Three quotes arrive for the same brief and they are ten thousand, forty thousand and ninety thousand. Nobody is being dishonest. They have quoted three different jobs, because the brief did not say which one it wanted.
A refresh, a rebuild and a rescue are separate pieces of work with separate prices, and knowing which you need before you ask is worth more than any negotiation. Here is how to tell, and what each actually costs.
Website redesign cost covers three different jobs
| Refresh | Rebuild | Rescue | |
|---|---|---|---|
| What changes | How it looks. Colours, typography, imagery, layout of existing pages, sometimes the words. | How it is built. New foundation, new structure, content reorganised, everything reconsidered. | Whether it works at all. Fixing what is broken, insecure, unmaintained or abandoned. |
| What stays | The underlying system, the structure, the addresses, most of the content. | The domain, the brand, and whatever content earns its place. Little else. | As much as possible, because the point is to stabilise rather than to improve. |
| Illustrative range | $4,000 to $15,000 | $20,000 to $80,000 | $3,000 to $25,000 |
| Timeline | 2 to 5 weeks | 2 to 5 months | 1 to 6 weeks, depending on what is wrong |
| Risk to search visibility | Low, addresses are not changing. | Real, and this is where cheap quotes cut corners. | Usually low, and sometimes it recovers visibility already lost. |
Illustrative ranges from the kind of work we quote, not a price list. The gap between refresh and rebuild is wide because they are genuinely different projects, not because one supplier is expensive.
A quick diagnostic
Answer these honestly and the answer usually announces itself.
- Can your team change content without a developer? If no, and that is the daily frustration, a refresh will not fix it. That is a rebuild.
- Is it slow on a phone on mobile data? If yes, find out why before deciding. Heavy images are a refresh. A structure that cannot be cached is a rebuild. Why is my website slow separates the two.
- Is anything actually broken, insecure or unsupported? Out-of-date platform, unpatched plugins, a developer who has vanished, warnings in the browser. That is a rescue, and it comes before anything cosmetic.
- Does the structure match how you sell today? If the navigation describes a business you were three years ago, no amount of restyling fixes it. Rebuild.
- Are people arriving and leaving without enquiring? Might be the site, might be the traffic. Diagnose before spending: visitors but no enquiries works through it in order.
- Do you simply dislike how it looks? If everything else is fine, that is a refresh, and it is a legitimate reason.
What a redesign can and cannot fix
Worth being blunt, because a lot of money gets spent on the wrong problem.
- It can fix a site that confuses people, one that is embarrassing to send to a prospect, one nobody on your team can update, one that is slow because of how it is built, and one that does not work properly on a phone.
- It cannot fix not enough people knowing you exist. That is a marketing problem and a new site does not solve it.
- It cannot fix an offer that is not compelling. The clearest page in the world describing a thing nobody wants will not sell it.
- It cannot fix a business whose enquiries are handled slowly. We have watched a rebuild fail to change anything because nobody replied to the form for four days.
- It will not, on its own, improve your rankings, and be wary of anyone implying otherwise. A rebuild is more often a risk to visibility than a boost to it, and protecting what you have is the real work.
Protecting your search visibility
This is the line item that separates a considered quote from a cheap one, and it is invisible until months later.
Your existing site has addresses that search engines have associated with particular questions, sometimes over years. A rebuild usually changes those addresses. Unless every old one is permanently forwarded to the new page answering the same question, that history is discarded and you have effectively launched a new site with none of the standing of the old one.
- List every address that exists today, not just the ones in the menu. Crawl the site and cross-check twelve months of analytics and search data, because pages people still land on are often ones you forgot you published.
- Pair each with a destination that answers the same question. Sending everything to the home page is the classic shortcut and it throws away most of what those pages had earned.
- Decide what should genuinely die and let it return a clean not-found, rather than forwarding it somewhere irrelevant.
- Test the map before launch, on a staging version, as a checklist you tick off. Dull, and the highest-value hour in the project.
- Watch for a fortnight afterwards. Expect some movement even when this is done properly; what you want to see is errors falling and the important pages returning.
Expect to see this as a named cost. If a quote does not mention redirects at all, that is the most useful thing you have learned about the supplier.
Why the same brief gets three different prices
Beyond the refresh, rebuild and rescue confusion, four things explain most of the remaining variation between quotes, and none of them are about one supplier being greedy.
- Who writes the content. The cheapest quotes assume you will supply finished text for every page. The most expensive include a writer. This single assumption can account for a third of the difference, and it is the one most often left unstated, which is also why projects stall: the site is built and waiting on words that nobody was assigned to write.
- Template, custom design, or a design system. Adapting a bought template, designing every page bespoke, and building a set of components your team can assemble into new pages forever are three different products. All three are legitimate. Only the third makes the next two years cheap.
- Whether the quote includes the fortnight after launch. Something always needs attention. A quote that ends at go-live is not wrong, it is just smaller than the project, and you will pay for that fortnight either way.
- Whether redirects and content migration are priced. Discussed above, and the most consequential omission because it is invisible until traffic drops.
The practical move when comparing quotes is not to negotiate the numbers. It is to ask each supplier the same four questions and see how the prices converge once everyone is quoting the same job. Frequently the cheapest quote becomes the most expensive.
Where the money goes in a rebuild
- Content. Consistently the biggest and most underestimated item, and the one that delays projects. Someone has to write it, and if that someone is you, the project waits.
- Design. Whether you are getting a template adapted, a custom design, or a proper system that can grow. Three different prices for three different things.
- Build. Turning the design into a site your team can actually run without help.
- Integrations. Booking, payments, your CRM, your accounting. Each one adds cost, and each is where estimates go wrong.
- Migration and redirects. Moving content across and mapping every address.
- The fortnight after launch. Something always needs attention. Budget for it and it is uneventful.
If the underlying complaint is really that the site looks tired rather than that it works badly, a design refresh is a much smaller job. If it is that nobody can safely touch it because whoever built it is gone, that is inherited systems and it is a rescue.
One sequencing point worth insisting on. Content should be written, or at least outlined, before the design is finished rather than after. Designing around placeholder text produces layouts that look elegant with three neat lines and awkward with the eleven your actual service needs, and the fix at that stage is either a rewrite to fit the design or a redesign to fit the words. Both are avoidable. Ask any supplier when they need your content and treat that date as seriously as the launch date, because in practice it is the one that determines it.
Not for you if
The most expensive redesigns we see are the ones where the brief said make it look modern and the actual problem was structural. Six months later it looks modern and nobody can update it, it is still slow, and the traffic is down because nobody mapped the redirects. Work out which of the three you need first, and the quotes stop varying so wildly.