Too many websites that get a full redesign didn’t need one. They needed a content refresh, a performance pass, and perhaps a new template on the same CMS. The real question isn’t ‘redesign or refresh’ in the abstract; it’s whether the problem you’re trying to fix lives in your content, your speed, your conversion path, or your site’s underlying structure, because only the last of those reliably justifies starting again. And the people best placed to advise you are usually the ones with the most to gain from the expensive answer.
This piece gives you a way to work out which problem you actually have before you brief anyone. Three tiers: when a refresh is enough, when a partial rebuild makes sense, and when a full redesign is genuinely justified. If you get the diagnosis right, you avoid paying rebuild money for a refresh problem, and keep everything the current site already does well.
The three tiers, escalating in cost and risk. A refresh keeps the platform and ranking URLs in place; a partial rebuild replaces the front end on the same foundation; a full redesign is reserved for when the foundation itself is the problem.
Why almost everyone you ask has a reason to say ‘redesign’
Three of the parties you’re likely to consult all have an incentive that points towards the biggest possible project. A redesign is a high-value engagement for an agency. A migration is a licence sale for a platform. And a brand-new site is simply more interesting to work on than a tidy-up. None of that makes anyone dishonest, but it does mean the default advice skews expensive, and you should price that bias in before you read a single proposal.
Look at the numbers from the agency’s side. In practice, a full custom redesign for a mid-market business is often a low-five-figure project at minimum, and £15,000 to £40,000 ($20,000 to $54,000) is a realistic planning range before ecommerce, integrations or complex migration work push it higher. A refresh of the same site, a content rewrite plus a performance pass and some design tweaks on the existing template, might be a few days of work, often sitting inside an existing retainer. One of those is a quarter you build a business around. The other is a Tuesday. When a client emails asking whether they ‘need a new site’, you can guess which answer most studios are quietly hoping for.
Platform vendors have the same pull in a different direction. A salesperson at an ecommerce or CMS company is measured on migrations and seat licences, not on talking you out of one. If you ask whether you should move to their platform, the honest answer is sometimes no, and that answer doesn’t appear on anyone’s commission sheet.
Then there’s your own team. The marketing lead who inherited a site they didn’t build is often itching to put their stamp on something. A redesign is a visible, ownable project. A quiet content rewrite that improves conversion is harder to put in a board update, even though it might be the better use of the budget.
None of this means you should distrust everyone. It means the cheapest, smallest fix has nobody arguing for it, so you have to argue for it yourself. A second opinion from someone who isn’t quoting for the work, an in-house developer, a fractional CTO, or simply a different agency asked to scope a refresh rather than a rebuild, is worth more than another redesign proposal. Before any of that, it’s worth knowing what a sensible monthly maintenance retainer should already be covering, because a good chunk of what gets sold as a redesign is work that should have been happening anyway.
What problem are you actually trying to solve?
Before you can choose between iterating and rebuilding, you have to name the problem in concrete terms. Almost every complaint about a website reduces to one of four things: the content is wrong, the site is slow, it doesn’t convert, or its structure can’t support what the business now needs. The first three are nearly always fixable without a rebuild. The fourth is the one that justifies starting over.
‘The site looks dated’ is the most common brief, and the least useful. Dated how? If the typography and imagery feel ten years old but the pages load fast and the structure makes sense, that’s a design refresh, not a teardown. If visitors can’t find what they came for, that’s an information architecture problem, which sometimes means a partial rebuild and sometimes just means reorganising the menu. Pin down the symptom and the right tier usually names itself.
Speed is the easiest to diagnose and the most over-treated. A slow site is a performance problem, and performance problems are fixable on the existing build far more often than not: image optimisation, caching, a lighter theme, fewer plugins, a better host. Google’s own documentation confirms that Core Web Vitals are used by its ranking systems, while making clear there’s no single page-experience signal and that relevant content matters more. So a slow site can cost you rankings at the margin, but the fix is a few days of optimisation, not £30,000 ($40,000) and four months. Rebuilding a site to make it faster is like buying a new car because the old one needs a service.
Conversion is the same story. If traffic arrives but doesn’t act, the problem is usually the page: weak copy, a buried call to action, a checkout with too many steps, a form nobody wants to fill in. These are testable, fixable changes on the live site. A redesign that ignores why the current pages underperform tends to ship the same mistakes in a nicer font.
