Website Redesign SEO Checklist: Keep Your Rankings Safe

Expert Quick-Start: A website redesign SEO checklist protects your rankings through five steps: benchmark current performance, crawl every URL, build a 301 redirect map, preserve on-page elements and content, then monitor Search Console daily for eight weeks after launch. Skip any step and you risk losing the traffic your old site earned.

Your website redesign SEO checklist matters more than your new design. A website redesign SEO checklist is the difference between a relaunch that grows your traffic and one that quietly deletes years of hard-won rankings overnight.

The pattern is brutally common. A business invests £15,000 in a beautiful new site, launches it on a Friday, and watches organic traffic fall 40% by the end of the month. Nobody changed the content on purpose. Nobody meant to break anything. The URLs simply changed, the redirects never went in, and Google treated the new site as a stranger.

Research compiled by Search Engine Land shows that the majority of businesses that redesign without an SEO plan see a measurable organic traffic drop within three months. The damage is avoidable, and every fix is cheaper before launch than after.

This guide walks you through the complete process: what to record before you start, how to build a redirect map that actually works, what to check on staging, what to do on launch day, and how to respond if rankings wobble anyway. Work through it in order and your new site will inherit everything the old one earned.

Why Website Redesigns Destroy Rankings

Google does not rank your website. It ranks individual URLs.

Every page on your current site has its own history: backlinks pointing at it, an indexing record, engagement signals, and a position in Google’s understanding of your site structure. When a redesign changes a URL without a redirect, all of that history points at a page that no longer exists.

A redesign without redirects is, in Google’s eyes, a mass deletion of your website followed by the launch of a brand new one. The new site starts from close to zero while the old rankings decay.

The most common ranking killers during a redesign are worth naming up front:

• Changed URLs with no 301 redirects. The single biggest cause of post-launch traffic loss. Old URLs return 404 errors and their authority evaporates.
• Thinned-out content. Designers love white space. If your 1,200-word service page becomes a 200-word page with big imagery, Google has 1,000 fewer words to rank you for.
• Staging noindex tags going live. The one-line tag that kept your staging site hidden gets copied to production and tells Google to remove your entire site from search.
• Lost title tags and meta descriptions. New templates frequently ship with default or duplicated metadata, throwing away years of optimised titles.
• Slower page speed. Heavy imagery, new JavaScript frameworks and unoptimised themes can push Core Web Vitals into the red.
• Broken internal links. Menus and in-content links that still point at old URLs create redirect chains and crawl waste.

Each of these is fully preventable. The rest of this checklist deals with them one by one.

Redesign, Rebuild or Migration: Know Which Project You Are Running

People use “redesign” loosely, but the SEO risk depends on exactly what is changing. Be honest about which project you are actually running, because the checklist gets heavier as you go down this table.

A pure reskin that keeps every URL, every heading and every word of content carries little risk. You still need the pre-launch technical checks, but you can skip the redirect mapping.

The moment URLs change, you are in redirect territory and the full checklist applies. If your URLs do not need to change, keep them. This is the single cheapest piece of risk reduction available in any redesign. Semrush’s migration research makes the same point: changing URLs introduces risk, so preserve them wherever possible.

A domain change stacks migration risk on top of redesign risk. If you can possibly avoid doing both at once, split them into two phases separated by at least two months.

Phase 1: Benchmark Everything Before Anyone Touches the Site

You cannot know whether the redesign hurt you if you never recorded where you started. Benchmarking takes an afternoon and settles every post-launch argument with data instead of panic.

Do this two to four weeks before the planned launch, and store everything in a shared folder the whole project team can see.

Record Your Search Console Baseline

Open Google Search Console and export the following:

1. Performance report for the last 16 months: clicks, impressions, CTR and average position, segmented by page and by query.
2. Your top 100 queries by clicks, with their ranking pages.
3. The Pages indexing report: how many URLs are indexed, and how many are excluded and why.
4. Core Web Vitals report: current pass and fail counts for mobile and desktop.
5. Any manual actions or security issues (there should be none, but check).

The 16-month query export is the crown jewels. Six months after launch, when someone asks “did we used to rank for this?”, this export answers it in seconds.

Record Your Analytics Baseline

In GA4, export at minimum 12 months of:

• Organic sessions by landing page
• Conversions and conversion rate by landing page
• Engagement rate and average engagement time for your top 50 pages

Flag your top 20 pages by organic conversions, not just by traffic. A page that quietly delivers three enquiries a month matters more than a blog post with ten times the traffic and zero enquiries. These are the pages that get white-glove treatment through the rest of this checklist. If you are unsure how your current pages actually convert, our conversion rate optimisation service covers exactly this kind of audit.

Record Your Rankings and Backlinks

Capture a rank-tracking snapshot for your target keywords, including local pack positions if you serve a geographic area. Then export your backlink profile from Ahrefs, Semrush or Search Console’s own Links report.

Sort the backlink export by linking domains per URL. Any page with external links pointing at it must survive the redesign with its URL intact or receive a dedicated 301 redirect. Those links are transferable assets, but only if you transfer them.

Crawl the Current Site

Run a full crawl with Screaming Frog or a similar crawler and archive it. This is your complete map of the old site: every URL, every title tag, every meta description, every heading, every canonical, every hreflang tag, and the internal link graph.

Combine the crawl with your Search Console and analytics exports to build one master spreadsheet. Every URL on the old site should appear in it, alongside its traffic, conversions, backlinks and target queries. This spreadsheet becomes the backbone of your redirect map in the next phase.

Phase 2: URL Structure Decisions

With your benchmark in hand, the next decision shapes everything that follows: what happens to your URLs?

Keep URLs Where You Possibly Can

Every URL you keep is a redirect you never have to build, test or monitor. It is also ranking history you never put at risk.

Challenge every proposed URL change. Designers and developers often change URL structures out of habit or platform convention, not necessity. “The new CMS prefers /services/seo/ over /digital-marketing/seo/” is a preference, not a requirement, and almost every modern CMS can be configured to match your existing structure.

