which ones you are guessing, not knowingPick the guess that is fatal if it is wrongBuild the smallest test aimed at just that one
PLATE I Choosing a minimum viable product: from a list of hopes to one small test.

So the first question is not what to build. It is which assumption to aim at. The right one carries two marks: you are guessing at it rather than knowing it, and the idea cannot survive if it is false. An assumption that is safe, or that you could be wrong about without much harm, can wait. Find the one that is both uncertain and fatal, and aim your smallest possible build at it.

Why is this true?

Why test the assumption most likely to be wrong, instead of the one that is easiest to build?

Because the cheapest thing you can buy is the knowledge that the idea is doomed, before you spend a year on it. Building the easy part first feels productive, but it leaves the fatal question unanswered — and answers it last, when it is most expensive to hear.

Ink That Thinks — guess first; the answer draws itself.
Sketch how sure you become that the riskiest assumption is true, as you spend more days building. Commit your pencil guess first.

02.557.51012.515020406080100days spent buildingcertainty about the riskiest assumption (%)
Drag across the axes to sketch.
PLATE II Certainty about the riskiest assumption versus days building — guess in graphite, truth in ink.
Retrieval Gate — answer before you continue 0 / 3

1.Among the assumptions behind a new idea, which one should your first build test?

2.In one sentence, state the riskiest assumption behind a paid weekly email that lists local weekend jobs for teenagers.

3.Without looking back: what is a minimum viable product, in your own words?

To find that assumption, write the idea as a short list of things that must be true for it to work. Then, beside each one, mark whether you actually know it or are only hoping. The riskiest is the hope whose failure would sink everything, and that is where the first build points. The build itself should be the least you can make that still forces a real answer — often no software at all, just you doing the work by hand for a handful of real people.

Choose the cheapest minimum viable product for a meal-planning service — the steps fade as you master them

1
Write the one assumption most likely to be false.
People will actually cook from a plan that someone else makes.
2
Name the smallest build that tests exactly that.
Hand-make one week of meal plans for five volunteers — no app.
3
State what result would count as a pass.
At least three of the five cook four or more of the planned meals.
ASSUMPTIONIF IT IS WRONG…TEST FIRST?People want weekly meal plansNo one signs upMaybePeople will cook from someone else's planThe whole idea failsYes — this onePeople will pay eight dollars a monthRevenue is smallerLaterGrocery delivery can be integratedA feature is missingLater
PLATE III The assumption that is both uncertain and fatal earns the first test.
Retrieval Gate — answer before you continue 0 / 3

1.Put the steps of choosing a minimum viable product in order, from first to last.

  1. List everything that must be true for the idea to work
  2. Mark which of those you are guessing rather than knowing
  3. Pick the guess that is fatal if it is wrong
  4. Build the smallest test aimed only at that guess

2.A full working product would take 90 days to build. A hand-run test of the same idea would take 3 days. How many days sooner do you get your first real evidence?

days

3.What does the word viable add to minimum product?

A minimum viable product is not a small version of the finished thing. It is a test, built just solidly enough to get an honest answer. Next, you will meet the cheapest build of all: doing the whole job by hand, for real customers, before a single line of the product exists.

Note

Stuck naming the riskiest assumption? The Atelier of Mind — practice for how you think — has a short drill on separating what you know from what you only hope.

Practice — new ink and old, interleaved

1.Rewrite 'Would this app be helpful to you?' as a discovery question. Which is best?

2.In a discovery interview, why do you ask what the person did the last time the problem came up, rather than what they would do in future?

3.You can reach 1,200 people, 15% of whom would buy, each spending $80 a year. What yearly market can you reach, in dollars?

$

4.At $240 per customer per year, what is the yearly revenue from 300 reachable customers, in dollars?

$

5.Which is the strongest sign that a problem is worth building on?

6.Which of these is a proper value proposition?

7.Why is a problem that happens daily usually worth more than one that happens once a year?

8.You want to test whether people will use a plant-watering reminder service. Which is the best minimum viable product?

9.You can reach about 400 dog owners in your town, and roughly 1 in 4 would plausibly pay for weekly walks. About how many paying customers is that, at most?

customers
The Call Slip — search everything Ctrl·K / ⌘K