Website Migration Risk Checklist

The state you need captured before a single DNS record changes.

• Export every indexed URL on the old siteSearch Console's Pages report plus a crawl of the current sitemap. This is the list every redirect has to account for.Google Search Console
• Screenshot current traffic, rankings and conversion numbersYou cannot tell whether a migration hurt you if you did not write down what normal looked like first.
• Take a full backup of the old site and databaseKept somewhere that survives the old host being cancelled, not on the old server itself.
• Confirm access to DNS, registrar and the old hosting accountA migration stalls fastest when nobody can log into the one account that controls the domain.

The single most common cause of a migration losing traffic.

• Every old URL maps to its closest equivalent, not the homepageRedirecting everything to the homepage is not a redirect strategy, it is a way to lose every page's accumulated authority at once.
• Redirects are 301, not 302302 tells search engines the move is temporary. It is not. Use 301 for anything permanent.
• No redirect chainsA to B to C costs speed and, past a couple of hops, search engines stop following it reliably.
• URL parameters and trailing slashes are handled consistentlyDecide once whether URLs end in a slash, and redirect every variant to the one you picked.

The window where things actually break, if they break.

• Lower DNS TTL 24-48 hours before cutoverA short TTL means a mistake gets fixed in minutes instead of the old value staying cached for a day.
• SSL certificate is live and valid on the new host before pointing DNS at itTest the new site on its raw IP or a staging URL first so the certificate is not the thing you discover broken live.
• Cut over during a low-traffic windowEven a clean migration benefits from fewer people hitting it during propagation.
• Monitor uptime and error rates through propagationDNS propagation is not instant everywhere. Watch it rather than assuming it finished.

The things that quietly stop working if nobody checks.

• Analytics and conversion tracking fire on the new siteConfirm in real time, not by waiting for tomorrow's report to come back empty.
• Tag manager container and all its tags moved overAd platform pixels are the ones nobody notices are missing until the campaign underperforms.
• Email deliverability (MX records) confirmed working after DNS changesA DNS migration that forgets MX records takes down company email, not just the website.
• Every form submits and lands where it used toTest end to end: submit, check the inbox, check the CRM if one is connected.

The first two weeks, where most missed redirects surface.

• Submit the new sitemap and remove the old oneTell search engines explicitly rather than waiting for them to notice on their own schedule.
• Watch the 404 report daily for the first two weeksEvery 404 with traffic is a redirect you missed. Fix them as they surface, do not batch it for later.
• Compare rankings and traffic against the baseline weeklyA dip in week one is normal. A dip still there in week four means something in the mapping was wrong.
• Do not cancel the old hosting until the new site has been stable for a monthKeeping it a few extra weeks costs little. Needing it back after cancelling costs a lot.