What Is a 404 Error? (And How to Fix It)

What is a 404 error and what causes it

The short answer to what is a 404 error is that it is the HTTP status code a web server returns when it can reach your site but cannot find the specific page or URL that was requested. The server is effectively saying "this domain exists, but there is nothing at this address." You have seen the result: a 404 Not Found page instead of the content you expected. The 404 is one of dozens of HTTP status codes, and it belongs to the 4xx family that signals a client-side problem — the request pointed at something that is not there.

It helps to separate the two halves of a URL request. The server responds 200 OK when the page loads normally, 301 or 302 when the page has moved, 500 when the server itself breaks, and 404 when the path simply does not resolve to a page. That distinction matters for SEO because Google treats each code differently: a 404 tells Google "stop expecting a page here," while a 301 tells it "the page moved, follow me."

Most 404 errors trace back to one of a few everyday causes:

- A deleted page — you removed a post or product and nothing was put in its place.

- A changed URL — you edited a slug, switched CMS, or restructured folders, and the old address no longer exists.

- A broken internal link — a link on your own site points to a page that has since moved or been removed.

- A mistyped or malformed link — a typo in an internal link, a bad link on another site, or a hand-edited URL.

- An expired or seasonal page — a sale, event, or listing that was taken down on purpose.

Not every 404 is your fault, either. Bots, scrapers, and old bookmarks constantly request URLs that never existed, and browsers routinely ask for files like /favicon.ico. Those requests generating 404s are completely normal background noise, not a problem to chase.

Soft 404s: the version Google actually dislikes

A soft 404 is a page that looks like a "not found" page to users but returns a 200 OK status code to search engines instead of a real 404. In other words, the content says "sorry, this page doesn't exist," but the server tells Google "everything is fine, this is a valid page." That mismatch is the version of the problem Google explicitly flags in Search Console, because it wastes crawl effort and pollutes the index with empty pages.

A true 404 is honest — it tells Google the page is gone. A soft 404 lies about it, and that is exactly why Google treats soft 404s as a quality issue worth reporting.

Soft 404s usually come from thin or empty pages that technically load: an out-of-stock product page with no content, an empty search-results or category page, a "no results found" screen, or a JavaScript app that renders an error message while the server still returns 200. Google's own algorithms try to detect these and will label them Soft 404 in the Pages report even though your server never sent a 404 code.

The fix is to make the status code match reality. If the page is genuinely gone, return a real 404 (or 410 Gone). If the page should exist but is thin, add real content or redirect it to a relevant page. Never leave a dead page returning 200 — that is the outcome Google penalizes most. You can confirm what code a URL actually returns by running it through the free SEO + GEO audit, which reports the real HTTP status alongside other technical issues.

Do 404 errors hurt SEO?

A few 404 errors do not hurt your SEO — Google expects them and treats them as a normal part of the web. Pages get removed, URLs change, and links rot; Google's documentation states plainly that returning a 404 for a page that no longer exists is the correct behavior and does not harm the rest of your site. So the common panic over a handful of 404s in Search Console is misplaced.

The real problems are specific and worth taking seriously. Mass 404s — hundreds or thousands appearing after a botched migration or CMS switch — signal that something broke, waste crawl budget on dead URLs, and can leave real content unreachable. Lost link equity is the other one: if a 404ing page had valuable backlinks or strong internal links, that ranking authority evaporates unless you redirect the URL to a live page. And broken internal links pointing at 404s hurt both crawlability and user experience, which is why finding and fixing them is a standard technical SEO task.

There is also a user-experience cost that indirectly affects rankings. A visitor who hits a dead-end 404 with no navigation and no helpful next step usually bounces, and a pattern of that behavior is a weak negative signal. A well-designed custom 404 page that offers search, popular links, and a path back into the site turns a dead end into a recoverable moment. So the honest answer to "do 404s hurt SEO" is: individually no, but the situations that produce lots of them often do.

How to fix a 404 error (decide first, then act)

Fixing a 404 error is a decision, not a single action: for each dead URL you choose to redirect it, restore it, or leave it. The mistake most people make is applying one fix to everything — blanket-redirecting every 404 to the homepage, which Google treats as a soft 404 and which frustrates users. Walk each URL through this decision instead:

Redirect it with a 301 when there is a clear replacement — a new URL after a slug change, a similar product, or a newer article covering the same topic. A 301 is a permanent redirect that passes most of the old page's link equity to the new one, which is why it is the right tool when the URL had backlinks or rankings. Make sure the target is genuinely relevant; redirecting to an unrelated page is treated as a soft 404. See 301 vs 302 redirects for which code to use.

Restore it when the page 404'd by mistake — an accidental deletion, a migration that dropped content, or a page that still has traffic, links, and search demand. If people and search engines still want that content, the best fix is to bring the page back at its original URL.

Leave it as a 404 when the page is truly gone and low value — an expired coupon, a discontinued product with no successor, a spammy old URL, or a page nobody links to. Returning a clean 404 (or 410 Gone to signal permanence) tells Google to drop it from the index, which is exactly what you want. Do not redirect these anywhere; a real 404 is the correct answer.

Whichever path you pick, back it with two safeguards: a custom 404 page that keeps lost visitors on the site with search and popular links, and a routine to fix internal links so your own pages never point at dead URLs. Our guide on how to find and fix broken links covers the tooling for the second one.

Finding 404s and keeping them in check

You cannot fix 404s you cannot see, so the first step is monitoring. Google Search Console's Pages report lists URLs excluded as "Not found (404)" and, separately, as "Soft 404" — start there, because those are the errors Google actually encountered while crawling. A dedicated crawler (Screaming Frog, Sitebug, or similar) then finds internal links pointing at 404s across your whole site, and your server or analytics logs reveal which dead URLs are still getting real traffic and are therefore worth restoring or redirecting.

Prioritize by impact, not by count. A single 404 on a page with backlinks and steady traffic matters far more than a thousand 404s from bots hitting random paths. Sort your list by inbound links and visits, fix the high-value URLs first with redirects or restorations, and let the low-value long tail return clean 404s. Chasing every 404 to zero is wasted effort — a healthy site always has some.

The efficient way to catch status-code problems before they spread is to audit the URL directly. Paste any page into the free SEO + GEO audit on the homepage and it reports the exact HTTP status a URL returns, flags soft 404s where the code and content disagree, and surfaces broken internal links — the same signals covered in why is my page not indexed when a 404 is the reason a page fell out of Google. Fix what it surfaces and your 404s stay in the harmless-background-noise category where they belong.