Someone has told you the site needs to move. A new platform, a new host, maybe a new domain. And somewhere in that conversation, somebody mentioned the traffic drop. That is usually where the decision stalls, because nobody can tell you how large the drop will be or how long it lasts.
Here is the most useful thing Google publishes about site moves, and almost nobody selling migrations in this market quotes it. In its own guide, Google says: "if you're also changing your site's design, information architecture, or content, we recommend you make those changes after completing the move, not at the same time." [1]
That single sentence is worth more than every migration checklist you will read this week. Move the site. Then redesign it. Changing platform and design together is exactly what turns a normal, recoverable wobble into a problem nobody can diagnose, because when the traffic falls you have no way of knowing which change caused it.
A migration starts from around AED 5,000 and rises with content volume and how much SEO preservation work is involved, with larger replatforms scoped in discovery. This guide covers the move types ranked by risk, redirects done properly, the pre-launch checklist, and what to watch afterwards.
Change one thing at a time
The reason to isolate changes is not superstition. It is diagnosis.
When only the URLs change, the list of things that can go wrong is short and every item is checkable. Redirects are missing, or chained, or pointing at the wrong page, or the sitemap still lists old addresses. You look, you find it, you fix it. When URLs, templates, navigation, titles and half the copy change in one launch, that list gets long enough that testing each possibility takes weeks, and by then the drop has compounded.
Google is direct about what to expect during the move itself. It says to "expect temporary fluctuation in site ranking during the move," and that "with any significant change to a site, you may experience ranking fluctuations while Google recrawls and reindexes your site" [1]. That fluctuation is the normal cost of a move. Your job is to make sure it stays fluctuation, and to be able to prove it when someone asks.
So sequence the work. Move the site with the content and structure it already has. Let it settle. Then run the redesign as its own project, with its own baseline. Our website redesign guide covers that second stage, and it makes the same argument from the other direction.
If the real question underneath all this is which platform to move to, decide that before anything else. Our comparison of WordPress, Webflow and Next.js and the Shopify versus WooCommerce versus custom comparison cover that decision properly.
The types of move, ranked by risk
Google splits site moves into two documented categories, and the split tells you almost everything about risk. A move without URL changes is a hosting, server or CDN transfer where every address stays identical [2]. A move with URL changes covers HTTP to HTTPS, domain changes including consolidations, and changes to URL paths [1].
The ranking below is ours, built on that split. Google does not publish a risk ladder.
| Move type | What changes | Risk |
|---|---|---|
| Hosting or CDN transfer | Servers and DNS, no URLs | Lowest |
| HTTP to HTTPS, same paths | Protocol only | Low |
| URL structure change, same domain | Paths and internal links | Medium |
| Domain change | Every URL, plus domain authority signals | High |
| Any of the above plus a redesign | Everything at once | Highest, and hardest to diagnose |
The bottom row is the one to argue about internally. It is not technically harder than a domain move. It is harder to recover from, because the evidence you need in order to fix it has been destroyed by the launch itself.
One nuance worth knowing: a platform change does not have to be a URL change. Move from WordPress to a new stack and keep every path identical, and you are doing the lowest-risk kind of move with a lot of build work behind it. That is often the right way to plan a replatform.
Redirects, done properly
Redirects are where migrations are actually won or lost, and there are only four rules that matter.
Use permanent redirects. Google recommends "HTTP permanent redirects if possible, such as 301 and 308" for a site move [1]. The distinction matters more than it looks. With a 301, Google states that "the indexing pipeline uses the redirect as a signal that the redirect target should be canonical." With a temporary 302, "Googlebot follows the redirect, but the indexing pipeline doesn't use the redirect as a signal that the redirect target should be canonical" [3]. A 302 left in place after a permanent move can leave Google indexing an address you have abandoned.
Map every URL individually. Google's five-phase process names "create a URL mapping" as its own step, before execution [1]. That is a spreadsheet job: every old URL, its new equivalent, and a decision for the ones with no equivalent. It is tedious and it is the single highest-value hour of the project. Redirecting an entire old site to the new homepage is the shortcut people take instead, and it throws away the specific page-level signal the redirect existed to carry.
Keep chains short. Googlebot can follow "up to 10 hops in a chain of multiple redirects," but Google advises "redirecting to the final destination directly" and keeping any unavoidable chain to "ideally no more than 3 and fewer than 5" [1]. Chains build up by accident when a CMS plugin rule, a server rule and a CDN rule each fire in turn. After launch, test a sample of old URLs and look at the full hop sequence, not only the final status code.
Be honest about deleted pages. Google's site move guidance says that if you are not moving all your old content, those URLs should "correctly return an HTTP 404 or 410 error response code on the new site" [1]. Its redirect documentation adds that where a genuinely relevant new page exists, you can "add a link pointing to the new page accompanied by a short explanation" [3]. Real equivalent, redirect to it. No equivalent, let it return an honest error. Never invent a match to keep the redirect count tidy.
