What customers say and what they do

You had an idea. Sensibly, you checked it before committing. You asked customers, or you ran a survey, or you talked it through with a dozen people who would know.

The response was good. Encouraging, even. So you built the thing, and then very few people bought it.

The natural conclusion is that people lied to you. They almost certainly did not. Something more useful than that was going on, and it is worth understanding because it decides how you check the next idea.

People are poor witnesses to their own future

When you ask someone what they think of an idea, you are asking them to predict their own behaviour in a situation that does not exist yet, while being polite to your face. Everybody is bad at the first part and most people are quite good at the second.

There is also no cost to saying yes. Agreeing that something sounds useful takes nothing from a person. Actually buying it means finding the money, changing what they currently do, and explaining the decision to someone else. Those are not the same question, and only one of them was asked.

So treat what people tell you as the weakest grade of evidence available. Not worthless. Weak. It tells you about their attitude, which is real, and almost nothing about their behaviour, which is what you are spending money on.

The ladder

It helps to have the grades straight, because the effort to climb them is much smaller than people assume.

• What people say about an idea. The weakest. Polite, cheap, and mostly attitude.
• What people say they will do. Slightly better and still an opinion about the future. Intentions and actions come apart badly.
• What people actually did. Much stronger. Records, past behaviour, what happened last time you offered something similar.
• What people paid for, or gave up something for. The strongest evidence you can get short of running the business. Money, a deposit, a diary slot, a signature.

Most decisions in small businesses get made on the first rung, because the first rung is the easiest to collect. The gap between rung one and rung four is where the losses happen.

The test that could never have failed

Now the less obvious problem, and the one that catches people who are already careful.

Before you check anything, ask what answer would have made you walk away. If there is no answer the test could have produced that would have stopped you, it was not a test. It was a search for permission, and it will always find some.

This is worth being strict about, because the shape of a real test and the shape of a reassurance exercise are identical from the outside. Both involve talking to customers. Both produce notes. Only one of them could have come back no.

It also explains a failure that looks like diligence. Someone does forty interviews, comes out with a list of two hundred problems, and makes no decision at all, because a list that long supports whatever you were going to do anyway. That is not research. It is a way of feeling busy while the decision sits untouched, and it is expensive in the one currency that matters, which is time.

What a real check looks like

The good news is that stronger evidence is usually cheaper and faster than the survey, not dearer and slower.

• Sell it before you build it. Describe the offer properly and ask for a deposit or a booking. If nobody will hold a slot, you have learnt the most important thing you could learn, for nothing.
• Put up the offer and count. A single page describing the new service, and a count of how many people enquire. Interest that costs someone an email is worth a hundred conversations that cost them nothing.
• Do it manually for three customers first. Deliver the thing by hand, badly, before you build the system. You find out whether anyone wants it and what it actually takes, in the same fortnight.
• Look at what already happened. Your own records hold real behaviour you have never examined. It is the cheapest rung-three evidence in the building.

Pick the smallest one that could still tell you no. The purpose is not to prove that you are right. It is to find out cheaply, while it is still cheap to change your mind.