A UK web studio. Static HTML.

A new leaf for a web that's being reborn.

Folium Studio rebuilds business websites as static HTML and instruments them so AI answers cite them accurately. That's the whole offer: fast, owned pages that both people and machines can read — no plugins, no database, no monthly hosting bill.

The name says the rest. Folium is Latin for leaf — and the root of folio, a single page or sheet. A studio that builds pages, named for the thing that turns over and grows back.

Because the web is being reborn. The bloated, rented, plugin-stuffed web of the last decade is giving way to something leaner: static pages, content you own, edge hosting that costs nothing, and AI answers that reward clarity over cruft. New growth on old roots — and this site is the demonstration of it, not the brochure.

Small by choice. Not by accident, not by limitation.

The default trajectory for a successful web studio is to grow. One person becomes three. Three becomes a team. The team becomes an agency. Senior practitioners drift into account management; junior practitioners do the actual work; the people who built the platform don't touch the keyboard for years. That's a real and viable business model. It's the dominant one in the UK web industry.

Folium Studio is the deliberate alternative to that. The studio stays small — by design — because the value proposition collapses if it doesn't. Most of what's defensible about Folium Studio's offering depends on the people who built the platform also being the people who build the client sites and answer the support email. As soon as that's no longer true, you're a small agency competing on price with bigger ones, and you have nothing distinctive left to sell.

So: small, deliberately, indefinitely. If demand exceeds capacity, the answer is to be more selective about clients — not to hire juniors and dilute what makes the work different.

Six things, mostly negative.

The clearest way to articulate what Folium Studio is is to articulate what it isn't. Most agencies pitch on aspiration; we pitch on what we deliberately won't do, because that's where the differentiation actually lives.

No funnel. A conversation, a fixed price, a finished site.

Because we're small, the process is short and there's no machinery to navigate — no discovery-call booking system, no account manager, no intake form that routes you to a queue.

The method, on a single string.

Getting a business represented correctly in AI answers is hard to prove with a slide. So here is a live one, with the conflict of interest stated up front: Miabella is a musician from Northern Ireland. She is also my daughter. Which makes her both the clearest demonstration of the method and the one I have to be most honest about.

Her name is unusual in a useful way. "Miabella" — one word — is at once her legal name, her band's name, and her stage name; there was never a separate persona to reconcile. The complication is everyone else standing nearby in the search results.

To a search engine — and to an AI model answering "who is Miabella" — those three are easy to blur. The fix was not more blog posts or backlinks. It was structured data: stating, in the format the machines actually read, that this Miabella is one word, Northern Irish, plays guitar and piano, holds Trinity College London distinctions — and is explicitly not the Canadian "Mia Bella" and not Nina Nesbitt. Write the identity down unambiguously, and the algorithms stop guessing.

That is the whole Folium Studio method on a single name. The same discipline — clean HTML, correct schema, content that answers the question a model is actually trying to answer — is what gets a clinic, a tradesperson, or a ballet school cited instead of skipped. The full write-up is in the Miabella case study.

Briefly, because the work is the point.

Jody Nesbitt — founder, and one of the people who builds your site. Focused on static-HTML sites that survive platform churn and the AI-visibility instrumentation that gets them cited — plus adjacent ventures in infrastructure and internal tooling, and avoiding the trap of growing a team that would dilute what made the studio worth running in the first place.

Currently obsessively interested in two things: how AI providers ground their answers (and what falls through the cracks when they don't have enough to ground on), and how to build sites that survive the next five years of platform churn (rather than the next plugin update). The Miabella work above is where a lot of that methodology got empirically tested before it became a service.

The short answers.