Legitimate reasons to change a URL are rare: fixing truly broken structures (session IDs, meaningless parameters), consolidating duplicate pages, or correcting a structure that actively confuses users. “The new one looks tidier” does not qualify.

If URLs Must Change, Change Them Well

When a change is unavoidable, follow these rules:

• Keep URLs short, lowercase and hyphenated
• Include the page’s primary keyword where natural
• Avoid dates in URLs for evergreen content
• Never include stop words you will regret (“the”, “and”, “best-2022”)
• Keep folder depth shallow; three levels is usually plenty
• Decide your trailing slash convention once and enforce it everywhere

Consistency matters more than perfection. A slightly imperfect structure applied consistently beats an elegant one applied inconsistently.

Phase 3: Build the 301 Redirect Map

This is the heart of the entire website redesign SEO checklist. The redirect map is a simple two-column spreadsheet: every old URL on the left, its new destination on the right. Simple to describe, and the most important deliverable in the whole project.

How to Build the Map

Start from the master spreadsheet you built during benchmarking. It already contains every URL on the old site. Now work through it row by row:

1. Direct equivalent exists: map old URL to its new equivalent. This covers most pages.
2. Page is being merged: map the old URL to the surviving page that best answers the same search intent.
3. Page is being retired with no replacement: map it to the closest relevant category or service page. Only use the homepage as a destination of last resort.
4. Page is staying at the same URL: mark it “no redirect needed” but keep it in the sheet so nothing gets forgotten.

Every single old URL must have a row and a decision. Not just the pages you like. Old blog posts, tag pages, PDF files, image URLs with backlinks, legacy landing pages from campaigns long dead. If Google has it indexed or anyone links to it, it needs a mapped destination.

Redirect Rules That Protect Equity

Use 301 (permanent) redirects, not 302 (temporary) ones. Google has confirmed that 301 redirects pass PageRank, and its own site move documentation treats them as the standard mechanism for URL changes.

Redirect old URL to final destination in one hop. Chains (old page redirects to an interim page which redirects again) leak crawl budget and slow users down. Audit for chains before launch and flatten any you find.

Match intent, not just topic. Redirecting a detailed “how to fix boiler pressure” guide to a generic “boiler services” page technically works, but Google often treats mass redirects-to-irrelevant-pages as soft 404s and drops the equity anyway. The closer the destination matches the original page’s intent, the more value transfers.

Never blanket-redirect everything to the homepage. It is the classic panic move and Google explicitly discounts it.

Warning: Redirects are not fire-and-forget. Keep your redirect rules live for a minimum of one year, and ideally permanently. Removing them after a few months resurrects every 404 you fixed at launch.

Where to Implement Redirects

Server-level redirects (in .htaccess on Apache, nginx config, or your host’s redirect manager) are fastest and most reliable. CMS plugins such as Redirection for WordPress are acceptable for modest volumes and much easier for non-developers to manage.

Avoid JavaScript redirects and meta refresh redirects for anything you care about. Google can process them, but slowly and unreliably compared with a proper server-side 301.

Phase 4: Preserve Your On-Page SEO Elements

Redirects protect your URLs. This phase protects everything that lives on the pages themselves.

Title Tags and Meta Descriptions

Your crawl archive contains every title tag and meta description on the old site. These were (hopefully) refined over years and Google has already rewarded them. The new site should launch with the same metadata on every carried-over page unless you have a specific, argued reason to change it.

The failure mode here is templates. New themes often generate titles automatically (“Page Name | Site Name”) and overwrite your carefully written tags with boilerplate. Check a sample of built pages on staging early, because if the template is doing this on ten pages it is doing it on all of them.

Heading Structure

Each page needs exactly one H1, and on carried-over pages it should stay substantively the same as before. Designers frequently demote the H1 into a styled div because it “looks better”, which silently deletes the strongest on-page relevance signal a page has.

Check that H2s and H3s survived too. Heading hierarchy is one of the main ways Google parses long content, and it increasingly feeds featured snippets and AI-generated answers.

Content Parity: Count the Words

Here is the check almost everyone skips. For your top 50 pages, compare the word count on the old page against the new build.

If a page that ranked with 1,400 words relaunches with 300 words and a hero video, expect the rankings to follow the words out the door. Modern, airy design and substantive content are not enemies; accordions, tabs and well-structured long pages let you keep both. Google indexes content in accordions and tabs, so use them rather than deleting copy to fit the design.

This is also the right moment to run your best pages through a proper on-page process rather than copy-pasting blind. Our SEO copywriting checklist covers the full page-level workflow if content is being rewritten as part of the project.

Images and Media

Image URLs are URLs too. If your images move from /wp-content/uploads/ paths to a new structure or CDN, image search traffic and any image backlinks need redirects like everything else.

Carry over alt text deliberately. It rarely migrates automatically between CMS platforms, and rewriting hundreds of alt attributes after launch is miserable. Export alt text as part of your crawl and reimport or re-enter it as pages are built.

Compress and serve modern formats (WebP or AVIF) from day one. The redesign is your one chance to get the media library right without retrofitting.

Structured Data

If the old site carried schema markup (Organisation, LocalBusiness, FAQ, Product, Review, Article), the new site needs equivalent or better markup at launch. Validate templates with Google’s Rich Results Test on staging.

Schema is easy to lose in a rebuild because it often lives in theme code or plugins that do not travel to the new platform. Inventory what exists now, and make its re-implementation an explicit ticket in the build, not an assumption.

Phase 5: Rebuild Internal Linking Deliberately

Internal links tell Google which pages on your site matter most. A redesign reshuffles them whether you plan it or not, so plan it.

Navigation Is Your Loudest Signal

Pages linked from the main menu receive the strongest and most frequent internal link equity on the site. If the redesign drops a service page out of the navigation, its rankings will feel it within weeks.