Then leave the redirects alone. Google's recommendation is to "keep the redirects for as long as possible, generally at least 1 year," because "this timeframe allows Google to transfer all signals to the new URLs" [1]. One year is a floor, not a finish line.
The pre-launch checklist
Most of this list comes straight from Google's two site move documents. The rest is the baseline work that makes the aftermath measurable.
- Crawl and export the current site. Every URL, its status, its title and description, its structured data. This is the reference you will compare against for months.
- Record the baseline. Search Console indexed page count, three months of the Performance report, sitemap status and Core Web Vitals [12]. Analytics organic sessions, conversions and top landing pages. Without this, no post-launch number can be judged.
- Build the URL map. Old to new, every row decided, before development finishes.
- Test the new environment before cutover. Google recommends verifying that Googlebot can reach the new infrastructure with the URL Inspection Tool in Search Console, and checking that firewall or denial-of-service protection is not blocking it [2].
- Keep the test environment out of the index properly. Google suggests a temporary test hostname carrying a noindex directive [2]. Note the mechanism: "for the noindex rule to be effective, the page or resource must not be blocked by a robots.txt file," because a blocked crawler "will never see the noindex rule" [6]. And robots.txt is not an indexing control at all. Google warns not to use it "as a means to hide your web pages from Google Search results," since a disallowed page "can still be indexed if linked to from other sites" [5].
- Check the production robots.txt within minutes of launch. A blanket disallow copied from staging is the classic launch-night disaster, and it is silent.
- Point canonicals at the new URLs. Google chooses a canonical from several signals and treats your rel=canonical as "a hint, not a rule" [7]. Self-referential canonicals on the new site, and no leftovers aimed at the old domain.
- Rebuild and revalidate structured data. Schema on most sites comes from a theme or plugin that will not exist on the new platform. Google warns that "pages might break after deployment due to templating or serving issues" and recommends the Rich Results Test before launch and the rich result status reports after [11].
- Regenerate and resubmit the sitemap. New canonical URLs only, absolute addresses, submitted in Search Console at cutover. Google caps each file at 50,000 URLs or 50MB, so larger sites need a sitemap index [10].
- Lower the DNS TTL a week ahead if DNS is changing. Google's advice is to lower it "to a conservative low value (for example, a few hours) at least a week in advance" [2].
- Submit the Change of Address, if it applies. Only for domain and subdomain moves, only after the redirects are live.
Bilingual sites: the reciprocity trap
If your site runs in English and Arabic, one item on that list carries more risk than the rest.
hreflang is bidirectional. Google states plainly that "if two pages don't both point to each other, the tags will be ignored" [8]. Not degraded. Ignored. A URL change moves the English address and the Arabic address of every pair at the same moment, which means every single pair is exposed at once, and a partial update breaks the annotation rather than half-breaking it.
The failure is quiet. Google may start showing the English page to Arabic searchers, or treat the two versions as duplicates and pick one. Nothing errors. Test the return tags on both sides of a sample of pairs before cutover and again after, wherever you declare them, whether that is HTML tags, HTTP headers or the sitemap [8].
Our guides to Arabic-first ranking in Google UAE and RTL website design cover the tag set UAE sites usually need.
Domain moves and the 180-day window
If you are changing domain, one extra tool applies. The Change of Address tool in Search Console tells Google to "emphasize crawling and indexing your new site over crawling your old site" and forwards signals from the old domain [4]. You need both properties verified in the same account as domain-level properties, and the 301 from old homepage to new homepage has to be live first.
Google publishes a specific number here: the preferential treatment runs for 180 days. After that, Google "does not recognize any relationship between the old and new sites, and treats the old site as an unrelated site" [4]. So the tool is a boost with an expiry date, and the redirects, which should live at least a year, are what carry the rest. Google also suggests holding the old domain for around a year so it cannot be picked up and reused by someone else [4].
The tool does not apply to HTTP to HTTPS moves, to path changes inside one domain, or to www and non-www switches [4]. Those are redirect and canonical work. Moving to a .ae domain is a domain change like any other, with no special process; .ae is administered by the .aeDA under the TDRA, and first-level registration is open to everybody [13].
What to watch afterwards, and the numbers nobody should quote
Google's five-phase process ends on monitoring, and specifically on monitoring both the old and the new URLs [1]. That second half is what people forget. The old URLs are where you find out whether the redirects are actually firing.
Watch three things in Search Console. First, the Page Indexing report, for a spike in pages blocked by robots.txt or excluded by noindex, which points at a staging configuration that shipped. Second, indexed page counts against your pre-launch baseline. Third, individual old URLs in the URL Inspection Tool, to confirm they return a redirect rather than a 404.
