How do you choose what to sell online?

Start from repeat purchase and work backwards. Something bought monthly at a decent margin will build a business faster than something bought once at a spectacular one.

What decides it

Then filter for three practical things: can it be shipped without damage, is it hard to price-compare in ten seconds, and do you know anything about it that a competitor does not. Failing the third is survivable; failing the second is not.

The part that gets skipped

The best ideas usually come from a job somebody already had. You notice a problem because you lived inside it, not because you brainstormed.

A market that already spends money is easier than a market you have to convince exists. Competition is evidence, not a warning.

Test whether people will pay before building anything. A hundred people saying they like it is worth less than one paying.

Where an app fits

If the idea is a shop, the sequence matters: sell first, prove people come back, then build the app. An app before the customers is a cost with nothing to do.