Website migration

What a migration actually risks

Three things: the URLs that already earn visits, the information buried in pages nobody has read in years, and the forms that quietly deliver enquiries to an address only one person knows about. A redesign that ignores any of those costs more than it saves.

Before you change anything

Scan the current site and read what comes back. The scan is read-only and changes nothing, and it is the last moment where the cost of finding a problem is zero.

• Every page, and which ones nothing links to
• Every URL, and which ones have inbound links
• The services, locations and claims stated across the site
• Every form and where it currently delivers
• Titles, descriptions and structured data
• Tracking and consent tags currently running

URLs and redirects

Keep a URL if it earns traffic, even if you dislike it. Where a URL must change, the redirect is written before publish, not after somebody notices. A redirect added a month late has already lost most of what it was protecting.

What comes across

Content, images and their provenance, page structure, titles and descriptions, structured data, forms, and the business facts the site was asserting. That last one is the part a normal migration loses entirely — it arrives as text in pages rather than as facts anything can check.

Before the old site goes dark

Compare old against new page by page, submit every form and confirm delivery, and check the redirect map against a crawl rather than against the list you wrote. Then publish. Keep the old site reachable until the redirects have been observed working.

If you are not ready

Connect the existing site instead. You get the measurement, the consent handling and the insight without touching a URL, and the migration is still there when it makes sense.