Redirect chain

On this page 7

A redirect chain is a sequence of two or more consecutive HTTP redirects that a client must follow before it reaches the final URL that returns content. In SEO, and particularly in crawling, redirect chains matter because every extra hop costs a request, adds latency and increases the chance that a search engine crawler stops following the path before it reaches the page that should be indexed.

A single redirect from an old address to a new one is normal and healthy. A chain appears when redirects are layered over time, for example when the HTTP version of a page redirects to its HTTPS version, which redirects to the www host, which finally redirects to the address with a trailing slash. Each hop is valid on its own, but together they form avoidable overhead that crawlers, browsers and link signals must all absorb.

How a redirect chain forms and is followed

A redirect chain forms when a URL that is itself a redirect target also redirects somewhere else. The server answers the first request with a 3xx status code and a Location header, the client requests the new address, and the cycle repeats until a 200 OK response, an error or a loop ends it.

Chains rarely result from one decision. They typically accumulate through successive site migrations, protocol changes from HTTP to HTTPS, host normalization between www and non-www, trailing slash rules, lowercase rewrites and legacy campaign URLs. Each rule is added by a different team or at a different time, and nobody updates the older rules to point directly at the newest destination.

The table below shows a typical four-hop chain and the rule responsible for each hop.

HopRequested addressStatusRule that fires
1HTTP, bare host, /Shoes301Force HTTPS
2HTTPS, bare host, /Shoes301Force www host
3HTTPS, www host, /Shoes301Lowercase path
4HTTPS, www host, /shoes301Add trailing slash
FinalHTTPS, www host, /shoes/200Content served
Anatomy of a four-hop redirect chain

The fix for this example is to consolidate the rules so that any variant of the address reaches the HTTPS, www, lowercase, trailing-slash address in one single 301 response.

Why redirect chains hurt crawling and SEO

Redirect chains harm SEO through wasted requests, slower discovery and fragile signal transfer. Googlebot follows up to ten redirect hops in a single attempt and defers the rest to a later crawl, while other crawlers such as Bingbot may give up earlier. Even within that limit, each hop consumes a fetch that could have been spent on a new or updated page.

On large sites the cost compounds. When thousands of internal links point to URLs that start chains, a meaningful share of the crawl budget is spent on intermediate responses instead of content. Users experience the same problem as added latency: every hop on a mobile connection can add roughly 100 to 300 milliseconds, depending on network round-trip time, DNS resolution and whether a new TLS handshake is required for a different host.

The following table summarizes the main effects and who feels them.

EffectWho is affectedPractical consequence
Extra fetches per URLSearch engine crawlersLess crawl capacity for real content
Hop limit reachedGooglebot and other botsDestination not crawled in that session
Added latencyUsers and crawlersSlower page load and worse responsiveness
Diluted consolidationRanking systemsSlower or less reliable signal transfer
Higher failure riskAll clientsOne broken hop breaks the entire path
Effects of redirect chains on crawlers and users

Google has stated that 301 and 302 redirects pass PageRank without a fixed loss per hop, so the old fear of losing a percentage of link equity with each step is outdated. The real risks are operational: delayed recrawling, inconsistent canonical signals and the higher probability that one intermediate hop later breaks or becomes a loop.

Redirect chains are one of several redirect and crawl anomalies that look similar in audit reports but require different fixes. Distinguishing them prevents wasted effort and helps teams prioritize the issues that actually block crawling.

IssueDefinitionTypical fix
Redirect chainTwo or more hops before a 200 responsePoint the first URL directly at the final one
Redirect loopHops return to an earlier URL and never resolveRemove the conflicting rule
Single redirectOne hop to the final destinationUsually no action needed
Broken redirectA hop ends in a 4xx or 5xx responseRepoint to a live, relevant page
Crawler trapEndless unique URLs generated by parametersConstrain URL generation and linking
Redirect chains and neighboring crawl issues

A chain can turn into a redirect loop when a new rule sends the final URL back to an earlier one, and a chain becomes a crawl error when any hop fails. For that reason chains are best treated as an early warning of fragile redirect logic, not only as a performance issue.

Finding and fixing redirect chains

Redirect chains are found by crawling the site with a redirect-aware crawler, by reviewing server logs and by testing important URLs manually. Desktop and cloud SEO crawlers report every hop, its status code and the final destination, which makes it easy to export a list of chains sorted by length.

Log file analysis adds the crawler perspective, showing how often Googlebot actually requests intermediate URLs and whether it reaches the final page. For a quick check of a single address, a command such as curl -sIL prints each response header in order, revealing every hop.

The decision table below maps common findings to the recommended action.

FindingRecommended actionReason
Internal links point to a redirecting URLUpdate links to the final URLRemoves the request entirely
Old rules point to outdated targetsRewrite rules to the final URLCollapses the chain to one hop
Several normalization rules stackMerge them into one ruleProtocol, host and path fixed at once
XML sitemap lists redirecting URLsList only final 200 URLsSitemaps should contain canonical URLs
External backlinks hit long chainsRedirect the old URL directlyPreserves signals and speed
Decision table for redirect chain remediation
  • Prefer permanent 301 or 308 responses for moves that will not be reversed.
  • Keep the canonical tag, internal links and sitemap aligned with the final URL.
  • Retain redirects from old URLs for at least one year, and longer when backlinks still point to them.
  • Re-crawl after deployment to confirm that every variant resolves in one hop.

Common mistakes when handling redirect chains

The most frequent mistake is fixing the redirect rules but not the links that trigger them. Internal navigation, canonical tags, hreflang annotations and XML sitemaps that still reference old addresses keep feeding the chain, so crawlers continue to request the intermediate URLs.

A second mistake is removing old redirects too early during cleanup, which turns a slow chain into a broken path and a 404 error. A third is mixing temporary 302 responses into permanent moves, which can make search engines hesitate over which URL to treat as canonical. Finally, JavaScript and meta refresh redirects inside a chain are slower and less reliable than server-side responses and should be replaced with HTTP redirects wherever possible.

Explained without jargon

Imagine sending a letter to a friend who has moved several times. The post office forwards it to the first new address, then to the second, then to the third. The letter arrives, but late and with more chances of getting lost. A redirect chain is the same thing for web pages: the address a visitor or search engine asks for sends them to another address, and then another.

Each step takes time and effort. Search engines have limited time for every site, so long chains waste that time and can delay new pages from appearing in search results. The simple cure is to send everyone straight to the final address in one step.

A typical use case

An online furniture retailer with about 40,000 product URLs moved to HTTPS three years ago, later consolidated on the www host and recently changed its category paths. An SEO audit crawl found that 12,000 internal links pointed to old addresses that passed through three or four redirects before reaching a 200 page. Server logs confirmed that roughly a quarter of Googlebot requests landed on intermediate redirect URLs.

The technical team rewrote the redirect map so that every legacy address pointed directly to its final URL, merged the protocol and host rules into a single rule, and updated navigation, canonical tags and the XML sitemap. Within six weeks, logs showed that Googlebot requests to redirecting URLs fell below 3 percent, new products were indexed several days faster and average page load time on category pages dropped noticeably for mobile visitors.