Forum Moderators: Robert Charlton & goodroi
First i want to know is it true? IF yes, how does G distinguish between the user coming for information rather then product or purchasing something.
Google tracks every click. Look at the javascript.
To all those posting "Google couldn't/wouldn't/shouldn't do this" because it's "silly/irrelevant/difficult/unfair", consider the following:
Google would never build a SE based on PageRank because of all the link spam it would encourage.
Google would never allow pages to be hijacked in the SERPS because they are too clever.
Google would never break their link: search tool because it would confuse people.
Google would never delay ranking new sites because fresh is good.
As fclark posted, Google uses searchers behaviour to rank Adwords Ads, so it would be silly to say they would never use it to rank the organic SERPS.
Google tracks every click. Look at the javascript.
eden: Are you sure about this? I can see the onClick-Handlers only at their menu links and at the adwords - but not in their 'organic' search results. I am not very proficient with javascript, but i thought these handlers could only be used in elements and not for the whole page.
With adwords their goal is to make as much money as possible, while maintianing some customer satisfaction for both the surfer and the advertiser.
With the SERPs the goal is strict surfer satisfaction.
As for whether they are tracking you or not, it depends on your browser. If you are still foolish enough to be using IE, then you are probably being tracked most of the time. They probably figure you aren't going to object based on privacy issues if you still use IE as your primary browser.
As for whether they are tracking you or not, it depends on your browser. If you are still foolish enough to be using IE, then you are probably being tracked most of the time. They probably figure you aren't going to object based on privacy issues if you still use IE as your primary browser.
Oh, really? The primary user of IE today is the non tech savvie person, who uses it because it's part of their PC setup. What part of that means that they've waived their right to privacy?
Look at the privacy policy, Google reassures us that it doesn't track individuals use of the internet from it's toolbar or it's organic SERPS. They're a big company, they're not stupid enough to lie about something like that.
The privacy policy that they serve to me includes such statements as
Google may present links in a format that enables us to understand whether they have been followed. We use this information to understand and improve the quality of Google's search technology. For instance, this data helps us determine how often users are satisfied with the first result of a query and how often they proceed to later results.
and
Google collects limited non-personally identifying information your browser makes available whenever you visit a website. This log information includes your Internet Protocol address, browser type, browser language, the date and time of your query and one or more cookies that may uniquely identify your browser. We use this information to operate, develop and improve our services.
and
Upon your first visit to Google, a cookie is sent to your computer that uniquely identifies your browser. A "cookie" is a small file containing a string of characters that is sent to your computer when you visit a website. We use cookies to improve the quality of our service and to better understand how people interact with us. Google does this by storing user preferences in cookies and by tracking user trends and patterns of how people search. Most browsers are initially set up to accept cookies. You can reset your browser to refuse all cookies or to indicate when a cookie is being sent. However, some Google features or services may not function properly without cookies.
Since we are not talking about the toolbar, that privacy policy doesn't count. But since you mention it, here is another bit:
Google may collect information about web pages that you are viewing when the advanced functionality is enabled. However, this advanced functionality is optional, and can be easily disabled and re-enabled at any time (by selecting "Privacy Information..." in the Toolbar's "Google" menu.)
You really should read the privacy policies before you make such statements.
And when I wrote the above, I was referring to individual clickthrough tracking, including following the user as he went from search to site and back again, not aggregate user tracking and stat compiling.
I'm not talking about that, I'm talking about your amazingly sweeping statement that users of IE have no right to expect basic privacy rights.
Funny that you should be objecting to that, since it was not what I said nor meant.
What I said was that IE users are less likely to object.
And they are not privacy "rights", they are privacy issues. There are privacy rights, but this does not fall into that category.
And when I wrote the above, I was referring to individual clickthrough tracking, including following the user as he went from search to site and back again, not aggregate user tracking and stat compiling
The funny thing is, that even that is not mentioned in their privacy policy that you told me to read.
All they say is that they do not correlate the data to an individual person. They do not match my search pattern to my gmail account.
They make it absolutely clear that they can and do keep track of what your computer searches on, and that they can track which sites are clicked on at your computer.
They tell you how to block some of that if you wish.
They probably figure you aren't going to object based on privacy issues if you still use IE as your primary browser.
I fail to see why anybody should "figure" that.
European privacy laws mean that tracking an individual, and associating that individual with personal data, is illegal (all right, not all of Europe have implemented these laws yet, but many have and the rest are planning to), unless certain procedures are followed. So yes, there are basic privacy rights, enshrined in law. You can write what you like in a TOS or Privacy Policy, but you need to adapt to the legislation, and keep the user fully informed of all information recorded, tracked and stored.
The topic of this thread was: is Google tracking usage of sites linked to out of its organic SERPS, and using that information to alter it's results? I stand by my original statement that it doesn't, and indeed, can't.
Does Google track, store and utilise general user statistics in order to improve its results? Quite possibly, but then, any halfway interested webmaster does the same.
Does Google track me as an individual, linking up my gmail, adwords, adsense and organic SERPS usage to build up a profile of me and my internet usage? Well, they could, but it would most probably be illegal in some countries, and unethical anywhere.
