Which of these have you actually watched?

You have been doing this a long time. You know why people choose you, what they care about, what annoys them, and what they would pay more for. That knowledge is real and it was expensive to acquire.

Here is an uncomfortable exercise. Take any one of those things you know and ask a single question about it. Did I watch a customer do this, or did I work it out?

Both answers feel identical when you are standing inside the business. Something you saw three customers do and something you reasoned your way to over a decade sit in your head with exactly the same weight. Neither one arrives labelled. And the second kind is where expensive decisions go wrong.

Two columns, and only one of them is evidence

Take a sheet of paper and draw a line down the middle.

On the left, write the things you have watched. Not what a customer told you they would do. What you saw them do, or what the records show they did. Someone rang three times before booking. Four people in a row asked the same question about pricing. Seven of your last ten jobs came from the same two referrers.

On the right, write everything else you know. The reasons people choose you. What they value most. Why the ones who left, left. What they would be willing to pay for if you offered it.

For most owners the right column fills up fast and the left one is thinner than expected. That is not a failure. Running a business does not leave much room for watching it. But the gap between the two columns is the most useful thing on that page, and almost nobody has written it down.

Why the right column costs money

Nothing in the right column is necessarily wrong. Plenty of it will be right. The problem is that you cannot tell which parts, and you are spending against all of it.

Every significant decision a business makes rests on a small number of these beliefs. A new service rests on people wanting it. A price rise rests on what they will tolerate. A new location rests on why people travel to you. Hiring rests on where the work is going to come from.

When one of those turns out to be a belief rather than an observation, the money is usually already spent. Nobody finds out at the point of decision. They find out eight months later, from the numbers, which is the most expensive time and the least useful place to learn it.

So the question is not whether you are right. It is which of the things you are betting on has never been tested, and whether that happens to be the one carrying the most weight.

Agreeing on paper is not evidence

There is one specific trap worth naming, because it catches careful people rather than careless ones.

You work out a customer problem. You design something that solves it. It fits. On paper everything lines up, and the fit is genuine. That feeling of things clicking into place is real and worth having.

It is also not evidence of anything. It says your reasoning is sound. It does not say a customer will change what they currently do, pay your price, or pick you over the thing they are already using. Those are separate claims and each one needs its own proof.

A plan that fits together beautifully and a plan customers have voted for with their behaviour look identical in a document. They stop looking identical about a quarter after launch.

What to do with a thin left column

Do not try to fill it. You are not building a research programme and you do not have time for one.

Pick one line from the right column. Specifically, the one that would do the most damage if it turned out to be untrue. Not the most interesting. Not the easiest to look into. The one carrying the most money.

Then ask what would move it across to the left. It is usually smaller than people expect.

• Six conversations with recent customers about what actually happened, rather than what they think of you.
• A look at what your own records already say. Most businesses are sitting on the answer to at least one of these and have never had a reason to look.
• One small version of the offer put in front of real people to see whether anybody takes it, before the full thing gets built.

One belief moved from assumed to evidenced is worth more than twenty more items in the right column. You are not trying to know everything about your customers. You are trying to stop betting on the one thing you have never checked.