The google dance - will you survive ...

nemsis said:

Absolutely not, they don't have access to private registar data, the maximum they can get is your public customer id (and they use this id to track & de-index all your sites if you do something really wrong for example). They don't use phone numbers or adresses to locate your site, the main factor to rank your site in a specific country is the language (obvious..) and your backlinks (if you receive a lot of backlinks from german regional tlds for example (.de), you'll rank higher in germany) and your tld (use a local tld to rank better for local markets..)

Click to expand...

If this is so... I guess Americans, Australians, British, Scotts, Irish, and Canadians, are all speaking a different English that apparently has minor difference that only google can pickup on. What of the Spanish in the US, the french in Canada, and other countries where the official language does not equal their country? I am having trouble with the 'obvious'ness of this. Help a brother out.

Take a gander with the ICANN Registrar agreement regarding what registrars have access to, and how privacy options actually work. I'm not sure what this public customer ID is you are referring to. If you use privacy protection, there is no public ID to speak of. There is just your original data, and your public proxied data.

MisterGemini said:

If this is so... I guess Americans, Australians, British, Scotts, Irish, and Canadians, are all speaking a different English that apparently has minor difference that only google can pickup on. What of the Spanish in the US, the french in Canada, and other countries where the official language does not equal their country? I am having trouble with the 'obvious'ness of this. Help a brother out.

Click to expand...

I said it was the main factor, it's probably not the only one. But do you really think they parse phone numbers to geolocate sites ? I don't think so, they try to make algorithms that fit to a maximum number of sites, and maybe 20% of them have about pages, and maybe 0.01% of these 20% put their phone number in it.
I think they struggle to locate exactly a site (you can have for example a website hosted in the netherlands, written in french, for the belgium market). Maybe they just give little bonus to sites hosted on the same region, the only thing we can be sure 100% is that a regional tld give a good bonus for regional markets.

Take a gander with the ICANN Registrar agreement regarding what registrars have access to, and how privacy options actually work. I'm not sure what this public customer ID is you are referring to. If you use privacy protection, there is no public ID to speak of. There is just your original data, and your public proxied data.

Click to expand...

The public customer id is what your registar shows when you have the whois "privacy", and normally (depending on your registar) this customer id doesn't change so if you have 5 domains with the whois privacy on the same registar, the customer id shown in whois will not change.
OK, google is accredited by the icann but do you really think they check their whois data to rank billions of domains ? I don't think so.

Well, you may be right sirGemini, i'm not working for google but I think they always try to use "easy to get" signals.

TheHockeyGod said:

It was much too long and complicated to give a full answer here, but I have responded as to the several reasons why this doesn't make sense:

http://www.dotcult.com/does-google-analytics-affect-seo

Click to expand...

That's a pretty good article on your blog. I think that it's likely that google will use the data from analytics more than you suggest, but I agree with you that many people put way to much into the idea of a google conspiracy.

Many people have this feeling that anything Matt Cutts says must be a lie to mislead everyone. That's always seemed a pretty funny idea to me, blackhatters aren't the only ones that reading and viewing googles press releases. The statements Google makes are intended for the regular whitehat webmasters and seo people. It wouldn't be in googles best interest to give out false information in the way the conspiracy theorists would have us believe.

Google is in the information business, they want to provide the most relevant results to searchers. If they help webmasters to make it easier for them to sort through all the websites, it just makes their job easier. If they give outright lies then the very websites they do want to show as most relevant will be presenting the wrong information, and the blackhatters that ignore the information would be the ones getting the best position. So outright lies would have the opposite effect of that which would be in googles best interest.

It's much more likely that the information that google releases is accurate, but probably in most cases not the complete story. I think they give out enough information to help people make the types of websites and do the types of seo they're looking for, but at the same time hold back enough information to prevent people from knowing too much about how to game the system. It's not so much a conspiracy, as just good business sense on their part.

Since the data in analytics is available to Google, I doubt that they would totally ignore it. I think if you have a whitehat site then GA is totally appropriate to use. I think it gives the webmaster some good information in a fairly simple layout, and any information that google might use themselves is more likely to be beneficial to your site. If you have a real spammy type site, or are running blackhat campaigns using your site, or doing a lot of blackhat seo then it just makes sense not to use analytics. As MisterGemini said, there are a lot of alternatives to GA that in many cases are superior to it. If you have anything you don't want to draw Googles attention to then it makes sense not to use their tools.

edit-
a few people pointed me to TheHockeyGod's reply so I feel I should make a note here. I believe TheHockeyGod is refering to someone else when he mentioned GrayWolf in his article. I imagine he was refering to Michael Gray's blog, if not then he meant someone else altogether. Anyone that has read many of my posts on bhw know that none of my replies ever really side one way or the other in regards to the google conspiracy theories.

The name I use here is GreyWolf, which is of course completely different than GrayWolf. LOL

@iceinthebox - I think the google dance has more to do simply with the algorithm resorting your pages position in the serp based on the backlinks it found. The crawlers are constantly finding new links and when there is a large amount found all at once then it's going to cause fairly drastic jumps as it sorts and correlates the new data.

You may have noticed a correlation between traffic and how your site finally ranked in the serps, but I think there's a real good chance that the traffic and the ranking are both results of similar factors, and not necessarily a cause and effect.

The best thing you can do is make sure that your backlinking campaign simulates the natural viral growth that could be expected from a truly viral site.

I discussed this same concept in the SEO - Understanding the Simulation thread. If you do a good job of backlink building in the same manner as could be expected of a viral site then it's only to be expected that you would get some surges of traffic as well. Backlink building is really just site promotion. Since google came into the picture that site promotion has reached a level that most backlinks won't actually get any real visitors on their own, but if you have a complete backlinking campaign your bound to get many that actually will send visitors directly. So if you did a good job of backlinking that is likely to help your position in the serp, then at the same time your experiencing a google dance you can occassionally expect to recieve some surge in traffic as many of the backlinks are clicked on by real people as well.

Very very good answer by Graywolf. Makes sense in all aspects, and especially in the aspect that Google gives us plenty of info, but probably not all. The missing variables are the only question for us to figure out !