Anything is possible but before we jump to conclusions may I suggest we take a moment to look at the bigger picture.
Why would Google waste time with Google Analytics which is limited to a domain when it can look at data from Chrome users who visits hundreds of domains? Chrome has roughly 40% market share of all internet users (depending on which report you look at).
Ok, so now you are thinking that Google is probably using the data from Chrome. Why would Google bother using Chrome data when they can theoretically buy the data logs straight form the ISPs and get 100% of the internet users?
Bottom line there are many ways for Google to gain access to real usage data. I expect Google to slowly de-emphasize backlinks and emphasize usage signals in their secret formula since real usage is much harder to manipulate. Handicapping yourself by not using a quality free analytics tool that can help you to make smart improvements to your site is not wise IMHO.
[edited by: goodroi at 4:15 pm (utc) on Aug 9, 2013]
Well said. I think you put this topic to bed once and for all.
|Why would Google waste time with Google Analytics which is limited to a domain when it can look at data from Chrome users who visits hundreds of domains? Chrome has roughly 40% market share of all internet users (depending on which report you look at). |
Huh? Your logic makes no sense at all.
Why would Google prefer using information from 40% of the users (via Chrome) or even smaller percentage (via ISP logs, after all, they can't buy the logs from all the world's ISPs) when they can get information about 100% of the website visitors via GA for free?!?
|Handicapping yourself by not using a quality free analytics tool that can help you to make smart improvements to your site is not wise IMHO. |
Putting free script on every page of your website is not wise IMHO.
Nothing is free, especially when it comes to Google. Sure they love to waste all this storage, bandwidth and computing power because they are great guys :)
They've stated before the data isn't used....but they usually tend to say one thing and do the other so who knows.
Google Analytics does not have 100% adoption across the web.
According to these guys who actually did scientific research [law.berkeley.edu...]
Google has a presence on:
73% of the top 1000 sites
37% of the top 25,000 sites
I would guess that percentage gets significantly lower when you look at the top 1,000,000 sites.
Since most websites do not have a Google presence, Google can't get "100% of the website visitors via GA for free" regardless if you place Google Analytics on your single website.
If you have thousands and thousands of dollars then you can afford to buy your own analytics solution (I doubt most people reading this have that money). Not to mention that 73% of the top 1000 sites have chosen to allow a Google presence on their site, so just maybe it isn't all bad.
|"Google Analytics does not have 100% adoption across the web." |
And on some sites, even when they do use GA, they often have it implemented incorrectly (using an old version, or pasting the code TWICE per page, as I once did).
I think a lot of their metrics are taken directly from google search page or google search implemented on various sites.
No one is talking about adoption across the web. The point is that once a website implements GA - Google has 100% visitor information about that website, regardless of browser or ISP. The data already comes well-organized and ready to be used.
While they will not get 100% of the web, they certainly know 100% about that particular website and could use the information either directly (affecting that website) or indirectly (across the board).
Anyone who thinks that Google, having this perfectly structured data, will not use it, is simply naive.
P.S. Don't get me wrong, I'm not saying they only use GA data, but that they use ALL of them - GA, chrome, WMT, etc.
Individually? I doubt it. Google is not interested in things that don't scale. Individually you don't even exist. Until some set of patterns involving your site triggers an alarm somewhere that tells them you are ripe for a penalty.
Just for arguments sake, maybe if someone's site is penalized and they have filed a reconsideration request AND they have google analytics on their site MAYBE the analyst would take a look at it on an individual level for the reconsideration request?
However, I would tend to agree with netmeg: I don't think it is feasible on an individual level.
Also, I am not sure what exactly they would / could determine cleanly on an individual basis.
I know some sites have LOTS of events per page. So I think it would be pretty difficult for google to use this data effectively.
Just my two cents though.
Read their TOS to understand that the data isnít private to a website owner alone.
There are no free meals :) I can assume that they do store it all, manipulating it and crunch it endlessly.
