Taking my own product to its first hundred customers
Project 5 is mine. No client, no brief, and nobody else to blame for the roadmap.
That turned out to be the point. On client work I was responsible for a slice of the problem. Here I was responsible for the whole loop: understanding the problem, building something, getting it in front of people, charging for it, and supporting it afterwards.
Next.js, Node.js, PostgreSQL, Stripe
The first version was too good
I spent four months building the first version. It had accounts, billing, a settings page, an audit log and a dark mode.
It had two users.
I had optimised for the version of the product that would exist if it succeeded, rather than the version that would find out whether it could.
The features I had been proud of were irrelevant to the question I actually needed answered, which was whether anyone would pay for the core thing at all.
What worked instead
I cut the product down to one screen that did one thing, put a price on it, and spent as much time on distribution as on code.
- Wrote about the problem in the places the people with the problem already were. Slower than advertising, and the only channel that still works two years later.
- Talked to every one of the first thirty customers personally. Half the roadmap came out of those conversations.
- Raised the price twice. Both times, conversion went up rather than down.
What running it taught me
Support is product work. Every recurring question was a piece of the interface failing to explain itself, and fixing the interface was almost always cheaper than answering the question again.
Pricing is positioning. The cheap price was not attracting careful customers, it was attracting people who did not have the problem badly enough to care.
Where it is now
A little over a hundred paying customers, running at a level that covers itself and some of my time. Small by most measures, and the most useful thing I have built for my own understanding of the job.