There are three different projects people call “we need a new website”, and they cost very different amounts.
A refresh keeps the site and changes how it looks. New colours, new images, new copy, same structure and same platform. Weeks, not months.
A redesign keeps the platform and rebuilds the front end. New layouts, new templates, new structure, same CMS and usually most of the same content. A month or two.
A rebuild replaces the platform. New CMS, new templates, content migrated across, integrations rewritten. A quarter, sometimes more.
Most businesses ask for a rebuild and need a redesign. Some ask for a redesign and genuinely need a rebuild, usually because the platform is the thing holding them back. Getting this wrong is expensive in both directions.
This guide covers how to tell which one you need. For the process once you have decided, see our comprehensive guide to website redesign.
The question that decides it
Not “does the site look dated”. Looks are fixable at any of the three levels.
The question is: is the platform preventing you from doing things you need to do?
If the answer is no, you need a redesign, however bad the site looks. If the answer is yes, and the things you need are strategic rather than cosmetic, you may need a rebuild.
Seven tests for a rebuild
Score these honestly. Three or more is a real case for rebuilding. Fewer than three and a redesign will almost certainly serve you better and cost a third as much.
|
Test |
Why it points to a rebuild |
|
The platform cannot do something commercially necessary |
You cannot add the payment method, the booking system or the integration your business now requires |
|
Nobody can safely update it |
The original developer is gone, the build is custom, and every change risks breaking something |
|
The stack is unsupported |
Running an end-of-life PHP version or abandoned plugins is a security exposure, not a style problem |
|
Editing content requires a developer |
If marketing cannot publish a page without raising a ticket, the platform is costing you speed every week |
|
Performance problems are structural |
Page speed cannot be fixed because the architecture itself is the bottleneck |
|
It cannot support your next twelve months |
Multiple languages, multiple markets, a product catalogue, a members area, none of which the current platform handles |
|
Licensing or hosting is locked |
You cannot move, cannot export, or do not own the code |
Note what is not on this list: “it looks old”, “the previous agency was bad”, “we have new branding”, “it is not mobile friendly”. Every one of those is a redesign.
What each option actually costs you
Money is the obvious axis. It is not the one that catches people out.
|
Refresh |
Redesign |
Rebuild |
|
|
Typical duration |
2 to 4 weeks |
6 to 12 weeks |
3 to 6 months |
|
Content work |
Light edits |
Rewrite key pages |
Full audit and migration |
|
URL changes |
None |
Some |
Many |
|
SEO risk |
Minimal |
Moderate, manageable |
Significant, needs managing |
|
Integrations |
Untouched |
Mostly untouched |
Rebuilt and retested |
|
Internal time required |
Low |
Moderate |
High, and consistently underestimated |
The hidden cost of a rebuild is your own team’s time. Content decisions, approvals, data checks, integration testing. A rebuild that stalls usually stalls on the client side, not the agency side, and it stalls because nobody budgeted the internal hours.
That is worth saying before the project starts rather than in month four.
The SEO consequence, stated plainly
The level you choose changes your exposure.
A refresh carries almost no SEO risk. URLs stay, structure stays, content mostly stays.
A redesign changes templates and often some URLs. Risk is real but manageable with redirect mapping and a pre-launch checklist.
A rebuild changes nearly everything: URL structure, internal linking, page templates, often the content itself. This is where sites lose traffic, and the losses are usually avoidable rather than inevitable.
The single most common cause of post-launch traffic loss is a redirect map that was produced late, tested briefly, or not produced at all. Our SEO-safe redesign guide covers the full process, and the website redesign checklist covers pre-launch.
Treat it this way: the SEO work is part of the build, planned from the start, not a check performed the week before launch.
When a rebuild is the wrong answer
Three situations where businesses rebuild and should not.
The site does not convert. Rebuilding an unconverting site usually produces a better looking site that does not convert. The cause is more often the offer, the message or the traffic quality than the technology. Diagnose first, using the sequence in traffic but no enquiries.
