Forum Moderators: open
We have a number of similar sites on the same IP address and wondering if this would affect how they rank?
The content on the sites are not duplicate content, but they are all based on the same theme, but with different purposes. There are some keyphrases which relate to more than one site, but only one of the sites would show up in the rankings on the first page, with Googles changing algorithm, they seem to swap and change which site shows up for which keyphrase. One month it will be one site, and the next will be a different site? Is this because of the sites being on the same IP address? Would the sites stand a better chance ranking for the same keyphrases if they were on different IP adresses?
They are not duplicate sites as I said, they are all similar, but the purpose for each site brings a different solution.
Cheers
What you need to do is get many independent links to both sites
cabbie - This has been my theory on how to fix this, but on the sites I have that have been affected that I still control, I haven't been able to get the client off the dime to spring for some independent links, so it's just been a theory. Have you seen independent links actually fix the problem?
From what I've been observing, by the way, the sites don't have to have the same IP address. Consecutive or close class Cs seems to suffice. In the sites I've observed that have this problem, the other factor is that they have common sources of inbound links.
In one case, two sites... say widgetworld and gizmoworld... both say in the automotive area, are sister sites, and generally anyone who links to one links to both. One, say, specializes in automotive windshields, and the other is about brakes and tires (not actually the case, but this is illustrative of the situation). There's only one reciprocal link between the two sites. The services really don't belong on the same site, but right now they'd probably be ranking better if they did.
In another case... 20 or so brick and mortar businesses going for similar but not identical terms all have a common parent management company that links once to each, and each has a links page that links once to the other locations. No other crosslinking. Here too, only one site ranks for each of the terms. Some of these have pretty good inbounds, but overlap in geographical area... and often the directories that link to them link from the same categories.
All of the above, in both cases, are independently run businesses but with a common or overlapping management at the top.
I can see the validity of "clustering" results among sites for the same searches, as Google does with pages returned in serps... but when the sites are different businesses going after different terms, the favoring of one of the sites and the dropping of the other is sloppy to unfair.
I look at this as collateral damage in Google's war against bigger networks.
Beyond getting additional independent inbounds, I've considered changing the links between the sites to javascript, and changing servers as a last resort. I've also been thinking of dropping some of the inbound links to the sites, so no site has link sources that overlap with another's.
So I'm really resisting changing the nature of the sites, because I think in the long run that Google is going to have to sort this out, just the way they did with cross-linking way back... and some aspects of Florida more recently.
No can't say that I have.:)
In fact the sites that i had this happen to and I had it with a few different sites,one of each pair bit the dust eventually with google by getting a pr0.I don't know if this was for some other reason or not.
But I have 3 big sites on the same c block at the moment that exchange links and Target the same kws and sit side by side in the serps.
I put it down to them having separate identities in content and inbound links
what i see for this is that its multiple sites owned by the same person targeting the same keywords that are getting hit.
I dont see any problems with a network of sites all promoting totally different topics/themes/keywords linking to each other.
usually when you run an own server it should not be a problem to get different IP-adresses on different IP-C ranges for your sites , shouldn't it?
Also you can add a nameserver for every site like ns1.domain.com ns2.domain.com
then the only thing which makes the sites look like they are from the same owner are links from same sources and the whois.
And I really doubt google using whois data. (think it is impossible)
greg
I get the feeling that the idea of IBLs being from similar sites do have an effect somehow. Our sites attract links from other sites in the same field because information on our sites are helpful to others and it would make sense for people to link to all 3, however, there is one site that is hosted on our server which our other sites dont link to that is still of a similar nature but attracts different links, this site can still rank alongside one of the others for the same keyphrase, so it could make sense that IBLs have made the difference.
...there is one site that is hosted on our server which our other sites dont link to that is still of a similar nature but attracts different links, this site can still rank alongside one of the others for the same keyphrase, so it could make sense that IBLs have made the difference.
wigsy - Thanks.... that's an important bit of information. Now, I'm wondering whether anyone has climbed out of this problem by securing additional inbounds to an apparently nuked site.
Actually, I'm not sure whether the problem's been around long enough, that, with the apparent delay in crediting links, there's been enough time for anyone to have observed a recovery.
usually when you run an own server it should not be a problem to get different IP-adresses on different IP-C ranges for your sites , shouldn't it?
