The Error Page Google Keeps Requesting

Where they come from

A not-found report lists addresses that were requested and did not exist, and on any established site there are more than you expect.

Some are your own doing: a page removed, a link mistyped in your own navigation, an address changed during a rebuild.

Many are not. Another site linked to you and got the address slightly wrong, an old directory holds a page you deleted years ago, or a printed reference sends people somewhere that no longer exists.

And a substantial share is automated noise: scanners requesting addresses that never existed anywhere, looking for known vulnerable paths.

Those four categories need different responses, which is the whole of this.

The triage

• A page you removed that has a current equivalent: redirect it.
• A broken link on your own site: fix the link.
• A wrong link from a real site: redirect, and tell them.
• An old address with traffic: redirect to the closest match.
• Nonsense from scanners: ignore entirely.
• Anything nobody has requested twice: ignore.

The last two are most of the list on most sites, and treating them as problems is how businesses waste an afternoon making a report look tidy.

Not-found is a correct answer

The reassurance worth stating, because these reports read as failures.

A page that does not exist should return not found. That is the accurate response and it is what you want.

Having such addresses does not harm your site, and search engines will stop requesting them eventually once they consistently return nothing.

What causes actual harm is a page that used to exist, has moved, and returns not found instead of pointing at where it went, since that loses whatever the old address had accumulated.

So the question is never how many errors are listed, it is whether any of them represent something real that has been lost.

Fix your own links first

The category that is entirely within your control and the one worth clearing completely.

A broken link in your own navigation, in a page's body, or in a footer is a visitor reaching nothing on your own site, which is the version of this that costs you something directly.

Those are identifiable in the reports because the referring page is your own, and a free link checker will find them in one pass.

Fix the link rather than redirecting the target, since the link is the thing that is wrong.

This is also the category that recurs, particularly after a rebuild, so it is worth a check whenever the site changes.

A worked example

A business found around sixty not-found errors reported and assumed the site had a serious problem.

Reading the list took ten minutes.

About forty were addresses that had never existed, requested by automated scanners looking for administrative paths.

Eleven were old blog post addresses from a site version three years earlier, none of which had been requested by a person in months.

Six were their own navigation linking to a page that had been renamed during a rebuild, which was the real finding and had been broken for eleven weeks.

Three were links from a supplier's website with a typo in the address.

They fixed the navigation, redirected the three typos, told the supplier, and ignored the remaining fifty-one.

Do not redirect everything to the homepage

The shortcut that makes the report look clean and makes the site worse.

A blanket rule sending every missing address to the homepage means nobody ever sees a not-found page, which feels tidy.

It also means somebody who clicked a specific link lands somewhere unrelated and has to start again, which is a worse experience than being told plainly that the page is gone.

Search engines tend to treat those as soft failures rather than genuine moves, so the tidiness is cosmetic.

Redirect to the genuine equivalent where one exists, and let the rest return not found honestly.

Make the not-found page useful

Since some visitors will reach one regardless, it may as well work.

It should say plainly that the page could not be found, look like the rest of your site rather than a bare server message, and offer a route onward: the main sections, a search box, and your phone number.

That last item matters for a service business, since somebody who followed a broken link to a plumber will call rather than navigate if given the option.

Build it once and it handles every future broken link, including the ones on other people's sites that you will never know about.

Check that it actually returns a not-found status rather than reporting success, since a page that says not found while claiming success is the soft failure case.

Old addresses with real traffic

The category worth actively looking for, because it is where recoverable value sits.

Where a removed page still receives requests months later, something is still pointing at it: a link from another site, a directory entry, or a printed reference.

Those are worth redirecting to the closest current equivalent, since the traffic is real and arriving.

Where the referring link is from a site you can contact, ask them to update it, which is a better outcome than a permanent redirect and frequently succeeds because most people are happy to fix a broken link.

Ask politely, give them the exact replacement address, and expect roughly half to act.

The counter-case

These reports can consume attention out of all proportion.

A tidy report is not a business objective, and a site with sixty listed not-found addresses is not sixty problems, it is usually two or three.

There is also a risk in redirecting aggressively to clear a list, which creates rules nobody can explain later and occasionally sends people somewhere unhelpful.

The proportionate approach is to check this a couple of times a year, fix your own broken links, redirect anything with real traffic behind it, and ignore the rest.

Nobody has ever lost a customer because a scanner requested an address that did not exist.

The ten minutes

1. Read the list rather than counting it.
2. Fix broken links on your own site first.
3. Redirect old addresses that still receive traffic.
4. Contact anybody linking to you incorrectly.
5. Ignore scanner noise entirely.
6. Never redirect everything to the homepage.
7. Make the not-found page useful, with a phone number.

Step two is the only category that costs you visitors directly, and it is the one hiding inside a list that is mostly noise.

The wider tidy-up is covered in finishing the year with the site tidy.

Frequently asked questions

West Coast Media Solutions Inc. provides web design, web development, hosting, digital marketing, and business consulting to organisations across Canada, drawing on more than twenty-five years in the field.