I’ve learned to ask one question before writing a line of code. Can I deliver this value manually first?
Not as a prototype. Not as a wireframe. As an actual service. Find the people who would use the product, do the thing for them yourself, and see if they want it. If the answer is no when you’re offering to do it by hand, the answer is no when you offer to do it with software. You’ve just saved yourself months.
The best validation signal is someone who’s willing to wait while you do the work manually. They’re showing you the problem is real, the value is clear, and the price is acceptable. None of that comes through in a survey or a landing page test. It only comes through when someone agrees to a process that’s slow, awkward, and obviously held together with sticky tape.
Most of the products I’ve built were preceded by a manual version. Sometimes the manual version was mine, sometimes it was the founder’s, sometimes it was a customer’s existing workaround. In every case the platform didn’t create demand. It compressed effort that was already being spent. The best early stage technical hire isn’t someone who builds fast. It’s someone who asks whether the thing is worth building at all. The fastest way to ship the wrong product is to skip the question entirely, write a clean spec, and execute it beautifully.
If you can’t find anyone who wants the manual version, you don’t have a product. You have a hypothesis you haven’t tested.