That leaves structure. When the way the site is built genuinely can’t support what the business now does, no amount of refreshing gets you there. That’s the real trigger for a rebuild, and it’s worth being honest about how rarely it applies.
When a refresh is enough
A refresh covers everything you can change without touching the site’s structure or platform: rewriting content, updating the design within the existing template, improving page speed, fixing the conversion path, sorting out accessibility. If the foundations are sound and the site simply looks tired or reads badly, this is almost always the right call. It’s cheaper, it’s faster, and it protects more of the SEO equity you’ve spent years building.
The work usually looks like some combination of a content audit and rewrite, refreshed imagery and brand colours applied within the existing theme, conversion-rate tweaks on the pages that matter, a performance optimisation pass, and the security and dependency updates that should have been happening all along. Most of it can be done in weeks rather than months. A meaningful refresh of a small-to-mid site usually lands in the low thousands rather than the tens of thousands; £2,000 to £6,000 ($3,000 to $8,000) is a sensible planning range when the work is content, conversion and performance rather than a rebuild, and a lot of it can roll into an existing retainer.
The decisive advantage is what you don’t lose. Your URLs stay put. The pages that already earn organic traffic stay live. Provided the refresh doesn’t gut the content, you’re protecting a known source of search visibility rather than gambling on a new structure. For a lot of businesses, the most profitable web project they can run this year is a ruthless content rewrite and a speed pass on the site they already have, and nobody will ever pitch it to them because it doesn’t sound like much.
A fair amount of refresh work doesn’t need an agency at all. If you’ve got someone in-house who can update content and run a theme, you may be able to handle most of it yourselves and only buy in the specialist bits. We’ve written separately about when you don’t actually need a web agency; the redesign question is one of the places that judgement pays off most.
When a partial rebuild makes sense
A partial rebuild sits between a refresh and a full redesign. You keep the platform, the content, and ideally the URL structure, but you rebuild the front end: a new theme or template, a reworked information architecture, a fresh set of the key page types. This is the right move when the design or the structure is genuinely holding you back, but the foundations underneath are still sound.
In practice that often means a new theme on the same WordPress or CraftCMS install, a reorganised navigation and information architecture, and a rebuild of the templates that actually drive the business, the homepage, the main landing pages, the product or service templates, while the long tail of older pages carries on unchanged. The database stays. The content stays. The platform stays. Because you’re not replatforming, the migration risk is far lower than a full rebuild, and you can stage it page-type by page-type rather than flicking a switch on launch night and praying.
Agencies that inherit and stabilise existing builds see this pattern often: the brief says ‘new website’, but the diagnosis is a new front end on a serviceable back end, plus the content rewrite the client was trying to avoid. A partial rebuild lets you fix the genuinely broken parts without paying to replace the parts that work.
The trap to watch for here is scope creep dressed up as necessity. Once you’ve agreed to rebuild the homepage, it’s an easy sell to rebuild ‘while we’re in there’, and a £6,000 ($8,000) partial job quietly becomes a £25,000 ($34,000) full one. If you go this route, write down which templates are in scope and which are explicitly out, and treat any addition as a change request with its own line item, not a freebie.
When a full redesign is actually justified
A full redesign means a new platform or build, new structure and fresh foundations. It is justified when the foundation itself is the problem. The clearest triggers are a platform that can no longer be maintained or secured, a business model the current site physically can’t support, or a structure so broken that fixing it piecemeal would cost more than starting again. If your reason for rebuilding isn’t on that short list, be suspicious of it.
A platform at the end of its life is the cleanest case. If your site runs on a bespoke CMS the original developer abandoned, or a version of a framework that no longer gets security patches, you’re not choosing to rebuild so much as being forced to. The same goes for a WordPress build so dependent on abandoned plugins that routine updates become risky, and maintenance starts costing more than a clean rebuild.
The second real trigger is a genuine change in what the site has to do. Moving from a brochure site to ecommerce. Adding subscription billing, where something like a WooCommerce and Chargebee setup needs to be wired in properly rather than bolted on. Standing up a customer portal or login area the current architecture was never built for. Going multi-language and multi-site for international audiences, the kind of structural requirement that may justify CraftCMS, WordPress multisite, a headless CMS or another platform built for that complexity. When the business has outgrown the foundation, refreshing the paint doesn’t help.
