Does Google care if links to my site come from the same or similar IP?

I'm likely to become a reseller, and friends will link to me

[ThatAdamGuy]

Due to my own geeky interest in working with multiple domains and the fact that I have geeky friends, I'm very likely to be starting up my own reseller account on a particular Web host.

This will give me a fixed amount of space and bandwidth and the option to host unlimited domains for my friends.

I'm guessing that many of my friends will link to one or more of my existing sites, and I'll likely link back. And yes, I really am serious about this being friends, rather than simply being coy about setting up a bunch of dummy sites all linking to me.

I'm curious, though, whether the fact that these will all be VIRTUALLY hosted will matter. I'm guessing that Google must have something in place to penalize link farms (unscrupulous Webmasters buying up a bunch of domains and pointing them at each other), and thus take into account whether a group of domains pointing to each other are all on the same IP or IP range, right?

Or does Google not really care, preferring just to assume that a scamster has to share PR with himself anyway, so in the end it doesn't really matter?

In the end, I'm really trying to just help my friends (many of whom are artists with little cash), and I don't care about gaining substantial amounts of Google Juice. But I sure don't want to LOSE Google Juice on my main site!

Thanks in advance for any insight you're able to offer.

[ThatAdamGuy]

Heh heh, Doug, cute ;-)

All this stuff really can make one paranoid, can't it? I know in my heart that this is NOT what Google wants, but it seems to be an unavoidable byproduct.

Spammers (and deceptive Webmasters) are evil, persistent folks. I can understand Google not wanting to spell out all of their penalty policies exactly so as to avoid having the jerks perpetrate n-1 to stay just under the radar <sigh>.

[kaled]

Simple solution. On your site use links that Google cannot follow, on their sites ask them to link to you with standard html links.

This way you may gain a small advantage in Google ranking with virtually no risk.

As I have said in other threads, links created dynamically using the javascript document.write method are never likely to be followed by Google or any other search engine since the robot would have to run the javascript in order to detect the link.

Kaled.

[ciml]

> There is no real evidence that crossliking can actualy cause a penality of any kind.

A very controversial view, ogletree. I suggest reading back through [webmasterworld.com...]

I don't think we should read too much into papers and patents. They're excellent food for thought, not an indication of how Google works.

It's a long time since IPs were a daily topic here, and it's a long time since GoogleGuy indicated that IP addresses weren't a good indicator. Things can change.

[ogletree]

I agree with 2_much. You can crosslink all you want as long as you got a bunch of inbound links as well. Of course I had 4 sites that all linked to each other and for the most part the only backlinks they had were themselves and one of them had a DMOZ listing. They all had a PR2. It was that way for months until I got one site to link to all of them now they are all PR4's. None of them really have backlinks outside of themselves still. I rank very well for a ton of minor keywords. I don't rank at all for the big ones but I would need a bunch more SEO to do that. I am working on it.

[Trisha]

I mean, what are the odds that unrelated sites linking to each other would be in the same narrow IP range? ;)

In the city where I live I've noticed that many of the local businesses and organizations tend to use local ISP/hosting companies to host their sites. Many of them link back and forth to each other too, because they tend to link to other local businesses and organizations. There are only a handful of local ISP/hosting companies here. So I suspect a lot of these sites share IP's. There would be a lot of innocent sites getting hurt if Google ever starting giving some large penalty to sites just because their inbound links had the same IP address.

[rfgdxm1]

Trisha has a good point. A number of sites linking to each other sharing an IP happens in the real world for reasons other than cheating Google.

[magic77]

I agree with 2_much. In my genius, all our sites and the SEO sites were on my co-located server with all 60 or so cross linked. We had just started seeking inbound links from other sites not connected with us. Obviously, we were a little late to move in this direction.

I'm still not positive we triggered a red flag, or if a SEO competitor caused the problem. Some of our clients are in highly competitive markets, so this is a very strong possibility.

Believe it or not, I really didn't realize this would cause any kind of penalty. Of course, I wasn't part of this forum way back then (Oct. 2002). Thanks for the help.

[vitaplease]

There are too many factors that could play a role to solely blame crosslinking within the same IP range as the reason of possible Google penalty.

I'd say natural site(s) and link building should always play a role when considering these things.

I could imagine the following imaginary Google program:

Give me all clusters of sites (above 10 sites) of which 75% of external linking is between these sites.

Flag red if there is no theme correlation

Flag it extra red if one of the clustered sites has SEM/SEO or Webdesign all over it.

Add another shade of red if 80% of anchortexts equal the title of the linked to page.

Extra purple shades for same IP range

and variations here-f. One could easily think of a dozen more tell-tale signs of when linkbuilding is artificially done soley for reasons of ranking better....

Also, so called penalties in my opinion do not always have to be full blown long-term penalties.

Think counting external links as internal links, downgrade of Pagerank pass-through, averaging of Pagerank pass-over by the amount of links per site, temporary shock-treatment de-indexing, etc.

[Arnett]

I read a post from GG that said that when they axed DeepBot the ip link referencing went with it.