Interviews tell you what people say. The next question is what they will actually do, and you can usually answer it for a fraction of the cost of building the thing.
Tests in order of cost
- Ask for a pre-order or a deposit. The most honest signal there is.
- Offer to do it manually for one customer, at a price. Slow and completely informative.
- A landing page describing the offer, with a real call to action and a small ad budget.
- A waitlist, which is the weakest of these — an email address costs almost nothing.
The manual version deserves more credit than it gets. Delivering the outcome by hand for three customers teaches you the workflow, earns revenue, and proves demand before you write any code.
Make the test cost something
A test where saying yes is free measures curiosity, not demand. Ask for money, a calendar slot, or genuine effort — anything with a real cost attached to the yes.
Decide the threshold first
Write down what result would make you proceed before you run the test. "If three of ten say yes and two pay a deposit, I build it." Deciding afterwards means you will interpret whatever happens as encouragement, because you want to.
Try it
Pick the cheapest test that would genuinely change your mind, write your pass threshold, and run it this month.