301 Redirects: Using Redirects Correctly

A 301 redirect is one of the few SEO tools that is technically trivial and strategically consequential at the same time. A single line in the server configuration decides whether the visibility a URL has built up over years is kept or lost. In domain migrations, relaunches, and the consolidation of HTTP to HTTPS, success depends almost entirely on correctly set redirects. This article explains the technical differences between the redirect types, shows the implementation on common servers and in Next.js, covers the most common mistakes, and looks at the growing role of redirects for AI crawlers and GEO.

Why 301 redirects are business-critical for companies

URLs change; on the web that is the rule, not the exception. Products get renamed, content management systems replaced, domains merged, site structures reworked. Each of these changes creates an old URL that leads nowhere without a redirect. For search engines that means a 404 error, for users a dead end, and for the company the loss of rankings, traffic, and all external links pointing to the old address.

The scale of this problem is considerable. An international study of more than two million websites shows that 66.5 percent of the links examined have rotted since January 2013 (international), meaning they point to pages that no longer exist or have moved. Every one of these dead links could have been preserved with a clean 301 redirect. At the same time, redirect problems are widespread: in an analysis of 150,000 randomly selected websites, 7 percent of sites had redirect chains or loops with three or more URLs (international), a technical defect that wastes crawl budget and slows down load times.

For the Austrian B2B market, the economic importance of the website comes on top of that. According to Statistik Austria, around 30.8 percent of Austrian companies made sales via websites, apps, online marketplaces, or EDI in 2023, up from 26 percent the year before (Austria). If a relaunch without correct redirects cuts organic visibility in half, that hits revenue directly. This is exactly why the 301 redirect is not a technical detail for developers but a business decision.

How it works: 301 vs. 302 vs. 307 vs. 308

A redirect is an instruction in the HTTP header that returns a status code and the redirect target (Location header). Browsers and crawlers follow this instruction and load the target address. What matters for SEO is which signal each status code sends to the search engine.

• 301 Moved Permanently: The permanent redirect. According to its documentation, Google treats it as a signal that the redirect target should be the canonical URL and transfers the ranking signals to the new address. The rule is: 301 redirects tell Google that a page has moved permanently, which is why the new URL almost always gets indexed and link signals consolidate forward to the target (international).
• 302 Found and 303 See Other: Temporary redirects. Googlebot follows them but does not use them as a canonicalization signal. In practice, this means: 302 redirects keep the old URL in the index, and link signals consolidate backward to the source (international). They are only suitable when the redirect is truly temporary, for example for maintenance pages or A/B tests.
• 307 Temporary Redirect: The HTTP/1.1 equivalent of 302 that strictly preserves the HTTP method (such as POST). 307 also sends no canonicalization signal and is unsuitable for permanent moves.
• 308 Permanent Redirect: The modern, method-preserving variant of the 301. For SEO, Google treats it the same as a 301.

The official position is clear: according to Google Search Central, 301 and 308 send a canonicalization signal, while 302, 303, and 307 do not; an immediate meta refresh is treated like a permanent redirect (international). For any permanent move, a 301 (or 308) is therefore the correct choice. The common mistake of setting a 302 out of convenience gives away link equity and keeps the wrong URL in the index.

Implementation: Apache, Nginx, IIS, and Next.js

A 301 should happen server-side, not via client-side JavaScript. Server-side redirects take effect before the page loads, are unambiguous for crawlers, and cost no rendering time.

