Redirect chain
A redirect chain is a URL that redirects to another URL that redirects again, such as A to B to C. Each hop costs the visitor a round trip and the crawler a fetch. Google follows up to about ten hops in one crawl before it gives up and tries again later.
By Théophile Louvart, founder of Porteur · Updated 15 September 2026 · Markdown
What a chain looks like
http://yourproduct.com/old-page 301 -> https://yourproduct.com/old-page
https://yourproduct.com/old-page 301 -> https://yourproduct.com/guides/old-page
https://yourproduct.com/guides/old-page 200Three requests where one would do. The visitor waits for all of them, on a phone connection each one costs a round trip, and a crawler spends three fetches to read one page.
A redirect loop is the pathological case: A redirects to B and B redirects back to A. It never resolves, the browser gives up, and the URL leaves the index.
Where chains come from
- Rules that accumulate: http to https, then non-www to www, then an old slug to a new one, each added by a different person on a different day.
- A migration where the old URLs point at the URLs of the previous migration rather than at the live page.
- A CDN rule and a server rule that both fire, each doing half the job.
- Internal links and sitemaps that still point at the first URL of the chain, so the chain is walked on every crawl.
What it costs
| Hops | What happens |
|---|---|
| 1 | Normal. A permanent redirect passes the old URL’s standing to the new one. |
| 2 or 3 | Slower for the visitor, more fetches for the crawler. Worth flattening, not urgent. |
| More than 3 | Each hop is a place to break. Flatten them. |
| About 10 | Google stops following in that crawl and comes back later, so indexing of the target is delayed. |
| A loop | The URL never resolves and is dropped. |
How to flatten one
Read the chain
Request the first URL and follow the redirects, noting every status and destination. A redirect checker does this in one call.
Point the first URL at the last
Replace the rules so the first URL redirects straight to the final 200, with a 301 when the move is permanent.
Fix what feeds the chain
Update internal links, the sitemap, canonicals and any email or ad that still uses the old URL.
Keep the old rules
Other sites still link to the old URLs. Leave the single-hop redirects in place; they cost nothing once they are one hop.
Re-check after the deploy
A chain reappears the day someone adds a rule at the CDN. Check the important URLs after each migration.
Questions
Google follows about ten hops in one crawl before giving up. In practice anything past three is worth flattening, because each hop is latency for the visitor and a place for the chain to break.
Google has said since 2016 that redirects do not lose PageRank in themselves. The cost is crawling, latency and fragility, not a tax per hop.
A chain ends at a page that answers 200. A loop never ends: two URLs redirect to each other, the browser stops after a few tries, and the URL is dropped from the index.
Crawl the site and look at the redirect report, or check the URLs that matter one at a time with a redirect checker. The usual suspects are the home page over http and any URL changed in a migration.
No. Keep them as single hops for as long as the old URLs are linked anywhere, which in practice means permanently. They cost nothing once they point straight at the final URL.
Check my site, free
The free check follows the redirects on your site and names the chains worth flattening, in about thirty seconds.
- Free check, no card
- Read-only, your own accounts
- Readable by your agent
Read next
- Glossary301 redirect
- Glossary302 redirect
- Free toolRedirect checker
- GuideBuild a redirect map: one old URL, one new URL, tested
- GuideHTTP status codes for SEO: what each one tells Google
- GuideMoving from HTTP to HTTPS: the checklist that keeps rankings
- Guidewww or non-www: pick one, redirect the other
- GuideThe best AI visibility tools, and the manual method that costs nothing