Before sign-off on the new navigation design, list every page in the current menu, every page in the proposed menu, and account for the differences. Any page leaving the navigation needs replacement links from elsewhere: footer, related-content modules, or in-copy links from strong pages.

In-Content Links

Your body copy is full of internal links pointing at old URLs. If those URLs changed, every one of those links now travels through a redirect.

Redirects rescue the equity, but linking straight to final URLs is cleaner, faster and clearer. Run a find-and-replace across the content database for old URL patterns, then crawl staging and fix any internal link that returns a 301 or 404.

Anchor Text and Orphan Pages

Keep descriptive anchor text. “Read our guide to lead generation” beats “click here” for both users and relevance signals.

After the new site structure is loaded on staging, crawl it and compare the URL list against your master spreadsheet. Any carried-over page with zero internal links pointing at it is now an orphan. Orphan pages get crawled less, refreshed in the index less often, and slowly bleed rankings. Give every page you care about at least two or three internal links from relevant pages.

Phase 6: Technical SEO on the Staging Site

Everything so far has been planning and content. Now the new site exists on a staging server, and the technical audit begins. Nothing in this phase should wait until launch day.

Keep Staging Out of the Index, Properly

The staging site must be invisible to Google, and how you hide it matters.

The robust approach is HTTP authentication: a username and password on the whole staging environment. Search engines cannot crawl what they cannot access, and there is no tag to accidentally carry over to production.

The riskier but common approach is a site-wide noindex meta tag or a robots.txt disallow. It works, but it plants a landmine: the number one cause of catastrophic post-launch traffic loss is the staging noindex tag shipping to the live site. If you must use noindex on staging, make “remove noindex and verify” the first and loudest item on your launch-day checklist.

Check the reverse case too. If staging is not properly blocked and Google indexes it, you launch with a duplicate of your entire site already in the index.

Full Staging Crawl

Crawl the staging site exactly as you crawled the old site, then work through this list:

• Status codes: every carried-over page returns 200, every redirect returns a single 301, and nothing important returns 404 or 500.
• Titles and metas: spot-check against your old-site archive. Look for template boilerplate and duplicates.
• H1s: one per page, matching the old page where content carried over.
• Canonicals: every page self-references or points at its correct canonical version. Watch for canonicals still pointing at staging URLs.
• Robots directives: no stray noindex or nofollow on money pages.
• XML sitemap: generates correctly, contains only final 200-status URLs, and no staging hostnames.
• Redirect testing: load your full redirect map into the crawler in list mode and confirm every old URL resolves to its mapped destination in one hop.
• Hreflang: if you run international versions, confirm the tags survived and reference the new URLs.
• Pagination and parameters: category pagination, filters and search pages behave and are indexed (or excluded) intentionally.

The redirect list test is the one to be obsessive about. Testing all of them takes the crawler minutes and catches the mapping errors that would otherwise surface as 404 reports three weeks after launch.

Core Web Vitals and Speed

New builds are usually faster in demos and slower in production, because demos have three pages and production has your full media library, tracking stack and cookie banner.

Test staging with Lighthouse and PageSpeed Insights on your top templates: homepage, service page, blog post, contact. Focus on the three Core Web Vitals: LCP (loading), CLS (layout stability) and INP (responsiveness). Google’s own guidance on Core Web Vitals sets the thresholds: LCP under 2.5 seconds, CLS under 0.1, INP under 200 milliseconds.

Fix the big rocks before launch: oversized hero images, render-blocking scripts, fonts loaded without font-display swap, and layout shift from ads or embeds. Post-launch is too late, because Core Web Vitals data is collected from real users over 28-day windows, and a slow launch month sits in your field data for weeks.

Mobile Rendering

Google indexes the mobile version of your site, full stop. Audit staging on real phones, not just a resized desktop browser window.

Confirm that content visible on desktop is present on mobile (hidden-on-mobile sections are invisible to indexing for ranking purposes), tap targets are usable, and the menu exposes the same key pages as the desktop navigation.

Phase 7: Launch Day Checklist

Launch on a Tuesday or Wednesday morning, never on a Friday. You want your developer, your SEO and your hosting support all awake and available for the 48 hours that follow.

Work through this sequence in order:

1. Take a final backup of the old site and database, and confirm you can actually restore it.
2. Deploy the new site to production.
3. Immediately remove staging noindex tags and robots.txt blocks, then view source on the live homepage and key templates to verify.
4. Confirm the production robots.txt allows crawling and references the new sitemap.
5. Test the full redirect map against the live domain in list mode. Every old URL, one hop, correct destination.
6. Click through the main navigation and top 20 money pages by hand. Check content, forms, phone numbers and tracking.
7. Submit the new XML sitemap in Google Search Console.
8. Use the URL Inspection tool to request indexing for the homepage and your top pages.
9. Verify analytics and conversion tracking fire on the live domain (a real test enquiry beats trusting the tag manager preview).
10. Confirm SSL is valid across the site and HTTP requests redirect to HTTPS in one hop.
11. If the domain changed, submit the Change of Address in Search Console and verify the new property.
12. Check the site renders correctly logged out, in a private window, from a mobile connection.

Warning: Do not delete the old hosting environment for at least a month. If something is badly wrong, rolling back beats debugging in production while your traffic burns.

Phase 8: The First 48 Hours After Launch

The two days after launch decide whether small mistakes stay small. Assign someone to own monitoring, with time actually blocked out to do it.

Hour-by-Hour Priorities

In the first few hours, watch three things.

First, Search Console’s URL Inspection on your key pages. Fetch the live homepage and top money pages, and confirm Google sees indexable, canonical pages with your content rendered.

Second, real-time analytics. Organic visitors should keep arriving via redirects from old URLs. A sudden flatline in organic sessions usually means a robots.txt block, a noindex tag, or redirects failing at scale. Search Engine Land’s migration guidance is blunt about this: when traffic vanishes at launch, check robots.txt and leftover noindex tags before anything else.

