Information about where searchers hover their mouse pointers over different parts of search results, as well as advertisements and Google Onebox results, may be collected by the search engine to be used as ranking signals to determine in part how relevant those items may be seen by Google users in response to a search query.
When I view the contents of a web page, I often find myself moving my mouse pointer along the areas that I am viewing. There are a couple of reasons for this. One is that it makes it easier to focus upon the part of the page that I’m looking at. Another is that it’s easier to click upon a link that I find interesting if my pointer is near what I’m viewing.
According to Google, I may not be alone in this kind of behavior. Google may track mouse movements on its search results pages to help rank pages that show up in search results, to determine the quality of sponsored ads within those search results, and to decide whether or not showing onebox results such as maps or definitions or news or stock quotes is appropriate for some search queries.
When Google ranks web pages, it considers a wide range of ranking signals, such as how relevant a page might be to keywords used by a searcher, the quality and quantity of links pointing to that page, and user-behavior data collected about that page.
A number of patent filings and whitepapers from Google have told us that Google might collect a fair amount of user-behavior data about how we browse web pages such as; how long we might spend on pages, how far we might scroll down those pages, which pages we might click upon in search results, which pages we might not click upon, which links we might follow when we visit pages, if we print or bookmark or save pages, and more.
A newly granted patent from Google explores how they might look at how we move our mouse pointers on search results pages as a ranking signal.
Problems with Click Through Rates
One method of ranking search results based upon user-behavior is to see which pages people click upon when they perform a search at a search engine, and which ones they don’t. But, there’s a potential problem with that approach.
Let’s say that you search for Barack Obama’s birthday in Google, and the top result shows the birthdate in Google’s snippet. There’s no need to click through to the page, and your informational need is satisfied. Google probably wants that page to continue ranking well, but if the search engine relied upon click-throughs to measure the relevancy of a page for a query, then a lack of clicks to the page would make it seem to be not very relevant.
If Google instead looked to see that people hovered their mouse pointer over a snippet that contained the President’s birthdate, and then moved on to a completely different search, Google might consider that page to be very relevant to that query.
If that search result was the third or fourth listing in that set of search results, but a large number of people hovered their mouse pointers over that snippet, and then went on to a new set of queries, Google may rerank that particular search listing and move it to a more prominent place in the search results, possibly even to the first listing.
The Google patent is:
System and method for modulating search relevancy using pointer activity monitoring
Invented by Taher H. Haveliwala
Assigned to Google
US Patent 7,756,887
Granted July 13, 2010
Filed: February 16, 2005
A method and system of modulating search result relevancy use various types of user browsing activities. In particular, a client assistant residing in a client computer monitors movements of a user controlled pointer in a web browser, e.g., when the pointer moves into a predefined region and when it moves out of the predefined region.
A server then determines a relevancy value between an informational item associated with the predefined region and a search query according to the pointer hover period. When preparing a new search result responsive to a search query, the server re-orders identified informational items in accordance with their respective relevancy values such that more relevant items appear before less relevant ones.
The server also uses the relevancy values to determine and/or adjust the content of an one-box result associated with a search query.
The Value of Mouse Movements and Placement
When you are looking at a set of search results, Google may track where your mouse goes on their results page. They tell us that:
A typical user’s behavior is to move the mouse pointer (or any other pointing indicator) over or near a target informational item, keep the mouse pointer there for a period of time while the user reads the item’s information (e.g., title and snippet), and then click through the underlying link or move to another item.
Sometimes, a user may review multiple informational items responsive to a search query, moving a pointer over or near each of the informational items that the user reviews. These various pointer activities can provide another way to evaluate the user’s feedback with respect to a particular informational item.
The patent presents a couple of assumptions about how mouse pointer movements can be interpreted:
For example, a longer hover over a result may indicate a positive opinion about how relevant a listing on the results page might be to a query.
And, if someone moves their mouse pointer across a snippet line by line at a normal reading speed, it may indicate a higher level of attention to that result than if pointer was kept in a static position or moved randomly.
So, the speed and movement of a mouse pointer as well as where it is placed on a search result page might be tracked to see how much attention a search pays to different search results. If someone hovers over one sponsored listing, or ad, but not another, that might indicate more attention and interest in the ad hovered over. If a local map is shown, or a definition, or some other OneBox result, and the searcher viewing the page hovers over those OneBox results for a while, that could be an indication that the map or the definition or other OneBox listing was helpful.
Client Attention Data
The patent refers information about the tracking of mouse pointer movements as “client attention data,” because this kind of measuring of browsing activities can give the search engine an idea of how much interest there is in different parts of a search result page, and how much attention visitors paid to each of those parts. If there are similar patterns about how a large number of viewers interacted with a page, that data may provide some meaningful information that can be acted upon by the search engine.
The patent also tells us that it might give different weights in determining a relevancy value for mouse pointer movements based upon different areas of a result. If someone hovers over the title to a search result, that might carry a different amount of weight than if they hover over the snippet of a result.
This patent was originally filed in 2005, and it’s possible that Google may not be using the methods described, or tested those methods and have since moved on to other ways of tracking searcher’s attention on search results pages. It’s also possible that Google may be using those mouse pointer movements today.
Looking at how someone may move their mouse pointer across a page does provide more useful information to the search engine than just looking at what items on a search result page that someone might click upon or not click upon. We haven’t seen too many patent filings from the search engines that go into so much depth on how they might measure one specific type of user-behavior and interpret it, like we do with this patent.
This patent was also filed before Google introduced Universal Search, and when it mentions OneBox results, it means those special results from other data repositories from Google such as Maps, News, Stock Quotes, etc., that were often shown at the tops of search results rather than blended into search results. It’s possible that a mouse pointer tracking approach could be used by Google to see how effective or useful blended results might be when they are located in places other than at the top of a set of search results as well.
If one sponsored listing, or ad, is hovered over for a while by many viewers, while others aren’t, should that play a role in the placement of an ad? Is that kind of user-behavior part of the quality score for sponsored listings?
Added (7/13/2010 – 12:30): Interesting observations on the Acuity blog about a June 2 presentation at Eyetrack UX in Belgium by Google Senior User Experience Researcher Anne Aula – Eye Gaze Data and the Correlation With Mouse Movement.
Added (7/14/2010 – 9:43 am): A paper to be presented next week at SIGIR’10 in Geneva, Switzerland, from researchers at Emory University is also worth a close look – Ready to Buy or Just Browsing? Detecting Web Searcher: Goals from Interaction Data. It describes how user-behavior data such as mouse movements and scrolling on search results pages from search engines can be used to help understand the intent behind a search.