Building for yourself is a good start and a risky finish

Joshua Schachter built Delicious because he had thousands of bookmarks and no good way to keep them. Craig Newmark started Craigslist as an email list of events for friends in San Francisco. Michael Arrington, who founded TechCrunch and spent years watching new companies pitch, saw the same pattern often enough to call it general.

The advantage of an itch is information. A founder who has the problem knows how painful it is, what the existing workarounds look like, and which features actually matter. They can judge each version of the product by using it. They are also less likely to give up, because they want the thing to exist whether or not it makes money.

The weakness is sample size. An itch felt by the founder may be shared by millions or by a few hundred people with similar habits. Plenty of tools built by programmers for programmers found enthusiastic users and no business. Many large markets, such as hospital billing, freight logistics or insurance claims, are full of problems that most founders will never feel personally.

There is a quieter risk too. A founder who is also the main user tends to build for their own level of skill. Delicious made perfect sense to people who already thought in tags, and much less to everyone else. The product that scratches the founder’s itch precisely can feel foreign to a customer with the same problem and different habits.

The rule works best as a first filter. Start from a real problem you know well, then check early whether enough other people have it. The strongest signal is a workaround: people who have built their own spreadsheet, script or paper system to cope with the problem are already paying for a solution with their time. Ask twenty of them what they use now and what it costs them. If the answers are vague, the itch may be yours alone.