Redesign when your website is blocking a business goal you can measure, such as leads, sales, speed or credibility with the buyers you now want, not just because it looks dated. A good redesign keeps what already works (URLs that rank, content that converts), rebuilds what doesn't, and is scoped before it's priced. This checklist walks through each decision in order.
How do you know it's time to redesign?
Look for symptoms you can point to rather than a general feeling. The common ones: mobile visitors convert far worse than desktop; Core Web Vitals fail on your main templates; the team avoids updating the site because every change needs a developer or breaks something; the brand has moved on and the site still describes the old business; you're adding a service, market or language the structure can't hold; or sales calls start with apologies for the website. One symptom can often be fixed in place. Three or more, especially when they share a cause such as an ageing theme or a page builder stacked with plugins, usually mean the foundation is the problem. Write the symptoms down: they become the redesign's success criteria, and a provider who can't connect their proposal to them is selling a new look, not a fix.
Refresh, redesign or rebuild: which do you need?
These are three different projects, and confusing them is how budgets go wrong. A refresh changes the surface: typography, colors, imagery and a few layouts, on the same platform and structure. It suits a site that works but looks behind the brand. A redesign rethinks structure and content as well: navigation, page types, and the journey from landing to enquiry or checkout. It suits a site whose problem is clarity or conversion. A rebuild replaces the foundation, meaning a new platform, a move from a theme to custom code, or a fresh content model, and suits sites where speed, maintenance or flexibility is the bottleneck. Many projects combine two of these. Decide which one you're buying before you collect quotes: a refresh quote and a rebuild quote aren't comparable, and the cheapest proposal is often simply answering a smaller question.
What should you keep from the old site?
Keep anything that already earns its place, even if it looks old. That means the URLs that bring organic traffic, the pages that generate most enquiries or sales, content that ranks for terms you care about, and any copy your customers quote back to you. Pull twelve months of data from Google Search Console and your analytics and list the top pages by clicks, conversions and backlinks before anyone opens a design file. Those pages get redesigned carefully, keeping their URL where possible and their substance always. Anything that moves gets a one-to-one 301 redirect. The reverse is also true: pages nobody visits and outdated offers can retire, which makes the new site smaller, clearer and cheaper to build. Redesigns that lose traffic rarely fail on design; they fail because valuable pages were deleted or moved without a map.
What drives the cost of a redesign?
Scope drives the price, and five variables carry most of it. The number of unique page templates: ten pages built from three templates cost far less than ten one-off designs. Content: rewriting copy, producing new photography or 3D visuals, and migrating hundreds of old posts are often bigger line items than the design itself. Platform change: staying put is cheaper than rebuilding on a new stack, though a rebuild can save years of maintenance. Functionality: booking, configurators, member areas, and integrations with your CRM or inventory. And languages, since a properly adapted bilingual site is more work than a translated one. Brand work adds a layer if the identity is changing at the same time. We publish our own bands on the pricing page and break down market ranges in our website cost guide; this checklist's job is to get these variables written down before you ask anyone for a number.
What should the redesign brief include?
A one-page brief saves weeks of back-and-forth. Include the business goal and how you'll measure it, such as qualified leads per month, conversion rate or average order value. List the symptoms from the first step, the pages and URLs you're keeping, and the audiences the site must convince, in order of importance. State what's fixed (brand guidelines, platform, a launch date tied to an event) and what's open. Add three sites you admire with one sentence on why, plus three you don't. Note who writes the content and who approves it, since late content is the most common reason redesign timelines slip. Finally, give a budget range. It feels like showing your cards, but it lets a good provider propose the best version of the project for that money instead of guessing, and it filters out mismatches in the first call.
How do you protect SEO and AI visibility during a redesign?
Treat search visibility as a requirement, not a post-launch task. Before launch: a complete redirect map for every changed URL, titles and meta descriptions carried over or improved, structured data rebuilt, and the new site crawled on staging to catch broken links and orphan pages. Keep staging blocked from indexing, and remember to unblock the live site at launch, a mistake that happens more often than anyone admits. Make sure content is server-rendered so AI crawlers, many of which don't run JavaScript reliably, can read it, and that robots.txt doesn't accidentally shut them out. After launch: submit the sitemap in Search Console and Bing Webmaster Tools, watch daily for 404s and indexing drops in the first weeks, and compare top-page traffic against your baseline. A redesign is also the right moment to restructure key pages into question-led, citable sections.
How long does a website redesign take?
Timelines follow scope the same way price does. A refresh on the existing platform can be a matter of weeks; a full redesign with new content and a platform change takes longer, and the critical path is usually content and approvals rather than design or code. Build in time for the unglamorous steps: the content inventory, the redirect map, testing on real phones, and a launch window that avoids your busiest season. Plan a short stabilisation period after launch, too, when analytics, forms and integrations get checked against real traffic. On our own projects we set the timeline in the proposal after the discovery call, and we'd rather give you a realistic date than an optimistic one. If you're weighing a redesign now, start with the symptoms list: it's the fastest way to know whether you need one at all.
Sources
Thinking about a redesign? Send us your site and your symptoms list, and we'll tell you whether it needs a refresh, a redesign or a rebuild.