Now the part that most articles get wrong. There is no credible traffic-loss percentage. No Google document states one. Every figure circulating on this topic comes from agency marketing, and we will not add another. There is also no honest recovery timeline. Google's own language is "a small to medium-sized website can take a few weeks for most pages to move in our index; larger sites can take longer," and it states flatly that "there are no fixed crawl frequencies; how fast Googlebot crawls depends on the size of your site" [1]. Anyone quoting you a precise recovery date is quoting themselves.
Google publishes exactly two hard durations for site moves: keep redirects for at least one year, and the Change of Address window is 180 days [1][4]. Both describe how long you maintain your side of the move. Neither describes how long Google takes on its. Do not let anyone blur the two.
Hosting location is worth one clarification, because it comes up in every UAE migration conversation. Server location is one of several signals Google uses to work out who a site is aimed at, alongside country domains, hreflang and local content, and Google explicitly ignores geographic meta tags [9]. It is not a ranking factor in its own right. Speed is what matters, which is the same position as our UAE hosting guide [14]. Move hosting for performance and reliability, not for a ranking effect that is not there.
Real client stories
These are anonymised situations from migration work.
The launch that changed everything at once. A UAE business moved from WordPress to a new platform and launched a full redesign on the same night. New URLs, new navigation, new copy, new templates. Organic enquiries fell over the following weeks and nobody could say why. The redirects existed, so the obvious answer was ruled out early, and after that the investigation had no shape: the drop could have been the URL change, the removed pages, the rewritten titles, or the new navigation burying two service pages three clicks deep. It was, in the end, several of those. Splitting the launch would have cost one extra deployment and saved weeks of guesswork.
The staging file that shipped. A site went live on new infrastructure with the staging robots.txt in the deployment. A single disallow line stopped Google crawling the entire live site. Nothing looked broken to anyone using the site, which is why it survived launch checks. It was caught when the Page Indexing report began flagging pages blocked by robots.txt. The fix took minutes. Finding it took far longer, because nobody had opened the live robots.txt after cutover.
The Arabic half that stopped pairing. A bilingual site changed its URL structure and updated the English hreflang references correctly. The Arabic templates kept pointing at the old English paths, so the return tags no longer matched and the annotations were disregarded. Arabic search results started showing English pages. There was no error message anywhere, and the site had passed a visual check on both languages. Testing hreflang pairs in both directions is now the first thing checked on any bilingual move.
How SKIMBOX approaches migrations
We start by asking what is actually changing, and then argue for changing less of it at once. If a replatform can keep its URLs, we keep them, because the safest redirect is the one you never have to write. We build the URL map before development finishes rather than on launch night, capture a full baseline so the result can be measured against something, and treat the redirect audit and the hreflang check as launch gates rather than follow-up tasks. Where a redesign is also wanted, we offer to sequence it after the move, and we will explain why in the proposal rather than after the traffic drops.
A migration starts from around AED 5,000 and rises with content volume and how much SEO preservation work is involved, with larger replatforms scoped in discovery. Final pricing depends on scope. For what the site costs to build and run either side of a move, see our website development cost guide and the real cost of running a website in the UAE.
See our web development services and digital marketing services, or contact us for a plain assessment of what your move actually involves.
For related reading, see our guides on website redesign in Dubai, WordPress versus Webflow versus Next.js, best web hosting in the UAE, and how to rank on Google in the UAE.
References
[1] Google Search Central - Site moves with URL changes. developers.google.com/search/docs/crawling-indexing/site-move-with-url-changes
[2] Google Search Central - Site moves without URL changes. developers.google.com/search/docs/crawling-indexing/site-move-no-url-changes
[3] Google Search Central - Redirects and Google Search. developers.google.com/search/docs/crawling-indexing/301-redirects
[4] Google Search Console Help - Change of Address tool. support.google.com/webmasters/answer/9370220
[5] Google Search Central - Introduction to robots.txt. developers.google.com/search/docs/crawling-indexing/robots/intro
[6] Google Search Central - Block Search indexing with noindex. developers.google.com/search/docs/crawling-indexing/block-indexing
[7] Google Search Central - Consolidate duplicate URLs with canonicals. developers.google.com/search/docs/crawling-indexing/canonicalization
[8] Google Search Central - Localized versions of your pages and hreflang. developers.google.com/search/docs/specialty/international/localized-versions
[9] Google Search Central - Managing multi-regional and multilingual sites. developers.google.com/search/docs/specialty/international/managing-multi-regional-sites
[10] Google Search Central - Build and submit a sitemap. developers.google.com/search/docs/crawling-indexing/sitemaps/build-sitemap
[11] Google Search Central - Introduction to structured data markup. developers.google.com/search/docs/appearance/structured-data/intro-structured-data
[12] web.dev - Web Vitals and Core Web Vitals thresholds (Google). web.dev/articles/vitals
[13] TDRA - About the .ae national domain and the .aeDA. tdra.gov.ae/en/aeda/about-us/about-ae
[14] SKIMBOX - Best web hosting in the UAE, on hosting location and rankings. skimbox.co/resources/blogs/best-web-hosting-uae



