Redirects Are Working But Traffic Still Dropped

A redirect that checks out in a crawler proves one thing: the old URL now sends people and crawlers to the new one. It doesn’t prove the new page answers the queries the old one ranked for, that Google has settled on the new URL, or that the rest of the site still points at it. When traffic falls after a migration with working redirects, look at five things in order: whether the drop is the fluctuation Google says to expect, which URL Google now treats as canonical, whether the new page changed what it answers, what happened to internal links, and whether pages were merged. Redirect chains come last, as a speed and maintenance fix.

What a working redirect does

Google’s documentation on redirects says that with a permanent redirect, Googlebot follows the redirect and the indexing pipeline uses it as a signal that the redirect target should be canonical. Google keeps track of both the old and the new URL; one becomes the canonical, the other an alternate name, and which one depends on signals such as whether the redirect was temporary or permanent.

That is the core of the job. A redirect is one canonicalization signal among several. It doesn’t carry the old page’s fit for a query onto a new page that no longer has it.

First, check the timing

Google’s site move guide puts the wait before Google starts showing a medium-sized site’s new URLs in place of the old ones at a few weeks or more, and asks you to watch traffic on both sets of URLs through that window.

A drop in the first weeks, with old URLs still showing in results and new URLs appearing gradually, can be the move settling rather than a defect. A drop that deepens after the new URLs have replaced the old ones in results points to other causes, such as canonical selection, content, internal links or merges.

Check which URL Google chose

Run the new URL through the URL Inspection tool and read the Google-selected canonical in the indexed data. If Google still treats the old URL as canonical, other signals may be outweighing your redirect, or Google hasn’t processed it yet. Google’s guide to consolidating duplicate URLs calls rel="canonical" a strong signal and sitemap inclusion a weak one, and says linking consistently to the URL you consider canonical helps Google understand your preference. Look for those signals on your own site:

• internal links, navigation and breadcrumbs that still point to old URLs and rely on the redirect
• XML sitemaps that still list old URLs
• canonical tags on new pages that point back to old addresses

Update each one to the new URL so every signal agrees with the redirect.

Check whether the page changed what it answers

For example, imagine a short page that answered “how to reset X” in three steps ranked for that quick question. In the redesign it was redirected into a long guide that covers resetting X among ten other tasks. The redirect is flawless, but a searcher who wanted the quick answer now has to hunt for it.

The Performance report shows whether this happened. Filter by the old URL for a date range before the move and note its top queries. Then filter by the new URL, and the old one, for an equal range after the move and look up the same queries, since clicks can still land on the old URL while Google shows it. If your traffic is seasonal, compare each range with the same weeks a year earlier as well. Queries that kept their clicks moved successfully. If the canonical check was clean, queries that lost them point to where the new page answers a different question. Those queries need content: give them a page, or a section at the top of the new page, that answers them as directly as the old page did.

Check what happened to internal links

A restructure rewrites the internal link graph. Google’s guide to crawlable links says every page you care about should have a link from at least one other page on your site, and that the anchor text of internal links helps people and Google make sense of your site. A destination that used to be linked from navigation, category pages and related-content modules, and now isn’t, has lost those links whatever the redirect does.

Crawl the site before and after the move, or compare a saved pre-move crawl with the current one, and count internal links to your top landing pages. Restore links to pages that fell toward orphan status.

Check for merges

Merging several pages that ranked for different queries into one page is a consolidation decision, not a redirect detail. A merged page may not serve every query its sources answered. If two old pages ranked for different searches, redirecting both into one can give up part of what one of them earned. Keep distinct-intent pages distinct unless you decided to give those queries up.

Redirect chains

Google’s documentation on HTTP status codes says Google’s crawlers follow up to 10 redirect hops by default. A chain of two or three hops still reaches the page. Flatten chains anyway: each hop adds a request for every visitor, and chains grow quietly when a rule from an earlier migration was never updated to skip the middle step. Point every old URL straight at its final destination.

Change one thing at a time

When a launch bundles new URLs, a new design and rewritten content, a traffic drop has three suspects that are hard to tell apart. The site move guide recommends that large sites, where technically possible, move just a piece of the site first to test the effects on traffic and search indexing. The same logic applies to the variables: move URLs one to one with the content unchanged, let the move settle, then redesign or rewrite as a separate project. A drop is then easier to trace to one cause.

Frequently asked questions

Do 301 redirects lose PageRank?

Google’s site move guide answers this directly: 301 and other permanent redirects don’t cause a loss in PageRank. Its redirects documentation adds that a permanent redirect is a signal that the target should be canonical. Plan around the signals that decide which URL Google keeps: redirects, internal links, sitemaps and canonical tags all pointing to the new URL.

My redirects all work. Why would rankings fall at all?

Because a redirect moves visitors and crawlers and sends a canonical signal. Google still decides which URL is canonical, and the new page still has to answer the queries the old one did, with the internal links it used to have.

How do I tell a content problem from a signals problem?

Check the Google-selected canonical first. If Google hasn’t settled on the new URL, fix the signals. If it has and specific queries lost clicks, compare what the old and new pages answer for those queries.