Most products fix problems.
Some attend to wanting.

Some products don't start with a problem statement. They start with something a person wants, has wanted for a long time, and will keep wanting whether or not anybody builds for it.

Call it need if you want. Desire is closer. Need implies a deficiency to correct: food, shelter, safety, the floor under a life. What these products answer is not a floor. Nobody is failing to survive because they see their friends less than they used to. They just want something, steadily, and nobody has built for it, because wanting is hard to write into a brief.

The problem statement is a good instrument. It built most of the software in the world. But it has a shape, and the shape decides what can be made with it. A problem is something broken. You identify it, scope it, solve it, close it. Success means it stops being a problem.

Wanting doesn't close. You don't solve friendship. You don't finish wanting to know where you came from. These are conditions of being a person, and the most a product can do is meet one honestly, for a while, and be good enough to come back to.

There's a distinction worth making here, and it isn't about quality. It's about where the thing started.

The order matters more than the labels. Form comes second. Money comes third. Neither one is optional and neither one goes first.

Often the person who wants it is the one building it. That's not a smaller version of the practice. It's the cleanest one, because nobody had to be convinced the wanting was real.

Notice what the first two have in common. Somebody else handed it to you. A brief, a market, a deck. The origin is secondhand. That's the part worth avoiding, more than problems as such.

A painting has no job. Nobody needs one. All of it is surplus, and the surplus is most of what makes it a painting.

A product has a job. It has to work, it has to load, it has to not lose your data. The art is whatever it does past that, carried alongside the function without breaking it. That's harder than it sounds, and it's the first thing cut when a date gets close.

You find out later anyway. Not while you're using the thing, but after. Something stays. A feeling, or a memory of something the thing made possible, still sitting there a week on.

That residue can't be built directly. Anything designed to linger on purpose is advertising. You make the thing more carefully than the job required, and it stays with someone or it doesn't.

On money, plainly. These products aren't free, and they aren't trying to be enormous.

An artist who can't sell the work stops making it. That isn't a compromise of the practice, it's the practice. Every form that has lasted ran on patrons, on tickets, on subscriptions, on somebody finding a way to keep the maker fed. Software came to this late. It spent thirty years believing the only honest outcomes were free or vast.

Monetization funds the work. It doesn't direct the work. Thin enough to sustain, never thick enough to steer. That's the whole discipline, and it's the first thing to go when a product starts doing well.

None of this was affordable until recently. Building a considered thing took a team, which took a runway, which took an investor, and an investor needs a return big enough to justify the fund. That math quietly decided that anything which couldn't get enormous wasn't worth making.

That was a constraint of capital, not a statement about value. Plenty is worth making at a scale of thousands. What was missing was never the wanting on the other side. It was a way to build for it and keep it alive without asking anyone's permission.

That constraint is dissolving. One person can now make something genuinely good in about the time it used to take to write the pitch. What comes next isn't a smaller software industry. It's a different one, closer to how music and books and film have always worked: a lot of small, particular, sustaining practices, and no requirement that any of them get enormous to have been worth doing.