The Practice
Fig. ¶ · the practice, London 2026I run a practice of one. Six websites a year, each given the attention the work actually needs — thought through before it is built, and built by the same person who reads your emails. Smaller and slower than the industry around it, on purpose
The industry I came up in ran the wrong way
I started building websites in 2014, in London — mostly for small businesses that needed someone to think about what they wanted to be before what they wanted to build. The work was honest. The sites were fine. And after eight years it was clear the industry around me was running the wrong way.
Agencies got bigger, faster, more diluted. Templates replaced craft. Junior teams replaced direct relationships. The client who walked in wanting their site walked out with the agency’s site, slightly recoloured. I kept watching it happen — and kept not enjoying my own contribution to it.
Goliathus is the practice I should have been running for the last four years. Smaller. Slower. Six projects a year, each given the attention the work actually needs. I think this is what most premium clients have wanted all along. They simply couldn’t find it without a referral from someone they trust.
Eight years inside the industry. Then a decision: build the practice it should have been — smaller, slower, by hand
Mornings, mostly. One hand, extended by software
I work mornings. Six days a week during an active project, with Sunday off as a hard rule — anything else stops being sustainable. The work happens in long, uninterrupted blocks: three or four hours at a stretch. Fewer meetings than most agencies would call professional. More written updates than most agencies bother with.
I use AI agents to extend my reach. Not a marketing line — an operational fact. Claude Code drafts boilerplate while I think; automation handles what doesn’t deserve human attention. But the decisions stay mine. The editorial judgment stays mine. The client conversations are mine. The production work that would normally need junior staff is handled by software, supervised by me, and captured as reusable patterns for the next project.
What that buys: one person’s full attention for the length of a project, with the output of a small team. Faster than agencies of comparable price. Slower than agencies of comparable size. The combination is the point.
A practice, not an agency. A bet, not a growth story
This is a practice — in the way a doctor’s office or an architectural firm is a practice. A small, sustained operation built around one craftsman’s judgment, applied repeatedly to similar problems. It isn’t a studio in the usual sense, and it isn’t an agency at all
This is the bet I’m making with the next decade: that as AI tools collapse the cost of a competent website to near zero, a tier of clients will remain who want the work done with care. Small enough for one person to serve. Discerning enough to pay for the difference. The work that lives between “the AI did it” and “the team did it” is the work I’m interested in.
If that sounds like a small bet — it is. If it sounds like a slow bet — that too. I have no intention of growing this past one person plus agents. If demand exceeds capacity, the waiting list extends. If it falls short, I write more, and build more deliberately.
Every project here is given the whole of one person’s attention — kept small on purpose, so that it can be
Goliathus London