A website redesign checklist protects the traffic you already have. Before any design work starts, benchmark your rankings, export every URL, map old pages to new ones, and agree on what success looks like. Most redesign traffic losses come from skipped preparation, not bad design.
Most website redesigns lose organic traffic, and the loss is almost never caused by the new design.
It is caused by the fourteen small things nobody wrote down before the project started. A URL structure that changed without redirects. A staging site that shipped with noindex still in the header. Three years of blog posts quietly dropped because nobody could agree who owned them.
This website redesign checklist covers the 22 things to do before you approve a single mockup. You will get the benchmarking steps that let you prove whether the redesign worked, the SEO preparation that keeps your rankings intact, and the launch sequence that catches problems in hours instead of quarters.
Start with why redesigns go wrong, because the failure pattern is consistent.
Why Most Website Redesigns Lose Traffic
A redesign is a migration, and migrations break things.
When you rebuild a site, you usually change three systems at once: the design, the content, and the URL structure. Each one carries risk on its own. Changed together, without documentation, they produce a launch where nobody can tell which change caused the drop.
Google’s own guidance on site moves with URL changes treats this as a serious operation requiring redirect mapping, sitemap updates, and monitoring over weeks. Most agency redesign proposals in Bangkok treat it as a line item called “SEO transfer” with no detail attached. Ask what that line item actually contains before you sign.
The fix is unglamorous. You do the preparation work in weeks one and two, before anyone opens Figma.
Phase 1: Benchmark Before You Change Anything (Items 1 to 6)
You cannot prove a redesign succeeded if you never recorded where you started.
1. Export your full URL inventory. Pull every indexed URL from Screaming Frog, your XML sitemap, and the Search Console Pages report. The three lists will disagree. The gaps between them are pages you forgot you had.
2. Record 12 months of baseline metrics. Organic sessions, conversions, top landing pages, and branded versus non-branded queries. Screenshot the Search Console performance graph and store the CSV. Google only retains 16 months of data, so an export is the only durable record.
3. Rank your pages by organic value. Sort landing pages by sessions and by assisted conversions. The top 20 percent are non-negotiable. Every one of them needs a matching URL on the new site or a deliberate redirect.
4. Export your backlink profile. Use Ahrefs or Semrush to list referring domains and the specific pages they point to. A page with 40 referring domains cannot quietly disappear, however ugly it looks in the current design.
5. Capture your Core Web Vitals baseline. Record LCP, INP, and CLS from the Search Console Core Web Vitals report. Google’s thresholds are 2.5 seconds for LCP, 200 milliseconds for INP, and 0.1 for CLS. Set these as the floor the new build must beat.
6. Write down the definition of success. One primary metric, agreed in writing, before the project starts. “More modern” is not a metric. “Hold organic sessions flat and lift demo requests by 30 percent within 90 days” is.
Finish this phase before briefing a designer. It takes about a week and it is the only insurance policy you get.
Phase 2: Content and SEO Decisions (Items 7 to 13)
Every redesign is a content audit disguised as a design project, so run the audit deliberately.
7. Score every page: keep, merge, rewrite, or kill. Use traffic, conversions, backlinks, and strategic value as the four inputs. Pages with none of the four get killed. Pages with backlinks but no traffic get merged into something stronger.
8. Build the redirect map before the build starts. One spreadsheet, old URL in column A, new URL in column B, one to one wherever possible. Redirecting everything to the homepage is the single most common and most damaging shortcut in this industry.
9. Preserve your URL structure unless you have a reason to change it. Changed URLs cost you time and equity even when the redirects are perfect. If your current structure is clean, keep it and remove one variable from the project.
10. Migrate your title tags and meta descriptions deliberately. New CMS templates frequently overwrite hand-written metadata with an automated pattern. Export what you have, then check it survived. Meta descriptions should stay under 130 characters to Inspira standard.
11. Preserve heading structure on high-value pages. Designers often flatten H2s into styled paragraphs. Keep one H1 per page and keep the hierarchy intact, since headings still carry meaning for both crawlers and AI answer engines.
12. Audit and migrate your schema markup. Article, LocalBusiness, FAQ, and Service schema all live in templates that get rebuilt. Confirm your structured data ships with the new site rather than being rediscovered six months later.
13. Plan internal linking into the new architecture. A redesign that reduces navigation depth from four clicks to two changes how equity flows through the site. Decide that on purpose. If you need help identifying which clusters matter, our SEO audit process for Thai sites covers the mapping approach.
Do this work in a spreadsheet your developer can actually read. The redirect map is the deliverable that matters most.
Inspira Insight
At Inspira, we audited a Bangkok B2B site three months after a redesign that had dropped organic traffic by roughly half. The design was fine. The cause was a redirect map that sent 300 old blog URLs to a single new blog index page. Google treated those redirects as soft 404s and dropped the pages. Here is what that taught us: a redirect map is not a checkbox on a launch plan. It is the project.
