How to Find a Product Worth Building (Before You Write a Line of Code)

The signal that actually matters

Every maker has a folder of ideas. Most are good ideas. Almost none are businesses, and the difference is rarely the idea itself.

There is only one reliable signal: someone is already paying for something like it, and keeps paying. Praise is not a signal. A thread with hundreds of upvotes is not a signal. Money changing hands, month after month, is the signal.

Look at what already earns

The temptation is to chase what is hot right now. By the time a category is visibly hot, the window for a small, unhurried builder has usually closed.

What you want instead is a quiet tool that has been collecting the same small subscription for two or three years. Longevity proves two things at once: the problem never goes away, and the price is acceptable.

Three free places to look for these: public revenue boards that verify earnings through the payment processor, AI tool directories that catalogue niche tools with traffic estimates, and maker communities where people publish real monthly numbers and how their first customers arrived.

Write it down, then judge it cold

For each promising product, note six things: what it is called, who it is for, when they would use it, what problem it removes, roughly what it earns, and how it charges. Six lines is enough.

Then judge it against yourself, not the market. Can you differentiate inside that niche? Can you build the first version alone, in the time you honestly have? Two yeses mean the idea has legs; one no means it stays in the notebook.

Thirty minutes a day, three to five products logged, replaces inspiration with evidence within a few weeks.

Copy the demand, not the product

Almost nothing that earns steadily is a new invention. The usual pattern is plainer: a proven paid tool exists for one group of users, and the same need is still unmet for a different group, at a different moment of the day.

So migrate the demand rather than copying the app. Take a tool built for office workers and ask what it looks like for students. Then compete on the details the original never bothered to get right.

You can see the same shape in ordinary categories. Parents have paid for bedtime stories for as long as there have been bookshops — the demand is settled, only the form is open. That is the gap LittleStory works in: a photo, a short voice note and one sentence become a personalised, illustrated story, read aloud in a voice you record yourself. Not a new need, just an easier one to reach at 8pm on a Tuesday.