Google patents search that tracks your mouse moves Google has patented a system that displays search results and ads based on where you move your mouse. Mountain View first filed for the patent — dubbed a "system and method for modulating search relevancy using pointer activity monitoring" — in February 2005, and the US patent office rubber-stamped the application earlier this …
"Should Google start taking such behavior, we can assume that somewhere, someone will attempt to game the system. You might call it hover fraud."
You could also assume that various lawsuits would ensue (except, of course, in the UK, where invasion of punters' privacy seems to be not only tolerated, but rather de rigeur...ref. Phorm), and that access to ixquick would explode. (I hope the Dutch have enough servers lying around...)
Yes, the onMouseover event will tell them where my pointer is, but it won't tell them where my eyes are. I tend to move the mouse out of the way of what I'm reading and use the scroll wheel to move up and down the page, then I move it to one of the links on the page, one of the form elements, the back button or the address bar as appropriate when I want to go somewhere else or leave a comment.
or whether it does. It works like this:
Google: Hey, Mr. Advertiser - we have this cool new thing: if a (l)user's mouse hovers over your link for more than X, we'll serve them up a whole bunch of ads for your product! All for just $1 per ad!
Adverdroid: I just ejaculated in my pantaloons. Where do I sign?
As long as they are gauging a surfer's "interest" by the mouse hovering over a particular link or bit of a search snippet, there is absolutely no reason for them not to set a threshold past which they assume that you're sufficiently interested to automatically open another window.
"Well, our research shows that users who hover over a search result for 'X' amount of time are likely to click on it -- we're just saving them time.
"... and (artificially) increasing your click-through rates. Pay up!"
When I'm bored, or idly reading without 100% interest, I whizz my mouse all over, randomly select/unselect text... drives my mother mad if I'd doing something for her.
But, come on... monitoring where the user moves their pointer? F*** off already.
That flowchart represents just about every dynamic mouseover menu I've ever coded, literally, line by line. I'm not in the ad business, but it's the same technique.
I can say I had the idea well before google, as I'm sure plenty more have well before me.
It's neither novel nor does it constitute more than a half hour of R&D. What benefit does the world get by granting this patent? Absolutely ridiculous.
Chrome isn't the issue, since this can already be done via Javascript in any browser. With Google Analytics already linked-in on millions of pages, a modification of that scripting would automatically extend the functionality to them, if they so desired. I doubt it will happen that way, but I also doubted Sarah Palin would be more than a blip on the political radar. And she's gone from blip to blob.
For the luddites who have Javascript turned-off in their browsers, bravo, and keep using tin snips to open your canned peaches.
This can all be done with ajax and such, and good adblocking or javascript disabling will foil it, yeah, but there's a bit of paranoia that says, "What's stopping the same logic from sneaking into Chrome at the binary level, where it reports even if you have adblocking or the page doesn't even use Google?"
There is also some useful information obtainable by determining which finger is used to point at a certain advertisement or search result. Index finger = casual/normal use. Pinkie = for more detailed work, smaller ads. Ring finger = dimwit user. Middle finger = Microsoft or political ad.
All browsers with javascript and an xml object can do this - basically any modern browser.
I think you'll find some sites you have visitied do this too, if it is of interest to them they will collect it.
I noticed an insurance site sent off an ajax call to collect my email address the second I'd finished typing it (it becomes valid = ajax call). So I didn't even need to hit the submit button for them to have my details.
I thought it was a good idea, this is no different - you don't want Google to know what you are doing, run adblock and noscript. If that's too hard turn the computer off.
Interesting that. I remember some discussions with a client a long while ago who wanted to do just that. We came to the conclusion that it might be a bit iffy, privacy/DPA-wise. In clicking a submit button you are consciously sending that data through, can agree to any terms and conditions, etc; but sneakily grabbing the email before I have necessarily decided I definitely want to send it doesn't seem right...
I'm sure some laywer could formulate a "in viewing this website you are agreeing to our terms which allow us to do what we want" type thing, but we weren't comfortable with it.
