Redirects: 301, 302 and why the difference matters
A redirect sends visitors and search engines to a new address. Permanent or temporary changes what search keeps. Which to use, and the chains to avoid.
The short answer
A redirect is an instruction that says “the thing you asked for is now over here” and sends the visitor on automatically. The number attached to it tells search engines how to treat the old address. A 301 means the move is permanent: transfer everything you knew about the old address to the new one. A 302 means temporary: keep the old address, this is a detour. Using the wrong one, or none, is how businesses lose years of search standing in an afternoon.
Permanent or temporary
| 301 or 308 (permanent) | 302 or 307 (temporary) | |
|---|---|---|
| Says | This page has moved for good | This page is elsewhere for now |
| Search engines | Transfer the old address’s standing to the new one and index the new one | Keep the old address indexed and treat the new one as a detour |
| Use for | Rebuilds, renamed pages, retired products pointing to their replacement, domain changes, http to https, www variants | Maintenance detours, short campaigns, A/B tests, geographic detours |
| Mistake | Using it for something that will move back | Using it for a permanent move, so the new page never inherits anything |
When a site is rebuilt
Every address that existed before either keeps its address or gets a permanent redirect to the one specific page that replaced it. Retired pages redirect to the closest relevant page, not to the homepage. The collection of those decisions is the redirect map, and it is the document that protects the search visibility the old site earned.
- List every old address from the sitemap, Search Console and server logs.
- Decide the destination for each: same address, a specific new page, or the closest relevant page for retired content.
- Generate the redirects from the map, as configuration that deploys with the site.
- Test after launch: every old address returns a 301 straight to its final destination, no chains, no loops.
- Watch Search Console for weeks: crawl errors and redirect issues surface there first.
Chains and loops
A redirect that points to a page that redirects again is a chain. Each hop is a round trip for the visitor and a small loss of standing for search. Chains form when a target moves and the old redirect is not updated. Fix them by pointing every redirect at the final destination. A loop, where addresses redirect to each other, breaks the page entirely and shows up as an error in browsers; it is usually a migration mistake.
What this means for you
If addresses change, for a rebuild, a rename or a new domain, insist on a redirect map, permanent redirects for everything that moved, direct destinations with no chains, and a test afterwards. If your traffic dropped after a change, missing or temporary redirects are the first thing to check, and correcting them recovers most of what was lost.
Frequently asked questions
We changed our page addresses and traffic dropped. Why?
Almost certainly missing or temporary redirects. Search engines had years of standing attached to the old addresses; without a permanent redirect to the new ones, that standing was not transferred and the old addresses now return errors. Adding correct 301 redirects recovers most of it, the sooner the better.
Is a redirect bad for speed?
One redirect is a small extra round trip. A chain of them, old address to newer address to newest, is several round trips and a sign nobody maintained the map. Redirect directly to the final address, and update old redirects when a target moves again.
Where should redirects live?
In a configuration that deploys with the site, generated from a maintained map, and enforced at the edge so they are fast. Not in a plugin that can be disabled, and not scattered across the content system where nobody can list them.
Sources
- Google Search Central: Redirects and Google Search (accessed 2026-09-11)
- MDN Web Docs: Redirections in HTTP (accessed 2026-09-11)