| 4:47 am on Nov 30, 2006 (gmt 0)|
Google is a registrar.
| 4:48 am on Nov 30, 2006 (gmt 0)|
That session caused quite a buzz almost immediately. I think what we are seeing is partly related to the fact that Google is now officially a registrar and they can access Whois data. But there are many other ways for them to identify (or highly suspect) common domain ownership.
I noticed in the session that Matt first asked IF the person also owned those domains - he did not exactly declare it as a fact. He was mostly noting signs that the domains were related and that some of them were quite spammy in their apparent purpose.
| 4:57 am on Nov 30, 2006 (gmt 0)|
I'd say pattern recognition plays a huge role. Fingerprints. Same aff code. Same Adsense ID. Same template. Same backlinks. (doesn't have to be interlinking) etc.
| 5:12 am on Nov 30, 2006 (gmt 0)|
>>Google is a registrar.
True, but that just tells Google who the domain registrar is. And ďThatí is where the domain info is stored, your domainís reseller.
Blocked means blocked.
Google has no more info than we do using whois.
Google has other ways, for example Adsense publisher code. AdWords accounts and Google analytics, and last but not least, Webmaster tools.
| 5:28 am on Nov 30, 2006 (gmt 0)|
Adsense ads! Haha, i forgot about that one
| 5:49 am on Nov 30, 2006 (gmt 0)|
and different adsense codes, accessed from same ip addresses, same mac address etc
| 5:57 am on Nov 30, 2006 (gmt 0)|
>>>How could he identify that the guy was the owner of these 50 sites, in just a few seconds, then?
I was at the session but I didn't get the guy's URL to do a backlink search. It was unclear, but I assumed those URLs were linking to his domain. But because I didn't have the URL I was unable to check it out for myself.
Doesn't Matt commonly do a backlink check at these sessions?
[edited by: martinibuster at 6:37 am (utc) on Nov. 30, 2006]
| 6:27 am on Nov 30, 2006 (gmt 0)|
ohh google does know much more then that, take a look around and see how much info they just get from a unknown user of the net, when you now have signed a domain name, they know much more, thats also why I always suggest to webmasters not to use anything related to a search engine, special google, special american, not be cause of wanting to hide or spam something, just to make sure they not suddently drop you sites of some kind of reason. Just a little advice.
| 6:40 am on Nov 30, 2006 (gmt 0)|
|nippi wrote: |
"...same mac address etc."
how would big G know your mac address?
| 7:09 am on Nov 30, 2006 (gmt 0)|
So, who has 50 spammy, similar sites hosted on 50 different webhosts, on 50 different IP addresses, all in different IP address ranges, located in 50 different datacenters?
OK, now... using 50 different site structures, 50 different editors/CMSs/pidgeon-English generators, 50 different affiliate IDs, etc.
50 different sites, all different in every way.
Answer: almost nobody.
| 7:24 am on Nov 30, 2006 (gmt 0)|
There are some networks out there doing just fine with 50+ domains. I'm not convinced that there is any penalty for owning 50+ websites. Anybody have convincing evidence?
| 7:36 am on Nov 30, 2006 (gmt 0)|
No, and Matt never said there was a penalty. When he started his review he noticed something and he wanted to give the site owner a bit of a warning. The specifics are known only to Matt and the site owner, but from what I saw Matt's point was clearly made.
There is no penalty just for owning a lot of domains. Let's not let another SEO Myth get out of the barn here, please. No penalty for owning lots of domain names. None. Zip. Nada. Have I made the point?
Penalties only come from how you use domains, not from the fact that you own them.
| 8:35 am on Nov 30, 2006 (gmt 0)|
From what he wrote, there was a hint of something that I have suggested in the past.
Matt wrote in his blog:
|Having lots of sites isnít automatically bad, and having PPC sites isnít automatically bad, and having whois privacy turned on isnít automatically bad, but once you get several of these factors all together, youíre often talking about a very different type of webmaster than the fellow who just has a single site or so. |
Lots of things don't cause automatic penalties, but they do raise flags, especially during any sort of manual check.
I don't even bother trying to hide from Google what sites are mine. I only have 2 domains with privacy protection, and Google knows that they are mine. I don't bother with different hosting for every domain.
