Website Migration SEO: What We're Learning Rebuilding vimanadigital.com

The short answer

Website migration SEO is the work of keeping rankings when a site is rebuilt or moved. Rebuilding vimanadigital.com, we inventoried 989 URLs, kept 935 at the same address and redirected 54 one-to-one, 47 of them templated town pages. It's the method we sell on our website migration SEO page, shown from the inside.

Why rebuild a website that already ranks?

Because parts of it were ranking and parts of it were quietly dragging the rest down. Ubersuggest shows the number of keywords this site ranks for falling from around 300 in November 2025 to around 80 in September 2026. Over the same period our best pages held: a 16-month Search Console export shows the Derby SEO cluster alone collecting roughly 31,000 impressions, with "SEO agency Derby" averaging position 4.1, and "web design Long Eaton" at position 2.0.

So the brief was awkward. Move to a new platform, redesign everything, and do it without touching the pages that pay the bills. That is the definition of a migration with something to lose, and it's why we treated our own site exactly as we'd treat a client's: inventory first, design second.

If you're planning a redesign yourself, the checklist we're following is on the website migration SEO service page. This article is what it looks like when you run it on your own site, with the numbers left in.

What does a URL inventory look like for a small business website?

Bigger than you'd think. People guess their site has "about 30 pages". Ours has 21 standard pages, and 989 URLs. Here's the breakdown from our inventory file, which has four columns: path, type, action and redirect target.

Three things the inventory taught us before a single page was designed:

• 81% of the URLs are media. Images have their own addresses, Google Images indexes them, and other sites link to them. Renaming every image path throws that away for nothing. Ours keep their paths.
• A site accumulates junk nobody meant to publish. Five standalone form URLs were indexable. An author archive existed for a single author. A page lived at both /IOS-app-development/ and /ios-app-development/. None of these would show up on a "list of pages", and all of them need a decision.
• Links point at pages that don't exist. One town page linked to /seo-agency/, which has never been a live URL. You only find that by crawling, not by clicking around.

What is the one-to-one URL rule, and why did we break it 54 times?

The rule we set at the start: every URL that earns anything keeps its address, character for character, including the trailing slash. Backlinks, Google's index history, citations on directories and the links in our own Google Business Profile all point at the old addresses. Changing them for tidiness is paying to lose something.

Google's own site-move documentation says a medium-sized site can take a few weeks for most pages to move across in its index, recommends keeping redirects in place for at least a year, and treats a blanket redirect of old pages to the homepage as a soft 404 rather than a transfer of signals (Google Search Central, "Site moves with URL changes"). So "we'll just redirect everything to the homepage" was never an option.

The seven exceptions in the original inventory are all duplicates or non-pages: the five form URLs go to the contact page, the author archive goes to the About page, and the uppercase duplicate goes to its lowercase twin. Each redirect has its reason written in the redirect file, so whoever maintains the site in three years knows why.

Then came the bigger decision.

Why are we consolidating 47 location pages?

Over the years the site grew 38 pages of the form /seo-agency-{town}/, eight of the form /web-design-{town}/ and one for Dubai. They were built to rank for town-name searches, and for a while some of them did. Reading them side by side in October 2026 was uncomfortable.

• Every SEO town page is 1,540 to 1,577 words long (Blackpool, 1,697). Every web design town page is 1,872 to 1,914. The only town-specific content is the H1, the image alt text and one sentence: "covering {town}, Nottingham and all surrounding areas."
• Several carry the wrong image. The Worthing page shows the West Bridgford hero photo with West Bridgford alt text.
• Google's spam policies, updated 28 August 2026, describe doorway abuse as "having multiple domain names or pages targeted at specific regions or cities that funnel users to one page" and "creating substantially similar pages that are closer to search results than a clearly defined, browseable hierarchy." That is a description of these pages.
• The data agrees. Across the whole small-town cluster, the 16-month Search Console export shows roughly 6,500 impressions and two clicks. Around 20 of the pages don't appear in the top 500 queries at all.

The honest complication is that a few of them do get seen. Worthing alone collected about 5,850 impressions across 14 queries, ranking between 8 and 31, for a town 200 miles from our office that we have never served in person. Beeston and Stapleford, which are genuinely a few miles from Long Eaton, rank first to third for their searches. Keeping the URLs live but near-identical would have meant launching a new site with 47 doorway pages on it. Deleting them outright would have wasted whatever the Beeston and Stapleford pages had earned.

The decision, approved on 9 October 2026, was consolidation with one-to-one 301 redirects, each page sent to the existing page that already serves that search properly:

That leaves 12 location pages, three per city for Nottingham, Derby, Leicester and Long Eaton, each being rewritten with content that is actually about that city. Whitespark's 2026 Local Search Ranking Factors survey puts "a dedicated page for each service" as the single highest-scoring factor in localised organic results (47 practitioners, November 2025). A dedicated page per town with the same words on it is not what they mean.

