HTTP redirects and SEO

1Overview

Redirects exist so that a URL can keep working after content moves. A server answers the request with a 3XX status code and the new address in the Location header, and the client requests that address instead [2][11].

For search engines, redirects do two jobs: they lead crawlers to the new URL, and they tell Google which URL to treat as canonical. Google lists redirects as a strong canonicalisation signal, alongside rel="canonical"[7].

2History

The 301 Moved Permanently and 302 Moved Temporarily codes were defined in HTTP/1.0 in 1996 [3]. Because many clients changed a POST request to a GET when following a 302, HTTP/1.1 in 1999 added 303 See Other and 307 Temporary Redirect to make the behaviour explicit [4].

The permanent counterpart that preserves the request method, 308 Permanent Redirect, was standardised in 2015 [5]. All of these are now defined together in RFC 9110, published in 2022 [2].

3Status codes

4Ways to redirect

Google lists its preferred methods in order [1]:

• Server-side redirects (HTTP 3XX): the clearest signal, seen by every client before any content loads [1].
• Meta refresh: an instant meta refresh is interpreted as a permanent redirect, a delayed one as temporary [1].
• JavaScript redirects: only seen after the page is rendered, so use them only when server-side redirects aren’t possible [1][13].

Browsers may also redirect internally. When a site uses HTTP Strict Transport Security, the browser upgrades http requests to https itself, before contacting the server [10]. This doesn’t remove the need for a real 301 from http to https for crawlers and first-time visitors.

5How Google processes redirects

Googlebot follows redirects as part of crawling. It follows up to 10 hops; if it doesn’t receive content within that limit, Search Console reports a redirect error [6].

A redirect whose final target returns an error is treated as that error, so the original URL’s signals are lost [6]. Google then picks a canonical among the URLs involved, using the redirect type together with other signals [7][12].

6Chains and loops

A redirect chain is a sequence of redirects, A → B → C, usually created by successive migrations or overlapping rules for protocol, host, case and trailing slashes. Each hop adds a round trip for users and another fetch for crawlers.

A redirect loop, A → B → A, never resolves; browsers stop with an error. The remedy for both is the same: rewrite each rule to point directly at the final URL, so every old URL resolves in one hop.

7Site migrations

For a move that changes URLs, Google recommends preparing a mapping from every old URL to its new equivalent, implementing permanent server-side redirects, and updating internal links, canonical tags and sitemaps to the new URLs [8].

Google advises keeping the redirects for as long as possible, generally at least a year, so that signals are transferred and users with old links still arrive [8]. For a move to a new domain, the Change of Address tool in Search Console tells Google about the move directly [9].

8Redirect, canonical or 404?

9Common misconceptions