Beta Position Eleven is live and still being built.
Home › Writing › SEO

The site migration list that actually prevents disasters

Site migrations lose traffic. Not always, and not irreversibly, but often enough that "we replatformed" is the most common answer to "when did it start declining".

The damage nearly always comes from the same short list.

1. Redirects

Every URL that had value must redirect, once, to the closest equivalent page.

Once. Chains of two, three or four redirects lose signal and waste crawl. Map old directly to final.

To the closest equivalent. Redirecting everything to the homepage is technically a redirect and functionally a deletion. Google treats a mass redirect to an irrelevant page as a soft 404.

301, not 302. A temporary redirect says "keep the old URL indexed", which is the opposite of the intent.

The list to check against is not your new sitemap. It is every URL that received an impression in the last sixteen months, which comes from Search Console and nowhere else.

2. The staging directives

noindex, Disallow: /, and canonicals pointing at the staging host. All three are correct on staging and catastrophic in production, and all three ship regularly.

Check the live site's HTML and HTTP headers immediately after launch. X-Robots-Tag in a header is invisible in page source.

3. Measurement continuity

The analytics tag, the container, and the conversion tags all have to survive the rebuild. They frequently do not, because they live in a template nobody thought of as content.

The symptom is a traffic cliff on launch day that is actually a measurement cliff. Which is arguably worse, because you cannot tell whether the migration hurt you.

4. Content parity

Pages that ranked, rewritten to be shorter and cleaner, stop ranking. The new design usually reduces text — that is often what "modernising" means in practice — and the words removed were doing work.

Compare the old and new versions of your top twenty pages. If word count halved, expect a decline and decide whether you accept it.

After launch

Submit the new sitemap. Keep the old one available for a while so Google can see the redirects. Then watch coverage and impressions daily for two weeks — not sessions, impressions, because it moves first and is not affected by whether the tag survived.

Read next

Reading about it is slower than looking

Run the free audit on your own site. Public data, about ten seconds, no account.

Audit my site

Stop reporting what happened.

Start telling clients what to change — and prove it moved.

Get started
Beta · help shape it

What would make this worth paying for?

Everything on this site works today. It is early, so the roadmap is still genuinely movable — tell us what is missing and we will build it. Testers who give us something useful get a lifetime deal on whatever plan they end up on.

No account needed. Email is optional.