A 404 Not Found response means the server has nothing at the requested URL. Every site has some, from mistyped links to deleted pages, and Google has said plainly that 404s on URLs that should not exist do not harm a site. They become a problem when important pages return 404 by accident, when valuable backlinks point at dead URLs, or when error pages return 200 instead.
When a 404 is the correct answer
If a page has been removed and there is no close replacement, returning 404 (or 410 Gone) is the honest response. Google drops the URL from the index after recrawling it. Redirecting every dead URL to the homepage instead tends to be treated as a soft 404 anyway and gives visitors a worse experience.
When a 404 needs fixing
Look at a 404 more closely when any of these is true.
- The URL is linked from your own pages, sitemap or navigation
- The URL has external backlinks or still gets organic or referral traffic
- It was removed in a migration and has a clear replacement
- It is a page that should exist and is failing because of a bug or deploy
Finding the 404s that matter
Search Console's Page indexing report lists URLs Google found returning 404. Your crawler shows internal links to 404s. Backlink tools show external links to dead pages, and analytics or server logs show which dead URLs still receive visits.
Prioritize dead URLs with backlinks or traffic. Redirect those to the closest relevant page with a 301. Fix internal links at the source. Leave the rest as 404s.
Building a useful 404 page
The custom error page should still return a 404 status. Inside it, help the visitor recover: a search box, links to the main sections, docs and pricing, and a plain sentence saying the page could not be found. For SaaS sites, a link to the docs search and the status page is often the most useful thing on it.
// Next.js App Router: calling notFound() renders not-found.tsx with a 404 status
import { notFound } from "next/navigation";
export default async function Page({ params }) {
const doc = await getDoc((await params).slug);
if (!doc) notFound();
return <Article doc={doc} />;
}Example
Recovering links from a deleted resource hub
An imagined email marketing SaaS, Sendloom, deleted an old "resources" section during a redesign. A backlink tool later shows dozens of referring domains pointing at those dead URLs, including a few industry publications.
The team maps each deleted resource to its closest current guide and adds 301s for the ones with links or traffic. The rest stay as 404s, which is fine.
Common mistakes
Panicking about every 404 in Search Console
Google reports 404s for URLs it has ever seen, including junk and typos. Most need no action.
Soft 404s from client-side routing
A JavaScript app that renders a not-found view while the server returns 200 sends the wrong signal. Make the server return a real 404.
Redirecting all 404s to the homepage
It confuses visitors and search engines alike. Redirect only where a genuinely relevant replacement exists.
Questions
Not on their own. Google has said 404s for URLs that should not exist are normal. The cost comes from broken internal links, lost backlink value, and important pages accidentally returning 404.
Either works. 410 Gone says the removal is deliberate and may be processed a little faster, but the practical difference is small.
You can ignore them. Google recrawls known URLs occasionally and keeps reporting them for a while. Once nothing links to them, they matter even less.
