# How to redesign your website without losing SEO

*Published: 2026-08-12*

*Author: Griffo — Senior Brand & Digital Designer*

Redesigns fail SEO for boring, preventable reasons. Here is the migration protocol we run on every project: benchmark, map redirects, keep content parity, and watch Search Console for 90 days.

Yes, you can redesign your website without losing SEO, but only if you treat the redesign as a migration rather than a facelift. Benchmark what ranks today, map every old URL to a new one with 301 redirects, keep the content on your top pages intact, remove the noindex tag from staging before launch, and watch Search Console for the following 90 days.

That paragraph is the whole method. The rest of this article is how to actually do each step, in the order we run them on client projects, plus the ways we have seen redesigns go wrong when a step gets skipped.

## Can a website redesign hurt your SEO?

It can, badly, and the damage rarely comes from the design itself. Rankings are attached to specific URLs, the content on them, the internal links between them, and technical signals like structured data. A redesign typically changes all four at once, which forces Google to re-evaluate a site it had already made its mind up about.

The risk scales with how deep the project goes. A visual reskin that keeps the same URLs, content, and CMS is low risk. A replatforming project that moves you from an old CMS to a new stack, restructures the URLs, and rewrites the copy is a genuine migration, and migrations fail when nobody owns the SEO side of them. If you are weighing that bigger jump, our comparison of [custom builds and templates](/posts/custom-website-vs-template) covers when replatforming is worth the disruption in the first place.

One distinction worth holding onto: fluctuation is normal, loss is not. Google recrawling a changed site produces a few weeks of wobble in positions and impressions. A cliff that appears at launch and never recovers means something specific broke, and it is almost always findable.

## What should you benchmark before touching anything?

Before the first design file opens, record what the current site has earned. You cannot recover what you never measured, and post-launch debugging without a baseline is guesswork.

The benchmark has four parts. First, a full crawl of the live site with a tool like Screaming Frog: every URL, its title, meta description, headings, canonical tag, and status code. This crawl becomes the left column of your redirect map later, so keep the export. Second, Search Console data: top queries and top pages by clicks and impressions, downloaded for at least the last 12 months so seasonal patterns do not fool you. Third, an analytics export of your most valuable pages by traffic and by conversions. These are not always the same pages, and the conversion list is the one clients forget. Fourth, a record of the invisible layer: structured data markup, hreflang tags if you operate in several languages, and which pages carry backlinks (any backlink tool's top pages report will do).

Annotate the launch date in your analytics the day you go live. Ninety days later, when someone asks "did the redesign work?", you will be glad the line exists.

![56](https://media.secondseason.studio/magnific_recreate-img1_vQLD1jYa47-d7a23e51f01e70d8.webp)

## How do you build the redirect map?

A redirect map is a spreadsheet with two columns: every URL on the old site, and the URL on the new site that replaces it. Every row becomes a 301, a permanent redirect that tells Google the page has moved and passes its accumulated authority to the new address.

The rules that matter:

Map one to one wherever possible. The old service page redirects to the new service page, not to the services overview and never to the homepage. Google treats a mass redirect of unrelated URLs to the homepage as a soft 404, and the ranking value evaporates. If a page has no true equivalent on the new site, redirect it to the closest relevant parent, and if nothing is even close, let it 404 honestly rather than faking relevance.

Include everything the crawl found, beyond just the pages in your navigation. Old blog posts, PDFs, image URLs that rank in image search, and parameterized URLs all carry value or traffic. This is why the pre-redesign crawl matters: your sitemap lies about what actually exists.

Flatten chains. If the old site already had redirects from an earlier migration, point those original URLs directly at the final new destination instead of stacking hop on hop. Chains leak crawl budget and occasionally break.

And a caution we give every client considering it: changing the domain at the same time as the redesign is combining the two riskiest moves in SEO into one launch. Sometimes a rebrand forces it. If so, use the change of address tool in Search Console and plan for a longer recovery window. If the domain change is optional, do it in a separate project, months apart.