I interlink them, but not excessively, just enough to give them a starting boost. Let google value those links however they want. The real links eventually come if you have good content.
| 10:23 am on Nov 30, 2006 (gmt 0)|
There is no penalty for owning many domains, but if you're running a network of domains in any particular vertical you need to think about how not to get flagged as a network.
There are many advantages to running networks as opposed to sites, just avoid leaving an obvious footprint. I'd say banner tags, site code/templates, cms use, css filenames, hosting, content, backlinks, whois, and possibly even layout can all mark your family of sites as incestuous.
The problem is that we all do things in a certain way out of habit and necessity and it can become very inefficient when you need to deviate just in order to obfuscate.
| 10:48 am on Nov 30, 2006 (gmt 0)|
This discussion has evolved from how Matt found the other sites to what is wrong with having that amount of sites.
Matt said that there was nothing wrong with having that amount of sites. His point was that having 50 sites meant that:
|"each domain doesnít always get as much loving attention, and that can really show." |
| 2:33 pm on Nov 30, 2006 (gmt 0)|
Also if all the sites are hosted on the same same IP adress, you can suspect (and only suspect) the webmaster to be the same. Domain Tools provides a great tool for identifying those sites.
[edited by: Frederic1 at 2:34 pm (utc) on Nov. 30, 2006]
| 3:56 pm on Nov 30, 2006 (gmt 0)|
I think that there is not any "penalty" however, I do think that they take multiple variables and calculate them in to the signals sent to Google and thus many of those domains do not hold nearly the amount of value they may have had at one time.
I know of a group of domains sitting on the same IP, same whois, and they are interlinked together. If i run a link operator on Google we all know it is only going to give a sample. Whats in that sample? For that group of domains that are linked together guess what that sample is? You guessed it, the suspect sites. Even full well knowing that Y! site explorer has a considerable amount more backlinks listed the link operator in Google is basically telling that domain that your strongest links are suspect as they are coming one place.
| 4:07 pm on Nov 30, 2006 (gmt 0)|
How else might G find out?
- if the webmaster uses the Google Toolbar
- if Google secretly owned the DomainsByProxy or had access to its private data
- if the sites use Google Analytics (after all, the few IP's that hit a site the most are likely the owners)
| 4:14 pm on Nov 30, 2006 (gmt 0)|
If I can go to a web site and scan a domain to find out all of the domains that reside on that IP, Google must have something more advanced internally. Yes? No?
| 5:49 pm on Nov 30, 2006 (gmt 0)|
True G must have something of its own.
But I feel that its not that hard for an individual to find out about the the network. Live.com has an excellent facility, by using ip:
you can find it out.
Also Domain tools is an excellent resource.
| 7:46 pm on Nov 30, 2006 (gmt 0)|
To answer some of the points brought up in the last few mesages:
It's in the Privacy and agreement policy at Domains By Proxy that they will not reveal your confidential data to a 3rd party unless subpoenaed by law enforcement.
And if Google had a financial/ownership interest in DBP, they would have to reveal that as a public company because of SEC reporting requirements.
If you interlink heavily and share domains on one ip then it's fairly obvious you have a network going. If you don't do that then it's less obvious. It only take a little imagination figure out ways to obscure that you have a network going.
| 8:16 pm on Nov 30, 2006 (gmt 0)|
This could happily cut the other way, as well.
Why would Google NOT use the reputation of the owner/webmaster/designer/hosting company/data center etc. etc. to help determine position for a new site?
That's a broad list of people/companies involved in websites that I gave above. Any and all can have an influence in the authority and legitimacy of a site.
Hosting policies, datacenter policies, cost of service, etc. etc. can tell you a lot about a site. Why would Google NOT use this information?
Hosts collect certain types of sites. Some sites ban adult material. Some are adult-only. Some cater to clueless newbies. some PREY on clueless newbies. (Anybody at those hosts clearly have no experience, or they would know better...) Some are too expensive for amateurs. Some provide great service at a fair price, and attract smart webmasters. Why would Google NOT use this information?
What technology is used to create, run, and maintain the site? What is the reliability record of the technology? What is the error rate in using the technology? (e.g. some programming languages are easier to make mistakes in than others). What are the demographics of the users of the technologies? (Some technologies tend to be used only by big businesses, because they are too expensive for anyone else, or because they meet a specific set of needs.) Why would Google NOT use this information?
Going further out there, does the owner have a criminal record? Civil judgements? Financial stability? What is the owner's credit rating? (To the extent permitted by law) why would Google NOT use this information?
