404 Errors: What They Mean and When They Hurt SEO

Three questions decide which 404s are worth fixing. Most of them are fine to leave.

If you landed here because a page just told you "404 Not Found," here's the whole answer: the website loaded fine, the server understood your request, and the page you asked for doesn't exist at that address. Nothing is broken on your end.

If you landed here because Search Console is showing you a list of 404s, the honest answer is usually "less trouble than you think" — and the rest of this page covers which ones are the exception. It sits inside our SEO guides hub, with the rest of the technical foundations.

What a 404 error actually means

404 is an HTTP status code — a three-digit number every web server sends with every response, telling the browser how the request went. 200 means "here it is." 301 means "it moved permanently." 404 means "not found."

The 4xx family means the problem is on the request side, not the server side. A 404 is the server saying: I'm working, I understood you, there's nothing at that address.

You'll see it written a dozen ways — "404 Not Found," "Error 404," "HTTP 404," or a custom page with a joke and a search box. Same status code underneath.

Why you're seeing one

Four causes account for almost every 404 on the web:

• A typo in the address. One wrong character in a URL you typed or pasted, or a link cut in half by a chat app or an email client.
• An old link. The page existed when someone linked to it — a blog, a forum, a Google result, their own menu — and has since been moved or renamed without a redirect.
• The page was deleted on purpose. A discontinued product, a finished event, an old admissions cycle. The 404 is doing its job.
• A migration that dropped URLs. The usual cause of a sudden wall of 404s — a redesign or replatform where the new URL structure didn't map to the old one.

If you're the visitor: check the URL for a typo, delete everything after the domain name and navigate from the homepage, or Google the site's name plus what you were after.

Do 404 errors hurt your SEO?

Not in the way most people assume. Google's documentation on HTTP status codes describes what happens in plain terms: for Search, "the indexing pipeline removes the URL from the index if it was previously indexed. Newly encountered 404 pages aren't processed."

That's the whole mechanism: the URL drops out of the index and Google crawls it less often. No penalty is described anywhere in that documentation, and nothing in it says a 404 drags down the rest of your site. The real cost isn't a ranking hit — it's what you lose at that URL:

• The rankings and traffic that page had. Those visits stop — a loss you chose when you deleted the page.
• The value of any inbound links pointing at it. The one that genuinely costs money: a page other sites linked to, returning 404, is a dead end for that equity.
• The visitor standing on the error page. Somebody with intent just hit a wall. Whether they leave or find what they came for is a design decision, covered below.

Which gives you the only triage rule you need: a 404 matters if something was pointing at it — a link, a ranking, an ad, a printed QR code. A 404 nobody was using is housekeeping.

404 vs 410 vs soft 404 vs 301: which one is correct

Four responses, four meanings. Getting this right is most of the job.

Redirect behaviour per Google's redirects documentation; 4xx and soft-404 handling per its HTTP status codes documentation.

Soft 404s are the ones to actually worry about

A soft 404 shows the visitor an error message but returns a 200 OK to the crawler. The human sees "sorry, nothing here"; the machine is told everything is fine.

Google is direct about the cause — "if the content suggests an error for Google Search, an empty page or an error message, Search Console will show a soft 404 error" — and the Page indexing report gives the fix: "We recommend returning a 404 response code for truly 'not found' pages."

Three common ways sites generate them without noticing:

• A "no results" page that returns 200. Empty internal search results, an out-of-stock filter combination, a category with nothing left in it.
• A JavaScript app rendering an error into a 200 shell. Common on headless and single-page builds, where routing is client-side and the server never learns the route was invalid.
• A blanket redirect to the homepage. Covered next — the most popular soft 404 of all.

They're worse than plain 404s for one reason: a real 404 resolves cleanly and Google comes back less often, while a soft 404 keeps getting crawled. Google's crawl budget documentation says as much — "soft 404 pages will continue to be crawled, and waste your budget."

The redirect decision tree

The instinct after seeing a 404 list is to redirect everything somewhere. Resist it — a redirect is correct only when a genuine replacement exists for that specific page.

• Is there a page that gives this visitor what the old URL promised? Not roughly — actually. If yes, 301 to it. A discontinued running shoe goes to the closest current model or the running-shoe category, not the front page.
• Was the content merged into another page? 301 to the page that absorbed it — the one case where several old URLs legitimately point at a single new one.
• Is it gone with no equivalent? Let it 404, or use 410 if you're certain it will never return. A closed admissions cycle at a Pune CBSE school has no replacement; sending it to the homepage helps nobody.
• Does the dead URL have inbound links? Try harder on the first question before giving up — and email the referring site to update the link, since a corrected link beats a redirect every time.

