The B2B website redesign guide.

When a redesign is actually justified

A redesign is justified when the site no longer matches how your buyers research, when the platform blocks changes your team needs weekly, or when measurable performance (enquiries, rankings, speed) has decayed and patching costs more than rebuilding. It is not justified because leadership is bored of the design.

That last point is worth sitting with, because internal fatigue is the single most common trigger for a redesign and the single worst reason for one. Your executive team sees the homepage several times a day for three years. Your buyer sees it twice, eight months apart, while comparing you to four competitors. The people most bored of a website are structurally the least qualified to judge whether it is working.

Run the maths first. If your site produces 20 qualified enquiries a month and a rebuild plausibly lifts that 30%, the project pays for itself quickly. If you cannot state the current number, that measurement gap is the first project, not the redesign.

How do I know if it is a redesign or a content problem?

Look at where visitors leave. If they arrive and bounce within seconds, the problem is usually positioning or speed, and both are fixable without a rebuild. If they read three or four pages, spend real time on them and still do not enquire, the problem is that the site never answers the questions that decide the purchase — also a content problem, and a cheaper one.

A redesign is the right call when the structure itself blocks the fix: when there is nowhere to put the pricing page, when the CMS cannot model your products, when adding a case study means a developer ticket. Structural limits justify structural work. Everything else is an argument for editing.

How long does a B2B redesign take?

A focused B2B site runs roughly six to twelve weeks from kickoff to launch. Discovery and structure take two to three weeks, design two to four, build three to five, with testing and migration overlapping the end. Sites with large product catalogues, multiple languages or software features run longer, and should be phased rather than stretched.

The variable that moves the date is almost never the build. It is content and approvals. Projects slip because copy sits with a subject-matter expert for three weeks, or because a stakeholder who was never in the kickoff appears at sign-off with fundamental objections. Name your decision-maker before you start and give content its own deadline, ahead of design, not alongside it.

Sequence: strategy, structure, design, build

The reliable order is buyer research, then information architecture, then wireframes, then design, then build. Every failed redesign we have audited skipped to design. Our process page documents the five-stage version we run, including what you receive at each stage.

Write the page-by-page intent map before wireframes: every page gets one buyer question to answer and one action to drive. Pages without a question get cut, which is how redesigns end up smaller and more effective than the sites they replace.

The reason the order matters is that each stage is a cheap place to be wrong. Changing the sitemap during structure costs an afternoon. Changing it during design costs a fortnight of rework. Changing it after build costs the redirect map, the analytics setup and half the trust of the people who signed off the last version. Front-loading the arguments is not bureaucracy; it is buying your mistakes at the lowest price they will ever be available.

What does the research stage actually produce?

Three artefacts you can argue with: a picture of how your buyers currently research (from sales calls, search data and AI-answer checks), a list of the questions that decide the deal, and a map of which competitors already answer them well. None of it is a persona document with a stock photo and a fictional name.

The output that matters most is unglamorous: the objections your sales team handles on every call. Those are the pages. A B2B site that answers the five objections your salespeople are tired of repeating will outperform one built around a brand narrative, because it is doing the job the buyer actually arrived to have done.

Should we redesign the brand at the same time?

Only if the brand is genuinely the problem, and it usually is not. Bundling a rebrand into a website project doubles the number of decisions, doubles the number of stakeholders and makes it impossible to attribute the result afterwards — if enquiries rise, you will never know which change did it.

If both are needed, run brand first and let it settle, then build the site against a finished identity. If the brand is merely tired rather than wrong, a tightening exercise inside the web project (type, colour, photography direction, a consistent voice) delivers most of the benefit for a fraction of the disruption.

Protecting your rankings during the switch

Rankings are lost through skipped redirects, not through redesign. Inventory every URL with traffic or links, map each to its successor one-to-one, and ship the redirect map with the launch, not after it.

Carry over title tags and metadata where pages keep their intent, keep your XML sitemap current, and re-crawl the site the week after launch to catch what broke. We treat this as part of the build scope, never an optional extra.

The failure mode is specific and predictable. A site launches on a Thursday, traffic drops 40% by the following Tuesday, and everyone blames the new design. Almost always the cause is that a few hundred URLs changed shape and the redirects were either skipped or pointed at the homepage in bulk, which search engines treat as a soft 404 rather than a move. It is recoverable, but recovery takes months and the trust cost inside the business is worse than the traffic cost.

• Full URL inventory from analytics and Search Console
• One-to-one 301 map, no blanket redirects to the homepage
• Metadata and schema carried over or deliberately improved
• Internal links updated to point at the new URLs, not through the redirects
• Post-launch crawl and Search Console check within seven days

Will a redesign hurt my SEO?

A redesign with a complete redirect map, preserved metadata and equal-or-better content typically holds rankings and then improves them, because the new site is usually faster and better structured. A redesign without those things reliably loses traffic, and the loss can take six months to recover.

The honest framing is that redesign is not the risk factor — URL change is. If the URLs stay identical and the content stays as good, the SEO risk is close to zero regardless of how different it looks.

How long before we should judge the results?

Give it a full quarter before drawing conclusions, and expect a dip in the first two to four weeks while engines recrawl and reindex. Judging a launch on week-one analytics is how good projects get reversed by nervous stakeholders.

What you can judge immediately is on-page behaviour: are people reaching the pages you built for them, are they scrolling, are the enquiry forms being started. Those signals arrive within days and tell you whether the structure works, long before the traffic numbers settle.

What a B2B redesign costs

UK pricing clusters in bands: template work below £3,000, credible studio builds £4,500 to £25,000, larger agency builds beyond that. Our websites start at £4,500 including strategy, design, build and SEO foundations; software and portals are scoped separately from £18,000. The pricing article in this series breaks down the whole market honestly.

Within the studio band, the variable that moves the number is structure rather than page count. Ten pages with a clean content model is a smaller project than a 200-item product catalogue that needs filtering, specification tables and comparison logic, even though the second one might have fewer designed templates. Ask any bidder to price the data, not the pages.

Budget for content properly, because it is the most commonly underfunded line and the most commonly missed deadline. Whether it is written internally, by the studio or by a specialist, someone has to do it, and pretending otherwise is how a six-week build becomes a five-month one with a half-empty site at the end of it.

After launch is where the return lives

A redesign sets the baseline; compounding comes from shipping improvements weekly against measurement. Plan for a post-launch programme (our B2B SEO retainer starts at £1,000 a month) rather than a launch party and silence. Sites that improve continuously, like Quartix's partner platform at +128% organic sessions year on year, are maintained as engines, not monuments.

The first ninety days are the highest-leverage period the project will ever have. You have fresh analytics, a live audience and a team still holding the context. Use it: watch where people stop, read the search queries that now land on you, and fix the three things that are obviously wrong before they harden into how the site is.

Decide who owns the site before launch, not after. The most expensive outcome in this entire process is a good site with no owner, which decays quietly for two years until someone declares it tired and the whole cycle starts again. A named owner with a small monthly budget is the difference between an asset and a liability.