See: I have a scroll wheel on my mouse. I tend to put the mouse well to the right of the window, away from the text I'm reading, so I can SCROLL with it without obscuring the text I'm reading.
Bet I'm not the only one.
All that said, once you think of something obvious, you might as well patent it even if your proposed implementation is useless; just so you stop anybody else doing anything genuinely useful with it.
Way to go USPTO. ....Hey that rhymes. Maybe we should put it on a shirt?
Where your mouse is located is not relevant.
Google does not do "if your mouse is at X,Y we will display Z. They chuck raw data into a big number cruncher and from there on it is only a matter of time until it will identify patterns that predicate you doing something of interest to their "customers" (they use Bayes stats so using the word "correlate" is not correct here).
Then you will be fed an offer you supposedly cannot refuse by one of their customers. Voila. Job done. Breath in, generate revenue, breath out, generate revenue yet again.
Not sure everyone posting comments has understood what Google are talking about, it's nothing to do with automatically sending you to a site just by hovering the mouse, not is it like a JS dropdown menu.
It's just about them collecting another stat on their search results (and will probably apply this to analytics too)
My interpretation of this is they'll have an area around each link/search result, when you move your mouse into one of these areas it'll start a timer, when your mouse has been there for a certain minimum period it'll send a tracking message, via ajax probably, to Google. I think I could probably code this in 10 minutes so i'm sure someone somewhere has prior art on this one?!
"Not sure everyone posting comments has understood what Google are talking about,"
I don't think you understand what everyone's posting about. :) The thing is that many people put the mouse where they're not looking so that it doesn't get in the way. In my case that would be on the far left, or right, or white space, or scroll bar. Where does the "area around each link" come into it?
Sorry, I was only really talking about a handful of comments that seemed to misunderstand the implementation.
By the sounds of it, they will only be collecting data about certain zones on a page, I assume those zones will be links in search results and the like, so people like you and me who would rather move the mouse pointer out of the way while we read/scroll through search results will probably not give them much in the way of tracking data anyway.
Why would you assume anything of the sort? There is no reason to beleive that Google will limit itself to any particular region of the screen. After all, if Google is able to determine that we move our mouses to the left/top/bottom/right of the screen in order to read the page we are viewing, then they will likely start putting ads there so that they can trigger those ridiculous expanding ads, thus blocking our view of the information we actually want to obtain with that which we do not.
Mouse position isn't nearly accurate enough. The next step is to take over the user's webcam and track their eye movements. Also monitor their expression. Pity there's not yet an easy way to measure their pulse rate etc.
After all, that's what the user is for - to respond to ads. They'll obviously be perfectly happy to not bother about this silly 'privacy' thing as long as they can be monetised.
I was going to post exactly the same thing.
I would say 99% of the time if my mouse 'hovers' over a link then I'm going to click it. I can't see how that information would prove usefull.
Though I can see a point when genetically engineered Google monkeys follow everyone around and report our movements back to HQ - 'He's looking for socks - quick sell him socks!'
This post has been deleted by its author
Out of interest, I noticed my default habit is to click a link and then move the mouse out of my eye line of the text I am reading and use it primarily to scroll until I spot a link I want to click on. I also notice where I move it, usually to the right, effectively hovers over where a lot of websites situate ads and other auxiliary items that I guess Google will therefore falsely think I am interested in.
Isn't there already a sort of prior art in this case seeing as how there's lots of informational sites out there that have words in their text as links and as soon as you hover over them, up pops another ad?
I suppose also I might as well stop hoping to own a dev shop someday because by the time I have a product, every line of code we write will be covered by at least one patent troll.
Can I patent the hello world example?
I thought there would also be prior art, however after reading the patent fully I now understand that it's specifically relating to changing the order of the search results weighted by which results a user has hovered their mouse over on previous searches, which probably hasn't been done before!
I use a Kensington trackerball mouse with a ball about 3 inches in diameter. It is easy for me to find I'm spinning and rolling it idly with my fingers while I am staring into space, thinking. The cursor makes many interesting, erratic, random journeys. Will I keep on doing this if Google puts their new patent into practice? You betcha!