Third, server logs or your host’s error dashboard. A spike in 404s tells you which redirects are missing while the problem is hours old instead of weeks old.

Day One and Day Two

Re-crawl the live site in full. Compare against your staging crawl; production deployments have a talent for reverting one template or config file.

Re-test the complete redirect list against the live domain. Then check page speed on live templates, because production tracking scripts, chat widgets and consent banners often drag performance below what staging showed.

Keep a running log of every issue found and fixed, with timestamps. If rankings dip later, this log separates “known issue, fixed in day one” from “something we have not found yet”.

Phase 9: Weeks One to Eight, Monitoring and Recovery

Even a flawless launch produces turbulence. Google has to recrawl every old URL, process every redirect, re-index every new page and re-evaluate the site as a whole. For most small and mid-sized sites this settles within four to eight weeks; larger sites can take longer.

What Normal Looks Like

Expect rankings to fluctuate rather than fall off a cliff. Individual keywords may drop a few positions and recover. Impressions often dip slightly before recovering as redirects get processed.

Normal turbulence is a wobble measured in positions. A problem is a drop measured in pages of results, or a fall in indexed pages that keeps getting worse.

Track weekly against the benchmark you recorded before launch:

• Indexed page count in Search Console versus your expected page count
• Clicks and impressions week over week, segmented by your top 20 pages
• Ranking positions for your benchmark keyword set
• 404 errors reported in the Pages report (fix by adding redirects to your map)
• Core Web Vitals field data as it accumulates for the new templates
• Organic conversions, not just organic sessions

Redesigns are also the natural moment to reset your routine SEO housekeeping. Our ongoing SEO checklist covers the weekly and monthly cadence that keeps a site healthy long after the launch dust settles.

If Rankings Drop and Keep Dropping

Work the diagnostic ladder from the top:

1. Indexing first. Is the site (or a template) noindexed or blocked? Check robots.txt, meta robots and X-Robots-Tag headers.
2. Redirects second. Sample your highest-traffic old URLs. Are they 301ing to the right pages in one hop? Any 404s or 302s?
3. Content third. Compare a dropped page against its old version in the Wayback Machine. Did the word count, headings or target keyword usage change materially?
4. Internal links fourth. Did the dropped page lose its navigation placement or in-content links?
5. Performance fifth. Did Core Web Vitals or hosting response times degrade badly?
6. External factors last. Check whether a Google algorithm update rolled out near your launch window before blaming the redesign for everything.

Most post-redesign drops trace back to items one to four, and most are reversible within a crawl cycle or two of the fix. The traffic usually returns; the panic-driven rebuilds people commission during week three are what cause permanent damage.

The 2026 Layer: AI Search Visibility

A redesign in 2026 has one more surface to protect that the classic checklists ignore: AI answers.

ChatGPT, Gemini, Perplexity and Google’s AI Overviews all cite and link to source pages. Our own analytics show a distinct AI referral channel now, and those visitors convert well above the average organic session. If your pages are being cited today, those citations point at specific URLs, and the same rules apply: redirect them properly or lose them.

A few AI-specific checks to add to the standard list:

• Keep the question-and-answer structure that gets content cited: clear H2/H3 questions, direct answers in the first sentence or two beneath them.
• Preserve FAQ content and its schema markup through the rebuild.
• Confirm your robots.txt decisions about AI crawlers (GPTBot, ClaudeBot, PerplexityBot and friends) are deliberate and carry over to the new site.
• After launch, re-test how the new pages surface in AI tools for your key commercial questions.

Structured, well-organised content wins in AI answers for the same reason it wins in classic search: it is easy to parse and safe to cite. If AI visibility is on your radar, our guide on how to appear in ChatGPT goes deeper, and our AI SEO service covers the full generative search workstream.

WordPress Redesigns: Platform-Specific Notes

Most of the sites we rebuild run WordPress, and it has its own repeat offenders.

Permalink settings are the classic one. A new theme or an agency habit changes the permalink structure from /%postname%/ to something date-based (or vice versa), instantly changing every post URL on the site. Lock the permalink structure decision early and put it in writing.

Watch these too:

• Theme-generated metadata overriding your SEO plugin’s titles and descriptions. Decide which tool owns metadata and disable the other.
• SEO plugin migration. Moving from Yoast to Rank Math or similar mid-redesign multiplies risk; migrate settings with the official importer and re-verify titles, canonicals and schema.
• Attachment pages. Confirm they redirect to the parent post or file rather than publishing as thin standalone pages.
• Tag and category archives. New themes resurface archives that were deliberately noindexed years ago. Re-apply the rules.
• Plugin-generated URLs from old sliders, portfolios and page builders. They are in your crawl archive; map them like any other URL.
• Caching and optimisation plugins. Re-test speed after they are configured on production, not before. On SiteGround, that means re-checking once SG Optimizer settings are applied.

None of these are exotic. They are simply the specific ways WordPress lets the generic mistakes happen.

Domain Changes: The Extra Rules

Everything above assumes you are staying on the same domain. If the redesign coincides with a rebrand and a new domain, several extra rules apply, and the tolerance for error shrinks.

Google’s documentation on site moves with URL changes is the canonical reference here, and it is worth reading in full before a domain move.

The additional steps:

1. Verify both the old and new domains in Search Console before the move.
2. Implement page-to-page 301 redirects from every old URL to its equivalent on the new domain. Domain-level redirects to the new homepage throw away most of your equity.
3. Submit the Change of Address tool in Search Console from the old property.
4. Update your backlinks where you can. Contact the sites behind your most valuable links and ask them to point at the new domain directly. Most will not bother, which is fine; the redirects carry the rest.
5. Update every citation you control: Google Business Profile, directories, social profiles, email signatures, ad accounts.
6. Keep the old domain registered and redirecting for years, not months. The old domain lapsing and being bought by a spammer is a very bad day.

