Tracking what you click on has been one of the fundamental pieces of Web analytics. But your clickthroughs represent only part of what you actually do online. Eye-tracking studies have often been seen as the best way to determine what people are actually thinking as they browse, but these sorts of experiments – until recently – have been either technology- or cost-prohibitive for many people.
But now researchers at Microsoft may have found an easier way to track where people are looking as they browse the Web. The new process doesn’t actually utilize eye-tracking hardware, but rather uses the position of the cursor as a stand-in – where your cursor moves, where you hover, and of course sometimes where you click. According to their research, the cursor’s position as actually a pretty good sign of what you’re looking at and what’s important, particularly when it comes to search results.
By looking at cursor data at scale in conjunction with click data, the researchers contend that this information could help improve the search experience. Cursor data can be captured for uncommon search queries, for example, where relevance is hard to gauge because of infrequent clicks.
Also, it can help identify what’s described as “good abandonment,” when the search itself satisfies the user’s query and a clickthrough is unnecessary. (This is different than a bad search abandonment where the user didn’t find anything they were looking for.)
Before starting off a panic here about another new way in which websites are tracking us, it’s important to note that the results of these studies are based on cursor movements at scale. In other words, this isn’t about tracking an individual user’s intents, but rather about analyzing those based on a large number. That might mean that this sort of thing becomes part of a standard analytics package.
But for now, it’s just a proof-of-concept.
Photo credits: Flickr user Stuart Pilbrow