• Apache (.htaccess / mod_rewrite): For individual URLs, Redirect 301 /old /new is enough. For patterns, you use mod_rewrite, for example RewriteRule ^old/(.*)$ /new/$1 [R=301,L]. The R=301 flag forces the status code, and L ends rule processing.
• Nginx: Efficient and fast via return 301 https://www.example.at$request_uri; inside a matching server or location block. More complex cases can be handled with rewrite ... permanent;. The return variant is preferable to the rewrite variant because it is evaluated faster.
• IIS (Windows Server): Via the URL Rewrite module or the web.config with a <rule> and redirectType="Permanent" (301). Permanent is the default value there, but it should still be set explicitly in the <rule>: the examples in Microsoft's URL Rewrite Module configuration reference use redirectType="Found" (302), and anyone who copies them only redirects temporarily. URL Rewrite does not support a 308. The separate HTTP Redirection module (<httpRedirect>) returns a 302 when nothing is specified and needs httpResponseStatus="Permanent" for a 301. From IIS 10 onward, it also sends a 308 with PermRedirect.
• Next.js: In next.config.js via the asynchronous redirects() function with permanent: true, which returns a 308. For dynamic or conditional logic, for example depending on cookies or geo data, the Proxy (proxy.ts, middleware.ts up to Next.js 15) with NextResponse.redirect() is the right fit. Client-side redirects via useRouter or window.location should be avoided for SEO-relevant moves because they only take effect after rendering and are unreliable for crawlers.

Regardless of the platform, every redirect should point directly to the final target and not to another redirect. This is exactly where the most common technical problems arise.

Strategy: avoiding redirect chains and planning domain migrations

A redirect chain occurs when URL A redirects to B, B to C, and only C to the target. Every additional hop costs time, puts load on the crawl budget, and worsens TTFB and with it the LCP value (Largest Contentful Paint). A redirect loop is the special case in which the chain leads back to itself and the page can no longer be reached at all.

Google gives a clear upper limit and recommendation here. According to Search Central, Googlebot can follow up to 10 hops in a redirect chain, but Google recommends redirecting directly to the final target and, if that is not possible, keeping the chain short, ideally no more than 3 and fewer than 5 (international). In practice, this means flattening chains to a single hop. If old mappings cause A to point to B and B to point to C, A should be rewritten to point directly to C.

In a domain migration or website relaunch, the redirect strategy is the most important success factor. A proven approach includes several steps:

• Create a 1:1 mapping: Every old URL gets a defined new target with the highest possible topical match. A full crawl of the old site (for example with Screaming Frog) provides the starting list.
• Document the redirect map: The old-to-new mapping is recorded in a spreadsheet and tested before go-live. It is the basis for the server configuration.
• Monitor in Search Console: For a pure domain change, the Change of Address tool helps. After the migration, the Page indexing report and the Crawl stats report show whether old URLs redirect cleanly and new ones get indexed.
• Plan a recovery period: After a migration, rankings take weeks to months until all signals have been transferred. Fluctuations during this phase are normal.

Google explicitly recommends keeping the redirects in place for a long time. They should generally remain in place for at least 1 year so that Google can transfer all signals to the new URLs, including recrawling and reassigning links on other websites (international). Only after that can you safely remove old rules.

Link equity, HTTPS consolidation, and how redirects work with canonicals

The real SEO value of a 301 lies in transferring link equity, meaning the ranking power built up through backlinks. This transfer comes with a condition, though: topical relevance between source and target. If you redirect a specific product page wholesale to the homepage, Google often does not treat it as a real redirect. A documented case study shows that redirects to the homepage or to less relevant pages can be treated as soft 404s, in which case link signals are not passed on and the redirected URLs lose their rankings and traffic (international). The rule is therefore: always redirect to the closest topical equivalent, not to the homepage.

A classic must-have use case is consolidating URL variants. A website should be reachable under exactly one canonical variant, and all other variants redirect there via 301:

• HTTP to HTTPS: All unencrypted requests are permanently redirected to HTTPS. This is standard today: according to the Web Almanac 2024, 89 percent of desktop pages and 88.9 percent of mobile pages used the HTTPS protocol (international). In addition, the HSTS header makes browsers enforce HTTPS directly and skip the redirect hop from then on.
• www vs. non-www: One of the two variants is chosen as canonical, and the other redirects to it via 301. This prevents Google from indexing two separate versions of the same page.

This 301 consolidation is backed by consistent canonical tags: the target page points to itself via rel="canonical", and hreflang annotations reference only the final HTTPS URLs. Redirect, canonical, and hreflang must name the same canonical target; otherwise, conflicting signals arise.

