Key takeaways
- "Not found (404)" means Google requested a URL it discovered on its own, without a sitemap or indexing request, and the server said the page does not exist
- Google's own documentation says 404s are normal, do not harm the rest of the site, and need no action for pages that are gone with no replacement
- The cases that need fixing are a page you still want, a page that moved and should redirect, and a URL your own site still links to
- Redirecting every 404 to the homepage is worse than leaving it; Google treats that as a soft 404
- A crawl finds the broken internal links that keep sending Google to dead URLs, which Search Console does not show
"Not found (404)" is the status people most often try to drive to zero and least often need to. Google's own documentation is unusually direct that a 404 is a normal part of the web and that most of them need nothing from you. But a minority do, and the report does not sort them for you. This post explains what the status reports, which 404s to leave, the three cases to fix, and how to find the broken links on your own site that cause them, which is the part Search Console cannot show.
What "Not found (404)" in Search Console means
Google's definition, from its Page indexing report documentation:
"This page returned a 404 error when requested. Google discovered this URL without any explicit request or sitemap."
The second sentence matters. Google found the URL on its own, through a link somewhere or its own memory of a page that used to exist. If the URL had been in your sitemap, the report would say so in the sitemap-specific variant of the status. Either way, the server answered correctly that the page is not there.
Note
This post is one of a series on the Page indexing report. Search console indexing issues covers every status and which ones to act on.
What Google says about 404s
Google's guidance has been consistent for years: a 404 for a page that does not exist is the right response, 404s do not count against the site, and there is no need to fix them unless the URL is one you care about. Googlebot will recheck known 404s occasionally, with decreasing frequency, and drops them from the index.
That means the right default is to leave this status alone. The report can list hundreds of 404s on an old site, from pages deleted years ago, misspelt links on other sites, and URLs that never existed, and none of them need touching.
The three cases that need fixing
1. A page you still want
If a URL under this status is a page that should be live, something broke it: a deleted post, a changed slug, a CMS migration that dropped a path, a case-sensitivity change on a new server. Restore the page at its original URL, or if it has moved, redirect the old URL to the new one.
2. A page that moved and should redirect
When a page has been replaced by another, a 301 redirect from the old URL to the new one preserves the traffic from every link that still points at the old address and passes most of its ranking signals. Redirect to the closest equivalent: the new URL for the same page, or the parent category if the page itself is gone. Point it at the final address, not at another redirect, or you start a redirect chain. The page with redirect status is what a correctly redirected old URL looks like in the report afterwards.
3. A URL your own site still links to
A 404 that external sites link to is their broken link. A 404 that your own navigation, content, sitemap or canonical tags link to is yours, and every crawl of your site walks Googlebot into it. These are the 404s worth finding first, because they are the ones you can fix completely.
The report does not list the referring pages for a 404 in one place; inspecting each URL shows a referring page where Google has one. On a site of any size that is where a crawl of your own site is faster.
What not to do
- Do not redirect every 404 to the homepage. Google's documentation calls this a soft 404 and treats it as such: a page that says "not found" in effect while returning a 200 or a redirect. Visitors who followed a link to a specific page get the homepage instead, which is a worse experience than an honest 404.
- Do not return 200 for a "page not found" message. That is a soft 404 too, and it gets the error page indexed. The soft 404 issue page covers why a site that does this at scale creates a crawl trap.
- Do not block 404 URLs in robots.txt. It stops Google seeing that the page is gone, which keeps the URL in Google's memory longer, not shorter.
- Do not mark every 404 as fixed in the report. Validation asks Google to recrawl the URL. For a page that is genuinely gone, that confirms nothing and delays nothing.
Tip
Make the 404 page useful. A clear message, a search box and links to the main sections turn a dead end into a second chance. It does not change what Google does, but it keeps the visitor.
404 versus 410
A 410 Gone tells crawlers the removal is deliberate and permanent. Google treats 404 and 410 almost identically, and its documentation says both lead to the URL being dropped. A 410 may be processed slightly faster. If your platform makes it easy to return 410 for deliberately deleted pages, use it; if not, a 404 is fine.
How to work through the list
- Export the URLs from the status in Search Console.
- Sort out the ones with traffic or links. Cross-reference with the Links report and with analytics for the old URLs. Those get redirects.
- Check each remaining URL against your current site. A URL that is a near-match for a live page, a changed slug or a moved section gets a redirect. A URL for a page that is gone with no replacement stays a 404.
- Find and fix your own links to any of them. This is the step the report cannot do for you.
- Leave the rest. They clear from the report on Google's schedule as it rechecks them.
Finding the broken links behind the 404s
Search Console tells you which URLs returned 404 when Googlebot requested them. It does not tell you which of your own pages sent Googlebot there, and on your own site that is the only fix that is both complete and under your control. The full routine is in how to find and fix broken links across your site.
Crawl Cove crawls every page on your site and follows every link. The broken-internal-link check reports every page that contains a link to a URL returning 4xx or 5xx, with each dead destination listed, so you can see both which URLs are broken and exactly which pages need their links changed. The status-code check lists every crawled page that returned a 404 or another error, and malformed-links catches hrefs that are not valid URLs at all, which produce 404s for addresses that never existed. sitemap-non-indexable-url flags sitemap entries that return a 404, so Google is not being invited to dead pages. For the site-level case where the server returns 200 for anything, soft-404-trap fires once rather than reporting every fake page.
The Live Index Status panel calls Google's URL Inspection API for any page and reports the page fetch state and last crawl, so you can confirm a restored page has been seen. Run a full SEO audit to see every broken internal link on the site, with every page it lives on, in one crawl. For a quick check of a single page's links, the broken link checker on our tools page tests every link on one page, with nothing to install.
Wrap-up
"Not found (404)" is mostly Google telling you about pages that are gone, and Google itself says that is fine. Spend your time on the three exceptions: pages you still want, pages that moved and should redirect, and URLs your own site still links to. Fix those with restores and single-hop redirects, never with a blanket redirect to the homepage. Then leave the rest, because the only thing wrong with a 404 for a page that no longer exists is the time spent worrying about it.
Frequently asked questions
- What does "Not found (404)" mean in Search Console?
- Google's definition is that the page returned a 404 error when requested, and that Google discovered the URL without any explicit request or sitemap. Google followed a link, old or new, to a URL that no longer exists.
- Do 404 errors hurt my SEO?
- Google's documentation says no. A 404 for a page that is gone is the correct response and does not affect the ranking of other pages. It matters only when the missing page is one you want, or when your own site still links to it.
- Should I redirect every 404 to the homepage?
- No. Google treats a redirect to an unrelated page as a soft 404, and visitors who followed a link to a specific page get dropped somewhere else. Redirect to the closest equivalent page when one exists; otherwise leave the 404.
- Why does Google keep crawling a URL I deleted months ago?
- Because something still links to it, on your site or elsewhere, or because Google periodically rechecks known URLs in case they return. The report shows the last crawl; it does not mean Google is wasting much effort on it.
- Should I use 410 instead of 404?
- Google treats both as "gone". A 410 is slightly more explicit and may be dropped a little faster, but the difference is small. Use 410 if your platform makes it easy for deliberate deletions; do not worry about it otherwise.