The traffic has fallen. Find out why before rebuilding. If the cause is an algorithm update, a competitor, or a technical issue, a new site does not fix it and may make it worse by resetting whatever was working.
A new team dislikes the old site. Common and expensive. A new marketing lead inherits a site they did not choose and proposes replacing it. Sometimes correct. Often it is a redesign dressed up as a rebuild, and the business pays three times the necessary amount.
The test in each case is the same: name the specific thing the current platform prevents. If you cannot, you do not need a new platform.
When a redesign is not enough
The opposite error, which is less common but more damaging because it wastes the redesign budget too.
If you redesign a site whose platform genuinely cannot support the business, you spend the money, launch, and hit the same wall within a year. Then you rebuild anyway, having paid twice.
The signal is usually that your requirements list includes things nobody can explain how to do on the current stack. If the answer to “can we do X” is repeatedly “not really, but we could work around it”, that accumulation of workarounds is the platform telling you something.
A practical decision sequence
- List what you need the site to do in the next twelve to eighteen months. Specific capabilities, not adjectives.
- Ask your developer which of those the current platform supports. Get it in writing.
- Score the seven rebuild tests above.
- If fewer than three, scope a redesign. Spend the difference on content and conversion work, which usually returns more than a new platform would.
- If three or more, scope a rebuild, and budget the internal hours honestly.
- Either way, plan the SEO work from day one, not before launch.
Frequently asked questions
What is the difference between a website redesign and a rebuild?
A redesign keeps the existing platform and rebuilds the front end: new layouts, templates and structure on the same CMS. A rebuild replaces the platform itself, with content migrated across and integrations rewritten. A redesign typically takes six to twelve weeks; a rebuild takes three to six months.
How do I know if I need a new website or just a redesign?
Ask whether the platform prevents you from doing something your business needs. If it does not, a redesign will serve you better at a fraction of the cost, however dated the site looks. If the platform blocks necessary capabilities, cannot be safely updated, or is unsupported, a rebuild is justified.
Will a new website hurt my Google rankings?
It can, and the risk rises with the scope of change. A refresh carries almost no risk. A redesign carries moderate risk, managed with redirect mapping. A rebuild changes URLs, templates and often content, which is where sites lose traffic. The losses are usually avoidable with planning rather than inevitable.
Is it cheaper to redesign or rebuild?
A redesign is substantially cheaper, both in agency cost and in internal time. The hidden cost of a rebuild is your own team’s hours on content decisions, approvals and testing, which is routinely underestimated and is the most common reason rebuild projects stall.
My website does not generate leads. Should I rebuild it?
Usually not as the first step. Poor conversion is more often caused by traffic quality, unclear messaging or weak trust signals than by the technology. Diagnose the cause first; rebuilding an unconverting site tends to produce a better looking site that still does not convert.
How long should a website rebuild take?
Three to six months for most business sites, longer with complex integrations or a large content migration. If a timeline of a few weeks is quoted for a genuine platform change, confirm what is being left out, usually content migration or integration work.
Can I keep my content when changing platforms?
Yes, and you should audit it rather than migrate it wholesale. A rebuild is the natural moment to remove pages that earn nothing and consolidate pages that compete with each other. Migrating everything including the dead weight is a missed opportunity.
What is the most common mistake in this decision?
Choosing a rebuild for reasons that a redesign would solve. The test is whether you can name a specific capability the current platform prevents. If the reasons are all about appearance or about disliking the previous build, it is a redesign.
If you want help deciding
We will tell you when a redesign is enough, including when that means a smaller project than you came in expecting. That conversation is cheaper for both sides than discovering it in month three.
For the scoping conversation, see website redesign in Singapore or web design in Singapore. If you are still establishing whether the site needs work at all, signs your business website needs a redesign is the place to start.