Then what do you call this (code taken from google search results page):Javascript:
function clk(el,ct,cd) {if(document.images){(new Image()).src="/url?sa=T&ct="+escape(ct)+"&cd="+escape(cd)+" &url="+escape(el.href)+"&ei=s9wYQsKkHMnwwAGisbDNCA";}return true;}
Search Result Links:
<a href=... onmousedown="return clk(this,'res',7)">
I agree - this is tracking, obviously. But they don't put it _always_ on search results links - sometimes they do, sometimes they don't. So it's more likely they check the traffic from time to time, in order of veryfying how relevant they are perhaps.
There are so many problems with using traffic patterns for ranking, that it is beyond ridiculous.
quoting from bigdave's quote of the google privacy policyGoogle may present links in a format that enables us to understand whether they have been followed. We use this information to understand and improve the quality of Google's search technology. For instance, this data helps us determine how often users are satisfied with the first result of a query and how often they proceed to later results.
I think it's safe to say that google does track based upon the observed javascript and the description of the usage in the privacy policy.
Furthermore, it is possible to measure by proxy the time spent on page and the usefulness of a page under certain *rules*. eg.
a/ if a user clicks through and does not come back it might be defined as a successful search
b/ if a user click through and comes back in the same session and clicks through to another result the time spent on the external site can be guessed and it can also be surmised he was not totally satisfied with the result
c/ if a user does (b) and then further narrows his search terms, it may be surmised that the search was not satisfactory because he is expending effort to refine his search
d/ if a user does (b) and (c) and then bails without clicking, there is a problem. because he has done one of the following:
(i) used a bookmark to navigate away
(ii) typed in a destination in the address bar, perhaps to another search engine
(iii) closed the browser
(iv) gone to the kitchen for some spam that is more filling
Yes there are problems, but it may be that they are willing to accept the limitations and gaming. To paraphrase, ... it is beyond ridiculous ... to declare unequivocally that they do not use this data.
+++
One hell of a database they've got there if they're linking up the two. Guess they're not running Access ;)
I just can't see anybody wasting that much processing power when there are so many other ways to evaluate a page.
I just can't see anybody wasting that much processing power when there are so many other ways to evaluate a page.
I wouldn't track every single action either. That's why they hire statisticians who can properly sample a fraction of these searches to derive statistically significant answers to a correctly constructed null hypothesis.
Of course, there are more advanced machine learning statistical techniques available than simple hypothesis testing... but they all work with sampling.
European privacy laws mean that tracking an individual, and associating that individual with personal data, is illegal
And if you read their privacy policy (that you told me to read) then you would know that they specifically *do not* do that. And I certainly never suggeszted that they do.
The topic of this thread was: is Google tracking usage of sites linked to out of its organic SERPS, and using that information to alter it's results? I stand by my original statement that it doesn't, and indeed, can't.
They certainly can. I doubt they use it for anything more than QA, but they can if they want to.
Does Google track me as an individual, linking up my gmail, adwords, adsense and organic SERPS usage to build up a profile of me and my internet usage?
They specifically state that they do not.
Yes there are problems, but it may be that they are willing to accept the limitations and gaming. To paraphrase, ... it is beyond ridiculous ... to declare unequivocally that they do not use this data.
And did I state they "do not use this data", or was my suggestion that they were not using it for ranking.
They would not collect it if they did not think there was a use for it. And it is easy to see QA uses. There are just too many problems to use it as part of the algo itself.
They might even be using it to help decide if their snippet creator is doing a good job or not.
And it is easy to see QA uses. There are just too many problems to use it as part of the algo itself.
Who is to say that QA isn't part of their algo? And that QA isn't automated in any way?
There are so many problems with using traffic patterns for ranking, that it is beyond ridiculous.
...
On the other hand, it makes a lot of sense to use that information, compined with human review, to do QA on their results.If a link is never clicked, it is worth looking at the result and the site.
...
And did I state they "do not use this data", or was my suggestion that they were not using it for ranking.They would not collect it if they did not think there was a use for it. And it is easy to see QA uses. There are just too many problems to use it as part of the algo itself.
You really cannot have it both ways. If the QA that you allude to changes the placement of a single result by so much as one position or eliminates it altogether, it is by definition a change in ranking. It does not have to be part of any particular algo to be part of the process and the end result.
In the end the algo does not result in referrals, the final end result of all algos, manual intervention and QA activities producing the page displayed to the searcher produces referrals.
+++
Javascript:
function clk(el,ct,cd) {if(document.images){(new Image()).src="/url?sa=T&ct="+escape(ct)+"&cd="+escape(cd)+" &url="+escape(el.href)+"&ei=s9wYQsKkHMnwwAGisbDNCA";}return true;}
Search Result Links:
<a href=... onmousedown="return clk(this,'res',7)">
Which information, and how much is returned to Google as a result of the onmousedown line?
I take it they can see that I clicked on www.example.com.
From their cookie, they know I'm the same guy who clicked on www.widget.com a little earlier.
Is there anything ELSE they can tell?
Suppose I clean out my cookies (being paranoid) between each page visit,
or disable cookies entirely for the session? -Larry
Unless they're doing something in the /url file, I don't think that they're bothering with cookies, so they're not tracking individuals.
All of which adds up to what they say in their PP - they track user trends for QA (do a lot of people click on "similar searches"? Do a lot of people go to the next page?). Hey, they've got to be able to prove it when they claim that 80% of users never get past the first result page, or that 60% click on the first 3 listings.
Unless you know how these numbers look for your industry and your kewords, the fact that on average 50% go past the 1st page means absolutely nothing.
