Should we hire a design studio, a freelancer, or an in-house designer for our website redesign?

All three routes can produce a good site at launch. We have inherited sites built by freelancers, by in-house designers and by other studios, and at the point of handover it is often hard to tell which was which. The differences appear somewhere between month six and month twelve, when the site is asked to do something nobody scoped: a second product line, a careers section, a pricing change that needs a comparison layout, a campaign page due Thursday.

That moment is the real test, and it is a test of ownership rather than talent. Someone has to decide whether the new page type is built from existing components, or whether a new component enters the system and gets documented. When nobody owns that decision, the answer defaults to “copy the closest page and override the bits that look wrong”, and the site starts accumulating one-off styling. In our experience that is what makes a two-year-old redesign feel tired: not the visual design ageing, but forty pages of small deviations from it.

So the comparison worth making is not price against quality. It is which of the three routes gives you a person or artefact that survives to month nine, and what each one costs you when it doesn’t.

• Freelancer: strong at month one, breaks at the first page type nobody scoped.
• In-house designer: owns the site indefinitely, but system maintenance is always the task that gets bumped for this quarter's campaign.
• Studio: builds the system properly, then goes quiet the moment the engagement or retainer ends.

Based on projects we have quoted, taken over or been shown competing quotes for, a UK marketing site of roughly 10 to 25 pages sits in these bands. Treat them as the middle of the market rather than the extremes; you can always find cheaper, and enterprise agencies start where studios stop.

The numbers that catch people out are not the build fees. They are the second-year costs: the freelancer who is no longer available at the rate you agreed, the in-house designer who needs a front-end developer to ship anything, the studio retainer you cancelled in month four because nothing was happening.

One cost applies to all three and is almost never budgeted: your own team’s time. Whichever route you take, the content decisions, approvals and copy remain yours, and that is usually the constraint that sets the timeline rather than the designer’s capacity.

• Freelance designer: £4,000 to £15,000 for design and build, six to ten weeks. Usually one person doing both, which is why it is cheap and why it is fragile.
• Freelance designer plus freelance developer: £10,000 to £25,000. Better output, but you are now the project manager and the person who resolves disagreements between them.
• Studio: £18,000 to £60,000, typically including a component set, tokens and documentation. Timelines vary widely by studio and scope, from a couple of weeks to a few months. Ongoing support is a separate arrangement, usually £800 to £3,000 a month.
• In-house designer: £45,000 to £70,000 salary plus employer costs, plus a developer to build what they design, plus three to five months before the first page ships. Cheapest per year at scale, most expensive for a single redesign.

The freelancer failure mode we see most often is a site that was scoped as a page list rather than a system. Nine pages were agreed, nine pages were designed, nine pages were built, and each one is genuinely good. Then the business needs a tenth page of a kind that was never discussed, and there is no rule for how it should look, because the rules were never written down. They lived in the freelancer’s head, and the freelancer’s contract ended.

What we then inherit is a site where the header, footer and typography are consistent and everything between them is bespoke per page. Spacing is hand-tuned in each template. Two buttons look identical but are built twice with different padding. Adding a page costs a designer’s day rather than an afternoon of assembling existing parts.

This is not a criticism of freelancers as practitioners; some of the best design work we take over came from one. It is a scoping problem. A single person paid for nine pages will sensibly deliver nine pages, and a system is more work than nine pages. If you take this route, buy the system explicitly or accept that the site will only ever have the pages you paid for.

An in-house designer is the only route where someone is still there in month nine, which is why it looks like the safe choice. The failure mode is different: the site drifts, because keeping the design system tidy is never the most urgent thing on that person’s list. There is always a campaign, a deck, a trade show banner or a product launch with a date attached, and refactoring three near-identical card components has no date attached.

We have picked up in-house-built sites where the token file is genuinely good and genuinely six months out of date. Nine greys are defined and fourteen are in use. The button component exists but three pages use a hand-written one because the component didn’t support an icon on the day someone needed an icon. Each of those decisions was rational under deadline pressure; together they mean the system no longer describes the site.

The other constraint is breadth. Website redesigns need design, front-end engineering, content structure, accessibility and performance work. One in-house designer covers one and a half of those. If they are your only route, they need a developer beside them and explicit protected time for system maintenance, written into their objectives rather than assumed.

Studios, including us, have a specific failure mode: the site is at its best on launch day and receives no design attention afterwards. The system is built, the components are documented, the handover happens, and then the studio’s attention moves to the next client. Six months later your team has added pages, and whether those pages respect the system depends entirely on how well the documentation was written and how motivated your team is to read it.

The honest version of this is that a design system nobody on your side owns will be ignored within two quarters, no matter how well built. We have handed over systems and come back a year later to find them used exactly as intended, and we have handed over systems that were bypassed within a month. The difference was never the quality of the documentation. It was whether one named person on the client side had responsibility for saying yes or no to new components.

The corollary: if you hire a studio, do not treat the retainer as insurance against staleness. A retainer only helps if there is scheduled work in it. “Ad hoc support as needed” reliably means nothing happens until something is broken.

Each route can work. Each one needs one specific thing to exist in writing before the engagement ends, and in our experience its absence is what predicts the site going bad. Ask for these by name during negotiation, not at the end.

If you cannot get the relevant artefact from the person you are hiring, that is useful information about which route you are actually on.

• Freelancer: a written page-type inventory that says how a new page is built from existing components, plus the design file with named, reusable components rather than flattened page mockups.
• Freelancer pair (designer plus developer): a named owner of the boundary between design and code, written into the contract, so nobody has to arbitrate whose job the gap was.
• Studio: a documented component set with tokens, and one named person on your side who has been trained on it and holds the yes-or-no authority on new components.
• In-house designer: protected maintenance time in their objectives, expressed as a recurring block rather than a percentage, and a second pair of hands for front-end work.

Take the freelancer route when the site is under roughly 15 pages, changes a few times a year, and you can live with paying someone each time a new page type appears. It is genuinely the right answer for a lot of small businesses and the wrong answer for anything expected to double in size.

Take the studio route when the site has multiple page types, more than one person editing it, and a two-to-four-year horizon before the next redesign. You are paying for the system and the documentation more than for the pages, so judge studios on what they hand over rather than on their portfolio screenshots.

Take the in-house route when the website is one of several ongoing surfaces the business needs designed, and there is enough work to fill a role. A single redesign does not justify a hire; a redesign plus product UI plus campaigns plus continuous iteration does. Many companies land on the sensible hybrid: a studio builds the system, an in-house designer or marketer maintains it, with a small scheduled block of studio time each quarter for the things the system does not yet cover.

If you are weighing this up and want a second opinion on which route your site is actually asking for, send us the URL and a rough page count. We will tell you plainly if a freelancer is enough, because a site that only needs nine pages is not a project we should be quoting for.