What happens when somebody is sent to your site with a web page address which doesn’t exist? Maybe it did exist once, or maybe it’s a mistake. What a human visitor should see is a nice message explaining that there’s a problem, as I discussed last month. But what a website crawler (such as Google’s) should see is something called a “404 response code”. if it doesn’t, you’re leaving yourself open to some real SEO problems.
This is because if it doesn’t get a “404”, Google will think it’s found a proper page, and try to index it. But all these missing pages it erroneously adds to its database will have the same content (the message you’re displaying to human visitors), so it’ll see lots of duplicate, low content pages. Trust me, you don’t want this.
Unfortunately I’ve seen a few examples recently of brand new websites where this has been set up wrongly. Visitors are presented with a helpful message, but this is done by sending them to a “proper” page which does not return a “404”.
If you’re now wondering: “is this me?”, then there’s a quick test for that here, kindly provided by a company called Web Site Advantage. You should also go into your Google Webmaster Tools account and see what’s currently being reported. Finally, there’s some background information from Google here.