Site Migration Playbook: Don't Lose Rankings in a Redesign

Site migration is the single fastest way for a UAE brand to lose 30 to 60% of organic traffic overnight. It is also entirely preventable. Every major migration failure we have audited (and we have audited plenty from the cleanup side) traces back to the same handful of missed steps.

This is the exact playbook we run for every replatform, redesign, or domain change we oversee at Digi Soft Rank, plus the reference document our clients use when their internal or external dev teams execute migrations. Nazir has led migrations across UAE brands including enterprise Shopify-to-custom, WooCommerce-to-Shopify, WordPress replatforms, domain changes, and multi-site consolidations. Every stage below has been ordered from experience about which step, if skipped, produces the worst downstream failure.

Read it once. Apply it in order. Skip nothing.

What is a site migration in SEO?

A site migration is any change that alters URLs, technology stack, domain, or site structure enough that Google needs to re-crawl and re-understand the site. Common triggers: rebrand with domain change, replatform from Shopify to custom (or WordPress to Shopify), redesign that restructures URLs, HTTP to HTTPS conversion, subdomain to subdirectory consolidation, merger of multiple properties, and CMS upgrade with URL implications. Every migration carries SEO risk; the risk scales with the number of URLs changing and the authority they carry. This playbook is the framework we use to keep that risk under control.

Stage 1: Pre-migration crawl

Everything else depends on this step. Skip it and you cannot recover what you did not measure.

• Full crawl of the current live site. Screaming Frog or Sitebulb, JavaScript rendering enabled.
• URL inventory export: every URL with status code, meta title, meta description, H1, canonical, hreflang, structured data types, redirect chain if any.
• Organic traffic per URL from Google Search Console Coverage and Performance reports. Last 12 months minimum.
• Backlink profile exported from Ahrefs, Semrush, or Majestic. Every referring domain to every URL.
• Ranking snapshot from AccuRanker or SEO tool of choice. Top 500 keywords minimum with current position.
• Baseline metrics documented in one spreadsheet: total organic sessions, sessions per top-20 URL, revenue attribution, average position, indexed URLs count.

This baseline is what you compare against 30, 60, and 90 days post-launch. Without it, "recovery" is guesswork.

Stage 2: Prioritise pages by revenue

Not every URL deserves the same migration attention. Top 20% of URLs generate roughly 80% of organic value.

• Tag each URL with its 12-month organic traffic, 12-month organic revenue (if attributable), backlink count, and business criticality.
• Priority 1: URLs in top 20% of organic revenue. Every P1 URL gets white-glove treatment.
• Priority 2: URLs with meaningful backlinks or category pages driving downstream traffic.
• Priority 3: Remaining indexed URLs.
• Priority 4: Low-value URLs where you can accept some risk (thin tags, old blog archives, redirect-only pages).

Prioritisation lets you focus staging validation, redirect verification, and post-launch monitoring on the pages that carry the business. Everything else gets standard treatment.

Stage 3: Map every URL

Before any URL changes, every old URL needs a mapped destination on the new site.

• Spreadsheet with two columns: old URL, new URL.
• No unmapped URL ships. Every P1 to P3 URL has a specific new destination.
• No mass redirects to homepage. Homepage 301 for missing URLs is the surest way to lose rankings. Google reads mass-redirects-to-homepage as "the old page no longer exists" and drops rankings.
• 404 for genuinely retired URLs only when there is no meaningful equivalent on new site. Even then, prefer a soft-related 301 over a hard 404.
• Category page consolidations documented explicitly.
• Wildcard rules for URL structure changes (e.g., /products/{slug} to /shop/products/{slug}).

The URL map is the single most important artefact in the migration. Every downstream stage references it. For enterprise sites, the same discipline is documented in our enterprise technical SEO template under Pillar 9 (migration risk).

Stage 4: Preserve on-page elements

The temptation during migration is to "improve while migrating". Resist. Migrate first, improve second.

• Titles, meta descriptions, H1s preserved on P1 and P2 URLs. Improvements come in phase 2 after ranking stabilises.
• Body copy preserved on high-authority URLs. Rewriting during migration invites both algorithm re-evaluation and user experience mismatches.
• Images with alt text preserved. Alt text is often overlooked in migration and is easy to lose.
• Internal linking preserved in structure. If old site had a "related posts" block linking A to B, new site should maintain that linking.
• Schema markup preserved. Every Article, Product, Organization, FAQPage, BreadcrumbList block on old site must render on new site.

The rule that saves most damage: never migrate URL structure and content simultaneously. Do platform migration first, preserving everything else. Do content improvements in phase 2 after 90-day recovery has stabilised. The complete SEO Checklist we published covers what "preserved on-page" actually means at the checklist level.

