The case for custom websites in 2026.

Let's start with the uncomfortable admission. Template platforms got really, really good. Webflow, Framer, Squarespace, Shopify — the gap between what they produce and what a bespoke build produces has shrunk every year for a decade. So what are we still doing here?

We should answer that honestly, because it matters. If a template can solve your problem, a template should solve your problem. We don't exist to generate work for ourselves.

Here is what we have actually learned.

What templates got right

Responsive design is solved. Nobody should be hand-coding breakpoints for a standard content site in 2026. Page builders ship with sensible defaults, clean grid systems, and typography that isn't embarrassing.

Forms, CMS, hosting, SSL, deployment, image optimisation, basic SEO, analytics — all of it is plug-and-play. You can launch a perfectly fine brochure site in a weekend. If that's what you need, go do that. Don't hire a studio.

And crucially: a well-made template site can perform excellently on search, look presentable on mobile, and serve its owner for years. Template doesn't mean bad. The insistence that "custom is always better" is mostly agency marketing, and we should stop believing it.

What they still can't do

Four things. We think about these in this exact order when a project comes in.

1. Performance at the edges.

Templates carry their generality as weight. A page builder has to handle ten thousand possible layouts, so every page ships the code for a hundred layouts you never use. You end up with four megabytes of JavaScript to render a paragraph and an image.

For most businesses this doesn't matter. For a site that's trying to rank for competitive keywords, convert visitors from paid ads, or serve a market where half your visitors are on a slow connection in a car — that four megabytes is money. You'll pay for it every single month in lost sessions, worse quality scores, higher bounce rates. A hand-built site at 80 kilobytes with the same visuals will out-convert it on a long enough timeline. We've measured this.

2. Distinctiveness.

Every industry has a dominant template aesthetic now. Restaurants all have the same hero video. Dentists all have the same "smile" stock image grid. Consultancies all have the same geometric accent shapes. Open any competitor's site and your visitor's brain classifies it in under two seconds: this is one of those.

You can't escape that with another template. You can only escape it by doing the thing templates are built to prevent: small, deliberate, cumulative oddness. A custom typeface pairing. A grid that breaks at an unexpected moment. An interaction that no tool ships with by default. None of these are possible in a page builder — not really, not at the fidelity that matters — because templates are designed to prevent exactly that kind of oddness.

Templates are designed to prevent the kind of oddness that makes a site memorable. That's a feature for 99% of use cases and a problem for the last 1%.

3. Long-tail maintenance.

A five-year-old template site is usually not the site it started as. The platform was acquired, the theme was deprecated, a plugin conflicted with an upgrade, the CMS changed URL structure, the migration tool didn't work on one of your product categories. We've been hired to untangle this exact situation maybe twenty times now.

A hand-built static site from 2021 with clean HTML and CSS will run forever. Literally forever. There is no maintenance treadmill because there is nothing to maintain. You'll update content, you'll occasionally add a page, you'll redesign when you want to — but you won't be forced into anything by a platform you don't control.

4. The specific backend that matters to you.

This is where most of the actual custom work lives. If your business has one genuinely unusual workflow — a booking system with dynamic seasonal pricing, a quote builder with three hundred SKUs, a multi-location scheduler, a niche regulatory form — the template can handle the brochure around it, but the backend needs to be yours.

And once the backend is custom, the front-end might as well be too. The cost of bespoke on top of bespoke is much less than the cost of custom interacting with template, which is a nightmare to maintain.

When custom doesn't make sense

We turn these projects down or redirect them:

• A simple personal portfolio or brochure. A Webflow or Framer template will look fine, cost a fraction, and serve you for years.
• A standard e-commerce shop with under 50 products. Shopify does this better than we ever will. The tooling, the payments, the inventory, the shipping integration — all of it.
• A blog where the goal is "just ship posts." Use Ghost or Substack. You're a writer, not a studio client.
• A landing page for a campaign that lives for six weeks. Framer. Done.

Our heuristic

When someone gets in touch, we ask one question before we start talking about the work: will this site still need to be here, unchanged at its core, in five years?

If yes — custom. The up-front cost amortises over a long time, and you don't want to rebuild this site every eighteen months because a platform shifted under you.

If no — don't pay for custom. The up-front cost doesn't justify the lifespan. Go template.

Most of our clients fall clearly on one side. The ambiguous ones are usually better served by a blunt conversation about what they actually want than by our sales pitch. A studio that genuinely tries to be useful will send a handful of these to Webflow every year and sleep well.

The custom work is there for people who need it. It's not better by default. It's better when the answer to that one question is yes.