Yes netmeg, it might be 'not individually' per site but on a collective basis. After all, what could be easier for them than taking a bunch of websites on the same vertical and explore their performance using GA data?
It could be that user engagement is monitored and used against the site. For example, an increase in traffic also means an increase in bounce rate, or a low page per visit which might bring the websiteís ranking down.
Google is a business that uses creative ways to grow its bottom line. They arenít a charity organization. Instead of convincing myself how honest Google is, Iíd rather think that Google is using this data against me than to be wrong and have it hurt me.
Making my own judgment call on this, I decided to remove GA completely. Without GA installed Google wonít be able to track (directly) visitors BR, P/V or % Exit rate on my sites.
I'd say, being safe is better than being sorry. Besides, you can find other statistic tools out there that provide the same features as GA does.
Please correct me if I am wrong.
|for some articles where GA was saying I had an 80% bounce rate, it turns out that approximately 65% of the people were reading to the bottom of the main content, and another 10% were reading to the very bottom of the page (past the comments section). |
Bounce rate doesn't mean the page stinks. It can equally well mean that the page answered the visitor's question so thoroughly and comprehensively, there was no need to go blundering around the site looking for more. Which, in turn, might mean that the search engine has done its job.
You could have every last one of a page's visitors reading all the way to the bottom of the page, triggering 20 Analytics events along the way. If they don't move on to a second page it's still a 100% bounce rate on paper. It's a meaningless number unless you can see inside the user's head.
Because Analytics isn't 100% adopted, I don't think it gives Google a very complete picture of the web, and therefore I doubt it factors directly into the algorithm (which fits with their claims). I think they COULD use it for testing limited datasets - they've never said they don't, and it could be useful for that. Then the results from those tests might impact some factors in the algorithm. But is Google going to take something that tons of important websites don't use and try to reshape the SERPs directly with it? I don't think so.
On another thread on roughly the same topic [webmasterworld.com...] I posted about a week ago...
|Google would want to use information that would be relatively uniform over its entire index. Clearly, not everyone has Google Analytics installed, nor are the installations uniform, so that in itself would make it a bad choice to use. |
I doubt that they use it even for manual audits, but I'm sure they're in some way looking for big data correlations on the installed base that might help them confirm patterns they're seeing in other ways.
Perhaps they check aggregate correlations on types of sites to suggest ways of improving both Analytics and the algo. Too "noisy", though, and too invasive, as a ranking metric.
I don't think the data collected by GA is at a sufficiently detailed or granular level to be an accurate indication of even an individual site (unless they collect much more detailed stats than are displayed in GA). I use another tracking system which is much more accurate but it has a limited functionality so I use that for the headlines and GA as a rough guide to everything else.
I agree that Google is collecting usage data from somewhere and employing it in the algo to a certain extent which may increase as their confidence in it grows. Whether its Chrome, ISPs or what I don't know or care really. I just know if I were Google I'd want usage data and I'd pull out all the stops to get as much as I need for it to be meaningful (and try to prevent rival search engines getting it).
Just look at Googles efforts in the mobile market. Why do they want everyone to use an Android device? Data collection is my bet. As the world migrates from the desktop to the mobile device they face new challenges in accessing data. Data is like oxygen to search engines. If you can grab all of it and starve rivals of it, the rivals will eventually die.
Former Google CEO Eric Schmidt once referred to Google's drive of acquiring data as "pushing the creepy line". Assume that Google gathers data from every source available and that they cross reference it in ways you haven't thought about. It's what they do.
|Bottom line there are many ways for Google to gain access to real usage data. |
That shouldn't worry you when it comes to website traffic unless your site is doing something shady. The added data has an equal chance of benefiting the site as it does harming it's rankings, imo.
Eric's comment happened at a time in which he visited North Korea, was demoted from CEO and separated from his wife of 30+ years. Something wasn't going well for him so who knows if his views reflected Google practice at the time or not.