Site refresh + SEO migration

Keep all 3,087 referring domains. Ship the rebuild with no traffic dip.

I read your 19 mockups and crawled your live site before quoting. This is the control room for the migration itself: every redirect, every verification source, every rollback trigger, built against your real URLs. Not a generic checklist, and not a rebuild of your designs.

Stack WordPress + Elementor + Yoast 27.6 Cayenne 404 218 RDs recoverable Subdomains Docusaurus, preserved

No production change until staging proves out.

ask-your-sensors capability preview custom Claude-demo widget

Pick an operational query

This proves the typed-query widget your homepage hero uses, rebuilt as a custom Elementor element. Generic sample queries, your live build runs against the Sensor Lakehouse.

Response

The redirect map

13 URLs

The single highest-risk artifact, built once in week 1 before any production change. Old URL, one disposition, new URL. Every current status below was checked live, not assumed. Click a row for the reasoning.

13 shown 6 single-hop 301 6 restyle in place 0 intentional 410 218 RDs reclaimed on the Cayenne 404

Old URL · live status

Disposition

New URL

Type

Five-source verification

T+24h / 30 / 60 / 90

No regression is measured against a baseline frozen on day one, across five parallel sources. The baseline is agreed before any cutover, so the post-launch comparison is not my word against yours. The after column stays empty on purpose: those numbers get measured live after cutover, never guessed here.

What no-regression excludes Drops caused by URLs we intentionally 410'd, and normal seasonal swings measured against the prior-year window, are not regressions. That distinction is written into the plan so the 30-day check is fair to both sides. Everything else, an unexplained indexed-page drop or a referring-domain loss, opens a same-day remediation ticket.

Rollback is a button, not a hope

named triggers

Each trigger is defined before cutover, with the action it fires. Two are hard rollbacks, one is a review. Early ranking wobble is normal, so it does not trip an auto-rollback.

Load the risk into week 1

M1 to M4

By the time anything touches production, every decision is already made and tested on staging. The cutover itself is boring on purpose. Click a milestone for its scope.