The Pre-Migration SEO Checklist: 11 Things to Export Before You Touch Anything

September 23, 2026

Every site migration that goes badly goes badly for the same reason: somebody started making changes before they had a complete record of what existed. Once the old site is gone, you cannot reconstruct what you had, which means you cannot tell what you lost or how to get it back.

This is the export list. Do all of it before anyone merges anything.

1. Every indexed URL

Search Console's Pages report, filtered to indexed, exported in full. This is Google's view of your site, which is the view that matters.

2. A full crawl of the live site

Screaming Frog or equivalent. URLs, status codes, titles, meta descriptions, H1s, canonicals, meta robots, word count. This is your site's view of itself.

3. The XML sitemap, as a file

Not a link to it. The actual file, saved.

4. Top pages by organic traffic, twelve months

From analytics, not from Search Console alone. You need the pages that produce revenue, not just the ones that get impressions.

5. Top queries by page, sixteen months

Sixteen, not twelve, so you have a year-over-year comparison available after launch. Search Console caps at sixteen months and it disappears as it ages.

6. Current rankings for your target keyword set

A dated snapshot from whatever rank tracker you use. Without a pre-migration baseline, every post-launch conversation becomes an argument about whether a drop is real.

7. Backlink data

Every referring domain and the specific page it points to. When you change URLs, you need to know which ones have external links worth preserving with a precise redirect rather than a lazy one to the homepage.

8. The existing redirect rules

Sites accumulate redirects over years. If you replace the server config wholesale, you break chains that were carrying value from URLs retired in 2019.

9. Structured data currently in use

Export the JSON-LD from your template types. Redesigns routinely drop schema because it lives in a template nobody thought to port.

10. robots.txt and any noindex directives

Saved as files, with a note about why each rule exists. Half of them will turn out to be load-bearing.

11. Core Web Vitals field data

The current values, per page group, from Search Console. Field data has a 28-day rolling window, so the pre-launch numbers vanish if you don't capture them.

Then build the map

With all eleven exports in hand, build the redirect map: old URL, new URL, one row per URL, no blanks. Every old URL gets a destination that is genuinely equivalent in content. Not the homepage. Not a category page. The equivalent page, or the closest thing that exists.

Test the map against staging before launch. Test it again the hour after launch. Then watch Search Console's coverage report daily for two weeks.

The one that catches everyone

Staging environments get indexed. Before launch, confirm the staging site is behind authentication, not merely noindexed – and after launch, confirm the production site is not still carrying the staging robots rules. That single inverted check is responsible for more migration disasters than every other item on this list combined.

Migration work is boring, and the payoff is entirely invisible when it goes right. That's the job.

Reference: https://deepbhardwaj.com/seo-expert-in-mumbai/