The short answer
404 errors do not hurt your SEO on their own. Google removes a 404 page from its index, crawls it less over time, and does not penalise the rest of your site for it. A 404 only costs you when the dead URL still has value pointing at it: external backlinks, internal links from pages that matter, or an audience still arriving from search.
So the job is not to fix every 404. It is to find the few with inbound links or search demand and 301 redirect them to the most relevant live page, to turn soft 404s into honest status codes, and to leave the true orphans for Google to forget.
- Fix 404s with backlinks or internal links first. Ignore dead URLs nothing points to.
- A soft 404 returns 200 on an empty page. That wastes crawling: return a real 404 or 410.
- Use 301 for permanent moves, redirect to a relevant page, and keep the chain short.
Do 404 Errors Hurt SEO?
This is the question that drives most panic, and the answer is reassuring. Google's own documentation on HTTP status codes is clear: when a page returns 404, Google drops it from the index and stops processing it, and it gradually crawls that URL less and less. There is no site-wide penalty for having 404s. They are a normal part of how the web changes.
The damage is specific and local. It happens in three situations, and only these:
- The dead URL has external backlinks. Links still pass authority, and a link pointing at a 404 passes it into a dead end. Reclaiming that is the single highest-value 404 fix you can make.
- Your own pages link to it. An internal link to a 404 breaks the flow of link value through your site and drops a reader onto an error page. Both are easy to fix once you know the link exists.
- There are a great many of them. A huge volume of 404s, often from a broken template or a bad migration, can lead Google to spend crawl effort rediscovering dead URLs instead of your good ones.
Everything else is noise. A URL you deleted on purpose, that no one links to and no one searches for, returning a 404 is the system working as intended. As Google's John Mueller puts it in the video below, old 404s are generally fine to leave alone.
The 404 Triage Framework
Sort every 404 into one of four buckets by what points at it. Work from the top. Most of your list will land in P3.
URLs that still have external backlinks. Every day they 404, the authority those links pass lands on a dead end. 301 redirect each one to the most relevant live page.
URLs linked from your own high-traffic pages. The broken internal link breaks the flow of link value and sends readers to a dead end. Update the internal link, or 301 the old URL.
Soft 404s: pages returning 200 with no real content. Return a proper 404 or 410 for the ones that are genuinely gone, or add real content to the ones that should exist.
True orphan 404s with no backlinks, no internal links and no past search demand. Google will crawl them less and drop them on its own. No action needed.
Redirect Codes: Use the Right One
When a 404 does need fixing, the status code you return tells Google what you mean. Get this wrong and the fix either fails to pass value or gets ignored.
301
Moved Permanently
The default for any permanent URL change. Google treats the target as the new canonical and ranks it in place of the old URL.
302
Found (Temporary)
Only when you genuinely plan to restore the original URL. It does not pass the canonical signal, so it is wrong for permanent moves.
404 / 410
Not Found / Gone
For pages that are gone with no replacement. 410 (Gone) signals a deliberate removal and is dropped from the index a little faster than 404.
One rule ties these together: redirect to a page that actually replaces the old one. Google may treat a redirect to an irrelevant page, such as sending every removed URL to the homepage, as a soft 404, which means it gets ignored. When there is no good replacement, a clean 404 or 410 beats a forced redirect. For the mechanics of implementing redirects without creating chains or loops, see our guide to finding and fixing broken links.
How to Find the 404s That Matter
The goal is not a complete list of every 404. It is a short list of the ones with something pointing at them. This order gets you there with free Google tools.
- 1
Open the Page Indexing report
In Search Console, go to Indexing, then Pages. Open the 'Not found (404)' status to see the URLs Google discovered that now return 404, and open 'Soft 404' to see pages returning 200 with no real content.
- 2
Separate the 404s that have inbound links
A 404 only leaks value if something points at it. Check which dead URLs still have external backlinks or internal links from your own pages. Those are the ones worth a redirect.
- 3
Check for lost search demand
In the Performance report, look at whether the dead URL used to earn clicks and impressions. A URL people were still finding in search is worth redirecting to its closest live equivalent.
- 4
Fix in priority order, then confirm
Work P0 to P2 from the triage framework. After you redirect a URL, inspect the new target to confirm Google can crawl it, has indexed it, and picked the canonical you intended.
A Worked Example
Illustrative scenario, not client data
A store migrates to a new platform and the Page Indexing report lights up with 2,400 new "Not found (404)" URLs. The instinct is to redirect all 2,400. Here is how triage changes that.
- Backlinks (P0): a check of the dead URLs against known inbound links turns up 18 old product pages with editorial links from blogs and press. Each one gets a 301 to its closest live product or category. Eighteen fixes, most of the value.
- Internal links (P1): around 60 of the dead URLs are still linked from the main navigation and a few popular guides. The links are updated to point at the new URLs. No reader hits an error from inside the site.