Expect a domain move to take longer to settle than a same-domain redesign. Two to three months of turbulence is normal even when everything is done correctly.

E-Commerce Redesigns: Extra Moving Parts

Online shops carry every risk above, multiplied by catalogue scale and revenue sensitivity.

Product URLs churn constantly (discontinued lines, renamed products, variant handling), so e-commerce redirect maps are bigger and messier. Prioritise ruthlessly: your benchmark data tells you which product and category pages actually earn organic revenue, and those get manual, intent-matched redirects. Long-tail discontinued products can follow rules-based redirects to their parent categories.

Faceted navigation deserves its own review. Filters and sort parameters can generate millions of crawlable URL combinations. Decide which facets deserve indexable landing pages (colour and category combinations with real search demand) and which get canonicalised or blocked, and make sure the new platform enforces the decision.

Watch these on staging:

• Product schema (price, availability, reviews) valid on the new templates
• Out-of-stock and discontinued product handling (keep the page and offer alternatives rather than 404ing anything with links or traffic)
• Category pagination indexable and linked
• Site search results pages noindexed
• Checkout and account pages excluded from crawling but not breaking the crawl

Revenue makes the monitoring phase easier in one way: you will hear about problems fast. Pair Search Console monitoring with daily organic revenue tracking for the first month.

Working With Designers and Developers: Making SEO a Requirement

Most redesign traffic disasters are not technical failures. They are project management failures. The SEO work never made it into the brief, so nobody scoped it, priced it or owned it.

Fix that with paperwork, not vibes.

Put SEO in the Contract

The build contract or statement of work should name the SEO deliverables explicitly:

• URL structure to be preserved except where agreed in writing
• Full redirect map implemented and tested before launch
• Metadata, headings, body content, alt text and schema carried over per the crawl archive
• Core Web Vitals targets on named templates
• Staging environment blocked from indexing, with removal of the block on the launch checklist
• Post-launch support window for redirect and indexing fixes

When these are contractual line items, they get built. When they are a hopeful email from three months ago, they get dropped the moment the timeline slips.

Involve SEO at Wireframe Stage, Not QA Stage

The cheapest moment to fix an SEO problem is before it is designed. A navigation that drops your best service page, a template with no room for body copy, a design that turns your H1 into a background image: each costs minutes to fix in a wireframe review and weeks to fix after build.

Give whoever owns SEO a formal review gate at wireframes, at first template build, and at staging sign-off. Three reviews, each an hour or two, prevent the majority of everything this guide warns about.

Agree Who Owns the Redirect Map

The redirect map fails when it falls between chairs. The SEO builds it, the developer implements it, and each assumes the other tested it.

Name one owner. That person confirms in writing, before launch, that the full map has been implemented and tested against staging with a crawler. “Tested” means the list-mode crawl described earlier, not clicking five URLs by hand.

The Complete Website Redesign SEO Checklist, Condensed

Everything in this guide, in one list you can paste into your project tracker.

Before Design Starts

• Export 16 months of Search Console data (queries, pages, indexing, Core Web Vitals)
• Export 12 months of GA4 organic landing-page and conversion data
• Snapshot keyword rankings, including local pack positions
• Export the backlink profile and flag URLs with external links
• Full crawl of the current site, archived
• Build the master URL spreadsheet (URL, traffic, conversions, backlinks, target queries)
• Classify the project accurately: reskin, redesign, rebuild or migration
• Write SEO requirements into the build contract
• Book SEO review gates at wireframe, template and staging stages

During Design and Build

• Challenge every URL change; keep URLs wherever possible
• Build the 301 redirect map covering every old URL
• Carry over title tags, meta descriptions, H1s and heading structure
• Enforce content parity on top pages (compare word counts)
• Carry over alt text and image optimisation
• Re-implement structured data on the new templates
• Rebuild navigation without dropping money pages
• Update in-content internal links to final URLs
• Block staging via HTTP auth (or noindex with a documented removal step)

Pre-Launch on Staging

• Full crawl of staging; compare against the old-site archive
• Test the complete redirect map in list mode: one hop, correct destinations
• Validate canonicals, robots directives, sitemap, hreflang
• Lighthouse and Core Web Vitals on all key templates
• Mobile rendering and content parity check on real devices
• Rich Results Test on structured data templates
• Fix orphan pages found in the staging crawl

Launch Day

• Backup old site with verified restore
• Deploy, then remove noindex and robots blocks immediately and verify
• Test the redirect map against the live domain
• Submit the XML sitemap; request indexing on key pages
• Manual click-through of navigation and top 20 pages
• Verify analytics, conversion tracking and SSL
• Change of Address in Search Console if the domain changed
• Keep the old environment available for rollback

Post-Launch

• Monitor Search Console, analytics and 404s daily for two weeks
• Full re-crawl of production in week one
• Track rankings, indexed pages and conversions weekly against the benchmark
• Add redirects for any 404s that surface
• Re-check Core Web Vitals field data as it accumulates
• Keep redirects live for at least a year
• Re-test AI answer citations for key commercial questions

A Realistic Timeline: Fitting SEO Into the Redesign Project Plan

Most redesign projects run eight to sixteen weeks. Here is where the SEO work slots in without slowing anything down, based on a typical twelve-week build.

Two things make this timeline work in practice.

The SEO tasks run in parallel with design and build rather than gating them. The redirect map, for example, gets drafted while templates are being built, because both depend on the same agreed sitemap.

The review gates are short and scheduled in advance. An hour at wireframes and an hour at template stage is enough to catch nearly everything expensive. What kills projects is discovering problems for the first time during week eleven QA.

If your project is already at week nine and none of this has happened, do not skip to launch. Pause, run the benchmarking and redirect mapping now, and push the launch date. A two-week delay is invisible in a year’s time; a 40% traffic loss is not.