The third is technical debt that’s reached the point where piecemeal fixes cost more than a rebuild. This is a real situation, and it’s also the one most often invoked without evidence, so make the agency show their working. If they can quantify the maintenance burden, the recurring breakages, the hours lost to a fragile codebase, fine. If ‘too much technical debt’ is doing the heavy lifting in a proposal with no numbers behind it, it’s a vibe, not a diagnosis.
Notice what isn’t on the list. ‘It looks old’ isn’t here; that’s a refresh or a partial rebuild. ‘A competitor relaunched’ isn’t here. ‘New marketing director wants a fresh start’ isn’t here. A full redesign at this tier is often a low-five-figure project at minimum, and integrations, a platform move and content work can push it much higher. It takes months, and it carries real migration risk. That price is worth paying when the foundation is the problem. It’s a poor trade when it isn’t. If you’re weighing a platform move as part of this, our breakdown of WordPress versus Webflow for a business website is a useful place to sense-check whether the destination platform earns the migration.
What does redesigning unnecessarily actually cost you?
The cost of an unnecessary redesign isn’t only the fee on the invoice. You put your search visibility at risk, you can lose rankings during the migration, you discard content that was already working, and you spend months of internal time on a project that delivers little the old site didn’t. A redesign that goes wrong can leave you measurably worse off than the tired site you started with.
The SEO risk is the one buyers underestimate most. A rebuild changes URLs, restructures the information architecture, and often quietly drops pages that were earning steady organic traffic. Done without a meticulous site-move and redirect plan, that traffic may not move cleanly to the new pages; some of it can disappear, and recovery can take months. The bigger your existing organic footprint, the bigger the downside, which produces the uncomfortable rule that the sites most worth being careful with are exactly the ones agencies are keenest to redesign. Ask who owns the redirect map before anyone touches a wireframe.
Content is the next thing to go. Rebuilds ship on a deadline, the rewrite gets deprioritised when timelines tighten, and a surprising number of redesigned sites launch with less and weaker content than the version they replaced. The thing that was actually wrong, the writing, survives the project that was supposed to fix it.
There’s also a simple asymmetry worth keeping in front of you. A refresh that underwhelms costs you a few thousand pounds and a fortnight. A full redesign that underwhelms costs you tens of thousands, a quarter of internal attention, and possibly a chunk of your search visibility. When the downside is that lopsided, the burden of proof should sit firmly with the bigger project.
One more possibility is worth naming. Sometimes the urge to redesign isn’t really about the site at all. It’s frustration with an agency that’s gone quiet, or a supplier who never delivers on time, and ‘we need a new website’ is the socially acceptable way to express it. If that’s the actual problem, a new site won’t fix it, and you’d be better served working out whether to renegotiate with or switch your current agency first. A redesign with the wrong partner is just a more expensive version of the same disappointment.
How to pressure-test a redesign proposal
If an agency proposes a full redesign, make them justify it against the cheaper alternatives before you sign anything. The questions are simple, and a good agency will welcome them. The answers tell you whether you’re buying a solution to your actual problem or funding someone’s preferred project. Whether you’re talking to a larger firm like Engage Interactive, a specialist studio such as The Web Kitchen, or a sole freelancer, the proposal you get back is shaped partly by what that firm is set up to sell, so the questions matter more than the logo.
Ask what specifically can’t be fixed on the current platform. This forces the conversation off ‘it’s outdated’ and onto something concrete. If the honest answer is ‘nothing, really, it just needs work’, you’ve learned what you needed to.
Ask what happens to your existing URLs and rankings, and who owns the redirect plan. A team that has a clear, costed answer has done this before and respects what you’ve built. A team that waves the question away is telling you how your launch will go.
Ask whether you could get most of the way there with a new template and a content rewrite, and watch the reaction more than the words. The reaction to being offered a smaller fee is unusually informative.
Ask for the refresh quote alongside the redesign quote. Not instead of, alongside. Any agency confident in its redesign recommendation should be happy to show you the cheaper option it’s advising against, and explain why. The refusal to scope the smaller job is itself an answer.
The agencies worth hiring will give you a straight answer to all of these, sometimes talking themselves out of the bigger project in the process. That’s the signal you’re dealing with a partner rather than a vendor. The ones who can’t tell you what specifically is broken are quoting for the redesign because it’s the product they sell, not because it’s the site you need.
