Search Console "Redirect error": Loops, Long Chains and Bad Targets
Short answer: "Redirect error" means Googlebot followed a redirect but never reached a usable page. The common reasons are a loop where two URLs send visitors back and forth, a chain longer than Google will follow, a redirect to an address that is empty or malformed, or a final target that fails. Diagnose in phases: list the affected URLs, trace each hop with a header check, identify which rule created the problem, then correct it so every old address reaches its final page in one step.
What Google follows and where it stops
Googlebot follows redirects, but not endlessly. Google documents that it will follow up to ten hops from the starting address before giving up, and a chain that long is already poor practice. Anything that does not resolve to a final page is recorded as a redirect error. That status is different from "Page with redirect", which is the normal listing for a URL that correctly forwards elsewhere and is not indexed itself.
So "Redirect error" always indicates something broken, whereas "Page with redirect" is typically fine.
Phase 1: gather the evidence
- In Search Console open Indexing, then Pages, and click the "Redirect error" row. Export the sample of URLs.
- Note whether they share a pattern: all
http://versions, all from one folder, all with a trailing slash, all old product addresses. - Open two or three in a browser. A loop shows as "too many redirects" or "ERR_TOO_MANY_REDIRECTS". A bad target shows as a 404 or a blank page.
Patterns point you to the cause faster than reading single URLs.
Phase 2: trace every hop
Use a header check so you see each step rather than the browser's final result.
- Run
curl -IL https://example.com/old-pageand read the sequence ofHTTP/lines andlocation:headers. - Or use a redirect checker in an SEO crawler and look at the chain column.
Write the path out. An example of a loop:
/shop/ -> /shop (301)
/shop -> /shop/ (301)
An example of a long chain from a redesign:
http://example.com/blog -> https://example.com/blog -> https://www.example.com/blog -> https://www.example.com/news/ -> https://www.example.com/insights/
Phase 3: match the symptom to the cause
| What the trace shows | Likely cause | Fix |
|---|---|---|
| Two URLs bounce between each other | Trailing-slash rule in the server fights a CMS setting or plugin | Choose one form, remove the conflicting rule in either place |
| Endless HTTPS to HTTP and back | Proxy or CDN terminates SSL while the origin forces HTTPS again | Set the SSL mode correctly or tell the app it is behind a proxy |
| Four or more hops before a page | Several migrations stacked on top of each other | Rewrite every old address to point at the final destination |
| Redirect to a blank or invalid location | A pattern rule produced an empty or relative path | Correct the rule and test with sample addresses |
| Final page returns 404 or 500 | Target removed or server error | Restore the page or redirect to a relevant live alternative |
| Redirect depends on cookies or language | A rule sends crawlers without cookies back to a chooser page | Make the default content reachable without a cookie or location rule |
Phase 4: repair properly
- Decide the canonical form of the site: protocol, host with or without www, and trailing slash. Write it down.
- Apply one rule per rule class, in one place. Mixing redirects in the CMS, a plugin, the
.htaccessfile and the CDN is the usual route to loops. - Flatten chains. If A goes to B and B goes to C, change A to go straight to C. Keep a spreadsheet of old and new addresses when you redesign so you can compile it in one pass.
- Use 301 for permanent moves and keep them in place for at least a year, longer where old links remain on other sites.
- Fix internal links so your own pages link to final addresses and not to ones that redirect.
- Update the sitemap to list final URLs only.
Phase 5: verify and tell Google
- Re-run the header trace on a sample of each pattern group and confirm a single hop to a 200 response.
- Test the same URLs in Search Console's URL Inspection live test.
- In the "Redirect error" report press "Validate fix". Google will recheck the sample and report the result.
- Check again after two or three weeks. Some old addresses are recrawled slowly.
A short illustrative example
A UK physiotherapy clinic changes its permalink structure, then installs a caching plugin that forces trailing slashes while the theme strips them. Every blog post now loops. The trace shows the pair of 301 responses. The clinic removes the slash rule from the plugin, keeps the theme's behaviour, flushes the cache and retests. The report clears after validation.
Habits that prevent a repeat
- Test redirects on a staging copy with a sample of fifty old URLs before launch.
- After any change to SSL, CDN or permalinks, re-check the homepage and a few deep pages with a header trace.
- Keep a single redirect map and name an owner for it.
FAQ
How many redirects in a row are acceptable?
Aim for one. Google follows a limited number of hops, and every extra step slows visitors and weakens the signal passed on.
Is "Page with redirect" an error?
No. It lists addresses that forward elsewhere as intended. It only needs attention if a URL there should have been indexed.
Do 302 redirects cause this error?
Not by themselves. Use 302 or 307 only for genuinely temporary moves; the error arises from the loop or failure, not the code.
Why does the browser work but Google fails?
Browsers may carry cookies, language settings or cached redirects that crawlers do not. Test with a clean curl request.
If a migration has left your redirects tangled, tell us about the site through our enquiry page.