What we're giving up: the Worthing impressions, knowingly. A page ranking 8th to 31st for a town we don't serve produced no business in 16 months, and it was part of the reason the site read as templated. We'd rather lose it than carry it over.

What rules are we following for the redirect map?

The redirect file is 50 rows: 47 town pages plus three rows for the seven URLs above (one wildcard row covers all five form URLs). Every row follows the same rules, which are the ones we'd apply to a client site:

• Nearest equivalent, never the homepage. A Derbyshire SEO page goes to the Derby SEO page, not to /seo/ and not to /.
• Same service, same type. Web design town pages go to web design city pages; SEO town pages go to SEO city pages. Mixing them would send a "web design Buxton" visitor to a page about SEO.
• Out of area goes national. A town we can't serve locally is sent to the national service page, which says plainly where we work.
• No chains. Every redirect target is a live page on the new site, never another redirect. A redirect that lands on another redirect wastes crawl budget and leaks signals.
• One reason per row. The third column says why, with the approval date. Redirects outlive the people who set them.
• Tested before launch. Every row is run against the staging build. "Spot-checking a few" is how a missed product page costs a shop its Christmas.
• Kept for at least a year, and in practice indefinitely for anything with a backlink.

What has to be identical on launch day?

The redirect map protects the URLs that change. The bigger risk on this rebuild is the 935 URLs that don't change but could still lose content on the way across. For every page we're checking:

• Title tag, meta description and H1, kept or deliberately improved, never reset by a template
• Body content: the Derby SEO page in particular isn't being "tidied up". Google ranked the words, not the layout
• Image file names and alt text on the 806 media URLs
• Internal links, updated so nothing on the new site points through a redirect
• Structured data rebuilt to match what's on the page
• robots.txt and the XML sitemap regenerated for the new platform, with the old sitemap kept ready to resubmit so Google crawls the redirects quickly
• Search Console baseline exported before launch: the 500-query report, impressions and position per page, and the "Web" search type, which is where Google counts AI Overview clicks (Google Search Central, December 2025). Bradgate Financial Solutions, a client site, is cited inside an AI Overview for a Leicester mortgage search, and we don't want to discover after launch that a content change cost a citation like that on our own site.

Two of the rebuilds in our portfolio are reminders of why this list exists. LJM Solutions came to us after their previous site was compromised and had to be rebuilt from the ground up while keeping the business findable. And the review below is from a client whose site was rebuilt on the same principle.

What will we publish after launch?

This is the part most migration articles skip, because the agency has moved on by the time the data exists. We're committing to the comparison in advance. At 2, 4, 8 and 12 weeks after launch we'll compare against the baseline:

• Positions and impressions for the Derby SEO cluster and "web design Long Eaton", the two things we most needed to protect
• Total ranking keywords (Ubersuggest), against the ~80 starting point
• Impressions and clicks on the nine redirect targets, to see whether the consolidated pages inherited anything
• 404 errors and redirect chains reported in Search Console
• Organic enquiries, which is the only number that matters

Results: To confirm: post-launch Search Console results. If the Derby cluster dips, we'll say so and say why.

Frequently asked questions

Should I redirect old pages to my homepage when I redesign?

No. Google treats a blanket redirect to the homepage as a soft 404, so the old page's ranking signals don't transfer. Each old URL needs its own 301 to the page that now serves the same purpose. On our rebuild, 47 town pages map to nine specific targets and nothing points at the homepage.

How many URLs does a small business website actually have?

Usually far more than the page count suggests. Our site has 21 standard pages and 989 URLs once images, case studies, category archives, tag pages, forms and feeds are counted. Crawl the site and export Search Console before any redesign, because the pages nobody remembers are the ones that get lost.

Is it safe to remove location pages that still get impressions?

It depends on what the impressions produce. Our 47 town pages collected roughly 6,500 impressions and two clicks in 16 months, and matched Google's doorway-page definition. We consolidated them with one-to-one 301s to the city page that serves the search properly, rather than deleting them, so any value they held moves with them.

How long should 301 redirects stay in place after a migration?

At least a year, which is Google's own recommendation, and indefinitely for any URL that still has links pointing at it. Redirects cost nothing to keep. Ours are recorded in a file with a reason and approval date against every row, so they survive future changes of platform or developer.

Will moving to a new platform hurt my rankings?

Not by itself. Rankings drop when URLs change without redirects, content is trimmed, or titles and headings are reset by the new theme. If every page keeps its address, content and metadata, and the few that change are redirected one-to-one, a platform change is a low-risk migration. The checklist is on our website migration SEO page.

Planning a redesign of your own?

Send us your current website address and the reason for the change. We'll tell you what's at risk, what the URL map would involve, and whether the migration can be planned into the build rather than bolted on afterwards. Call 0115 647 0444, email contact@vimanadigital.com, or request a free pre-migration audit. If we're also building the new site, the migration work is planned into the web design or web development quote.

5.0 from 50 Google reviews.