Testing the business model

You now have a whole business model on paper. But remember: every box is an assumption. This lesson is about turning assumptions into knowledge — testing the model against reality before you bet everything on it. It is cheaper to discover a flaw now than after two years and an empty bank account.

Find the riskiest assumptions

You can't test everything at once, and you don't need to. The key is to find the assumptions that are both most uncertain and most important for the model to work. These are often called "leap-of-faith" assumptions: the things that must be true, or everything falls apart.

Go through the nine building blocks and ask for each: "How sure am I of this, and how bad is it if I'm wrong?" Boxes that are both uncertain and critical get tested first. Two questions usually sit at the top: Will anyone actually want this (the value proposition)? And will they pay enough for it to be worthwhile (revenue and unit economics)?

Cheap experiments for each building block

An experiment doesn't have to be a finished product. The goal is to learn as much as possible for as little money and time as possible. Some examples:

• Value proposition: a simple landing page that describes the promise, and a "buy" button that measures how many click.
• Customer segment: ten in-depth conversations with people in the segment before you build anything.
• Channel: a small ad campaign with a fixed budget to see whether people react at all.
• Willingness to pay: a pre-sale or a letter of intent — will anyone put down money or a signature before the product exists?
• Delivery: do it manually first (one customer, hands-on) to see whether the value actually arises, before you automate.

Every experiment should have a clear success criterion set in advance: "If at least 20 of 100 visitors click buy, we move on." Without a threshold decided before the test, you always interpret the result in your own favor.

Evidence vs. opinions

The most important habit in validation is separating evidence from opinions. "This is a good idea" is an opinion — even when it comes from an investor or a friend. "17 of 100 left their email to be notified at launch" is evidence.

Be especially skeptical of compliments. People readily say your idea is nice because they want to be kind. What counts is what they do: do they leave money, time or contact details? Action is truth; politeness is noise. Watch out for confirmation bias too — the tendency to hear what you want to hear. Design your tests so they can prove you wrong, not just right.

A simple trick is to agree with yourself in advance what a no would look like. If you can't describe a result that would make you change your mind, you are not testing — you are looking for applause. The best founders treat their ideas as something to be challenged, not protected, and they are more relieved than disappointed when a cheap test reveals a flaw early.

Decision gates: continue, adjust or stop

After each experiment you stand at a gate with three roads:

1. Continue (persevere): the evidence supports the assumption — move on and test the next one.
2. Adjust: something held, something didn't — change a building block and test again.
3. Stop or pivot: the evidence clearly says no — then something fundamental must change, the topic of the next lesson.

The brave thing is not to push on regardless, but to let the evidence decide. A founder who tests honestly and turns when the numbers say so gets further than one who falls in love with their own plan.

Do this now

Pick the single assumption that is most critical and most uncertain in your model. Design one cheap experiment that can test it this week, and write down the success criterion before you start: what has to happen for you to count the assumption as confirmed? Run the test, and let the result — not the hope — decide the next step.