Facebook wants to peer a little deeper into how you spend your time on a PC or mobile device.
According to a new report in The Wall Street Journal, the company could begin siphoning up data related to how long a user keeps the Facebook newsfeed visible on a smartphone or tablet, or even how a PC cursor moves over content. That collected information could find its way into Facebook’s broader data-analytics efforts, and possibly inform how it sets ad rates or builds new features.
Facebook analytics executive Ken Rudin told the newspaper that the tests would enhance the company’s store of behavioral data, but that any real implementation is months in the future—if it ever happens at all. “I can’t promise that it will roll out,” he said. “We probably will know in a couple of months.”
Along with Google and some other large Internet firms, Facebook is one of the world’s largest collectors of online behavior, to the point where it needs to build an impressive amount of custom software to keep up with it all.
Given all that data sloshing through its systems, Facebook executives also have some real concerns about figuring out the most efficient ways to get actionable insights from massive datasets, rather than spend too much time deciding which tools are useful and which are not. At the recent Strata + Hadoop World conference in New York City, Rudin suggested that companies’ reliance the popular Hadoop framework was useless unless it translated into very real results. “The problem is that Hadoop is a technology, and big data isn’t about technology. Big data is about business needs,” he said. “In reality, big data should include Hadoop and relational [databases] and any other technology that is suitable for the task at hand.”
In order to handle all its data, Facebook has built platforms such as Corona, a scheduling framework with an integrated cluster manager for tracking nodes in clusters and the free resources in play across the network. It also employs teams of analysts with PhDs and employees trained in best data practices. But there’s always a need for better information to improve its operations—hence, these new, micro-focused tests.