Orphaned URLs nobody listed
A sitemap is not an inventory. Landing pages, PDFs, event pages and campaign URLs live outside it, and those are the ones that quietly return a 404 after cutover.
Migration runbook · one chance to get launch week right
SEO Migration Services
That Keep Traffic After Launch
A replatform, a domain move or a rebuild is the riskiest week in a site's life. Our SEO migration services are built for exactly that week: every URL accounted for, every 301 tested, every canonical pointing at the new host before DNS changes. Migration day is when traffic is easiest to lose, and also when it is easiest to save.
- Full URL inventory with priority scoring
- 301 redirect map, written and QA'd before launch
- Canonical, robots and sitemap re-submission
- Internal link and parameter cleanup
- Search Console verification and 30/60/90-day watch
No ranking promises. A crawlable, redirect-clean launch with a documented rollback plan.
Old site · today
/shop/industrial-pumps-2.php
/category.php?id=14&sort=price
/blog/2019/old-blog-post
/index.html?utm_source=old
New site · launch day
/products/industrial-pumps
/products/industrial-pumps?sort=price
/insights/choosing-a-pump
/
Where the traffic goes
LEAK 01
LEAK 02
LEAK 03
LEAK 04
LEAK 05
LEAK 06
Migrations rarely collapse on day one. They leak for three weeks.
The pattern is consistent: clicks hold for a few days, then a category page 404s, a filtered URL clones itself into forty variants, and a paid landing page starts competing with its own redirect. By week three the drop is being described internally as an algorithm penalty. It is almost never a penalty.
Redirect chains three hops deep
Old URL to staging URL to new URL. Every extra hop wastes crawl budget and dilutes the signal you were trying to preserve.
Canonicals pointing home
Default templates often canonicalise to the homepage. Thousands of product pages then ask Google to ignore them, and Google complies.
Parameter and faceted page sprawl
Sort, filter and tracking parameters multiply overnight. Without rules, the new site invites Google to crawl combinations instead of products.
Internal links still hard-coded
Navigation is updated, body copy is not. Hundreds of in-content links keep pointing at URLs that now redirect, so equity arrives one hop late.
Verification left on the old host
If Search Console verification, analytics and the sitemap reference die with the old domain, you lose the ability to see the problem while it is still small.
The ledger
Old URL, new URL, and the status each one has to return
Every URL gets one row and one decision. Not "redirect everything to the homepage" — that single shortcut is responsible for more post-migration loss than any other choice. The table below is the format we hand to developers, with the status we expect to see verified in staging before go-live.
| Old URL | New URL | Rule | Expected |
|---|---|---|---|
| /shop/industrial-pumps-2.php | /products/industrial-pumps | 1:1 301 | 301 → 200 |
| /category.php?id=14 | /products/category/pumps | Pattern rule | 301 → 200 |
| /category.php?id=14&sort=price | /products/category/pumps?sort=price | Canonical to clean URL | 200 + rel=canonical |
| /blog/2019/old-blog-post | /insights/choosing-a-pump | Consolidate | 301 → 200 |
| /landing/webinar-2021 | — | Retire | 410 |
| /index.html | / | Host-level 301 | 301 → 200 |
| /de/produkte | /de/produkte | hreflang pair kept | 200, reciprocal |
Rows are illustrative. Your ledger is generated from a full crawl plus server logs, then reconciled against Search Console's indexed URL list.
Deliverables
What the engagement actually puts in your hands
The work is documentation-heavy by design. If the mapping exists only inside someone's head, the launch will not survive a staffing change or a developer shortcut.
01 URL inventory and priority tiers
We combine a full crawl, XML sitemaps, server logs and Search Console exports into one list, then tier every URL by organic clicks, inbound links and commercial value. Tier-one pages get individual treatment; long-tail pages get pattern rules. The output is a spreadsheet your team can audit line by line.
02 The 301 map, plus QA against staging
Rules are written for the web server or edge layer, not improvised in a plugin. We test the top few hundred redirects in staging, look for chains and loops, and confirm the final status code is a single 301 to a 200 page. Anything that needs a 410 gets an explicit decision, not a silent 404.
03 Canonicals, robots and sitemaps
Self-referencing canonicals on every indexable template, a staging block that never ships to production, and a sitemap index that reflects the new URL structure. We submit the new sitemap the same day and keep the old one available long enough to catch stragglers.
04 Internal linking and parameter rules
In-content links, navigation, breadcrumbs and schema URLs all get replaced at the source. Faceted and tracking parameters receive a rule set — canonical, noindex or robots exclusion — so the new crawl budget goes to pages that can rank rather than to sort permutations.
05 Verification and the 30/60/90 watch
Search Console, analytics and any hreflang pairs are re-verified on the new host. We compare clicks and impressions against a frozen baseline at day one, day thirty, sixty and ninety, and keep a live issue log until every tier-one page is confirmed indexed with its intended canonical.
