Web development

What a 404 and a Soft 404 Mean for Your Website

A shop counter with a laptop showing a short missing-page message and a paper list of old links beside it

A soft 404 is a URL that responds as if the page exists, then shows an error, an empty body, or a message that says the page is missing. Google treats that as a failed page even though the server said success. Its HTTP status guide says a 2xx response is considered, not guaranteed to be indexed, and empty or error content can be treated as a soft 404 [1].

What is a normal 404?

A 404 means the server says that URL is not there. Google groups 404 and 410 with other 4xx responses: the content is ignored, and URLs that were indexed and then start returning those errors are removed from results over time [1]. A 404 is not a special penalty. It is the honest answer for a page you deleted and do not intend to replace.

Why is a soft 404 worse than saying the page is gone?

Because the server claimed success. Google says empty content and error content on a 2xx response can be read as a soft 404 [1]. A styled "sorry, nothing here" page that still returns 200 looks, to the crawler, like a real page with nothing on it. The URL can sit in the index as a thin result instead of dropping the way a clear 404 does.

Does sending every old URL to the homepage count?

Google's site-move guide says not to redirect many unrelated URLs to the homepage. That pattern is treated as a soft 404 [2]. One old product sent to the closest new product is a redirect. Fifty dead blog posts sent to the homepage is a dump. The homepage is not a substitute for a page that no longer exists.

When should you redirect instead of 404?

When there is a real replacement. A 301 or 308 is a permanent redirect and a strong signal that the new URL is the one to keep. A 302 or 307 is temporary and a weak signal [1]. Google's redirect guide says server-side 301 and 308 are the permanent canonical signal, and 302, 303, and 307 are temporary and are not that signal [3]. Use a redirect when the new page answers the same request. Use a 404 when nothing does.

What about a chain of redirects?

Google says its crawler can follow up to 10 hops, and it prefers a direct redirect. The site-move guide says to avoid chains, ideally three hops or fewer and under five [2]. A product that moved three times and still points at the first old URL makes the crawler walk the whole history. Point the old URL at the current page.

How long should a real redirect stay up?

Long enough that people and Google both finish using the old address. Google's site-move guide says to keep redirects generally for at least a year [2]. The Change of Address tool, which is only for a domain or subdomain move, asks you to keep redirects for at least 180 days and recommends paying for the old domain for at least a year [4]. A path change on the same domain does not use that tool. The redirect still has to stay.

Can a sitemap put a dead URL back?

A sitemap is a discovery hint, not a command to index [5]. If the sitemap still lists a URL that now 404s, you are asking Google to fetch a page you already said is gone. Remove retired URLs from the sitemap when you remove them from the site. Leaving them in does not force them to stay indexed. It just keeps sending the crawler to an error.

What if you meant to hide the page, not delete it?

Hiding and deleting are different. A noindex rule has to be on a page Google can fetch. If robots.txt blocks that URL, Google may never see the noindex, and the URL can still appear in results [6]. A disallow is not a 404. A 404 is not a noindex. Pick the one that matches what you want: gone, replaced, or present but not listed.

  • Gone for good, no replacement — return a real 404 or 410.
  • Replaced by a specific page — 301 or 308 straight to that page.
  • Many unrelated URLs — do not send them all to the homepage.
  • A pretty error on a 200 — that is the soft 404 pattern.
  • Still in the sitemap — take the dead URL out.

Where does this show up in an ordinary site?

Old campaign links, renamed service pages, and products you discontinued. A website build that changes URLs without a map creates these. So does a redesign, which is why the signs you need a redesign should include a list of URLs you intend to keep. Maintenance is where a quarterly check catches a soft 404 before it sits for a year. The launch checklist should include one pass of the old addresses.

What is the one check that finds them?

Open three old URLs you know you changed. If the browser lands on a useful new page with one hop, the redirect is doing its job [3]. If it lands on the homepage, or on a friendly error that still loaded as a normal page, you are in the soft 404 pattern Google describes [1] [2]. Fix the map. Do not add another paragraph of "page not found" on a 200.

Sources & references

  1. Google Search Central: HTTP status codes and network errors.
  2. Google Search Central: Site moves with URL changes.
  3. Google Search Central: Redirects and Google Search.
  4. Google Search Console Help: Change of Address tool.
  5. Google Search Central: Learn about sitemaps.
  6. Google Search Central: Block search indexing with noindex.

Status codes and Search Console tools are described on the linked pages. Check the current page before you change live redirects.

If old links land on a blank page or the homepage, we can map which addresses should stay and which should go.

Talk to us arrow_right_alt