Stage 5: Build the redirect map

Every old URL redirects to its new equivalent via 301 (permanent) or 308 (permanent, method-preserving).

• 301 vs 308: functionally similar for SEO. 308 preserves HTTP method (POST stays POST); 301 does not require that. For most SEO migrations, 301 is fine.
• One-hop redirects only. Every redirect goes from source to final destination in one step. Chains (A → B → C) dilute PageRank and waste crawl budget.
• HTTPS enforcement in redirect logic. Every HTTP URL redirects to HTTPS equivalent, then to final destination.
• www vs non-www consistency. Choose one; enforce via redirect.
• Trailing slash consistency. Choose one; enforce via redirect.
• Multi-market redirects must preserve hreflang cluster integrity. See our international SEO deep-dive for how hreflang breaks silently during multi-market migrations.

Test every P1 redirect on staging before production launch. A P1 URL that returns 404 or redirects to homepage costs revenue from day one.

Stage 6: Pre-launch staging test

Never launch a migration blind. Everything gets tested on staging first.

• Full crawl of staging site with Screaming Frog. Compare URL inventory to expected.
• Redirect testing. Sample of P1 URLs verified redirecting correctly.
• Schema validation. Rich Results Test on P1 URL templates.
• Hreflang cluster validation. All variants correct on multi-market sites.
• Core Web Vitals baseline on staging. Should be at least as good as live.
• Robots.txt check. Should not block money pages.
• XML sitemap check. Should reference all indexable URLs, no noindex pages included.
• Structured data emitted server-side, not JavaScript-injected.
• 404 handling. Custom 404 page in place with useful navigation.

Password-protect the staging site to prevent Google from indexing it (or add noindex to all staging URLs). Indexed staging environments are a separate SEO disaster we regularly clean up. Our technical SEO team runs full staging validation for enterprise migrations as part of the pre-launch package.

Stage 7: Launch on the right day and time

Launch timing affects recovery speed.

• Launch Tuesday or Wednesday morning UAE time. Gives 2 to 3 business days before the weekend for issue detection.
• Not Friday. UAE weekend starts; issues discovered late compound.
• Not before a public holiday. Same reason.
• Coordinate with Google: submit the new sitemap immediately after launch.
• Notify major stakeholders in advance: paid media (retargeting audiences may need updating), CRM (URL parameters may change), analytics team (tracking may need adjustments).

Our web development team runs the pre-launch coordination call with client dev, product, and marketing 48 hours before launch for enterprise engagements. Ninety minutes on Monday saves the discovery of preventable issues on Friday night.

Stage 8: Immediate post-launch actions

The first 48 hours after launch are critical. Every issue detected and fixed in the first 48 hours costs a fraction of what fixing it in week 2 does.

• Full crawl of live site within 2 hours of launch. Compare to staging crawl.
• Verify redirects firing on random sample of 50 P1 URLs.
• Verify schema rendering on P1 URL templates via Rich Results Test.
• Check GSC coverage report for immediate issue signals.
• Monitor server response codes in log files for 4xx or 5xx spikes.
• Check Core Web Vitals via Chrome UX Report and PageSpeed Insights. Full framework in the Core Web Vitals fixes that moved the needle.
• Verify tracking (GA4, Meta Pixel, Google Ads) fires on new URLs.
• Test conversion flows end-to-end: contact form, checkout, WhatsApp, phone click.

Have the SEO lead available in the first 48 hours to triage issues as they surface. Most catastrophic migration losses are preventable in the first 24 hours if someone is watching.

Stage 9: 90-day recovery monitoring

Ranking recovery is not linear. Expect a dip in weeks 1 to 4 as Google re-crawls, re-indexes, and re-evaluates.

• Weekly organic sessions vs baseline tracked for 90 days.
• Weekly ranking position on top 100 keywords.
• Weekly GSC coverage report monitoring for coverage decay.
• Weekly log-file analysis for enterprise sites: is Googlebot crawling the new URLs? See enterprise SEO measurement discipline for how portfolio-level attribution replaces keyword tracking at scale.
• 30-day, 60-day, 90-day checkpoints with formal review.
• Investigate any 15%+ traffic drop immediately. Common causes: broken redirect chain, missing hreflang cluster, schema dropped on key template, robots.txt error.

Recovery timelines: best case 4 to 8 weeks to full recovery, common case 3 to 6 months, worst case (major mistakes) 12+ months or never full recovery. The 90-day window is where you catch what needs catching before it becomes permanent damage.

The 5 migration killers

Every catastrophic migration failure we have audited traces back to one or more of these five.

