Loading...
AI can generate a competent website in an afternoon. When "good enough" is free, the only thing worth paying for is a site that doesn't look generated.
Something has changed in the last eighteen months, and most business owners haven't clocked it yet. You can now describe a website to a machine — a plumber in Crewe, a clinic in Nantwich, a Shopify store selling candles — and get back something that works. Real pages, a tidy layout, copy that reads fine, a contact form that submits. Wix builds it for you. Framer generates it. You can talk a general-purpose AI through the whole thing over a lunch break.
And here's the uncomfortable part: it's fine. Not good. Fine. The bar for a passable website has collapsed, and it's still falling. Our honest read is that within a couple of years, most new small-business websites will start life inside an AI builder rather than on a designer's desk.
That's not a threat we're bracing against. It's the single best thing to happen to real design in a decade. Let us explain why.
For twenty years, a competent website was an advantage. If your site loaded, looked current and explained what you did, you were ahead of the half of your market still running something built in 2014. "Good enough" won business.
That advantage is evaporating. When anyone can generate a competent site in an afternoon, competent becomes the floor — the price of entry, not the edge. The economics are simple and they're not on the side of the average website: the moment something becomes free and instant, it stops being worth paying for. The value migrates to whatever the machine can't do.
So the real question for a business owner in 2026 isn't "can I get a website cheaper and faster?" You can. It's "if my website looks exactly like the output everyone else can now generate for free, what is it actually doing for me?"
Here's the trap in generated design, and it's structural, not fixable with a better prompt. An AI builds by averaging. It's trained on millions of existing sites, so it produces the statistical middle of everything it has seen — the same hero layout, the same three-column feature grid, the same rounded cards, the same stock-photo mood, the same tidy, confident, anonymous tone. It is engineered to produce the expected thing. That is the opposite of what a brand needs.
The result is a web where everyone in your category looks the same. The candle shop looks like the other candle shop. The consultancy looks like every consultancy. And when you look like your competitors, you've handed the customer a decision made purely on price — because you've given them no other reason to choose. Sameness doesn't just look cheap. It makes you invisible, and invisible is expensive.
We build for UK businesses who are trying to be chosen, not just found. For them, a website that blends in is a website that isn't working — however clean it looks.

The generated web: eleven competent, forgettable sites — and one built on purpose.
Design isn't decoration. Good design is a sequence of decisions — what to say first, what to leave out, where the eye goes, what to make big and what to make quiet, which single idea the whole page is built around. Art direction is having a point of view and the discipline to commit to it.
An AI has no point of view. It has an average. Ask it for "a bold, distinctive site" and it gives you the average of every site ever described as bold — which is, by definition, not distinctive. The very mechanism that makes it fast is the mechanism that makes it generic. You cannot prompt your way out of the middle, because the middle is where the model lives.
This is the whole reason we say our work is art-directed, not generated. Not as a slogan — as a description of where the value actually comes from. A person deciding, on purpose, that this brand should feel calm and expensive while that one should feel fast and loud. That judgement is the product. It's also the one thing the software can't hand you.
The lazy take is that AI is coming for designers. The truer take is that AI is coming for average design, and it's going to be ruthless about it. Templated, undirected, could-be-anyone websites are about to be worth roughly what it costs to generate them: nothing.
But when the baseline becomes free, the premium on the exceptional goes up, not down. Photographers didn't become worthless when every phone got a good camera — the ones with a real eye got more valuable, because the flood of competent snapshots made a genuinely directed image stand out more. The same correction is coming for the web. When every competitor has a generated site, the business with a deliberately designed one doesn't look 10% better. It looks like it's playing a different sport.
That's the bet we're making, and we're comfortable making it in public: the generated web makes art direction the scarcest, most valuable thing on it.
None of this is a refusal to use the tools. That would be daft, and it would be dishonest — we use AI every day. It drafts, it explores, it handles the repetitive middle of a build so the studio's hours go where they matter. Used well, it's a brilliant junior that never tires.
The distinction that matters is simple: AI belongs in the process, not as the product. A generated site is a machine's average, shipped. An art-directed site is human decisions, with the machine doing the fetching and carrying underneath. One is what the software wanted to make. The other is what you needed to say. Same tools, opposite outcomes — and the customer can feel the difference even when they can't name it.
You don't need our eye to spot it. A few honest tells:
If your site fails those, the fix isn't a better prompt — it's a point of view. That's the whole job, and it's the way we design a website: every build art-directed by a person, then engineered to be fast and findable (Awwwards-listed studio; sites that hold a 93+ performance score and load in under two seconds). The speed and the SEO are table stakes we take seriously. The distinctiveness is the part nobody can generate for you.
AI has made a competent website free. That's genuinely good news — for your competitors who were going to settle for competent anyway. It quietly retires the tired, template-filled site as a category. What it can't do is give your business a point of view, a decision, a reason to be chosen over the identically-competent shop next door. The web is about to be flooded with the average of everything. In a flood, the thing that stands out is the thing built on purpose. That's still a human job. We think it's about to become the only one worth paying for.