The blanket-to-homepage habit deserves its own warning, because Google names it explicitly in the site move guidance: "Don't redirect many old URLs to one irrelevant single URL destination, such as the home page of the new site." The documentation notes this "can confuse users and might be treated as a soft 404 error."

So redirecting 800 dead product URLs to the homepage doesn't remove your 404s — it converts them into a worse problem.

Don't hide them in robots.txt either. Google's crawl-budget documentation contrasts the two — "a 404 status code is a strong signal not to crawl that URL again," whereas "blocked URLs will stay part of your crawl queue much longer, and will be recrawled." What robots.txt does and doesn't control is a separate guide.

How to find the 404s that matter

The Page indexing report in Search Console is the screen — it separates "Not found (404)," "Soft 404" and "Page with redirect" into their own buckets with example URLs. The full tour of the tool is in our Google Search Console guide; here we need those three rows.

Two sources it won't give you:

• Your server logs or hosting analytics. Every 404 request that actually happened — including ones from an old app, an email campaign or a printed QR code. Real humans hitting real dead ends.
• A backlink tool's broken-page report. The dead URLs other sites still link to. The highest-value 404 list you'll look at, and usually the shortest.

Triaging a 404 list without wasting a week

Work top-down. Most of the list ends in "leave it," and that's the correct outcome.

1. Open Search Console › Pages and export the "Not found (404)" and "Soft 404" buckets separately. Different problems, different fixes — don't merge them into one column.
2. Fix every soft 404 first, regardless of traffic. These are misconfigurations: make the page return a genuine 404, or put real content on it if it should exist.
3. Cross-reference the list against inbound links. Any dead URL with external links pointing at it goes to the top of the redirect queue.
4. Cross-reference against your own site. A 404 still linked from your navigation, a post or a sitemap is an internal fault — fix the link at source, don't paper over it with a redirect.
5. Check for lost traffic, then redirect. If the URL had impressions or clicks recently, it earned a 301 to the closest live equivalent — one-to-one, and logged, so the next person inherits a map rather than a mystery.
6. Leave the rest. Search Console keeps listing 404s long after they stop mattering. A stable, non-growing 404 count is a healthy site, not a to-do list.

This sweep is one line item in a broader technical pass — our SEO audit checklist shows where it sits. If you'd rather someone else ran the migration mapping and redirect map, that's part of what our SEO services cover; a technical-led SEO retainer in India typically runs ₹25,000–₹1,50,000+ a month depending on site size and scope.

Design a 404 page that recovers the visit

The status code is for machines; the page is for the human who is now stuck, and a default server error screen loses almost all of them. Keep the 404 status code (a helpful page returning 200 is a soft 404) and make the page useful:

• Say what happened in one plain line. "This page doesn't exist any more." No jargon, no blame, no apology.
• Put a search box on it. The highest-recovery element, because the visitor already knows what they wanted.
• Link to three or four most-wanted destinations. Top categories for e-commerce; main service pages and contact for a services business. Not the whole sitemap.
• Show your phone number. A visitor to a clinic chain or a services business who can't find the page will often just call — if the number is in front of them.
• Keep your normal header, footer and breadcrumb trail. A lost visitor needs orientation, and breadcrumbs give a working path back up the hierarchy — one of the quieter arguments for breadcrumbs.

The crawl budget caveat, stated honestly

You'll read that 404s "waste crawl budget." For nearly every site reading this, it isn't a real concern.

Google's crawl budget documentation is specific about who the advice is for: "large sites (1 million+ unique pages) with content that changes moderately often (once a week)" and "medium or larger sites (10,000+ unique pages) with very rapidly changing content (daily)." A 300-page brochure site with 40 old 404s is nowhere near it.

Where it does apply — a marketplace, a classifieds site, a D2C catalogue churning thousands of SKUs — the same documentation says to "return a 404 or 410 status code for permanently removed pages," because "a 404 status code is a strong signal not to crawl that URL again." What actually consumes crawl budget at that scale is covered in our crawl budget guide.

When letting it 404 is the right answer

Deleting pages is normal. Not every removal needs a redirect:

• Expired, time-bound content. A 2023 event page, a closed application window, a festival offer that ended. No current equivalent exists.
• Thin pages you pruned on purpose. Redirecting 200 near-empty tag archives somewhere reintroduces the mess you just cleaned up.
• Spam URLs and injected junk. Pages that were never yours should 404 or 410; redirecting them passes a problem along.
• URLs that never existed. Typos, bot probes, malformed links. Search Console lists them; they need nothing from you.

A site with zero 404s isn't well-maintained. It's usually a site that redirects everything to the homepage.