301 redirects for AI crawlers and GEO

With the spread of AI crawlers, the 301 takes on a new function: enforcing canonical content toward systems such as GPTBot, ClaudeBot, or Googlebot in its AI role. The volume of this traffic is growing fast. According to Cloudflare, requests from GPTBot rose by 305 percent and from Googlebot by 96 percent between May 2024 and May 2025; total AI and search crawler traffic grew by 18 percent on a fixed customer base (international). These bots run into the same outdated and moved URLs as classic crawlers, and here too, redirect chains waste resources.

Cloudflare shows a concrete use case with an edge solution that enforces canonical signals toward AI crawlers via redirects. When a verified AI crawler hits a page with a non-self-referencing canonical tag, the edge server returns a 301 to the canonical URL before outdated content is served. As a result, in the first seven days, 100 percent of requests from AI training crawlers to pages with a non-self-referencing canonical tag were redirected via 301 instead of being served outdated content (international).

For GEO, this means that clean 301 redirects and consistent canonicals ensure AI systems capture and cite only the currently valid version of a piece of content. Where edge or CDN redirects are available, this behavior can be implemented closer to the user and with fewer resources, complemented by IndexNow, which actively notifies search engines about changed URLs.

Common mistakes and audit

Redirect errors are rarely dramatically visible, but they add up to a considerable loss of visibility. The most common patterns can be checked one by one.

• Blanket redirect to the homepage: The classic soft 404 case described above. Every redirected URL needs a relevant target.
• Redirect chains and infinite loops: Multiple hops and loops, often caused by rules stacked on top of each other over several relaunches. Solution: flatten to one hop.
• Mixed content after the HTTPS switch: If HTTPS pages still load resources over HTTP, browsers block them. HTTPS consolidation must cover all internal links and assets.
• Client-side workarounds: Meta refresh and JavaScript redirects instead of server-side 301s. Google does treat an immediate meta refresh as permanent, but it is slower and more error-prone than a clean HTTP 301.
• Temporary codes for permanent moves: A 302 where a 301 belongs keeps the old URL in the index.

Specialized tools have become established for the audit. Screaming Frog crawls the site and lists redirect chains, loops, and faulty status codes including the hop path. Ahrefs and Sitebulb provide additional reports on redirect issues and affected backlinks. Finally, Google Search Console shows which URLs Google actually classifies as redirected, indexed, or soft 404.

GDPR and tracking aspects in the DACH context

In the DACH region, redirects also raise data protection questions. Redirect tracking, meaning routing users through an intermediate tracking domain to measure clicks, can process personal data and pass on cross-domain identifiers. Such methods generally require a legal basis and, depending on how they are set up, consent under the GDPR. Purely technical 301 redirects for URL consolidation are a separate case: they serve correct delivery and are not primarily designed for user tracking. If you use cross-domain measurement during domain migrations, document the data flows and align them with your consent management so that you do not trigger processing that requires consent without having obtained it.

301 vs. canonical tag: the decision matrix

301 redirects and canonical tags are often confused, but they solve different problems. Both signal a preferred URL, but only the 301 physically sends users and crawlers to another address.

• Use a 301 when: The old URL should no longer exist. For migrations, relaunches, deleted pages with a replacement, and HTTPS or www consolidation. The user always lands on the new address.
• Use a canonical when: Several similar or identical pages should remain accessible, but only one should be indexed. Typical for filter and parameter URLs, print versions, or identical content under several paths. Both pages remain accessible to the user.

The rule of thumb: if the old URL should disappear, it is a 301. If both URLs should remain reachable but only one should rank, it is a canonical. A canonical is also only a hint that Google can follow but does not have to, while a 301 is a binding instruction.

Going further

The 301 redirect is the backbone of any clean URL strategy. Used correctly, it transfers link equity, consolidates variants, and makes canonical content unambiguous for crawlers, AI systems, and users. The key principles are quick to sum up in practice: for permanent moves always 301 or 308 instead of 302, server-side instead of client-side, directly to the topically matching target instead of the homepage, without chains or loops, and kept in place for at least one year. Before every relaunch, a full crawl with a documented redirect map pays off, followed by monitoring in Search Console. Useful follow-up topics are the canonical tag for controlling duplicate content, hreflang for multilingual sites, HSTS configuration for enforcing HTTPS, and IndexNow and AI crawler management for a future-proof technical SEO architecture.