Why we don't do templates — and never will

When you're building a website product, templates are the obvious move. They're the thing users expect. They're what every competitor offers. They de-risk the first experience — give someone something to start from, something that already looks good. The logic is airtight.

We spent weeks making the case for templates internally. And then we made the decision not to build them — not now, not later, not ever. Here's exactly why.

Templates are a head start that becomes a prison

The problem with templates isn't how they start — it's how they end. You choose a template because it looks close to what you want. But close is not the same as right. And the closer you get to right, the more you run into the template's constraints.

The font is almost right but not quite. The section spacing is off. You need a layout that the template's grid just doesn't support. You want to change something that's tied to something else you don't want to change. And six hours later, you've spent your entire afternoon fighting the tool instead of building your site.

The template isn't the problem. The assumption embedded in the template is the problem: that your site should look like one of these options, within these constraints, at this level of customisation. That's not a head start. It's a ceiling with a nice lobby.

What we built instead

Instead of templates, Ullbek builds from a brief. You describe what you want — your business, your aesthetic, your audience, your feel — and the agent produces a first draft that's unique to you. Not a template with your logo swapped in. A site designed, from scratch, for what you actually said.

No ceiling. Ever.

The other reason we'll never do templates: they impose a ceiling on complexity. Templates have to be maintainable. They have to be understandable by the template editor. They have to be limited to the set of features the template system supports.

Ullbek writes code. Real, full-spectrum HTML, CSS and JavaScript. Which means there is no ceiling. If you can describe it, it can be built. A portfolio with a custom cursor. A calculator embedded in a service page. An interactive pricing comparison. A micro-animation on scroll. None of those require a "template for that" — they just require describing what you want.

Templates would undermine that. They'd make us a better version of what already exists. We're not trying to be better than the existing options. We're trying to make the existing options irrelevant.