Find a real itch
I only start things I'd use myself. The first filter isn't market size on a slide — it's whether the problem keeps nagging me.
I find problems worth solving — then ship the smallest thing that actually works.
No roadmap theatre. A tight loop from itch to shipped, repeated until something sticks.
I only start things I'd use myself. The first filter isn't market size on a slide — it's whether the problem keeps nagging me.
The smallest build that proves the core loop. No settings page, no onboarding, no landing page until someone actually wants it.
Ten real users beat a thousand impressions. I watch exactly where they get stuck and delete everything that isn't the point.
Ship weekly, talk about it, and let real traction — not my opinion — decide what earns the next month of work.
Three bets currently eating my evenings. The live versions live on the home page.
An AI website builder that designs and publishes a complete site in about 30 seconds.
02 · BuildingA marketplace to ask a neighbour to cook for you — delivered, or made fresh in your kitchen.
03 · EarlyOn-demand home cleaning — a vetted cleaner booked in a few taps.
The rules I keep coming back to.