How many redirects does Google really follow before it splits the crawl?

What does it actually mean to 'treat as a single request'?

When Googlebot encounters a URL that redirects to another, it automatically follows the chain of redirects to reach the final URL. Mueller clarifies here that Google follows up to five consecutive redirects without interrupting this process. These five hops are treated as a single crawl operation, counted only once in your crawl budget.

Beyond this limit, the process fragments. Each additional redirect becomes a separate request , which mobilizes more server resources and disproportionately consumes your crawl budget. For a site with thousands of pages, this difference is not trivial — it can delay the indexing of priority content.

Why does Google impose this limit of five redirects?

The main reason is performance . Following infinite chains of redirects would expose Googlebot to redirect loops, configuration errors, or malicious cloaking attempts. By setting a threshold at five, Google strikes a balance between technical flexibility and protection against abuse.

This limit also corresponds to on-ground reality. In 99% of healthy configurations, you never exceed three redirects — and if you have five, it's usually a sign of a problem with architecture or poorly cleaned up successive migrations. Google leaves you some leeway but does not endorse chaos.

Does this rule apply to all types of redirects?

Mueller does not specify whether the distinction between 301/302/307/308 influences this treatment. In practice, the type of redirect has no impact on the counting of hops — it’s the number of consecutive HTTP requests that matters. Whether you chain five 301s or a mix of 302 and 307, Google stops the clock at the fifth hop.

However, JavaScript or meta-refresh redirects are a different story . They require client-side rendering, hence a different processing method. Mueller speaks here only of classic server-side HTTP redirects. If your chain mixes server and JS redirects, it is even more costly in resources — and potentially not counted the same way.

• Google follows up to 5 consecutive HTTP redirects as a single request
• Beyond that, each additional jump becomes a separate request , wasting crawl budget
• This limit applies to all server redirects (301, 302, 307, 308)
• JS or meta-refresh redirects do not follow this logic — they require rendering and are heavier
• In a clean architecture, you should never exceed 2-3 redirects

Is this statement consistent with on-the-ground observations?

Yes, and it's even one of the rare times Google gives a precise number. Empirical tests show that Googlebot really follows up to five hops before treating subsequent redirects as fragmented requests. This behavior is observed in server logs: a chain of four redirects generates a single crawl entry, while a chain of six generates multiple.

That being said, Mueller says nothing about crawl speed or lost PageRank . Even if Google 'follows' the five redirects, each hop dilutes the link equity transmitted a little more. Tests show that beyond two redirects, you start to lose juice — even if technically, the final URL is indexed properly.

What nuances should be added to this rule?

First point: Mueller talks about 'consecutive' redirects . If your architecture mixes server redirects and client redirects (JS, meta-refresh), the counter does not work the same way. Google must first render the JS page, which adds latency and consumes more budget. A chain of three server redirects + one JS can cost as much as a chain of six classic redirects.

Second nuance: this limit concerns crawling, not necessarily final indexing . Google can very well index the destination URL even if the chain exceeds five hops — but the process will be slower, less prioritized, and the link equity diminished. [To be verified] : no official data specifies whether a chain of seven redirects completely blocks indexing or simply delays it.

In what cases does this rule not strictly apply?

Google may raise this limit for very high authority domains — it's a plausible hypothesis but not confirmed by Mueller. Sites like Wikipedia or Amazon have almost unlimited crawl budgets. If their technical architecture imposes six redirects on some pages, it's likely that Google still follows them. But for 99% of sites, don’t count on it.

Another edge case: temporary redirects (302, 307) are not always treated the same way as permanent ones . If you chain five 302s, Google may decide to recrawl the chain multiple times to check if it changed — which doubles the actual budget consumption. Mueller does not specify this point, but it's consistent with observed behavior in logs.

What should you do to clean up your redirect chains?

First, audit your site with Screaming Frog or Sitebulb by enabling the 'Redirect Chains' option. These tools automatically detect chains of two or more redirects and list them for you. Export the report, sort by chain length, and focus on those exceeding two hops.

Next, fix the chains by redirecting directly to the final URL . If A → B → C → D, replace with A → D and B → D. This reduces user latency, preserves link equity, and frees up crawl budget. In 80% of cases, these chains are the result of poorly documented successive migrations — no one ever cleaned up afterwards.

What mistakes should be avoided when managing redirects?

Never redirect a redirect — this is the golden rule. If you migrate a site that has already been migrated, don't just add another layer of redirects on top of the existing ones. Trace the full chain, identify the original source URL, and redirect it directly to the new final destination.

Second common mistake: forgetting external backlinks . You can fix your internal redirects, but if thousands of inbound links point to a URL that redirects twice, you're wasting equity. Contact the most significant referring sites to update their links — or use disavow if those links are toxic anyway.

How can I check if my site complies with this limit?

Inspect your server logs to spot crawl patterns. If you see Googlebot making the same request multiple times on a redirect chain, it's splitting the process — a sign that you're exceeding the limit. Compare with your Search Console reports: if some pages take weeks to be indexed while being crawled, redirect chains are often to blame.

Also, use the 'URL Inspection' tool in Search Console to manually test your problematic URLs. Google will show you the indexed final URL and the redirect path followed. If you see more than two hops, that’s an immediate red flag. Prioritize correction based on the current organic traffic of those pages.

• Audit the site with Screaming Frog to identify all redirect chains
• Fix chains of 3+ redirects by redirecting directly to the final URL
• Check external backlinks and update links from important referring sites
• Inspect server logs to detect split crawl patterns
• Test critical URLs in Search Console to validate they are indexed without friction
• Document each migration to avoid stacking layers of redirects over time
  In practice, aim for zero or a maximum of one redirect on your critical URLs. Beyond two hops, you start losing performance, crawl budget, and link equity — even if Google technically follows up to five. These technical optimizations can be complex to implement alone, especially on sites with a history of multiple migrations. Engaging a specialized SEO agency can quickly identify problematic chains, prioritize fixes based on business impact, and automate monitoring to prevent new chains from appearing after each overhaul.