Enthusiasm is the cheapest thing a prospect can give you, and it looks exactly like validation.
A founder posted this week that they had built a requirements management tool, found four design partners who matched their ideal customer profile perfectly, showed them the product, and watched all four say they loved it. Months later none of the four have paid. The post reads like a confession. It should read like a receipt.
Why do customers say they love a product and then not buy it?
Because the two acts have completely different prices.
Saying you love something costs a person nothing. It costs them less than nothing, actually, because it makes an awkward conversation pleasant and gets them out of the meeting on good terms. Buying costs money, an internal argument with someone who controls a budget, and the risk of being the person who brought in the tool that didn't work.
When one behaviour is free and the other is expensive, you cannot use the free one to predict the expensive one. That is not a flaw in your customers. It is arithmetic.
Are design partners a reliable validation signal?
They are reliable about one thing and misleading about everything else.
A design partner is reliable evidence that you can describe a problem in a way that a specific kind of person recognises. That is genuinely useful and most founders cannot do it. What a design partner is not is evidence of budget, urgency or priority, because the arrangement almost always removes exactly those three things. You gave them the product for free, you asked for feedback instead of a decision, and you made being encouraging part of the deal.
Then you go back six months later and ask for money, and you are surprised that the relationship you built on kindness produces kindness.
What is the difference between enthusiasm and demand?
Enthusiasm is a reaction. Demand is a reallocation.
Enthusiasm happens when someone hears a description of a better world and agrees that it would be better. Demand happens when someone takes resources away from something else and gives them to you. The first is an opinion about the future. The second is a decision about this quarter.
Almost every false validation signal in existence is enthusiasm being read as demand: the waitlist, the beta, the survey, the friend who said they would definitely use it, the AI score that reviewed your pitch and liked it. All of them measure agreement, and agreement is available in unlimited quantities at no cost.
What counts as a costly signal?
Anything the person could not have done casually.
Money is the obvious one, and the cleanest, because nobody spends it by accident. But it is not the only one. Somebody who built a workaround has already paid, in evenings. Somebody who switched from a tool they were already paying for has paid, in switching cost and internal risk. Somebody who made an introduction has paid, in reputation. Somebody who filled in your ugly spreadsheet twice without being chased has paid, in attention, which for a senior person is the scarcest currency they have.
One founder put it well this week: a complaint is not an idea, the workaround is the signal. Complaining is free. Building the thing yourself, badly, in a spreadsheet, at eleven at night, is not.
What should you ask someone who loved it and did not pay?
Ask what they are doing instead, right now, today.
Not what they would do, not what they think of your pricing, not whether they would buy it at a lower number. Ask what happens on Tuesday when the problem occurs, who does what, how long it takes, and what it costs them when it goes wrong. If the answer is a real process with real hours in it, you found a market and you have a positioning problem. If the answer is a shrug, you found a nice conversation.
The gap between those two answers is the whole business, and it is entirely findable before you build anything.
Does that mean positive feedback is useless?
No. It means positive feedback answers a different question than the one you asked.
Use it for what it is good at. Enthusiasm tells you the message lands, the category is legible, the demo is clear. Those are real wins and they compound. Just stop reading them as a forecast, because they were never a forecast, and the invoice will eventually explain that to you at a much worse moment.
Key takeaways
- Saying you love something is free. Buying is expensive. Free behaviour cannot predict expensive behaviour.
- Design partners are strong evidence that your description of the problem lands, and weak evidence of budget, urgency or priority.
- Enthusiasm is a reaction. Demand is a reallocation of resources away from something else.
- A costly signal is anything the person could not have done casually: money, switching, an introduction, a workaround they built themselves.
- Ask what someone does today when the problem occurs, not what they would do with your product.
FAQ
Why do design partners say they love a product and then not pay for it?
Because the design partner arrangement deliberately removes cost, urgency and the budget decision. You get accurate feedback on whether the problem is recognisable and no information at all about whether it is worth paying to solve.
Is positive feedback ever a valid validation signal?
It is valid evidence that your framing of the problem is clear and that the category makes sense to the listener. It is not evidence of demand, because giving positive feedback costs the person nothing.
What is a costly signal in customer validation?
Any action that took the person real money, real time, real switching effort or real reputation. Payment is the clearest one. A self-built workaround, a paid competitor being abandoned, or an unpaid introduction all count.
How do I test demand before I have a product to sell?
Look for what people are already doing about the problem without you. Existing workarounds, existing spend on adjacent tools and the hours currently going into the manual version are all evidence that predates your product entirely.
What should I do with four design partners who never converted?
Go back and ask each one what they actually do when the problem occurs now. Their current process is the data. Whether they liked your demo is not.