The Cost of Getting It Wrong: What Traffic Loss Actually Means

It is worth putting numbers on the risk, because “we might lose some rankings” never wins a planning argument against “the new site looks amazing and we want it live for the trade show”.

Take a modest example. A service business gets 5,000 organic sessions a month, converts 2% of them into enquiries, and closes a quarter of enquiries at an average customer value of £2,000. That is 100 enquiries and £50,000 of new business a month riding on organic search.

A botched relaunch that cuts organic traffic 40% for four months costs that business roughly £80,000 in lost pipeline. The full SEO workstream described in this guide, done properly by an agency or a competent freelancer, costs a small fraction of that. Redesign SEO is not an optional extra on the project; it is insurance on every pound of organic revenue the old site currently earns.

The recovery maths is worse than the loss maths. Rankings lost through 404s and thin content do not spring back when you fix the cause; Google recrawls, re-evaluates and restores gradually, and competitors occupy the gap in the meantime. Businesses that lose positions for competitive commercial terms often spend a year of content and link work winning back what a missing redirect file lost in a weekend.

The Five Most Common Redesign Mistakes We See

After enough post-mortems on relaunches gone wrong, the same handful of failures keep appearing. These are the patterns to check your own project against.

The Friday Afternoon Launch

The site goes live at 4pm on Friday. The developer is offline by six. The noindex tag nobody spotted sits on the live site all weekend, and by Monday, Google has begun dropping pages from the index. Everything about the failure was ordinary; only the timing made it expensive.

Launch midweek, in the morning, with the people who can fix things on call.

The Beautiful Ghost Town

The new site wins design awards internally. It also has a quarter of the old site’s content, because every page was cut down to fit the new minimal aesthetic. Rankings slide over six weeks rather than overnight, which makes the cause harder to see: no errors, no 404s, just far less for Google to rank.

Content parity checks on the top 50 pages catch this before launch. Accordions and tabs let long content live inside clean design.

The Redirect Map That Covered “The Important Pages”

The team mapped the 40 pages everyone remembered and skipped the 600 blog posts, old landing pages and PDFs. The long tail was most of the site’s organic entry points, and it all went to 404. Traffic fell by half, and the failure looked mysterious because “we did the redirects”.

The map covers every URL in the crawl and Search Console export, or it is not finished.

The Platform Switch Nobody Mentioned to SEO

Marketing commissioned a redesign; the development agency quietly moved the site from WordPress to their preferred framework. URL patterns, metadata handling, schema and image paths all changed by default. Nobody flagged it because nobody classified the project as a migration.

Classify the project accurately on day one, in writing, using the table at the top of this guide.

The Post-Launch Silence

The site launched cleanly and everyone moved on. Nobody watched Search Console, so the 404 report quietly filled up over a month as Google recrawled deep pages the map had missed. Each individual missing redirect was trivial; a month of them compounded into a real loss.

Monitoring is part of the launch, not an afterthought. Two weeks of daily checks, then weekly for two months.

Measuring Success: What a Good Relaunch Looks Like in the Data

The redesign succeeded when the data says so, not when the stakeholders like the design. Define the success criteria before launch so everyone judges the same scoreboard.

A healthy relaunch, measured at the eight-week mark, looks like this:

• Indexed pages within a few percent of the expected count, with 404s trending to zero
• Clicks and impressions at or above the pre-launch benchmark
• Benchmark keywords within a position or two of where they started, with no losses among the top 20 money terms
• Core Web Vitals passing on all key templates in field data
• Organic conversion rate flat or improved (a redesign that keeps traffic but halves conversion rate has still failed)

From week eight onwards, stop comparing against the old site and start banking the upside. A redesign done on this checklist usually beats the old site rather than merely matching it: faster templates, cleaner structure, consolidated content and better internal linking are all ranking tailwinds once the turbulence passes.

Set a 90-day review with three questions. Did we keep what we had? What improved? Which pages deserve investment next now the platform is stronger? That last question turns the redesign from a defensive exercise into the start of a growth plan, and it is where disciplined SEO work after launch compounds the value of everything the rebuild protected.

Local SEO During a Redesign

If your business serves a geographic area, the redesign touches your local visibility too, and local signals have their own failure modes.

Location pages first. Many local sites rank through a set of town or service-area pages, and these are exactly the kind of “thin-looking” pages a redesign tidies away. Check your benchmark data before letting any of them go: a location page with 30 clicks a month across five towns is often the difference between appearing in nearby searches and not. Carry them over, redirect any that consolidate, and keep their unique local content intact.

Your Google Business Profile links next. The website link on your profile, and any links on individual products or posts within it, may point at URLs that are about to change. Update them at launch rather than leaving them to ride a redirect.

Then the wider citation footprint. Directories, industry bodies, chambers of commerce and local news links all reference specific URLs. The redirects will catch them, but update the ones you control, starting with any that point at pages being merged or retired.

Two structural checks for the new build:

• NAP consistency: name, address and phone number rendered as crawlable text (not just in an image or map embed) on the contact page and footer, matching your Google Business Profile exactly.
• LocalBusiness schema: carried over or added to the new templates, with the same address details, opening hours and geo coordinates.

Local rankings respond to the same fundamentals as everything else in this guide. Keep the URLs, keep the content, keep the signals consistent, and the map pack positions survive the relaunch with them.

Content Consolidation: The One Great Opportunity in Every Redesign

Almost everything in this guide is defensive. Consolidation is the exception: the one move that routinely leaves a site ranking better after a redesign than before it.

Most sites that have existed for five or more years carry content debt. Three blog posts covering the same topic from different years. A service described on two competing pages. Thin pages published for a campaign and never retired. Each of these splits relevance and links across multiple weak URLs instead of concentrating them in one strong page.

A redesign is the perfect moment to fix this, because you are already touching every page and already running a redirect process.

