Will this actually save someone an hour tomorrow?
If the answer is no, I don't build it. The best software I've made is measured in hours-not-spent, not features-shipped.
Xzist Digital is a tiny UK practice building software and AI tools for the construction trades. No office plants, no account managers, no decks. Just the people writing the code, talking to the people using it.
He ran a joinery. Good at the work, bad at the paperwork, paying a bookkeeper to untangle a year's worth of receipts every January. Off-the-shelf software didn't fit. Big consultancies wanted fifty grand up front.
I built him a quoting tool in a fortnight. It wasn't clever. It just matched how he worked. Three years later his Sundays are his own and I've been building those kinds of tools for the trade ever since.
I stay small because it's the only way to keep doing the work properly. Two or three live engagements at a time. No growth for its own sake. No VC. No roadmap to sell you a platform next year.
When the shape of a project is unclear, these are the three questions I come back to.
If the answer is no, I don't build it. The best software I've made is measured in hours-not-spent, not features-shipped.
Features invented in a meeting room almost always die on site. I build what I've seen someone struggle with, not what sounded clever in a brief.
Throwaway software is fine for a prototype, bad for a business. I write things to last and to be cheap to change later.
Who you'll talk to
Founder. Writes most of the code. Answers the phone.
Thirty minutes, no pitch. I'll ask a lot of questions. You decide if I'm the right person for what you're trying to do.