I think the message here is that there is a whole array of factors that Google could look at (whether or not they are doing so now) that haven't been previously identified as search quality factors.
| 8:40 pm on Nov 30, 2006 (gmt 0)|
> I only have 2 domains with privacy protection
Privacy protection may be helpful in situations other than trying to fool a search engine or an engineer. Since there may still be some klansmen hiding in the bushes, webmasters working on certain niches could be at risk of having a bullet signed on their names. If the owner of a group of sites never bothered in hiding footprints and has made it really easy to be identified either by same affiliate codes, certain html blocks, same adsense accounts, same IP address logged in Google accounts or the like... isn't this an indication that privacy protection is being used for a completely different purpose?
And how would Google categorize the owner of a group of sites that has been set up for circulation, rotation and filtering of traffic? Would not this have been done just for traffic reasons? So what kind of webmaster ends up this site owner being? An accomplished one, a spammer, one that focuses on traffic while protecting his life and business? I understand being on the defensive side but comments from everyone here have me confused.
| 9:24 pm on Nov 30, 2006 (gmt 0)|
I think, there is also possibility that SE can "flag" your inquiries when you are checking positions of all 50+ sites keyword pages in short amount of time using software like WP.
One IP address that is routinely checking rankings of the same sites every time, can create the repeating search patterns groups [specific IP + short time + always the same sites] that maybe are recognizable by SE.
Is it possible?
| 9:40 pm on Nov 30, 2006 (gmt 0)|
I believe it CAN cut the other way. I run 50 websites. I also am at least the technical contact on about 200 domains. All the sites are relatively diverse and owned by various companies. All on the same class C. about 10 of these have been looked at from a SEO point of view and of those 10, all rank on the 1st page of G and other search engines for relatively competative keywords. Some link to others, if related, some don't. All have different copy, logos, art all branded to the company. Some are templated somewhat the same, some are very unique. Non of the domain names have blocks on them via the registrar. Some are less then a year old, others over 10. some are e-commerce, some static... mix in a couple of forums and non-profits. No adult, no black hat, all on the up and up...
I see absolutey no problem in having G or any other entity know of these sites, know that I am the developer, no that I host these sites and handle email for many of these companies. So what! Some of you act as if it is something that has to be hidden. Some of these sites let me put a very generic tag at the bottom of their home page... "Developed and Hosted By MyCompany.com." I still see no problem. I even have a page that lists these sites for examples of some of our work. I still see no problem.
If you have nothing to hide, then there is nothing to hide.
| 10:11 pm on Nov 30, 2006 (gmt 0)|
Well said webdude :)
One follow up question; What if you do not have any thing to hide, but some of those domains that you have are in a position of being something that was once not a problem, but as years pass it becomes an issue as alogorithms and SE policies change and you keep your core focus on less than those 200 domains?
| 10:34 pm on Nov 30, 2006 (gmt 0)|
I would even go so far as to suggest that there might be advantages for Google to know that your sites are related.
The buzzword of the month seems to be "trusted". Let's suppose that there is a hardware store that you always deal with because you absolutely trust the owner. He only sells quality merchandise and he stands behind his products. Not only that, all the employees are incredibly helpful and willing to instruct you in how to make your repairs.
The hardware store owner loves to fish, and he opens a fishing store right next to the hardware store. When looking to buy some fishing equipment, wouldn't some of your "trust" in that business owner get carried over to the new store?
The same thing goes with the previous buzzword "authority". If someone has a good reputation for producing authority content on all their sites, why would it be bad for them to admit to owning them?
| 10:53 pm on Nov 30, 2006 (gmt 0)|
Same AdSense publisher ID, same ID in any number of other ad networks or affiliate programs. Same Google Analytics ID or third-party page counter ID (depending on how these work).
These would all be quite powerful, and aren't just limited to Google. They're all plainly visible in the HTML code, and any search engine or even an ambitious scraper could use it to make connections.
| 2:38 am on Dec 1, 2006 (gmt 0)|
OK, I've got the secret formula. Define NH="total number of hyphens" and ND="total number of domains owned".
If NH/ND > 1, then Google gives every domain owned by this person a trust penalty (TP).
You will note that this approach does not penalize an owner unless they average more than one hyphen per domain.
| This 89 message thread spans 3 pages: 89 (  2 3 ) > > |