Recovering old WordPress URLs after a migration

Client work. The client's name and dates are withheld, and I have not reproduced the original test records.

What they wanted

An old URL with a relevant replacement should reach it in one redirect. A retired or ambiguous page should not be sent somewhere unrelated just to make the number of 404s look better.

What was wrong

• Old URLs did not have reliable current destinations.
• The redirect rules needed reviewing page by page, because the old site had pages with no equivalent in the new structure.
• Canonicals, the sitemap and internal links had to agree with the replacement addresses, or the redirect would lead to a page that disagrees with itself.

What I did

• Mapped the recoverable old URLs to specific current pages.
• Excluded retired and unmatched pages from unsafe catch-all redirects, so they stay missing rather than landing on the wrong answer.
• Aligned the plan with the canonical URLs, the sitemap and the internal links.

How I checked it

I sampled old URLs to confirm each reached its intended destination in one hop. The set was also checked for chains, loops and soft 404s, and the live responses were compared against the agreed map.

What it doesn't prove

The redirects were checked technically. This summary does not report recovered rankings, traffic or indexing for every old URL.