For around three months, I have been tracking a persistent Redirect Error in Google Search Console (GSC), and for all of that time, every test I could run showed the redirects working perfectly. Even using GSC’s own live URL test reported them as fine. But the moment Google’s crawler tried to index them, they came back as Redirect Errors.
I tested as systematically as I could. Curl requests across multiple user agents all returned clean responses. I moved the redirects across completely different systems to rule out anything server-side: one site handles them through Cloudflare, one through .htaccess, one through a dedicated redirect plugin. Even WordPress’s own canonical trailing-slash redirects were being flagged, which is something WordPress handles automatically on every installation. I even changed server entirely during testing. Nothing I did, made a difference. The “Redirect Error” in GSC stayed the same regardless of what was producing the redirect.
A key part of the diagnosis came from comparing what GSC’s URL Inspection tool showed versus what Google actually recorded during indexing. The actual Googlebot crawl recorded “Failed: Redirect error.” The live test on the exact same URL, run immediately, showed the page as fully accessible with no issue. Google’s crawler and Google’s own inspection tool were producing different results on the same URL, which ruled out the redirect itself as the cause.
The most useful external evidence came from a thread on the official WordPress.com support forum. Redirect errors appearing on pages with no redirects at all, a different scenario entirely, but the pattern was the same: live GSC tests passing, actual crawl failing. Automattic staff investigated and found nothing wrong at their end, they concluded it appeared to be a Google-side issue. A different problem, pointing in the same direction.
The practical reality is that this keeps coming back. Several times a week I am going through affected URLs, making small changes, and requesting reindexing through GSC, waiting for Google to recrawl, and then checking whether the classification has changed. It doesn’t. The process of auditing this error has become a significant time drain, and that is before factoring in that each reindex request means waiting for Google to act on it.
I am still working through it. Everything points towards this being a Google-side behaviour, but because I can’t find evidence of this being more widespread, I have to keep searching, so answers on a postcard if you have any tips or ideas what may be causing it.
