Every action you take on Facebook can metamorphose into a statistic; assigned and categorized according to behavior. And no matter what industry you’re in, Facebook’s behavioral data looks mighty juicy. We’re hungry for data’s insight into human nature, as illustrated in Pew’s recent study: the American Life Project.
You could glean an image of social generosity just by looking through the report. Over the course of a month, some 40% of Facebook users sent out a friend request, while 63% received at least one. Personal messages follow a similar pattern, sending an average of 9 messages per month, but getting 12. And don’t get started on photo tagging. Thirty-five percent of Facebook users get tagged, but only do 12% of the tagging.
It seems a little backwards, to receive so much more than you give. But it’s not surprising, given you’re just one person to take all those actions, while multiple friends will collectively return more actions. That receiving action pool gets even bigger when you consider virtual connections beyond Friends. But there’s far more to understand about Facebook behavior than the first layer of the social graph.
Every major social outlet, online or off, provides a unique data set. But Facebook’s drawn a particular concourse for future potential. This is a prime target for emerging big data solutions, all contributing to the creation of Facebook’s a big data story. There’s a great deal Facebook wants to do with their data, serving up reports on ad returns, user demographics and relational results. An important platform leading the charge is Hadoop.
“If you look at Hadoop for as long as we have, you know it’s important–it’s a good choice,” says Pete O’Leary, VP of Customer Operations at Quantivo. “We get a lot of customers with a lot of data that have Facebook-like questions, but don’t have access to the resources. So they come to us.”
While Facebook juggles every facet of running a global network, an industry emerges from Facebook’s big data story, generating chapters all their own. Many fill in the gaps Facebook hasn’t gotten to yet. Focusing on the real-time aspects of relational data is a surge in behavioral analysis, understanding more about social interactions based on a given business’ needs.
Where behavioral analysis meets BI
Quantivo specializes in performance analysis solutions, with a hosted platform dedicated to the management of large volumes of data. It slims down the time you’re able to see the fruit of your labor and make changes accordingly. With a dearth of operational reports based on historical data, Quantivo’s also picked up on a trend that’s made its mark on BI. Tap past site performance, user purchases and responses for a deeper look at your specific user set. Quantivo lets you run public data sets against your own user data to better understand the customers you’ve got.
Not surprisingly, the Hadoop overlay in behavioral analysis is spawning a big data story as well. “It presents the best of both worlds,” O’Leary says. “Data coming from the world (not from your system) is in ways, beyond your control, but you can take that data and apply analytics to determine patterns. What are they saying about your brand?”
Quantivo is already seeing Hadoop proliferate through its own ecosystem, layering in MapReduce and other solutions to churn up the big data results they need. We know our needs for data will only continue to change. Now it’s a matter of building the flexibility to secure our fate.
Contributors: Saroj Kar
Kristen Nicole has also contributed to other publications, from TIME Techland to Forbes. Her work has been syndicated across a number of media outlets, including The New York Times, and MSNBC.
Kristen Nicole published her first book, The Twitter Survival Guide, and is currently completing her second book on predictive analytics.
Latest posts by Kristen Nicole (see all)
- The Land of Variables: IoT’s map to monetization - September 14, 2016
- Destroy to create: How one CEO innovates in object storage, open source - September 8, 2016
- Where’s the money in IoT? Start with real-time data - August 25, 2016