Results for one site link to completly different 2nd site

[sjhardy]

We host a number of sites for different companies on the same physical server. We have seen though that when searching for one company through G we are getting the right page title and summary shown in the results. However, the url and link from this are the root url for a completly different companies site on the server. There are no cross links between the sites and the content is completly different on both sites. Additionally the further links which appear below the entry in the results contain yet another completly different companies pages on a completly seperate URL to the others.

This is happening for a few of the sites on the server and we have found examples of this for other sites on other servers, indicating I would have thought that this is not some peculiarity with our method of hosting.

The companies are obviously very concerned since G is completly mis-representing them by using their page title (company name) and company desription as the snippet and then sending people off to a completly different site! It is only a matter of time before G start sending people you have searched for one companies web-site to a competitiors site.

[tedster]

I've spliced in the next thread -- sounds like it's related to the same issue.

[Regent Street]

My company runs websites for a number of clients and as you would expect, we have all of their sites hosted on a shared infrastructure. There is a common IP address for the clustered servers running these sites and for purposes of ease of maintenance etc they all run as seperate contexts, with CNAME entries for the relevant domains.

This has always meant that if we have contexts for widget, foo and bar then they would be referenced like this:

www.widget.com --> www.widget.com/widget
www.foo.com --> www.foo.com/foo
www.bar.com --> www.bar.com/bar

We've always understood that the domain names are synonymous with the main server name, so if somebody typed in

www.widget.com/foo

Then they would see the foo context. OK no problem there - and Google has been indexing these sites for years and most of them have respectable rankings in the SERPs.

Over the past few weeks, we have been getting intermittent problems of mangled SERPs. For instance, typing 'widget' into Google would normally bring up www.widget.com as the first entry.

Now what is happening is, we are still getting the page description for widget as the first entry, but the we are getting the title for foo, and the the link is for www.foo.com. Also some of the links below the desciption point to www.bar.com. All of these are our sites on the same infrastructure, but apart from that there is no actual relationship between them and they do not link to each other.

We are not doing anything dodgy here, each one of these clients is completely seperate and we are not trying any search engine tricks to try to increase traffic. If Google were penalising us I would expect the result to drop out of the rankings, but it still appears top of the SERPs.

This is naturally very concerning for us as clients are starting to notice this and assume it is something we are doing wrong. I have used the feedback form to try to contact Google but cannot believe that if this is a Google issue that other people are not having the same problem.

I would be keen to hear other people's experiences and opinions.

Thanks for your help.

[g1smd]

Set up redirects so that all folder accesses are redirected to the domain name (and the redirect retains the requested page name too). Make sure the redirect returns a "301" status code.

I'm currently looking at a site that has separate folders for separate sites, and has aliased multiple domain names to the server (but NOT to individual folders). Their situation is that every piece of content has 16 different URLs that are valid for it. Their duplicate content penalty is very very complex, much more so that yours.