404 page SEO is how a website handles addresses that no longer lead anywhere: which HTTP status code the server returns, what the visitor sees, and whether the old address should send people somewhere else. A 404 on its own does not lower your rankings. Handling it badly does cost you, through lost links, lost visitors and pages Google keeps crawling for nothing.
This guide sets out what Google actually does with a 404, based on its own documentation, then shows what we found when we tested our own site, serpwize.com, on 4 October 2026. Two of the twelve checks failed. We fixed both before publishing this page.
What Google Does When It Finds a 404
Google’s documentation on how HTTP status codes affect its crawlers, last updated in February 2026, is clear on the main points:
- The page drops out of the index. A URL that returns a 404 is not indexed, and one that was indexed is removed.
- Google keeps checking, less often. Crawling of that URL slows down gradually rather than stopping at once.
- Crawl rate for the rest of your site is unaffected. Client errors in the 4xx range, apart from 429, have no effect on how fast Google crawls the site.
- 404 and 410 are handled the same way. All 4xx codes except 429 tell Google the content does not exist.
So a site with some 404s is normal. The damage comes from three specific situations: a removed page that still had backlinks or traffic, a broken internal link that visitors keep hitting, and an error page that returns the wrong status code.
We Tested Our Own Site: 12 Checks, 2 Failures
We requested twelve kinds of address on serpwize.com and recorded the status code each one returned. This is the same check we run during a technical audit for a client.
| Address tested | Result before | Verdict |
|---|---|---|
| A page that never existed | 404, noindex | Correct |
| A blog post that never existed | 404, noindex | Correct |
| A service page that never existed | 404, noindex | Correct |
| Address without the trailing slash | 301 to the slash version | Correct |
| Same address in capital letters | 301 to the lower-case version | Correct |
| Image attachment page | 404 | Correct |
| Author archive (disabled) | 404 | Correct |
| Feed for a post that does not exist | 404 | Correct |
| Site search with no results | 200, noindex | Acceptable |
| Empty category archive | 200, noindex | Acceptable |
| /blog/page/999/ and other made-up page numbers | 200, indexable | Failed |
| What a visitor sees on the 404 page | One line of default theme text, no search, no links | Failed |
Failure 1: endless duplicate page numbers
WordPress answered any page number added to a normal page, such as /blog/page/999/ or /about/page/3/, with a working page and a 200 status. Each one was a copy of the real page. A canonical tag pointed back to the original, which limited the harm, but crawlers could still find an endless supply of duplicate addresses. These now redirect to the real page. The one section that genuinely uses page numbers, our case studies list, was left alone.
Failure 2: a 404 page that left visitors stranded
Our theme’s default 404 page said the page could not be found and offered nothing else apart from the main menu. We replaced it with a page that still returns a 404 status but includes a search box, links to the six most useful sections of the site, and a way to report a broken link.
An earlier audit of the same site, on 30 September 2026, found the other common problem: 16 internal links in “related reading” boxes pointing at pages that no longer existed. Each was replaced with a link to a live, relevant article. Internal broken links are the easiest 404s to fix and the most avoidable, because the dead link is on your own site.
Soft 404s: The Error Google Reports Most Often by Mistake
A soft 404 is a page that tells visitors nothing is there but returns a 200 status, which tells crawlers everything is fine. Google’s documentation says that when a 200 response looks like an error, an empty page or an error message, Search Console flags it as a soft 404. Common causes:
- A “not found” message built into a normal page template instead of a real 404 response.
- Empty category, tag or search result pages that are open to indexing.
- Out-of-stock or expired listings that keep the page but remove all the content.
- Redirecting every removed page to the home page. Google can treat a redirect to an unrelated page as a soft 404, so the redirect gains you nothing.
The fix is to return the status code that matches reality: 404 or 410 if the content is gone, a 301 to a close replacement if one exists, or real content if the page should stay.
Redirect, 404 or 410? How to Decide
| Situation | Best response |
|---|---|
| Page moved, or a close replacement exists | 301 redirect to the replacement |
| Page removed, but it has backlinks and a related page exists | 301 redirect to the related page |
| Page removed, no replacement, little traffic or links | 404 |
| Page removed for good and you want that stated explicitly | 410 (Google treats it the same as 404) |
| Product temporarily out of stock | Keep the page with a clear status and alternatives |
| Address that never existed | 404 |
The test for a redirect is relevance. Someone who clicked a link to a guide on keyword difficulty should land on a guide about keyword difficulty, not on the home page. If nothing on your site matches, an honest 404 with good navigation serves them better. Our guide to 301 redirect problems covers the redirect side in more detail.
What a Good 404 Page Includes
- A real 404 status code. Check it in your browser’s developer tools or any HTTP status checker. A good-looking page that returns 200 is a soft 404.
- A plain message. Say the page does not exist and suggest why: an old link or a typo.
- A search box. Visitors who arrived with a specific need can look for it directly.
- Links to your most useful pages. Main services, the blog or guides, and contact details.
- Your normal header and footer. The visitor should still recognise where they are.
- A way to report the broken link. Visitors often find broken links before your tools do.
Whether the 404 page also carries a noindex tag matters little. The 404 status already keeps it out of the index. WordPress SEO plugins such as Yoast add noindex anyway, which does no harm.
How to Find 404 Errors on Your Site
- Google Search Console. The Page indexing report lists URLs under “Not found (404)” and “Soft 404”. Look first at URLs that had traffic or that you link to yourself.
- A site crawl. A crawler such as Screaming Frog follows every internal link and lists any that end in a 404, with the page holding the broken link.
- Your backlink tool. Filter for links from other sites pointing at pages that now return 404. These are the ones most worth redirecting.
- Analytics. In GA4, filter page views by the page title of your 404 page to see which missing addresses visitors actually request.
- Server logs. Logs show every 404 request, including those from crawlers, with the referring page where available.
Frequently Asked Questions
Do 404 errors hurt SEO?
Not directly. Google states that 4xx responses do not affect how it crawls the rest of a site, and a missing page simply drops out of the index. The cost is indirect: links pointing at the dead page stop helping you, and visitors who land there may leave. Fix the 404s that had traffic, links or internal links, and leave the rest.
Is a 410 better than a 404 for SEO?
For Google there is no practical difference. Its documentation says all 4xx status codes except 429 are treated the same way. Use 410 if you want to state that removal is permanent, but do not expect faster de-indexing from it.
Should I redirect all 404 pages to my home page?
No. Google can treat a redirect to an unrelated page as a soft 404, so you keep the problem and add a confusing experience for visitors. Redirect only to a page that genuinely replaces the missing one.
What is a soft 404?
A page that shows an error or empty content but returns a 200 success status. Search Console reports these separately in the Page indexing report. Fix them by returning a real 404 or 410, redirecting to a relevant page, or adding the content the page is meant to have.
How many 404 errors is too many?
There is no safe or unsafe number. A large site can have thousands of harmless 404s from old campaigns or mistyped links. What matters is whether any of them are pages people link to, visit or reach from your own navigation.
Should a custom 404 page return a 200 status?
No. The page can look like any other page on your site, but the server must still send a 404 status, otherwise search engines treat it as a real page and may report it as a soft 404.
Want the same checks run on your site? Status-code testing is part of the free audit behind our professional SEO services.