## What is content parity, and why does everyone skip it?

Content parity means the new version of a page says substantially the same things as the old one: same topics covered, comparable depth, same title tag, same heading structure where it made sense, same structured data. It is the step everyone skips because it fights the natural energy of a redesign. The new design is clean and minimal, the old page had 900 words on it, and someone decides the words were clutter.

Those words were often the reason the page ranked. Google matched queries against that specific text for years. When the redesign replaces it with a hero image and three short value propositions, the page becomes beautiful and invisible at the same time. We have watched this exact trade happen, and it annoys us every time, because the fix is so unglamorous: keep the text.

The practical version of the rule: for your top pages by traffic and conversions (the benchmark list from earlier), launch with parity. Not pixel parity, content parity. The design can change completely; what the page communicates and the terms it uses should not. Once rankings have held steady for a couple of months after launch, improve the content deliberately, one page at a time, so you can see what each change does. Pages that earned nothing before the redesign are free territory; rewrite them from scratch.

Parity also covers the technical layer. If the old pages carried FAQ or product structured data, the new templates need to output it too. Structured data does not migrate itself, and a new frontend built by a team that never saw the old markup will silently drop it.

![43](https://media.secondseason.studio/magnific_recreate-img1_s7ZFy0kl8e-dda0396845a200bc.webp)

## What belongs on the launch-day checklist?

Launch day is where redesigns are won or lost, usually within the first hour. This is the list we run, in order:

1. Remove the noindex tag and any robots.txt disallow rules from the new site. Staging environments are blocked from search engines on purpose, and shipping that block to production is the single most common catastrophic redesign mistake.
2. Confirm the 301 redirects are live by testing a sample from your map: top pages, a few blog posts, one PDF, one deep URL. Check the status codes rather than trusting where the browser lands.
3. Crawl the new production site immediately. Look for 404s, redirect chains, missing titles, and accidental noindex tags on individual templates.
4. Check canonical tags. Every page should point at itself (or its intended canonical), not at the staging domain. Staging canonicals leak into production more often than anyone admits.
5. Submit the new XML sitemap in Search Console and confirm the old sitemap no longer lists dead URLs.
6. Validate structured data on the key templates with Google's Rich Results Test.
7. Confirm analytics and conversion tracking fire on the new site. A redesign that breaks tracking looks like a traffic drop even when rankings held.
8. Run a performance check on the main templates. New sites should be faster, but heavy hero videos and unoptimized fonts have reversed that expectation before.
9. Spot-check the site on a phone, on a real connection. Not the office wifi.
10. Leave the old hosting and DNS access alive for a few weeks. When something surfaces, and something usually does, you want the old site's configuration available for reference.

Print it, assign each line to a named person, and do not launch on a Friday. Nobody fixes redirect maps well at 11pm on a weekend.

## What should you watch for the first 90 days?

Monitoring is the phase everyone under-budgets, because the launch felt like the finish line. It is not. Google now has to recrawl the site, follow every redirect, and rebuild its understanding of your pages, and that takes weeks.

In Search Console, watch four reports. The page indexing report shows whether new URLs are being indexed and old ones are dropping out in an orderly way. The 404 report catches URLs your redirect map missed; fix these weekly, prioritized by whether the URL has backlinks or traffic history. The performance report shows impressions and clicks per page: impressions usually recover before clicks, so early impression recovery with lagging clicks is a good sign, not a bad one. Crawl stats show whether Googlebot is working through the site at a healthy rate.

Expect fluctuation for 4 to 12 weeks. Individual keywords will jump around while the aggregate trend matters more. Compare against the same period last year where you can, because seasonality has been misdiagnosed as migration damage many times.

There is also a 2026 layer to this. Visibility in AI search features runs through the same plumbing as classic search: Google's AI optimization guide, updated in June 2026, states that pages only need to be indexed and snippet-eligible to appear in AI features, with no special file, format, or writing style required. A redesign that breaks indexing therefore breaks your AI visibility too, at a moment when SparkToro and Similarweb data shows 68.01% of US Google searches ended without a click in early 2026. Being the source AI answers quote is worth protecting as deliberately as rankings. We cover how those systems choose sources in [our guide to showing up in ChatGPT and AI search](/posts/website-visibility-ai-search).

