Error 404
Error 404 is the HTTP status code returned when a server cannot find the requested page or resource at a given URL. The most universally recognised error on the web - usually shown as a “Page Not Found” message in the browser. Caused by deleted pages, mistyped URLs, broken links, or migrations that didn’t redirect properly.
Inevitable on any site online for a few years. The question is whether you catch and handle them well enough not to damage UX or SEO.
Where 404s come from
Four common sources:
Deleted content without redirects. An old blog post got removed. Anyone who still links to it lands on a 404. Backlinks pointing at the deleted URL stop passing link equity.
URL structure changes. Site redesign moved /blog/post-name to /articles/post-name without setting up redirects. Every existing link, bookmark, and search result for the old URL now 404s.
Mistyped URLs. Someone shared the link with a typo. Or copy-pasted a URL with a trailing comma. User error, but the site decides what they see when they hit it.
Crawlers following stale references. Old sitemaps, archived pages, scrapers all still point at URLs that no longer exist. Generates ongoing 404 traffic that doesn’t represent real user need but does waste crawl budget.
What good 404 handling looks like
Three things:
Useful 404 page. Don’t show a generic browser error or a stark “Page Not Found.” Show the brand, link to the homepage, suggest popular pages, offer search. The 404 page is also a content opportunity - a chance to recover the visit.
301 redirects for known-moved content. When you delete or move a page, set up a redirect from the old URL. Preserves SEO equity, preserves user experience for anyone landing via an old link.
Active monitoring. Search Console reports 404s Google has hit while crawling. Server logs show 404s real users are hitting. Reviewing both monthly catches issues before they become widespread broken-link problems.
Where 404 management goes wrong
Three patterns:
Mass 301-redirecting everything to the homepage. Migration shortcut: every old URL redirects to /. SEO terrible (Google treats this as soft 404), user experience terrible (user expected one page, got something completely unrelated). Don’t do this.
Ignoring 404s entirely. “It’s just a 404, not our problem.” Until you realise the most-linked-to article from 2018 has been 404ing for 18 months and you’re losing thousands of monthly visits.
Over-engineering the 404 page. Animated character, witty copy, custom illustrations. Cute the first time, slow to load, and most users still bounce. Better to ship a useful, fast 404 page than a memorable one.
An example
A solo content creator’s niche site had 410 articles published over four years. A migration two years ago changed the URL structure but only redirected high-traffic posts. They never audited the rest.
An audit found 187 articles still 404ing, including 23 with 50-2,000 backlinks each pointing at the old URLs. Estimated lost monthly traffic: 8,400 visits.
The fix: 187 individual 301 redirects to the new URLs (where the content still existed) or to the most relevant existing article (where it didn’t). Three weeks of batched work. Six months later, organic traffic to the affected URLs recovered to about 7,200 monthly visits - close to the previous level. Most equity wasn’t recoverable since Google had forgotten the old URLs, but UX was instantly fixed.
Related terms
- Canonical URL - an adjacent SEO directive in the URL-management family
- Duplicate Content - a related SEO concern often surfaced by the same audits that catch 404s
- Algorithm - the search systems that punish sites with persistent 404 problems
- Breadcrumbs (Navigation) - a UI pattern that helps recover from 404 landings
- Domain - the asset 404 management ultimately protects