404 page

A 404 page is the page a website serves when a visitor requests a URL that does not exist on the server, whether from a typo, an outdated link, or a page that was deleted without a replacement, and a well-designed one turns that dead end into a chance to keep the visitor on the site.

What is a 404 page?

Every website has a default 404 page, usually a plain, unbranded message supplied by the hosting platform or CMS. "404" is the HTTP status code a server returns when it cannot match a requested URL to any resource: a mistyped address, a bookmark to a page that moved, an inbound link from another site pointing at content that no longer exists, or a visitor following a broken internal link. Replacing the default page with a custom, on-brand one does not change why the visitor landed there, but it changes what happens next: instead of a dead end, the page can offer navigation, a search box, or links to the content the visitor was probably looking for.

How the 404 status code works

Before a browser renders anything, the server sends an HTTP status code describing what happened to the request. A correctly built 404 page returns the real 404 status in its response headers, not just a page that visually looks like an error while technically returning a 200 (OK) status, a mismatch known as a "soft 404".

Search engine crawlers read this header, not the page's design, to decide whether a URL should be dropped from the index. A soft 404 misleads crawlers into thinking a broken page is still valid content, which can leave dead URLs indexed indefinitely and waste crawl resources on pages that serve nothing useful. The same status code also matters for monitoring tools and uptime checks, which rely on it to tell a genuinely missing page apart from a page that loaded but happens to be empty.

What a good 404 page includes

• A plain-language message confirming the page was not found, without technical jargon.
• A link back to the homepage and to a handful of the site's most useful or popular pages.
• A search bar, so a visitor looking for specific content can find it without guessing at a new URL.
• The site's normal header, footer and branding, so the page still feels part of the site instead of a dead end.
• A short, on-brand tone rather than a purely technical error message, since this is often a frustrating moment for the visitor.
• A tracked event in analytics, so recurring 404 URLs surface for the team to fix rather than going unnoticed.

404 page vs soft 404 vs 301 redirect

Best practices and common pitfalls

The most common mistake is leaving the platform's default 404 page in place: generic, unbranded, and offering no way back into the site, which pushes an already frustrated visitor straight to the back button or to a competitor. A second common issue is the soft 404 described above, usually the result of a CMS or custom error page that was never configured to return the correct status code. It is also worth monitoring 404s in analytics or Search Console on a regular basis: a sudden spike usually signals a broken internal link, a bad migration, or an external site linking to a page that was deleted without a 301 redirect put in its place.

Impact on crawl budget and SEO

A 404 status is not, by itself, harmful to SEO: it is the correct, expected response for a URL that genuinely no longer exists, and search engines will eventually drop it from the index without penalizing the rest of the site. The real cost shows up in two places: crawl budget, since search engines allocate a limited number of requests to each site and a large number of genuine 404s wastes part of that budget on dead ends instead of real content, and lost ranking signals, when a page that should have been redirected was instead left to 404, discarding whatever backlinks and authority it had built up. That is the case for a 301 redirect instead of a plain 404, whenever a replacement page exists.

404 page at BeBranded

We design a custom, on-brand 404 page for every website project we build, and set up 301 redirects for any page that moves or is retired, so a dead link never turns into a lost visitor.