![17](https://media.secondseason.studio/magnific_recreate-img1_EbSdhr5uuO-6a5dbe009f78a41d.webp)

## Where do redesigns actually go wrong?

We get called in for redesign rescues a few times a year, and the failures repeat so reliably that we can usually guess the cause before opening Search Console. No numbers here, because every situation differs, but the patterns are real.

The staging block that shipped. A B2B services company launched a genuinely better site and vanished from search within days. The noindex tag from staging had gone live with everything else. Nobody on the project was checking Search Console, so it took weeks to notice, and the recovery took far longer than the mistake did. This failure is so common it is item one on our checklist above.

The homepage dump. An e-commerce client's previous agency had handled "redirects" by pointing several hundred discontinued category and product URLs at the homepage. Google treated them as soft 404s, the accumulated authority of years of links went nowhere, and the category pages that replaced them started from zero.

The content massacre. A professional services firm redesigned toward a minimal aesthetic and cut its service pages from long, specific explanations to a sentence and a button. The new site converted its visitors fine. It just had far fewer of them, because the text that had matched hundreds of long-tail queries no longer existed.

The aesthetic URL restructure. Midway through a project, someone decides the URLs should be prettier, so /services/brand-strategy becomes /what-we-do/strategy. No user asked for this, the redirect map doubles in size, and risk is added for zero benefit. If the old URLs are functional, keep them.

The ownership gap. The most common failure is not technical. The design agency assumed the client handled SEO, the client assumed the agency did, and the migration protocol belonged to nobody. Before any redesign starts, one named person should own the redirect map, the parity review, and the 90-day monitoring. If your agency cannot tell you who that person is, that silence is your answer.

A redesign done with this protocol is genuinely safe. We have run migrations where traffic held through launch week and grew afterward, because the new site was faster and better structured than the old one. The protocol is not exotic. It is a spreadsheet, a checklist, and someone accountable for both. If you want a sense of how this fits into a full project schedule, see [how long a website actually takes](/posts/how-long-does-a-website-take).

## FAQ

### How long does it take for rankings to stabilize after a redesign?

Expect some movement for 4 to 12 weeks while Google recrawls the site, follows the redirects, and re-evaluates the pages. If redirects and content are handled properly, positions usually return to their previous level within that window. A drop that persists past three months with no recovery trend points to a real problem, most often missing redirects or thinned content.

### Do I need to keep the same URLs when redesigning?

No, but keeping them is the single safest choice you can make. Every URL you preserve keeps its ranking history, backlinks, and internal link value without any redirect risk. Change URLs only when the old structure is genuinely broken, and then map every old address to its closest new equivalent with a permanent 301 redirect.

### Will changing my domain name during a redesign lose SEO?

A domain change is the riskiest move you can combine with a redesign, because Google must re-evaluate both the new site and the new domain at once. If you must do it, use Google's change of address tool in Search Console, redirect every old URL to its exact new counterpart, and accept a longer stabilization period. Where possible, separate the two projects.

### Should I rewrite my content at the same time as the redesign?

Not on your best performing pages. Changing design, URLs, and content simultaneously means you cannot tell which change caused a drop. Launch the redesign with content parity on the pages that earn your traffic, confirm rankings hold, then improve the content in a second phase. Pages that earn nothing today can be rewritten freely.

### Do 301 redirects pass the full ranking value of the old page?

Google has said for years that 301 redirects pass PageRank without loss, and in practice a clean one-to-one redirect preserves most of a page's authority. What loses value is sloppiness: redirect chains, redirects pointing at irrelevant pages, or hundreds of URLs dumped onto the homepage. Google treats those bulk redirects roughly like 404s.

Second Season designs and builds websites, and the SEO migration protocol above is part of every redesign we deliver, not an add-on. If your site is due for a redesign and the traffic it earns is too valuable to gamble with, [talk to us](/contact) about how we would run yours.