The consolidation workflow:

1. Group your URL inventory by topic and search intent. Your Search Console query data shows you which pages compete for the same terms.
2. For each group, pick the winner: usually the page with the most links and the best engagement, upgraded with the best material from the others.
3. Merge the content properly. Bring across the sections, data points and internal links that earned the losing pages their traffic.
4. Redirect the retired URLs to the winner. This is where the equity concentrates.
5. Update internal links so nothing still points at the retired pages.

Handled this way, consolidation compounds: the winning page inherits the links and history of everything merged into it, and Google gets one clear, deep answer where it previously saw three shallow ones.

Be conservative about what qualifies. Pages ranking for their own distinct queries are not duplicates, even if the topics feel adjacent. Consolidate on intent overlap shown in the data, not on tidiness.

Tracking, Tags and Consent: Do Not Launch Blind

Analytics continuity is not strictly a ranking factor, but lose it at launch and you lose the ability to know whether anything else in this checklist worked.

The redesign typically rebuilds the very templates your tracking lives in. Plan for it:

• Reinstall the tag stack deliberately. Google Tag Manager container, GA4 configuration, ad platform pixels and call tracking all need to be present on the new templates, once each. Duplicated containers inflate every metric and take weeks to untangle.
• Re-test conversions end to end. Submit a real form, make a test call, complete a test purchase. Confirm each lands in GA4, your CRM and your ad platforms. The tag firing in preview mode is not the same as the lead arriving where sales will see it.
• Carry over your consent setup. The cookie banner and consent mode configuration must work on the new build before launch, both for compliance and because consent signals shape what your analytics can measure.
• Keep event names stable. If the new forms or buttons fire differently named events, your year-on-year reporting breaks. Map old events to new ones, or better, keep the names.
• Annotate the launch. Mark the launch date in GA4 and your rank tracker so every future chart explains its own discontinuity.

One hour of tracking QA at launch protects every measurement decision the business makes for the next year. Skipping it is how relaunches end up “feeling quieter” with nobody able to prove why.

A Worked Example: Mapping a Small Site

Abstract rules are easy to nod along to, so here is what the redirect map actually looks like for a fictional 10-page plumbing site moving to a new structure.

Ten rows, ten decisions, nothing left to memory. Notice the patterns: the dated blog URL loses its date but keeps its slug, the two overlapping posts consolidate into one, the dead campaign page goes to its nearest relevant parent rather than the homepage, and even the PDF gets a destination.

Your site has more rows. It does not have different logic. Every redirect map, from ten pages to ten thousand, is this table scaled up, and the discipline of “every URL gets a row and a decision” is what separates the relaunches that hold their traffic from the ones that spend autumn writing recovery plans.

Score your current site’s technical health before you start, too. A free SEO audit style review of the existing site often surfaces problems worth fixing in the rebuild rather than faithfully migrating: redirect chains from the last redesign, orphaned sections, and crawl waste that the new structure can retire for good.

Hosting, HTTPS and Infrastructure: The Layer Under the Redesign

Redesigns often bundle a hosting move, and infrastructure changes have their own quiet ways of hurting rankings. None of them are complicated; all of them are forgettable at 11pm on launch night.

Server response time sets the floor for everything Core Web Vitals measures. If the new site moves to cheaper hosting to offset the design spend, test time-to-first-byte on staging under something resembling real load. A beautiful template on a struggling server still fails LCP.

Check the HTTPS setup end to end. The certificate must cover every hostname variant in use, www and non-www must resolve to one canonical version in a single hop, and old HTTP URLs must still redirect. Redesigns have a habit of resetting these rules to platform defaults, quietly creating two-hop chains (http to https to new URL) that waste crawl budget and dilute signals.

If a CDN enters the picture for the first time, confirm three things: it passes the correct status codes rather than caching 404s, it does not strip or override canonical and robots headers, and its firewall rules do not block Googlebot. A misconfigured bot-protection rule that challenges crawlers looks, from inside Search Console, exactly like the site went down.

DNS changes deserve respect too. Lower the TTL a couple of days before launch so the switch propagates quickly, and keep the old server answering until propagation completes everywhere. The hour where half the world sees the old site and half sees the new one is normal; make sure both halves work.

Finally, email. Moving hosts often means moving DNS zones, and it is depressingly common for a flawless site launch to break MX records and take the enquiry inbox down with it. The leads the redesign was meant to win arrive nowhere. Confirm mail flow the same hour the site goes live.

Forms, Gated Content and the Pages Nobody Remembers to Test

The commercial plumbing of a site (forms, thank-you pages, downloads, booking flows) sits slightly outside classic SEO, but a redesign breaks it just as readily, and broken plumbing turns preserved rankings into preserved traffic with no revenue attached.

Thank-you pages are the sneaky one. They are usually noindexed (correctly), which means nobody thinks of them as SEO assets, and they are frequently rebuilt with new URLs. Two problems follow. Conversion tracking that fires on the thank-you URL silently stops counting, and any historic paid campaigns pointing at old form pages start 404ing. Map thank-you pages and form URLs like everything else, and re-point conversion triggers at the new destinations.

Gated resources (guides, whitepapers, calculators, portals) need the same row-by-row treatment. The gate page carries the rankings and links; the asset behind it carries the value. Both need destinations, and the form connecting them needs an end-to-end test with a real submission.

Downloads and documents generally deserve a rethink during a redesign rather than a blind copy. PDFs rank, but pages rank better, update more easily and convert visitors that PDFs strand. Where a PDF earns meaningful search traffic, consider rebuilding it as a page and redirecting the file, exactly as the worked example above did with its pricing document.

Booking and quote flows should be walked through on launch day on a phone, on mobile data, as a stranger would. Not because Google ranks the flow, but because the whole point of protecting your rankings is what happens after the click. A relaunch that keeps 100% of its traffic and breaks its enquiry form has lost 100% of what mattered.