I disagree. We do our own in-house hosting (in contrast to hosting with a specialized company), and all we got was a range of 4 official IP-adresses from our ISP. He claimed that since IP-adresses are running out a company of our size has to give a very good reason so that the InterNIC (or whoever is responsible for assigning IP-adress ranges) would hand out more IP-adresses. They would heavily promote virtual hosts and IP-masquerading in order to stretch the available pool of IP-adresses as long as possible.
Getting different IP's on different networks is even a bigger problem if you do in-house hosting, since your ISP needs to set up a rather complex routing and so do you. PLUS the issues that you probably won't get granted SEVERAL small networks to be routed to you.
Well intresting question and some intresting comments from Robert and others too.I think the best answer would be.
You might get into trouble if google smells the blood.
Well, I undestand what you mean when you say
The content on the sites are not duplicate content, but they are all based on the same theme, but with different purposes.
If your network is caught by goolge you may get penalised. Try to avoid the following things:
1)Simmilar I.P's (first 3 octets)for the sites of the same feild.
2)Same pattern of inbound links pointing towards these sites.
3)Interlinking among the sites.
The basic reason behind all this excercise is that you shoudn't get caught as a group of affiliate sites trying to rule the SERPS on a feild. Hiltop algorithm has this role in the latest google algo to weed out such sites.
Affiliate sites are defined as follows –
Pages that originate from the same domain (www.ibm.com, www.ibm.com/us/, products.ibm.com, solutions.ibm.com etc.)
Pages that originate from the same domains but with different top level and second level suffixes (like www.ibm.com, www.ibm.co.uk, www.ibm.co.jp etc.)
Pages that originate from neighborhood IPs (first 3 common octet in the IP number like 66.165.238.**** is common)
Pages that originate from affiliate of affiliates (if www.abc.com is hosted on the same IP octet as www.ibm.com, then www.abc.com is an affiliate of www.ibm.co.uk even if they are on a different IP series)
Check out the above points and move accordingly.
1) Eliminate links to penalized domains
2) Decrease linking to same website network
3) Getting solid links in from quality sites on different networks
Hey, it's almost like having a legitimate site! I'm starting to sleep better and get fewer hate calls from competitors.
We are in the process of untangling our former spammy network and getting clean relationships with other networks.
It's difficult to get away from being xenophobic, but I'm starting to get used to it. Hope this helps a little.
We have also noticed improvements by switching simliar themes to seperate A classes and cleaning up content. Jury is still out on the full extent of results on this.
Simplest solution would be to ensure they don't both/all turn up for a particular KW search.....
Way back in another time, we had too many sites on a C block. We move some but not all, to different C's (and different hosts in some cases). Eventually we moved even more sites to different hosts.
Now, some sites co-exist on same C but only if cross-links are extremely limited, and outbound and inbound links are quite different. Make sense?
could you please explain what do you mean by
Consecutive or close class Cs?
Hi taj79 - I did a WebmasterWorld site search using Google for the phrase "same class c" and came up with 44 threads. This one has the most succinct definition I found...
from [webmasterworld.com...] ....
...for an ip address = 111.111.111.xall values of x will belong to the same c class
An example of a class C that's consecutive to the above would be 111.111.112.x
I discovered on one site where we requested different class Cs at the host, we got consecutive class Cs. I think it's a no-brainer for Google to notice this.
Talking to a tech at the host about the situation, I was told that the system they use automatically assigns IP numbers and they can't really control it. Not sure how much to believe this, or how much trouble it would be for them to override their automatic system. I think in the future I'll host sites that might be seen as related at different hosts. Google has good reasons for its filters, but the current application is nuking sites that in my opinion are completely legitimate.
Incidentally, in searching for "same class c," I stumbled across a post by SEOMike that relates to what we're discussing....
on [webmasterworld.com...] ....
Why moving hosting? Can't I just quit the redirection and change the content?Well, my experience, and research says that if a site has the same class C as another (aaa.bbb.CCC.ddd) the links between the two won't be counted by Google.
What is considered duplicated content exactly?
The same document?
The same texts?
Some common elements?
The same design but different texts?Well, this is kind of tough to really nail down. My rule of thumb is; if it looks a little different, and reads a little differently, you will be ok. Word for Word will definitely be picked up. Common elements will help establish that they are in the same industry.
