Redirects, 404 Pages and URL Management for SEO

Redirects send visitors and search engines from an old URL to a new one. Use a permanent redirect when a page has moved for good and a temporary one when the move is short-lived. Deleted pages should redirect only to a genuinely close match, otherwise return a real 404 or 410. A helpful 404 page, tidy URL structure and canonical tags complete good URL management.

Key takeaways

• URLs change during redesigns, migrations, renames and clean-ups, and every change can break links, bookmarks and search visibility unless handled.
• Permanent redirects are for moves that last; temporary redirects are for short-term situations.
• Avoid redirect chains and loops: each old URL should reach its final address in one step.
• Build a redirect map during any redesign before launch day.
• Not every deleted page needs a redirect; a redirect to an irrelevant page can confuse users and be treated as a soft 404.
• Keep URLs readable, lowercase and stable, and use canonical tags for duplicates.

Why do URLs change and what breaks when they do?

URLs change for ordinary reasons: a new platform, a rebranded product name, a restructured menu, a switch from http to https, or a clean-up of old content. When an address changes without a redirect:

• Links break. Other sites, emails, documents and social posts point to a page that no longer exists.
• Bookmarks fail. Returning visitors land on an error.
• Search visibility suffers. As far as is publicly documented, search engines need to find the new address and understand the relationship to the old one. Without a redirect, the signals attached to the old page, including links pointing at it, may not carry over smoothly.
• Analytics gets muddled. Traffic sources and landing pages split between old and new URLs.

See our guide to redesigning a website without losing SEO for the bigger picture.

What types of redirects exist, in plain words?

• Permanent redirect (commonly status 301, or 308). "This page has moved for good." Use it for restructures, renamed pages and domain changes. Search engines are told to treat the new address as the main one.
• Temporary redirect (commonly 302, or 307). "This page is elsewhere for now." Use it for short-term situations such as maintenance, a seasonal switch or testing.

As far as is publicly documented, search engines handle both permanent and temporary redirects, but permanent ones communicate intent more clearly. If you are unsure, ask whether the old URL is ever coming back.

What are redirect chains and loops?

A chain happens when A redirects to B, and B redirects to C. Each hop adds delay and a chance of failure. Chains build up over years of redesigns: http to https, then non-www to www, then an old path to a new path.

A loop happens when A redirects to B and B redirects back to A. Browsers give up and show an error.

Fix both by pointing every old URL directly at its current final destination. After each migration, update older rules rather than stacking new ones on top.

How do you build a redirect map during a redesign?

A redirect map is a spreadsheet pairing every old URL with its new one.

1. Collect old URLs. Combine a crawl of the current site, sitemap entries, analytics landing pages, Search Console data and backlink reports.
2. Prioritise. Mark pages with traffic, links or conversions.
3. Match each to its closest new equivalent. Same topic, same intent.
4. Decide the fate of the rest. Redirect, merge or retire.
5. Write the rules. Use exact matches first, then patterns for groups of similar URLs.
6. Test on staging. Crawl the old list and check where each URL ends up.
7. Launch and monitor. Recheck after going live.

What should you do with deleted pages?

Not every removed page needs a redirect.

• A close replacement exists: redirect to it.
• The topic is gone, but a parent category is relevant: a redirect to the category can work.
• Nothing relevant exists: let the old URL return a 404 (not found) or 410 (gone). As far as is publicly documented, both are acceptable ways of signalling that a page is no longer available.

Avoid sending everything to the home page. It is a poor experience, and search engines may treat mass redirects to irrelevant pages as soft 404s.

What is a soft 404?

A soft 404 is a page that looks like an error to users, for example "sorry, nothing found", but returns a normal success status to the browser. It confuses crawlers because the page says one thing and the server another. As far as is publicly documented, search engines try to detect these and may treat them as errors anyway.

Common causes are empty search results pages, out-of-stock pages with no content, and custom error pages that return a success response. Make sure real error pages return a real 404 status.

What makes a helpful custom 404 page?

Visitors who reach a 404 are lost, not hostile. Help them.