While you are in the forms, check the small print around them: privacy policy links, consent checkboxes and the destination CRM. Automated follow-up that used to trigger from form fields with particular names can fail silently when a rebuilt form renames them, and a well-run email nurture setup is only as reliable as the form data feeding it.

Paid Campaigns During the Switchover

Organic search takes weeks to settle after a relaunch. Paid campaigns break in minutes, which at least makes them honest.

Every live ad, sitelink and asset that points at a changing URL needs updating at launch. Redirects will rescue clicks in the meantime, but ad platforms periodically re-crawl destination URLs, and a destination that redirects (or worse, 404s) can get ads disapproved mid-flight. Export all final URLs from your ad accounts during benchmarking and treat them as another tab in the master spreadsheet.

Landing page quality changes cut both ways. A faster, cleaner page typically lifts Quality Score and conversion rate; a landing page that lost half its copy in the redesign does the opposite. Watch cost per lead alongside the SEO metrics in the fortnight after launch, and compare against the pre-launch baseline you exported.

If the budget conversation ever pits the redesign against ad spend, remember they compound. Paid traffic onto a rebuilt, faster, better-converting site costs the same per click and returns more per visit. Getting the Google Ads account and the new landing pages reviewed together in the same fortnight is one of the highest-yield hours in the whole project.

The Tools That Do the Heavy Lifting

You do not need an enterprise stack to run this checklist well. A small business can execute everything in this guide with the following kit.

Two usage notes matter more than the tool list itself.

Screaming Frog’s list mode is the workhorse of the whole project. Paste in your full set of old URLs and it tests every redirect in minutes: status code, destination, hop count. Learn this one feature and you have automated the most tedious and most critical QA task in the checklist.

Keep every export in one shared, dated folder. The value of benchmarks is being able to find them under pressure three weeks after launch, when someone senior is asking why enquiries dipped and you need the answer to be data rather than memory.

Should You Do This Yourself or Bring in Help?

An honest sizing guide, since the answer depends on the site.

Run it in-house when the project is a true reskin or a small redesign: under a hundred URLs, URLs staying put, content carrying over unchanged. The checklist in this guide, followed carefully, is enough. The work is methodical rather than difficult, and the tools are mostly free.

Get experienced help when any of the following is true: URLs are changing at scale, the CMS or domain is changing, the site has hundreds of pages or meaningful organic revenue, or nobody internally has run a migration before. The cost of support is small against the downside, and an experienced operator will also spot the opportunities a defensive checklist misses.

The worst option is the middle path: assuming the web design agency “handles the SEO” without anyone confirming what that means. Design agencies build sites; very few crawl-test redirect maps or benchmark Search Console data unless it is explicitly scoped. Ask the direct question, get the answer in writing, and fill the gap if the answer is vague.

A redesign is also the natural moment to look at the bigger picture. If the site is being rebuilt to win more business, the technical protection in this guide is table stakes, and the growth comes from what you do with the stronger platform afterwards: content, conversion paths and lead generation strategy built on top of it.

Professional FAQs

How long does it take for rankings to recover after a website redesign?

For a same-domain redesign with correct redirects, expect four to eight weeks of fluctuation before rankings stabilise. Domain changes typically take two to three months. If rankings are still falling (rather than wobbling) after four weeks, work the diagnostic ladder: indexing, redirects, content parity, internal links, then performance.

Will I lose SEO if I redesign my website?

Not if you protect it. Rankings are lost through changed URLs without redirects, deleted content, lost metadata, indexing blocks and slower templates, not through visual redesign itself. A redesign that keeps URLs, content and technical signals intact can launch with no meaningful loss, and often improves rankings once Google processes the faster, cleaner build.

Do I really need 301 redirects if only a few URLs are changing?

Yes. Every changed URL needs a 301 redirect to its new location, even if only five URLs change. Each old URL carries its own indexing history and backlinks, and a 404 throws that away. Five redirects take ten minutes to implement, so there is no economy in skipping them.

Should I redesign my website and change domain at the same time?

Avoid it if you can. Each is a manageable risk on its own; together they multiply and become harder to diagnose when something goes wrong. If the rebrand forces both, run page-to-page redirects, use Search Console’s Change of Address tool, and budget for a longer stabilisation period with closer monitoring.

How do I check if my web designer is handling SEO properly?

Ask for five artefacts: the pre-launch benchmark exports, the master URL spreadsheet, the completed redirect map, the staging crawl report, and the launch-day checklist. A team doing this properly produces all five without hesitation. If the answer is “our sites are built SEO-friendly”, that is a description of the templates, not a migration process, and you should assume the process does not exist.

How much does redesign SEO cost compared with the redesign itself?

For a typical small business site, the full protective workstream (benchmarking, redirect mapping, staging audits and eight weeks of monitoring) usually runs between 10% and 25% of the design and build budget, whether delivered in-house as staff time or bought from an agency. Set that against the pipeline value of your current organic traffic and it is nearly always the cheapest line on the project. The expensive version of redesign SEO is the one purchased after launch, as a recovery project, with the traffic already gone.

Protect First, Then Grow

A website redesign should be the best thing that happens to your search visibility this year, not the worst. The difference is process, and the process is not complicated: it is a checklist, followed in order, with names against each item.

The essentials one more time:

• Benchmark before anything changes: Search Console, GA4, rankings, backlinks and a full crawl.
• Keep URLs wherever possible, and build a complete 301 redirect map for everything that moves.
• Carry over what Google already rewards: titles, headings, content depth, alt text and schema.
• Audit staging like it is launch day, including the full redirect list test.
• Launch midweek, remove staging blocks immediately, and verify with your own eyes.
• Monitor daily for two weeks, weekly for two months, against your benchmarks.

Do that, and the new site inherits every ranking the old one earned, with a faster, better platform to grow from.

If you are planning a redesign and want the SEO side handled by people who do this routinely, we are happy to look at your project before anything gets built. Book a call with GLO and we will tell you exactly what your site needs protected, and where the relaunch can win you more than you have today.