Definition
When a server answers a request with a 301 status and a new location, it is declaring the move permanent. The related 308 code is also permanent, and Google passes signals through both in the same way. The temporary codes, 302 and 307, say the original address will return, so a search engine may keep the old URL indexed for a while; Google's documentation on redirects adds that a temporary redirect left in place long enough is eventually treated as permanent. Redirects set in the server or the content management system are the dependable form; Google's guidance prefers them to redirects made in page scripts or meta refresh tags, which it can follow only once the page itself has been fetched and processed.
What a permanent redirect transfers is the old URL's role. Links that other sites made to it, its history in the index and the internal links still pointing at it all resolve to the destination. That is why redirects matter most for pages that have earned links and rankings, and matter least for URLs nobody ever visited.
In practice
Three situations call for a permanent redirect:
- Migrations. A new domain, platform, protocol or URL pattern changes many addresses at once. The site migrations guide treats the redirect map, a table of every old URL and its new home, as the core artefact of the whole project.
- Consolidation. Two pages that compete for the same query are merged and the weaker one is redirected to the stronger.
- Single URL changes. A product renamed or an article given a clearer address. Changing a URL is an architectural decision in its own right, and the site architecture guide recommends changing URLs only with a reason and a redirect map.
After any of these, the internal links should be updated to point straight at the new URL rather than leaning on the redirect. A redirect protects the links a site does not control; the links it does control should be fixed at source.
Redirect, canonical tag or removal
| Situation | Tool | Why |
|---|---|---|
| The old URL is gone and a true equivalent exists | Permanent redirect | Users and signals move to the replacement |
| Both URLs must stay reachable but only one should be indexed | Canonical tag | A hint that consolidates without moving anyone |
| The content is gone and nothing replaces it | Not-found or gone status | Honest removal; the URL drops out |
A canonical tag is a request the search engine may overrule; a redirect is an instruction the browser obeys. The crawling and indexing guide explains how the two signals are weighed together and why a canonical pointing at a redirecting URL sends contradictory messages.
Chains, loops and irrelevant targets
Each of the common redirect faults wastes a different thing:
- Chains, where A redirects to B and B to C, cost an extra fetch at every hop and slow consolidation. They usually build up across successive migrations. The map should resolve every old URL to its final destination in one hop.
- Loops, where B redirects back to A, leave the page unreachable for everyone.
- Irrelevant targets, above all redirecting many removed pages to the home page or a broad category, are commonly treated as soft not-found errors. The signal is discarded and visitors land somewhere that does not answer what they were looking for.
- Removed redirects. Taking redirects down a few months after a move breaks every external link that still points at the old addresses.
Audits look for all four. The technical SEO audit guide lists redirect chains and loops among the findings a crawl surfaces, and the technical SEO audit checklist includes confirming that redirects resolve in one hop. When a page is published or its address changes, the web page publishing SOP checks that the live URL returns a normal response rather than a redirect.
Discontinued colours sent to the home page
Slabhollow Kitchen Worktops moves to a new platform with 1,400 old product URLs. The 1,080 current products are redirected one to one, but the 320 pages for discontinued colours are all redirected to the home page. Those pages had brought 2,100 organic visits a month from searches for named colours; within six weeks that falls to under 200, and Search Console reports many of the old URLs as soft 404s. The team rebuilds the map, sending each discontinued colour to the nearest current worktop in the same finish and adding a note on the product page. Ten weeks later the same searches bring 1,650 visits a month.
Related terms
Redirects and the Canonical TagTells search engines the preferred URL. are the two consolidation tools, one firm and one advisory. Every redirect hop is a fetch, so long chains draw on Crawl BudgetHow much a search engine crawls a site. on large sites. An Internal LinkA link from one page to another page on the same website. that points at a redirected URL sends its signal through a detour and should be updated, and each BacklinkA link from another site to yours. to an old address is the main reason the redirect has to stay in place for good.
Common confusions
- Permanent and temporary are not interchangeable. Temporary codes delay consolidation; use them only when the original URL really will return.
- A redirect is not a canonical. One moves visitors; the other only suggests which URL to index.
- Redirecting everything to the home page is not a migration. It is usually read as a soft error.
- Redirects are not a substitute for updated internal links. Fix the links the site controls.
- Redirects are not temporary scaffolding. Removing them later breaks the external links they protect.
