Building stopped being hard. That means a functioning product stopped being proof of anything except that you can build.
Does a working demo prove demand?
No. A demo proves the thing can exist. Demand is a separate claim about other people, and nothing you build on your own can test it.
There was a period when this distinction did not matter very much. Building was slow and expensive enough that finishing something was itself a filter. If you got a product out, you had probably survived enough contact with reality to have learned something along the way. The demo was a weak proxy for demand, but it was a proxy.
That proxy is gone. A founder can now produce a polished, functioning product in a weekend without speaking to a single person outside their own head. The demo has stopped carrying information about the market. It now only carries information about the builder.
Why did building stop being proof?
Proof requires that the thing being proved was at risk of failing. When shipping was hard, shipping proved something, because plenty of people tried and could not do it.
The constraint has moved. The scarce thing is no longer the ability to make software, it is knowing which software is worth making. A finished product no longer separates founders who understand their market from founders who do not, because both groups can produce one.
This is why so many launches now look identical from the outside. Two products with the same polish, the same landing page and the same demo video can sit on completely different amounts of evidence, and nothing visible tells them apart.
What is a demo actually testing?
A demo tests comprehension, not desire. When you show someone a working product, you find out whether they understand what it does. That is worth knowing and it is not nothing. But people routinely understand a product perfectly well and remain completely indifferent to it.
The trouble is that comprehension feels like approval. Someone says they get it and that it is clever, and it lands like a yes. It is not a yes. It is confirmation that your explanation worked.
There is a worse version of this. A good demo actively suppresses the signal you need. A person watching a smooth product is being led. They are answering questions you chose, in an order you chose, at a pace you set. Every moment where they would have got confused, given up, or quietly decided it was not worth the trouble has been designed out of the experience.
What counts as evidence instead?
Evidence is behaviour with a cost attached. The cost does not have to be money, although money is the cleanest version of it. It can be time, effort, reputation, or giving up something they already use.
Things that are evidence: someone cancels a tool in order to switch to yours. Someone hands over a card number. Someone puts their own name behind an introduction. Someone comes back a second time without being asked. Someone does the annoying setup work you were worried about.
Things that are not evidence: praise, a signup with no follow-through, a survey response, a friend saying they would definitely use it, a demo call that ended warmly, a waitlist number, an AI tool scoring your idea highly.
The dividing line is whether the person gave up anything at all. If they did not, you have learned what they say. If they did, you have learned what they do. Only one of those two predicts what happens after you launch.
How do you tell a compliment from a signal?
Ask what happened next. A compliment is complete in itself and leaves no trace. A signal has a consequence you can point at afterwards.
So the useful question after any positive reaction is not whether they liked it. It is what that person did in the seven days that followed. If the honest answer is nothing, then the reaction was information about your presentation, and not about your market.
Key takeaways
- A demo proves a product can exist. It does not prove anyone wants it.
- Cheap building removed capability as a filter, so a finished product stopped signalling demand.
- Demos test whether people understand you, not whether they want what you made.
- Evidence is behaviour with a cost attached: money, time, effort, or switching away from something else.
- Praise, signups, surveys and waitlist numbers are not evidence, because nothing was given up to produce them.
- After a positive reaction, the question that matters is what the person did in the week that followed.
FAQ
Is building a demo a waste of time?
No. A demo is useful for testing whether people understand what you made, and for the parts of an idea that only become clear once it exists. The mistake is treating the finished demo as an answer about demand rather than as a tool for asking better questions.
What if people say they love the demo?
Take it as evidence your explanation is clear. Enthusiasm about a demo predicts very little about payment, because liking something costs the person nothing. Look at what they did in the days afterwards instead.
Are signups evidence of demand?
Only weakly, and only if the signup cost something. A free email capture with no follow-through is closer to politeness than to intent. A signup where someone completed setup, invited a colleague or entered payment details is a much stronger thing.
Does this apply to B2B as well as consumer?
Yes, and it is often worse in B2B, because meetings are cheap and pleasant for the buyer. A friendly demo call with a company that never moves toward procurement is one of the most common false positives founders run into.
How much evidence is enough before building?
There is no fixed number and anyone offering you one is guessing. The practical test is whether you can name specific real people who did something costly in response to your idea, and say what they did. If you cannot name them, you do not have evidence yet.