Phase 3: Design, UX, and Technical Preparation (Items 14 to 18)
Design decisions have SEO consequences, and the time to catch them is at wireframe stage. This is where our web design team in Bangkok spends most of its planning effort.
14. Design mobile first, and test on real Thai networks. Thailand runs overwhelmingly on mobile. A hero section that loads beautifully on office fibre can take five seconds on 4G in Chatuchak. Test on a real device, not a browser emulator.
15. Keep primary content in the initial HTML. JavaScript-rendered content adds a rendering step for Google and may be skipped entirely by AI crawlers that do not execute JavaScript. If a paragraph matters for ranking, it should be in the server response. Our technical SEO team in Bangkok checks this at staging, not after launch.
16. Plan for Thai and English versions properly. If you serve both languages, decide on subdirectory structure and hreflang tags now, not after launch. Retrofitting a second language into a single-language build is expensive.
17. Set an image and asset budget. Agree maximum page weight per template before design begins. Specify WebP or AVIF, explicit width and height attributes to prevent layout shift, and lazy loading below the fold.
18. Keep the staging site out of the index. Password protect it. Do not rely on noindex alone, because noindex is the tag most likely to survive into production and take your entire site offline in search results.
Agree these constraints in the design brief. Retrofitting performance into a finished build rarely works.
Phase 4: Launch Day and the First 30 Days (Items 19 to 22)
The 48 hours after launch decide whether small problems stay small.
19. Run the launch-day crawl immediately. Crawl the live site the hour it ships. Check for the noindex tag, broken internal links, missing canonicals, and redirect chains longer than one hop. Fix these on the same day.
20. Submit the new XML sitemap and validate the old URLs. Upload the updated sitemap to Search Console and use the URL Inspection tool on your ten most valuable old URLs to confirm each redirect resolves where you mapped it.
21. Monitor Search Console daily for four weeks. Watch coverage errors, the Core Web Vitals report, and impressions on your top 20 landing pages. Google’s documentation on site moves notes that ranking fluctuations during a migration are expected, so read the trend, not the daily noise.
22. Keep the redirects live permanently. There is no expiry date. Old links keep sending traffic for years, and removing redirects at the next redesign compounds the loss from this one.
A redesign is not finished at launch. It is finished when the numbers you benchmarked in phase one have recovered or improved. If you are reading this after a launch that already went wrong, start with our guide to recovering from a traffic drop after a redesign instead.
Frequently Asked Questions
How long does a website redesign take?
A small business site typically takes 8 to 12 weeks from kickoff to launch, and a larger site with hundreds of pages takes 4 to 6 months. Roughly a quarter of that time belongs to preparation: URL inventory, content audit, and redirect mapping. Projects that skip the preparation phase appear faster, then spend the saved time recovering lost rankings after launch. We break the schedule down stage by stage in how long a website takes to build.
Will a website redesign hurt my SEO?
It can, and the risk comes from changed URLs, dropped content, and missing redirects rather than from the design itself. A redesign that keeps its URL structure, migrates metadata and schema, and maps every old page to a new one usually holds rankings steady. Expect some fluctuation for two to four weeks after launch even when everything is done correctly.
Should I change my URLs during a redesign?
Only if the current structure is genuinely broken, such as URLs containing session IDs, dates that no longer apply, or a deep subdirectory that limits eligibility for features like Preferred Sources. Changed URLs always cost time and some equity, even with perfect 301 redirects. If your structure is clean, keeping it removes one large variable from the project.
How do I know if my website needs a redesign?
Look at evidence, not age. Rising mobile bounce rates, failing Core Web Vitals, a CMS your team cannot update without a developer, and conversion rates below your industry benchmark are all real signals. A design that simply looks dated is a weaker reason than a site that costs you leads every week.
What should I benchmark before a website redesign?
Export your full URL list, 12 months of organic sessions and conversions, your top landing pages by revenue or leads, your backlink profile with target URLs, and your Core Web Vitals scores. Store these outside the platform being replaced. Without a baseline you cannot prove the redesign worked, and you cannot diagnose what broke if it did not.
Conclusion
The most valuable item on any website redesign checklist is the one nobody wants to pay for: the redirect map.
You now know that redesign traffic losses come from migration errors rather than design choices, that benchmarking has to happen before work starts because Search Console data expires, and that launch day is the beginning of the monitoring period rather than the end of the project. You also know which questions to ask an agency before signing, starting with what exactly their “SEO transfer” line item includes.
If you are planning a redesign this year, request a free pre-redesign audit from Inspira Digital Agency and we will build the URL inventory and redirect map with you.
