A 404 error does not hurt your site’s rankings on its own. Google’s own documentation says 404s “are not necessarily a problem” for pages removed without a replacement, and its search advocate John Mueller has called 404s and 410s “how the web is supposed to work.” What costs you is a narrower set: dead pages that still have backlinks, traffic or internal links. Below: what Google does with a 404, whether 410 vs 404 matters (barely), a five-question triage for what to fix, and the redirect habit that turns harmless 404s into soft 404s.
Do 404 Errors Hurt SEO? The Short Answer
No, not as a site-wide signal. Mueller made the point again on Reddit (reported by Search Engine Roundtable in January 2026): “404s/410s are not a negative quality signal.” A site with thousands of 404s in Search Console isn’t being marked down for them. Every site that removes pages picks up 404s, and Google expects that.
The damage, when there is any, is local to the URL: backlinks that now point at nothing, visitors landing on an error page, and broken internal links that send users and crawlers into dead ends. So the useful question isn’t “how do I get my 404 count to zero?” It’s “which of these 404s still carry links, traffic or internal links?” Usually that’s a small share of the list.
What Google Actually Does With a 404 (and a 410)
Google’s Search Central page on HTTP status codes is short on this point. It says “All 4xx errors, except 429, are treated the same.” Googlebot tells the indexing system the content doesn’t exist, and “the indexing pipeline removes the URL from the index if it was previously indexed.” For a URL Google has never indexed, “Newly encountered 404 pages aren’t processed. The crawling frequency gradually decreases.”
- A 404 removes a page from Google once it’s recrawled. You don’t need a noindex tag on it as well.
- 404s don’t slow your crawling. The same page says 4xx codes other than 429 (“too many requests”) “have no effect on crawl rate.”
- Google keeps checking. Mueller said in a March 2026 Reddit reply, reported by Search Engine Journal: “They’ll be recrawled for potentially a long time, a 410 won’t change that.” Old 404s in your crawl stats are normal.
410 vs 404: Does the Difference Matter?
404 means “not found”; 410 means “gone, on purpose.” SEOs used to say 410 deindexes faster. Google’s documentation lists both under the same rule, and Mueller has said “The difference in processing of 404 vs 410 is so minimal that I can’t think of any time I’d prefer one over the other for SEO purposes.”
| Question | 404 Not Found | 410 Gone |
|---|---|---|
| What it tells the client | Nothing exists at this URL | Something existed and was removed on purpose |
| Google indexing | Removed from index once recrawled | Same: Google treats all 4xx except 429 alike |
| Stops Google recrawling? | No, crawling slows over time | No, per Mueller (2026) |
| Default on most platforms | Yes (WordPress, Shopify and most frameworks) | Needs a server rule or plugin |
| When to use it | Almost always | Planned bulk removals or post-hack spam cleanup, where a deliberate record helps |
Our rule: if your platform already returns 404, leave it. Use 410 when it’s cheap and documents intent, not for a ranking effect.
When a 404 Error Actually Costs You
Four cases are worth your time.
- The page moved and nobody redirected it. The classic migration loss: a new URL structure or renamed category leaves old, linked, ranking URLs returning 404 (see our site migration SEO guide).
- The dead URL has backlinks from other sites. Links you earned point at nothing. A redirect to the right replacement gets that value back.
- You link to it yourself. Broken internal links in navigation, related-post widgets or old articles are fully in your control. Google’s help center says: “we recommend fixing only 404 errors that you link to yourself or list in a sitemap.”
- Your sitemap lists it. A sitemap should only contain live, indexable URLs.
What doesn’t cost you: URLs invented by scrapers, mistyped links on spammy sites, query strings bots made up, and pages you deleted on purpose that never had links. For those, 404 is the right answer.
Where to Find the 404s That Matter
- Search Console’s Page indexing report. The “Not found (404)” reason lists URLs Googlebot requested and got a 404 for. It’s noisy, because Google also finds URLs through external links and old sitemaps. Treat it as what Google is asking for, not a to-do list. See our index coverage guide for the other statuses.
- A site crawler. A crawl follows your own links, so every 404 it finds is a broken link you created, along with the page that contains it. This is the highest-value list.
- A backlink tool. Filter your pages by 404 status and sort by referring domains to see where redirects would win back link value.
- GA4 and server logs. Filter GA4’s Pages and screens report by your error page’s title to see which missing URLs real visitors hit. Logs show everything Googlebot requests.
Bing Webmaster Tools has its own Site Scan crawl that reports errors. SEO Rocket’s site audit runs a real crawler (Screaming Frog-backed, with JavaScript rendering) and lists 404s, broken internal links with their source pages, redirect chains and soft-404 candidates in one pass.
The 404 Triage: Five Questions per URL
Run each dead URL through these in order and stop at the first that applies. It’s part of a playbook proven across 1,000,000+ ranking pages, and it stops reflex redirects.
- Should this page exist? If a deploy broke it or it was removed by accident, restore it.
- Is there a true replacement? A successor product, an updated guide, the same content at a new URL. 301 to that exact page.
- Do you link to it yourself? Fix the internal link at the source. A redirect hides the problem; fixing the link removes it.
