404 page

A 404 page is the page a browser shows when it asks your server for an address that does not exist: a typo in the web address bar, an old link from another site pointing at a page you have since renamed or removed, a bookmark from years ago. The number comes from the HTTP status code the server sends back, 404, which means “not found.” Every site has this page whether anyone designed it or not; the only question is whether it was designed on purpose.

A default, undesigned 404 page is a blank, ugly dead end: a generic error message, no navigation, no way forward except the browser’s back button. A designed one keeps your branding, explains plainly that the page could not be found, and gives the visitor an obvious next step.

Why it matters to you

Every visitor who lands on a 404 page arrived wanting something. They clicked a link, typed an address, or followed an old bookmark expecting a specific page. A dead end at that moment loses them. A 404 page with your navigation intact, a search box, and a couple of links to your most important pages turns a wrong turn into a small detour instead of an exit.

It also matters for how search engines crawl your site. A 404 that correctly reports itself as not found tells a search engine’s crawl to move on and not index that address. A broken setup that returns the wrong status code, showing a normal-looking page while quietly telling the browser it does not exist, or the reverse, confuses that process and can hide or duplicate content in ways nobody intended.

What I do about it

I build a proper 404 page into every site, styled to match the rest of the site rather than left as a bare default, with a short explanation and links back to the homepage and the main sections. When I remove or rename a page during a rebuild, I check the old addresses against a redirect map first, so a visitor or a search engine following an old link lands on the right new page instead of hitting a 404 at all. The 404 page is the safety net for the addresses nobody anticipated, not the answer for ones I know are moving.

I also check, after launch, that the 404 page actually returns the correct status code to the browser, not just a page that looks like an error while technically reporting success. That distinction is invisible to a visitor and matters entirely to a search engine.

What it looks like in practice

Type a web address for your site that does not exist and, instead of a blank page or a generic browser error, you should see a page that still looks like your site: your header, a short message that the page was not found, and a couple of obvious ways to get back on track. Behind the scenes, that page is quietly telling search engines the address genuinely does not exist, so they stop trying to show it in results.