Online gaming companies are sizing up their top dogs to determine which players are worth the most, according to one analytics company executive.
Using a learning algorithm, online social gamers are being ranked by their estimated dollar value, a number that takes into account a user's contacts, specific transactions and other behaviors, according to Ninja Metrics, a company that performs predictive analytics.
RELATED: Retailers use big data in hopes of bigger sales
The algorithm can be used to calculate the so-called social value of users in nongaming environments too, said Ninja Metrics CEO Dmitri Williams.
But in order to be the most accurate and useful, it must be applied to sites and apps where transactions are frequent, like movie ticket or song sales, Williams told "Big Data Download."
"Because there's an algorithm here that has to learn something, the thing has to happen more than once or twice for us to really feel confident about it," he said.
RELATED: How Health Care Is Mining, Using Your Data
In online gaming, currently the only environment in which the algorithm is being used, the predicted value of individual gamers isn't perfectly accurate.
By comparing earlier predictions of a gamer's social value to actual outcomes, Ninja Metrics has found that its algorithm is 90 percent accurate on average, Williams said.
"Social value is useful in acquisition, retention and monetizing current customers. If you know a customer's lifetime value and their social value, you'd know who to go after. You'd know who to work harder to keep," Williams said.
The Ninja Metrics algorithm is just one of many that try to predict the value of individual Web users as influencers or consumers, however. Tech giants like IBM and SAP also have their own predictive analytics software used to boost the bottom lines of businesses big and small, online and offline.
-- Comments, Questions, Suggestions? Tweet Us @BigDataDownload