Just to add, none of out sites have been penalised, there isn't really anything to penalise us for as the sites I am talking about have no dodgy practices going on for higher rankings. As far as linking between these sites, it actually makes sense for these sites to interlink for the user, so really, it would be difficult in this case for them not to link to each other, so I guess, the best solution would be to look into options mentioned regarding the IP addresses.
the links between the two won't be counted by Google
I don't know that this has actually been proven (if anyone has tested it, please let us know). It seemed to me that until recently, they had counted.
What's happening here, though, is more than not counting... Some of the sites effectively drop out, even for their own names.
I think Google must be looking at a set of patterns, and that the IP address is one of the things that might raise a flag, but it might not be enough to do it alone.
hm well. We did so and it was no prblem at all.
We have our own server but it is not in house it is located in a data processing center. We have a company which does all rebooting and administration for us.
At the moment we have 16 IP Adresses , from which 3 are from different C-Ranges. We get the IPs at no extra cost.
We could have more IPs without doing anything, just telling the admin.
Thought of changing ISP? ;-)
greg
we got was a range of 4 official IP-adresses from our ISP. He claimed that since IP-adresses are running out a company of our size has to give a very good reason so that the InterNIC (or whoever is responsible for assigning IP-adress ranges) would hand out more IP-adresses. They would heavily promote virtual hosts and IP-masquerading in order to stretch the available pool of IP-adresses as long as possible.
I am not really sure how virtual hosts work but some of my sites are using them. Are the IP's "masqueraded" as is implied above? Can a spider still recognize the true IP?
Affiliate sites are defined as follows –Pages that originate from the same domain (www.ibm.com, www.ibm.com/us/, products.ibm.com, solutions.ibm.com etc.)
Pages that originate from the same domains but with different top level and second level suffixes (like www.ibm.com, www.ibm.co.uk, www.ibm.co.jp etc.)
Pages that originate from neighborhood IPs (first 3 common octet in the IP number like 66.165.238.**** is common)
Pages that originate from affiliate of affiliates (if www.abc.com is hosted on the same IP octet as www.ibm.com, then www.abc.com is an affiliate of www.ibm.co.uk even if they are on a different IP series)
Search for "Google" and the SERPs are dominated by a whole list of Google subdomains and international sites. Surely they realize there are legitimate reasons to not penalize what you define as affiliate domains.
NNY
>>>...the links between the two won't be counted by Google
I don't know that this has actually been proven (if anyone has tested it, please let us know).
Toolbar PR is flowing for sites sharing same ip.
As for whether it affects ranking for similar sites, I don't know. I don't like the idea of sharing ip's with similar sites, especially if they are interlinked. At best, I'd advise to not share links between them, if they're similar. Purely cautionary, take it with a grain of salt, as it's not based on any observed phenomenom.
At best, I'd advise to not share links between them, if they're similar. Purely cautionary, take it with a grain of salt, as it's not based on any observed phenomenom.
I've even thought about trying out javascript on some cross-links, in an effort to maintain the user experience, but I hate depending on javascript for navigation.
Too bad Google can't nuke the links rather than nuke the sites. I believe there is some variant of Hilltop involved in this, but, like Florida, it may be cutting too deep.
I have a server on which a lot of our websites are hosted.
We have three sites on this server which have PRs of 6.
Lets say Site A with PR6 mentions site B with PR 3 in the sister concerns page. It does not link to it. A mere mention, but contains the bare minimum keywords for which site B is optimized for. (thats the only way to describe site B, thats why the keywords exist on site A)
Even with no link to site B from site A, site A comes up for the keywords instead of site B!
ON top of this, a link from my PR6 sites works beautifully to any other IP giving the other site a boost, but does nothing for sites on my own server. On top of it, using the correct keywords in anchortext to link will just bring up the high PR site A on top instead of the low PR site B!
We are into hosting too - and now I can't host any of my clients on my own server as links from my own sites to them are worthless, and are great if they are hosted elsewhere.
The your external datacenter has several networks in a bunc. Tht's possible - yes.
I'm working with that ISP for over 7 years now and I have no intention at all to change. We have certain reasons why we need the setup the way we have. The only change I might do in the future is go for certain country specific sites towards local hosting in that country.
Less from the point of assessing LR/TSPR than from wresting back some control over their off-page ranking factors.