1. Broken redirect chains (old URL → intermediate → new URL). Each hop loses PageRank; multi-hop chains often lose the destination entirely to Googlebot.
2. Missing hreflang clusters after language variant restructure. Google serves wrong-market version to users; both markets drop rankings simultaneously.
3. Schema not preserved. Drops rich results overnight, loses AI Overview citations, kills any Product/Article/FAQ visibility built up over years.
4. Robots.txt accidentally disallowing new URLs. Nothing gets crawled. Common when staging robots.txt (which blocks everything) accidentally ships to production.
5. Canonicals pointing to old URLs on the new site. Google keeps trying to index the URLs that no longer exist, ignoring the new ones. Silent failure that takes weeks to detect.

For ecommerce brands considering platform migration specifically, killers 1, 3, and 5 account for most Shopify-to-custom traffic losses we have cleaned up. For the broader platform decision that leads to migration in the first place, see the WordPress vs Shopify vs custom decision matrix.

Common mistakes UAE brands make in migrations

1. Migrating without SEO oversight at all. "The dev team will handle it."
2. Migrating on a Friday or before a public holiday.
3. Redesigning content while migrating platforms simultaneously.
4. Mass-redirecting missing URLs to homepage.
5. Not building baseline metrics before migration.
6. Not documenting the URL map so post-launch verification is impossible.
7. Assuming Google will figure it out. It will not.
8. Panic redesigns in week 2 when traffic dips, adding change on top of change.
9. Ignoring our full SEO programme measurement discipline during recovery.

Tools stack for migration

• Screaming Frog SEO Spider: full-site crawls, redirect chain checks.
• Sitebulb: visualisation of internal linking, technical audits.
• Ahrefs: backlink profile export, ranking baseline.
• Semrush: ranking baseline alternative.
• AccuRanker: granular rank tracking during recovery.
• Google Search Console: coverage report, sitemap submission, ranking monitoring.
• Chrome UX Report: CWV baseline and post-launch.
• Google Sheets or Airtable: URL mapping and priority tracking.

For enterprise migrations with 100,000+ URLs, add log-file analysis and JetOctopus or DeepCrawl for scale. Our enterprise SEO team runs the enterprise-scale migration toolkit as part of embedded co-team engagements.

FAQ, answered

How long does migration recovery take?

Best case, 4 to 8 weeks to full traffic recovery. Common case, 3 to 6 months. Worst case (major mistakes), 12+ months or never full recovery. The variance is almost entirely explained by how many of the 5 migration killers went undetected pre-launch.

Can we migrate without an SEO team?

Only if you accept the risk. Migration without SEO oversight is like performing surgery without anaesthetics. Some things will work; the ones that go wrong will hurt and be expensive to fix.

Should we migrate URL structure and platform at once?

No. Do platform migration first, preserving URLs. Do URL restructure as a separate project 6+ months later. Combining doubles the risk surface and makes post-launch diagnosis nearly impossible because you cannot tell which change caused which problem.

What is the biggest cause of migration traffic loss?

Broken or missing redirects. Every URL that returns 404 or redirects to homepage instead of the correct new URL loses accumulated authority. This one failure mode accounts for the majority of catastrophic traffic losses we have cleaned up.

Should we deploy on a staging domain first?

Yes, always. Password-protect the staging site to prevent Google from indexing it (or add noindex to all staging URLs). Test everything on staging before production. Indexed staging environments are their own SEO cleanup category.

How much does a proper migration cost in Dubai?

Small site (under 1,000 URLs) SEO oversight: AED 8,000 to AED 20,000. Mid-market (1,000-10,000 URLs): AED 30,000 to AED 90,000. Enterprise (100,000+ URLs): AED 90,000 to AED 300,000. Enterprise with international preservation: AED 150,000 to AED 500,000+. The cost of doing it wrong is typically 3 to 10x the cost of doing it right. See our transparent pricing for tier benchmarks and our case studies for what recovery from failed migrations actually costs.

Should we redirect to homepage if we cannot find an equivalent page?

No. Prefer a category or related page. Homepage redirect for many URLs signals to Google that the URLs no longer exist and drops rankings.

Final recommendation

Never migrate without a mapped URL inventory, a 301 redirect map, and pre-launch staging validation. Every dollar spent on proper migration preparation saves 3 to 10 dollars in post-migration recovery cost.

If you have a migration coming up (rebrand, replatform, redesign, domain change) or you have just been through one and are watching traffic decline, book a discovery call with our web development team and we will audit either the plan or the aftermath on the first call, free. For redesign-triggered migrations specifically, our web design team handles redesigns with SEO oversight embedded rather than bolted on after launch.