We made a hot chocolate at Shklet properly. Real chocolate, not the sugary powdered kind everyone else was selling.
I was certain. Cold weather. People wanting something warm and rich. Nothing available at that quality anywhere near us. It made complete sense in my head, and I want to be clear that the reasoning was not stupid. Every part of it was true.
It did not sell.
Making sense is not evidence
This is the trap, and it catches experienced people more than beginners, because experienced people build better arguments.
I could explain exactly why it would work. The logic held together. What I could not do was point to a single person who had actually paid money for it, because the product did not exist yet and I had never tested the assumption.
A good argument for a product is not the same as demand for it. You can construct a flawless case for something nobody wants.
Test the downside, not the dream
When we start something, we imagine the good version. The queue. The reviews. The growth.
We do not imagine the version where it simply does not work, and that is exactly the version you need to test for.
So ask the uncomfortable question deliberately, before reality asks it for you. Not how well could this do but what would have to be true for this to fail, and is any of it true already?
Cheap tests answer that. Ask people who are not your friends. Sell a few before you make many. Put it in front of strangers and watch what they do rather than what they say.
Because if you find out on a small batch, it costs you a small batch. If you find out a year later, it costs you a year.
And failure is information, not disaster
The hot chocolate was not a catastrophe. It was one product that told us something true about our customers that we did not previously know.
That is what a failed test is. You have a plan B, you use the information, and you are not starting from a crisis. You are starting from exactly where you should have started.
The founders who get hurt are not the ones whose ideas fail. They are the ones who found out too late, because they never made the test cheap enough to run.