- Soft 404s (P2): about 200 emptied category pages return 200 with a "no products" message. They are set to return a real 404, so Google stops treating them as live pages.
- The rest (P3): the remaining ~2,100 URLs are discontinued variants and filter pages with no links and no search history. They are left to 404. Google crawls them less each week and drops them.
The list of 2,400 became roughly 80 real fixes plus one status-code change applied in bulk. That is the difference between a week of busywork and an afternoon that actually protects rankings. A ranking drop that traces back to a bad migration is one of the patterns in our guide to why Google rankings drop.
Which dead URLs still had demand?
Murkuz connects Google Search Console to Claude or ChatGPT, read-only. Ask Search_Analytics which of your URLs were still earning impressions before they broke, so you redirect those first. After you 301 one, point Inspect_URL at the new target to confirm Google has indexed it and picked the canonical you intended.
A Quick Self-Check
Every 404 URL with external backlinks has a 301 to a relevant live page, not the homepage.
No internal links on your site point to a 4xx URL.
No 200-status pages appear under Soft 404 in the Page Indexing report.
Redirect chains are collapsed: old URLs point straight to the final destination.
Genuinely deleted pages with no replacement return a clean 404 or 410, not a forced redirect.
Further reading
The best material on how Google handles broken pages and redirects, starting with Google's own documentation. Each one was read in full before it was added here.
Official docsGoogle Search Central
How HTTP status codes, and network and DNS errors, affect Google SearchThe primary source for how Google treats 404, 410, 301 and 302, and how it detects soft 404s. Confirms Google simply drops 404 pages from the index and crawls them less over time.
Official docsGoogle Search Central
Redirects and Google SearchExplains that a permanent (301) redirect tells Google the target should be canonical, that a temporary (302) does not, and that server-side redirects have the best chance of being read correctly.
Official docsGoogle Search Console Help
Page Indexing reportWhere you actually find your 404s. Shows the 'Not found (404)' and 'Soft 404' statuses, what triggers each, and that Google keeps retrying a discovered 404 for a while.
GuideAhrefs
301 Redirects: What They Are and How to Use ThemA practitioner walkthrough of implementing 301s and the nine mistakes that undo them, including redirect chains and loops. Useful when you move from deciding what to fix to actually fixing it.
GuideConductor
HTTP Status Codes: The Complete Guide for SEOA clear overview of every status code that matters for SEO, with a good section on why soft 404s are a problem for large sites that return 200 on empty pages.
Frequently Asked Questions
Do 404 errors hurt your SEO?
Usually no. Google removes 404 pages from its index and gradually crawls them less often, and it does not penalise the rest of your site for having them. The damage is specific: a 404 on a URL that still has backlinks wastes the authority those links pass, a 404 on a page your own site links to breaks the flow of internal link value, and a very large number of 404s can slow how often Google crawls your good pages. Fix the 404s that have inbound links or real search demand, and leave the true orphans alone.
What is a soft 404 and why does it matter?
A soft 404 is a page that returns a 200 OK status code but shows a 'not found' message or has effectively no content, such as an empty search result or a category page after every product was removed. Google flags these in the Page Indexing report because the status code says 'this page is fine' while the page says the opposite. That mismatch wastes crawling on near-empty URLs. The fix is to return a real 404 or 410 status for pages that are genuinely gone, or to add real content if the page should exist.
When should I use a 301, a 302 or a 410?
Use a 301 (moved permanently) when a page has a new home for good: Google treats it as a signal that the new URL should be the canonical one and should appear in search. Use a 302 (found, temporary) only when you genuinely intend to restore the original URL, because it does not pass that canonical signal. Use a 410 (gone) when a page is deleted with no replacement and you want it dropped from the index a little faster than a plain 404. Using a 302 for a permanent move is the most common mistake here.
How do I find the 404 errors that actually matter?
Start in Google Search Console under Indexing, then Pages, and open the 'Not found (404)' status to see the URLs Google discovered that now return 404. Prioritise by consequence, not by count: fix URLs that still have external backlinks first, then URLs linked from your own high-traffic pages, then soft 404s. A dead URL with no inbound links and no past search demand is safe to leave: Google will stop crawling it on its own.
Does a 301 redirect pass ranking signals to the new page?
Yes. Google treats a permanent redirect as a strong signal that the redirect target should be canonical, so the new URL is the one that appears in search. Google also says server-side redirects have the highest chance of being interpreted correctly, so prefer those over client-side tricks. Keep the chain short: redirect the old URL directly to the final destination rather than hopping through several URLs, which slows crawling and can dilute the signal.
Should I redirect every removed page to my homepage?
No. Redirect each removed URL to the most topically relevant live page, for example an old product to its replacement or its category. Google may treat a redirect to an irrelevant page, such as sending every deleted URL to the homepage, as a soft 404, which means the redirect is ignored and the original URL is treated as not found anyway. When there is no relevant replacement, it is better to return a clean 404 or 410 than to force an unrelated redirect.