Google is not the product, the information they provide is not the product, the links are not the product, YOU ARE THE PRODUCT!
You and all the information that makes up who you are, that is the product. To be harvested and sold off to the highest bidder. Same with FaceBook, Twitter and millions of others. The site and it's services are not the product, the harvesting of you, that is the really valuable product.
Google's search rankings are already crap because they are worked out on two things firstly Google's commercial interest and secondly by what other people have clicked on.
I'm not other people I'm me and I'm searching for what I want. Try searching for a specific phrase on google, say half a dozen words that you know appear on a particular web page but are not a common phrase. Now see how far they come down the Google ranking. Look at some of the stuff that comes above it that doesn't contain the exact phrase you searched for.
Google might be great if you wan't to buy a new laptop, but it's not much use when it comes to searching for information. Google see the interwebs as a commerce tool and little else. I see the net as a great source of information and, as such, seldom use Google. And before anybodu starts this is about usefulness not privacy so I don't see the point in using any of the proxies and the like that hide your identity from Google.
Who reads the screen with their mouse. I read the screen and then when I'm ready to click I move the mouse to the link I want to click. So my mouse stays where I last clicked. This will really skew their results since I may not even be mildly interested in the location of the mouse on the new page, but it will stay there until I have read the rest of the page.
People who read using their mouse probably trace the printed word with their finger tip and move their lips for the difficult words. What are Google going to call this tech? Google Sunreader?
Oh and those of you worrying about what you will do *when* Google implement this tech. How do you know they haven't done this already? Just because they have only recently applied for a patent doesn't mean they're not using it already. Get your tinfoil hats on.
Google have been monitoring user mousedowns (mouse clicks on results) in their search results since Nov 2003, so taking this one step further by monitoring mousemoves isn't really anything new.
To see what Google have been doing with mousedowns since November 2003, search on Google for:
google click monitoring onmousedown return clk
and you will find some articles about this.
Whether monitoring mousemoves is useful or not is another story. How many people read with their finger, (or the mouse cursor equivalent), and what about the new touch devices like the iPad and tablet competitors, and the touch phones etc, all of which do not have the equivalent of a mouseover (at least not when the finger is not in contact with the touch surface, i.e. not dragging).
Say I want to see who's posting a particular phrase over and over, on certain social sites. I google for "phrase" .
After opening up about 50 of these my browser slows down a little, so I may spend the threshold amount of time with the mouse over a result I have no interest in. Of course I am scrolling with the mouse over the centre of the links, to make it easy to open one when I find one for the site of interest.
Does this mean I have any interest in that site or what it stands for? nope, my interest is in the person(s) posting my keyword.
Now, if Google could come up with a Preference setting so it won't deliberately refuse to give me ALL the results (we have omitted similar results........ HEY I asked you to SEARCH, and I will decide if I like them or not). Another graduate of the MS "we know what you want' school.
Speaking of MS, they have prior art on this.....the Office menus that reorder their content according to what you used more often most recently............bad luck you are just about to switch to a different project with different requirements.
Google did not invent the mouse, so why should they have a patent related to mice? Or better yet, why should they have a patent related to what I do with a mouse? I know it is about tracking for marketing and not just using a mouse, but still......
One more reason to not use google or any of it's products or services.
The US Patent office is broken, someone needs to patent fixing it.
...and applying some common sense would go a long way to targetting adverts properly. Has anyone noticed when abroad that going to an English language Google site and typing in a search query returns adverts in the language of the country it thinks you happen to be in? Talk about missing the target when you don't even recognize the alphabet that the adverts are written in!
For sheer stupidity this has to rank alongside "Keyboard not found. Press F1 to continue".
This news, coupled with the recent extension of Google's images search result user interface (UI) to offer more information when the user mouseovers an image, does suggest that perhaps google might be changing its main search results UI to offer more information when the user mouseovers a result in the future.
That would be interesting if google are toying with the idea of changing the main results UI in this way, quite a different change to something that has been fairly static and lacking in the past in terms of a rich interactive user experience.
Although offering more information to mouseovers still ignores the touch screen phone and touch screen pad/tablet users where there are no mouseovers.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