• A clear, friendly message that says the page was not found.
• A visible site search box.
• Links to popular pages and main categories.
• A way to contact you or report a broken link.
• Your normal header and footer so the page feels like part of the site.
• The correct 404 status code behind the scenes.

Keep it light, on brand and useful. Humour is optional; helpfulness is not.

What are good URL structure principles?

• Readable. Use words people can understand, not strings of numbers.
• Lowercase. Mixed case can create duplicate versions.
• Hyphens between words. Easier to read than underscores or run-together words.
• Stable. Pick a structure you can keep. Do not put dates or authors in paths unless they will always be correct.
• Short but descriptive. Include the topic, not every word of the title.
• Few parameters. Parameters for sorting, tracking and sessions create many near-duplicate addresses.
• Logical hierarchy. Folders that reflect real sections of the site, in line with your navigation and information architecture.

How does canonicalisation handle duplicate URLs?

The same page can often be reached at several addresses:

• With and without a trailing slash
• With and without www
• Over http and https
• With tracking or sorting parameters
• With different capitalisation

Choose one preferred version of each. Then:

1. Redirect the variants permanently to the preferred one, where practical.
2. Add a canonical tag on each page pointing to the preferred address.
3. Use that same version in internal links, sitemaps and structured data.

As far as is publicly documented, canonical tags are treated as a strong hint rather than a command, so consistency across your signals matters. Our technical SEO checklist for business websites places this among other basics.

How do you monitor errors in Search Console?

Google Search Console reports pages it could not fetch, pages that are not indexed and reasons such as not found, redirect errors and soft 404s. Review those reports after every release and monthly otherwise. Look for sudden increases, high-value URLs in the error list and redirect problems. Our Google Search Console guide for business owners explains the reports in plain language.

How do you manage redirects at scale?

• Server rules. Fast and central; changes usually need developer access and a deployment.
• CMS tools or plugins. Friendly for editors and good for one-off changes; keep them tidy as lists grow.
• CDN or edge rules. Redirects run close to the visitor and can be quick, but they add another place to look when debugging.

Pick one primary location and avoid scattering rules across several layers. Keep an owner, a naming convention and a review schedule.

How do you test and document redirects?

1. Crawl the list of old URLs and record the status code, the final destination and the number of hops.
2. Check that each ends on a 200 page with matching content.
3. Test with and without trailing slash, with www, and over http.
4. Document every rule: old pattern, new target, reason, date and owner.

An imaginary example

Imagine a furniture retailer moving to a new platform. The team exports old URLs from a crawl and Search Console, matches its product and category pages to new ones, and merges several near-identical categories into one. Discontinued items with no replacement are allowed to return 410. After launch, a crawl reveals a chain from an old http address through two earlier rules, so the team rewrites the rules to go straight to the final page.

Common mistakes

• Redirecting everything to the home page.
• Forgetting to update internal links after changing URLs.
• Serving a friendly error page with a success status.
• Changing URLs without a map.

FAQ

What is the difference between a 301 and a 302 redirect?

A 301 is a permanent redirect, saying a page has moved for good. A 302 is temporary, saying the page is elsewhere for now. Use permanent redirects for lasting changes and temporary ones only when the original URL will return.

Should I redirect deleted pages?

Only if a closely related page exists. Redirecting unrelated pages to the home page is a poor experience and may be treated as a soft 404. If nothing relevant exists, a proper 404 or 410 response is fine.

What is a soft 404?

A soft 404 is a page that tells users it does not exist but sends a normal success status to browsers and crawlers. Search engines may flag it as an error. Make sure real not-found pages return a true 404 status.

How many redirects in a row are too many?

Aim for one. Every extra hop adds delay and a chance of failure, and long chains are best cleaned up. Point each old URL straight at its final destination and update old rules when you add new ones.

Related reading

• Website Redesign Without Losing SEO: A Safe Migration Plan
• Google Search Console Guide for Business Owners
• Technical SEO Checklist for Business Websites

Next steps

Crawl your site, list URLs that return errors or pass through more than one redirect, and fix the worst offenders first. Check the Search Console error reports and write the redirect rules you rely on in one shared document. If you are planning a migration and want a second pair of eyes, contact us and our SEO services team can help.