Most redesign requests we get start with a feeling rather than a fact. The site looks old. A competitor launched something slicker. Someone on the board said the homepage feels tired. Those are real observations, but they do not tell you whether you need a coat of paint or new foundations, and the gap between those two answers is usually a factor of five in budget. This guide separates the three options, prices each one honestly, and shows how to move without losing the search traffic you already have.
A website redesign is worth doing when the site blocks business outcomes: leads have dropped, the CMS fights every edit, mobile is broken, or the brand has moved on. If only the visuals feel dated, a refresh is cheaper and faster. Rebuild only when the platform itself is the problem.
What actually signals you need a redesign?
Start with evidence, not vibes. A site earns a redesign when it is measurably getting in the way of the business, and there are only a handful of signals that reliably prove that.
The first is conversion decay. Traffic is flat or growing but form fills, calls and quote requests are trending down over two or three quarters. That points at the experience, not the marketing. The second is edit friction: if publishing a new service page requires a developer, your marketing calendar is capped by someone else's availability. The third is mobile behavior. Check your analytics for the gap between desktop and mobile conversion rate. A small gap is normal. A gap where mobile converts at a third of desktop usually means the mobile layout was an afterthought.
Then there are the structural ones. If your service list has doubled since launch and the navigation still reflects the old business, visitors cannot find what you now sell. If the brand has genuinely changed (new positioning, new pricing, new audience) the site is arguing for a company that no longer exists. And if you cannot pass a basic accessibility check or the platform no longer receives security updates, the decision is made for you.
What does not justify a full redesign: a homepage hero you have grown bored of, one competitor's launch, or a new marketing hire who wants to put their stamp on things. Those get solved with a refresh.
Before scoping anything, we pull twelve months of analytics and list the top twenty pages by traffic and by conversions. If the pages that earn the business are all working and only the brand feels dated, we recommend a refresh and put the saved budget into conversion optimization instead. That conversation has cost us projects. It has also kept clients for years.
Refresh, redesign or rebuild: which one fits?
The three options differ in what they touch. A refresh keeps the site structure, the CMS and the templates, and changes the surface: type, color, imagery, spacing, some copy. A redesign keeps the platform but rethinks the information architecture, the templates and the content. A rebuild replaces the platform underneath as well, which means everything is new including the risk.
| Option | Typical cost | Timeline | Risk to traffic | Right when |
|---|---|---|---|---|
| Refresh | $3,000 to $12,000 | 2 to 5 weeks | Low. URLs and structure stay put | Pages convert fine, the look is dated, content is broadly right |
| Redesign | $15,000 to $60,000 | 8 to 16 weeks | Moderate. New URLs need redirect discipline | Navigation, messaging and page structure are all wrong for the business you run now |
| Rebuild | $40,000 to $150,000 and up | 14 to 30 weeks | High. New platform, new templates, new everything | The platform blocks you: no security updates, no integrations, unusable editing, or you are moving to ecommerce |
Those are US market ranges for small and mid sized businesses, and they move with page count, custom functionality and how much new content someone has to write. A ten page marketing site sits at the bottom of each band. A hundred page site with gated resources, a location finder and a CRM integration sits at the top or above it.
If you are genuinely torn between refresh and redesign, pick the refresh first. It ships in a month, you learn what actually moves your numbers, and that evidence makes the redesign brief far sharper six months later. If you are torn between redesign and rebuild, ask one question: can the current platform do what the plan requires? If yes, redesign. Platform migrations should be forced, never chosen for novelty.
Where does the money actually go?
Clients are often surprised that design is not the biggest line. In our own projects the split on a mid sized redesign tends to land near a third for strategy and content, a third for design, and a third for build and quality assurance. Content is the line people underestimate every time, because someone has to audit what exists, decide what dies, and write what replaces it.
Custom functionality is the other multiplier. A booking flow, a member area, a quote calculator or a live inventory feed each add meaningful hours. So do integrations that look small on paper: pushing form data into a CRM with proper field mapping and error handling is a day of work, not an hour. If budget is tight, cut features before you cut content quality. A plain page with sharp copy outperforms a clever page with filler.
How do you avoid losing rankings?
Traffic drops after a redesign are common and almost always self inflicted. They come from three places: URLs that changed without redirects, content that got shortened in the name of clean design, and technical details that never made it across. Work through this in order.
- Crawl the current site and export every URL that has traffic, backlinks or conversions. This is your protected list and nothing on it disappears silently.
- Map old URL to new URL for every single one. Where a page is being merged, point it at the closest replacement, not the homepage. Blanket homepage redirects are treated as soft 404s.
- Hold content parity on your top pages. If the old service page ranked with 1,200 words covering eight questions, the new one cannot be 300 words of atmosphere. Rewrite it better, do not shrink it.
