When is a redirect chain harmless vs a real problem?
- Published November 6, 2026
- Linked from 2 answers
- In a cluster of 3
A single redirect from an old URL straight to its replacement is not a chain and is not a problem. A chain starts when that target redirects again, and it becomes a real problem when it grows long, loops back on itself, or breaks partway. Google’s site move guidance says plainly: “Avoid chaining redirects”, because “Chaining redirects adds latency for users, and not all user agents and browsers support long redirect chains.”
What each extra hop costs
Every hop is one more HTTP request before the visitor or crawler reaches the page. That adds waiting time for people on slow connections and one more fetch for Googlebot. Google’s HTTP status code documentation sets the outer limit: “By default, Google’s crawlers follow up to 10 redirect hops”, and it adds that specific products’ crawlers may have different limits. Staying far below that limit is the point; the limit is where Google stops, not a safe target.
| Chain | Status |
|---|---|
| A to B, B returns 200 | A normal redirect |
| A to B to C, C returns 200 | Works, adds a request; worth collapsing on pages that matter |
| A to B to C to D and beyond | Slower for users, more places to break; collapse it |
| A to B, B back to A | A loop: the page never loads for anyone |
| A to B, B returns 404 | A dead end: everything that pointed at A is lost |
The two failures to fix immediately
A loop and a broken hop both stop the page from loading. A loop sends the browser and Googlebot in a circle until they give up. A chain that ends in a 404 delivers visitors and crawlers to a missing page, and Google drops indexed URLs that return a 4xx status. Find and fix either one the day you see it.
When a chain is not worth fixing
A two-hop chain on a URL that gets almost no visits and has no links pointing to it costs little. If the clean-up means editing a redirect map that other teams depend on, it can wait for the next scheduled change. The chains worth fixing first are on pages with traffic, pages with external links, and anything your own navigation or sitemap still points at.
How to collapse a chain
Point every old URL directly at the final destination, so A redirects to C without passing through B, and update internal links to use C. On pages with real traffic, plan the change with the checks in what can go wrong when you redirect a high-traffic page. If the chain comes from an HTTP-to-HTTPS move, whether HTTPS still gives a ranking edge puts that move in context. Use a crawler that reports redirect paths to list chains, then recheck after each migration, since a new redirect layered on an old one is how chains form.
Add a comment