06 Handover documentation
You receive the ledger, the rule set, a rollback checklist and a short runbook for future changes. When the next redesign happens, the process starts from a working template instead of a blank page.
Sequence
Five moves, in this order, before DNS changes
Order matters more than effort. Mapping before crawling produces guesswork; testing after launch produces a live incident.
01
Freeze the baseline
Export clicks, impressions, top queries and indexed URL counts for the trailing twelve months. Without a frozen baseline, every later comparison becomes an argument about seasonality.
02
Crawl both sites and reconcile
Old site, new staging site, sitemaps and logs. Reconcile the lists until nothing is unexplained — the pages that appear in logs but in no sitemap are usually the valuable ones.
03
Write the map and the exceptions list
One-to-one redirects where an equivalent exists, consolidation where two pages should become one, and explicit retirement where content has neither demand nor links. Ambiguity becomes a 404 later.
04
Test in staging, at scale
Recrawl staging with redirects active, count chains over one hop, verify canonicals, and click through the twenty pages that produce the most revenue. Fix, retest, then freeze the rules.
05
Launch, submit, and watch daily
Deploy, submit the new sitemap, request indexing for tier-one URLs, and monitor coverage daily for the first fortnight. Most recoverable problems show up as a coverage or canonical anomaly before they show up as lost clicks.
Launch date already booked? Start the map now.
We can run the URL inventory and redirect ledger as a standalone project while your developers finish the build. Early mapping is the cheapest insurance available.
Observed outcomes
What a clean cutover looks like in the data
Figures below come from client reporting; names are withheld. They describe what happened, not what will happen for you.
“A B2B equipment supplier, 8 weeks after a domain move: organic sessions back to 94% of the pre-move baseline, and above it by month four.”Industrial equipment · domain change
“A specialty retailer, 6 weeks after replatforming: 1,180 redirect rules live, 41 chains eliminated, indexed URLs down 62% but organic revenue flat.”Retail · platform migration
“A SaaS company, 90 days after a subfolder restructure: support-doc organic clicks up 38% because retired pages were consolidated instead of 404'd.”SaaS · URL architecture change
Engagement shapes
Scope-based quotes, sized to the size of the move
Migration audit
A frozen baseline, a full URL inventory, a written redirect ledger and a launch-day checklist your developers can execute. Best when the build is already underway and you need the map quickly.
Full migration support
Everything in the audit, plus rule implementation guidance, staging QA, sitemap and Search Console handling, internal link replacement, and daily monitoring through the first fortnight after launch.
Post-migration recovery
For launches that already happened. We diagnose the coverage and canonical anomalies, rebuild the missing rules, and work the backlog until tier-one pages are indexed as intended. Often paired with Technical SEO Services to close the crawl-level gaps the move exposed.
Quotes depend on URL count, template variety, language versions and how much of the mapping your team can implement. We size the work after the inventory, never before it — a number quoted without seeing the site is a guess wearing a suit.
Questions buyers ask
Migration FAQ
Will we lose rankings when we migrate?
Expect some volatility for one to four weeks while Google recrawls. The depth of the dip tracks redirect coverage, canonical accuracy and how many URLs you retire. We plan for a controlled dip; we do not promise a flat line.
How long until the new site recovers?
Most sites return to baseline within four to eight weeks when the map is complete. Catalogues with heavy faceting, or multi-language moves with hreflang, take longer — which is why the watch runs to ninety days.
What do you need from our side?
URL and log exports, the ability to add redirect rules at server or edge level, a Search Console role, and one developer contact who can ship rule changes. If direct access is off the table, a shared ticket queue works.
Should we pause paid search during the move?
Usually no. Keep brand and high-intent campaigns running, but audit their final URLs and sitelinks so they point at the new structure. Pausing hides the very demand signals you want to compare against.
What about blog posts and resources we drop?
Each one gets redirected, consolidated or retired with a 410. Dropping them to 404 is the one option we advise against, because those URLs often carry the links that hold up the rest of the site.
How do you keep working after launch?
Daily coverage checks for two weeks, then structured reviews at day 30, 60 and 90. Where the migration exposed broader weaknesses — thin templates, slow pages, duplicated category copy — those go into a separate Website Managed Services roadmap rather than being smuggled into the launch scope.
Who runs the map
Practitioners who have sat through launch night
Our migration lead has run cutovers across Magento, Shopify Plus, WordPress and custom headless builds, usually alongside an in-house developer who owns the deploy. We write rules, not slide decks, and we stay on the call through the first night.
Every engagement includes a named practitioner, documented decisions and a direct WhatsApp line during the launch window. If something breaks at 02:00, you are not filing a ticket into a queue.
Curious how this fits a wider programme? Review SEO Services for the ongoing operating model, or send us the old sitemap and we will tell you what it would take.