- Carry over titles, meta descriptions, headings, image alt text, structured data and canonical tags. These get rebuilt from scratch in new templates and quietly lost.
- Stage the new site behind a password, then crawl the staging build and diff it against the old crawl. Missing pages, missing metadata and broken internal links all show up here, which is the cheap place to find them.
- Launch, then immediately recrawl production, resubmit the sitemap and confirm redirects resolve in one hop rather than a chain.
- Watch for six weeks. Some ranking wobble in the first fortnight is normal. Positions that have not recovered by week six mean something in steps two to four was missed.
One more thing that gets skipped: keep analytics and conversion tracking working on day one. Plenty of teams discover in week three that form submissions were never firing on the new templates, and by then they have lost the comparison data that would have told them whether the redesign worked.
We treat the redirect map as a deliverable with a named owner, not a launch day task. It gets built during design, reviewed alongside the sitemap, and tested on staging. It is the least glamorous artifact in a website design and development project and the single biggest cause of post launch traffic loss when nobody owns it.
What does a redesign actually run like?
A typical eight to sixteen week redesign moves through discovery, information architecture, design, build, content load and launch. Discovery is where you agree what the site must achieve and which pages earn their place. Information architecture is where most of the value sits, because a clear navigation fixes problems that no amount of visual polish can.
Design usually starts with two or three key templates rather than every page. Get the homepage, a service page and a conversion page right, and the rest of the system follows from those decisions. Build follows once templates are signed off. Content load runs in parallel if your writing started early, and sequentially (slowly, painfully) if it did not.
The delays are predictable. Content arriving late, feedback coming from too many people at once, and mid project scope additions account for most overruns. Decide upfront who has final sign off, and give them a deadline for each round.
How do you decide this week?
Run four checks. Does the site block a specific business goal you can name? Do your best pages still convert? Can your team publish without a developer? Does the platform still get updates and support the integrations you need?
If everything passes except the way it looks, refresh. If the navigation, messaging and page structure are wrong for the business you run now, redesign. If the platform itself fails the test, rebuild, and accept the timeline that comes with it. Then set a success measure before work starts: qualified leads per month, cost per lead, or organic sessions to service pages. A redesign with no target is impossible to judge, which is how sites end up being redesigned again in eighteen months for exactly the same unstated reasons. Pairing the launch with a real traffic plan, whether that is performance marketing or steady content creation, is what turns a new site into new revenue.
Frequently asked questions
How often should a website be redesigned?
There is no fixed schedule, and calendar based redesigns waste money. Most business sites need a significant redesign every four to six years, with lighter refreshes in between. Use performance as the trigger instead: declining conversions, a CMS your team cannot use, or a site that describes a business you no longer run. If none of those are true, spend the budget on traffic and conversion work instead.
Will a website redesign hurt my SEO?
It can, but the causes are avoidable. Nearly all post launch traffic loss traces back to three things: URLs that changed without redirects, pages that got shorter in the name of cleaner design, and metadata that never made it into the new templates. Map every old URL to a new one, hold content parity on your ranking pages, crawl staging against the old site, then recheck production on launch day.
What is the difference between a website refresh and a redesign?
A refresh changes the surface: type, color, imagery, spacing and selected copy, while the site structure, templates and URLs stay where they are. A redesign rethinks navigation, page structure and content first, then rebuilds templates around those decisions. Refreshes take two to five weeks and carry very little risk to search traffic. Redesigns take two to four months and need proper redirect planning.
How much does a website redesign cost in the US?
For US small businesses, a refresh commonly runs $3,000 to $12,000, and a full redesign $15,000 to $60,000. Platform rebuilds start near $40,000 and climb fast once custom functionality is involved. The variables that move the number most are page count, integrations such as CRM or booking systems, and how much brand new content someone has to research and write.
Should I redesign or just improve the pages I have?
If your key pages still convert and your team can publish without a developer, targeted improvements usually win. Rewriting the two or three pages that carry your pipeline, fixing mobile layout problems and clarifying calls to action often produce more measurable gain than a full redesign, for a fraction of the cost. Redesign when the structure is wrong, not when a page is merely underperforming.
How long should I wait before judging a redesign?
Give search results six to eight weeks. Rankings commonly wobble in the first fortnight while the new pages are recrawled, and positions that have not settled by week six point at a redirect or content problem worth investigating. For leads, wait roughly a quarter so you have enough sessions to compare fairly. Judging a redesign on two weeks of data mostly measures noise.
The takeaway
Redesigns fail quietly more often than they fail loudly. The site looks better, everyone is pleased at launch, and six months later nobody can say whether it earned its budget. Avoid that by naming the business problem first, choosing the smallest option that solves it, protecting the pages that already work, and agreeing a success measure before a single screen is designed. The choice between refresh, redesign and rebuild should follow the evidence.