Website Redesign Checklist: 27 Steps Before You Launch
Most redesign damage is self-inflicted and predictable: URLs changed without redirects, ranking content cut, staging sites left blocked from Google. This checklist exists so none of that happens to yours.
This website redesign checklist exists because most redesign damage is self-inflicted and follows the same script: URLs changed without redirects, ranking content cut for looking dated, a staging noindex tag shipped to production. Sites lose 40% of their organic traffic in a launch week this way, and every item of it was preventable.
Below are 27 steps in five phases, from benchmark to post-launch monitoring. Copy them into a document, assign an owner to each, and check them off in order. The boring ones protect the most value.
How do you use this website redesign checklist?
Three rules. Every step gets a named owner; "the team" ships nothing. The phases run in order, because design decisions made before the audit are guesses wearing confidence. And the benchmarks in phase one are non-negotiable, since without them you cannot know in week two whether the redesign worked or quietly broke something.
Scale it honestly: a 5-page brochure site can compress phases two and three into a day, but no site skips the URL mapping, the redirects, or the post-launch monitoring. Those three are where the irreversible mistakes live, and they cost the least to do right.
Timing note: the audit phase takes a few days, and the two monitoring weeks at the end are fixed. Whatever the build schedule promises, the checklist adds roughly three weeks around it, and compressing those specific weeks is where launches go wrong. Assign the monitoring window to a named person before launch day, because afterward everyone believes the project is over.
Phase 1 (steps 1 to 7): what do you audit and benchmark?
1. Export your top pages and queries from Search Console; this list defines what must survive.
2. Record current rankings for your money keywords, with dates.
3. Capture Core Web Vitals scores as your speed baseline.
4. Crawl the site for a complete URL inventory, including images and PDFs people link to.
5. Identify converting pages in analytics, since traffic and revenue live on different pages.
6. Archive all current copy; you will want lines from it later.
7. Take a full backup, files and database, stored off the server.
Phase 2 (steps 8 to 13): what happens to content and structure?
8. Decide keep, merge, or cut for every page, using the phase-one data instead of taste.
9. Map every old URL to its destination in a spreadsheet: same URL, new URL, or gone.
10. Keep URLs unchanged wherever possible; an unchanged URL is a risk removed.
11. Write the new copy before design starts, so layouts serve real content instead of lorem ipsum.
12. Plan internal links between services, locations, and guides.
13. Draft titles and meta descriptions for every page while context is fresh.
Step nine's spreadsheet becomes the project's most consulted document: the redirect source in phase four, the QA sheet at launch, and the diagnostic tool if week-six numbers dip. Build it once, carefully, with a row for every URL from the phase-one crawl.
Phase 3: design and build (steps 14 to 19)
14. Design mobile-first; that is where your visitors are.
15. Set a speed budget, for example LCP under 2.5 seconds, and hold the build to it.
16. Test every form end to end, including where submissions actually arrive.
17. Keep one H1 per page and a logical heading hierarchy.
18. Compress and convert images; media weight is the most common speed regression.
19. Keep staging blocked from indexing, and write down where that block lives, because step 24 depends on remembering it.
Hold the speed budget in review sessions, where it is easiest to lose. Every added slider, font, and script arrives with a good reason, and the budget exists precisely because good reasons accumulate into a slow site.
Phase 4 (steps 20 to 24): how does the SEO migration work?
20. Implement a 301 redirect for every URL that changes, from the step-nine map, before launch.
21. Set canonical tags on the new pages.
22. Carry over or add structured data for your business and services.
23. Generate the new XML sitemap, ready to submit.
24. Remove the staging noindex at launch, then verify with a live crawl. This is the single most damaging step to skip, and teams skip it every week. Google's site move documentation covers the mechanics behind this whole phase.
Phase 5: launch and after (steps 25 to 27)
25. Crawl the live site on day one: broken links, redirect chains, missing pages, lingering noindex tags.
26. Submit the sitemap and watch Search Console daily for two weeks; indexing errors caught early are cheap, and the same errors caught in month two are not.
27. Compare rankings, traffic, and Core Web Vitals against your phase-one benchmarks at two, six, and twelve weeks. Small dips that recover within a month are normal crawl turbulence. Losses past six weeks mean something specific broke, and your URL map plus the benchmark data is how you find it.
Keep the old site's backup and the redirect map for at least six months. When a forgotten URL surfaces from an old ad or a directory listing, restoring its path takes minutes with the map and days of archaeology without it.
Why does the website redesign checklist not end at step 27?
A redesign is a reset, never a result. The site that emerges is faster and cleaner, and from that day it either keeps receiving attention or begins the same slide that made this project necessary. Rankings compound for sites that keep publishing; they erode for sites that freeze.
So add the standing item 28: someone owns the site monthly. New pages for what customers search, content refreshed when reality changes, speed watched. That is the arrangement WebsiteOS sells as a service, redesign included, with 2-3 new pages published every month for a flat monthly price. Run the project through proper redesign services or your own team; either way, the checklist that matters most is the one that repeats.
Frequently asked questions
What is the most important step in a website redesign?
URL mapping plus 301 redirects, steps 9 and 20. Changed URLs without redirects are the leading cause of post-redesign traffic loss, and the damage compounds while it goes unnoticed. A close second is removing the staging noindex at launch, a one-line mistake that has deindexed entire sites for weeks before anyone checked.
Do I need all 27 steps for a small website?
No. A 5-page site can compress the content and design phases into a day or two. What no site should skip: the Search Console benchmark, the URL inventory and redirect map, form testing, the staging noindex removal, and two weeks of post-launch monitoring. Those few steps carry nearly all the irreversible risk.
Will my rankings drop after a redesign?
Expect minor turbulence for two to four weeks while Google recrawls; that is normal and recovers on its own. Losses that persist past six weeks are not normal, and they trace to specific causes: missing redirects, cut content, slower pages, or indexing blocks. The phase-one benchmarks are what let you find and fix the cause instead of guessing.
How long should I monitor the site after relaunch?
Actively for two weeks: Search Console daily, checking coverage errors, redirect problems, and indexing status. Then checkpoints at six and twelve weeks against your pre-redesign benchmarks for rankings, traffic, and speed. Most launch-week problems surface in the first fortnight, and the ones that appear later are exactly why the twelve-week comparison exists.
Who should run the redesign checklist?
One named person owns the checklist end to end, whoever that is: you, your provider's project lead, or an operator service. Split ownership is how steps fall between chairs, and the steps that fall are always the invisible ones like redirects and noindex removal. If you hire the redesign out, ask who owns this list, and expect a name.
Your website, running itself.
Get a Free Quote