The Redirect And 404 Audit

A redesign moves URLs. A CMS migration changes them. A product line gets renamed. Each time, the old addresses stop working, and whatever authority they had goes with them.

Most businesses discover this months later, when the traffic has already gone.

What actually breaks

External links pointing at pages that no longer exist. The most expensive loss, because somebody else did the work of linking and you threw it away.

Internal links pointing at old URLs, which waste crawl effort and send visitors to error pages.

Rankings on URLs that were earning them, which do not transfer unless you tell the search engine where the page went.

Bookmarks and email links, which matter more than people think on a site with returning customers.

Old campaign URLs, printed on a flyer, in a signature, on a van.

The audit

Pull the old URL list. From the previous sitemap, from analytics covering the period before the change, from Search Console, and from a crawl of an archived copy of the site.

Pull the current URL list from the live sitemap and a fresh crawl.

Diff them. Everything in the first list and not the second is a candidate.

Prioritise by what the page had. Traffic, backlinks, and rankings, in that order. A page with ten referring domains is worth more attention than fifty pages nobody ever visited.

Crawl for internal 404s, which catches the links inside your own site pointing at nothing.

Read the Search Console coverage report, which lists what the crawler is finding and failing on.

Then redirect properly

One permanent redirect to the closest equivalent page. Closest equivalent, not the homepage. A redirect to the homepage passes almost nothing and tells the visitor you could not be bothered.

Permanent rather than temporary, because a temporary redirect tells the search engine the old URL is coming back.

No chains. A redirect to a redirect to a redirect loses value at every hop and slows the page down. Each old URL points directly at its final destination.

No loops, which break the page entirely and are easy to create when two people edit the rules.

Keep them. Redirect rules are not temporary scaffolding, they are permanent infrastructure, and deleting them in a later cleanup undoes the whole exercise.

When there is no equivalent page

Sometimes the content is genuinely gone and nothing replaces it. Then a 404 is correct, and a 410 is better where the removal is deliberate, because it tells the crawler to stop asking.

Redirecting an irrelevant page to a loosely related one is worse than letting it 404. It confuses the index and irritates the visitor.

Make the 404 page useful

Most 404 pages are a dead end with a joke on them.

A useful one says the page has moved or gone, offers the main service pages, offers search, and gives a way to contact somebody. A visitor who hit a broken link is still a visitor, and a dead end converts nothing.

Build the check into the process

The reason this problem recurs is that nobody owns it.

So it goes on the pre-launch checklist for any site change: old URLs mapped, redirects tested, sitemap updated, internal links repointed. And it goes on the quarterly list as a crawl for new 404s, because they accumulate from edits, deleted posts and changed product names.

Ten minutes a quarter prevents the thing that costs a quarter of recovery.

The one that catches everybody

Changing a blog URL structure. One setting, every article moves, and if nobody maps the old paths the entire archive disappears from search at once.

Which is the most common single cause of a sudden traffic drop that has nothing to do with an algorithm.

What this unlocks: the authority you already earned stays attached to the site instead of expiring quietly after a redesign.