- Does it have external links or recent traffic? With no exact replacement, redirect to a closely related page that meets the same need. If nothing is genuinely related, let it 404.
- None of the above? Leave it as a 404 (or 410). That’s correct behaviour, not an error to clear.
| Situation | Correct response |
|---|---|
| Page moved to a new URL | 301 to the new URL; update internal links |
| Discontinued product with a direct successor | 301 to the successor |
| Product temporarily out of stock | Keep it live (200) and show availability; don’t 404 it |
| Deleted thin post, no links, no traffic | 404 or 410 |
| Typo URL linked from another site | 301 to the intended page if the link is worth having |
| Junk URLs made up by bots | Ignore; 404 is correct |
Use our free 301 Redirect Generator to write the rules once you know which URLs need them.
A Worked Example: Triaging a Search Console Export
An illustrative case: a store retires a product line and restructures its blog. Weeks later, Search Console’s “Not found (404)” list runs into the hundreds and the owner wants to redirect everything to the homepage.
Instead, export the list and add three columns: referring domains (from a backlink tool), organic clicks over the past year (Search Console performance data), and whether a crawl finds internal links to the URL. The pattern is usually lopsided. Most rows have nothing: old parameter URLs, scraper typos, posts deliberately removed. They stay 404. Old blog URLs with backlinks map one-to-one to their new slugs and get 301s. Retired products with obvious successors get 301s to them. Internal links the crawl found get fixed at the source. The result is a short list, not hundreds of redirects, and the leftover 404s in the report are fine.
The Redirect-Everything Mistake
Redirecting every 404 to the homepage looks tidy, and it’s the most common way to make things worse. Mueller has called homepage redirects “not a great practice (confuses users)” and said Google mostly treats them as 404s anyway: “they’re soft-404s, so there’s no upside.” You lose the clear 404 signal, the visitor lands somewhere that doesn’t answer their question, and no link value moves. His advice: make a better 404 page instead.
Also avoid chains: if A redirects to B, which redirects to C, point A straight at C. Google follows up to 10 hops, but every hop is another point of failure. Turn off plugins that send all 404s to the homepage. Redirects should be decisions about specific URLs. Our soft 404 guide covers how Google decides a page is a soft 404.
404 Page SEO: Build a Custom Page That Returns the Right Code
A good 404 page is mostly a usability job, with one rule you can’t break: it must return a 404 status code. A friendly “Sorry, we couldn’t find that” page served with a 200 status is the definition of a soft 404.
- A plain statement that the page doesn’t exist, without blaming the visitor.
- A search box, the most useful element, because the URL already hints at what they wanted.
- Links to your main categories or most-used pages, chosen for people.
- Your normal header and footer, plus optionally a way to report the broken link.
To test it, load a URL that doesn’t exist and check the status in your browser’s developer tools (Network tab), or use URL Inspection in Search Console. If either shows 200, fix the server before the design.
Five Technical Traps That Turn 404s Into Soft 404s
- Apache ErrorDocument pointing to a full URL. Apache’s own documentation warns that an
ErrorDocumentset to a remote URL sends a redirect, so “the client will not receive the original error status code.” Use a local path. Our .htaccess Generator helps with the surrounding rules. - Single-page apps returning 200 for every route. A client-side router can’t change the HTTP status. Google’s JavaScript SEO guidance offers two fixes: a JavaScript redirect to a URL the server answers with a 404, or adding a
noindexrobots meta tag to error views. Better still, return a real 404 from the server or edge. - Empty templates for missing records. A product URL with a bad ID renders an empty template with a 200. Return 404 when the record doesn’t exist.
- Blocking dead URLs in robots.txt. If Googlebot can’t fetch the URL, it never sees the 404. Let it crawl.
- Using 404 for pages you still want. A seasonal or out-of-stock page that 404s for a few weeks drops out of the index and has to win its rankings back.
404 Error FAQ
How many 404s are too many?
There’s no threshold. Google doesn’t count 404s against a site. What matters is whether any of them have links, traffic or internal links pointing at them. One broken URL with strong backlinks matters more than thousands of junk ones.
How long until Google drops a 404 page from search?
Google removes it once it recrawls the URL and sees the 404. How fast that happens depends on how often Google crawls that page. For urgent cases, Search Console’s Removals tool hides a URL quickly, but Google says those requests last only about six months, so the page still needs to return 404 or 410.
Should I fix 404s from other sites’ broken links?
Only when the link is worth keeping. If a relevant site links to a mistyped or old URL, a 301 to the right page captures it. Reclaiming those links is also a link-building tactic in its own right; see broken link building.
Do 404s waste crawl budget?
For most sites, no. Google says 4xx codes don’t affect crawl rate. See our crawl budget guide for when crawl budget starts to matter.
Make 404 Checks Routine, Not a Cleanup Project
Teams that never have a 404 problem still have 404s. They catch the expensive ones early. Check internal links after every deploy, map redirects before any URL change, and review linked 404s once a quarter. Re-running SEO Rocket’s site audit after each release means the crawler, not a client email, tells you when a template starts returning soft 404s. A 404 error is a status code doing its job. Make sure it only shows up where a page really is gone.
