The churn rate in the US telecom industry was 1.28 to 2.19% in third quarter, 2014, according to Statista. When you multiply these rates against the millions of telecom subscribers, lost revenues, and the job of acquiring new customers, revenues become major issues. Churn is the primary reason why telecom companies that provide wireless and other services are looking to real-time big data analytics for a competitive edge.

“We serve telecom companies in 51 different countries around the world,” said Abhay Doshi, vice president of products and marketing for Flytxt, a big data analytics solution provider for Communication Service Providers (CSPs). The purpose is to monetize this big data intelligence in telecom up-selling and cross-selling to customers.

“A telecom has abundant customer data stored in databases that is available to it,” explained Doshi. “This data comes in the form of information that is stored in customer relationship management (CRM), billing, and order system transactional data, and it provides an excellent historical resource on customers. However, if you want to know what a customer is doing at a present point in time and you want to respond to that behavior, you need real-time, actionable data.”

Flytxt uses in-memory, real-time analytics technology developed by VoltDB, which uses specialized SQL databases that are tuned for both subject matter and performance, and that outperform traditional databases. The combination of in-memory processing, the scalability of SQL, and full streaming capability enable VoltDB to process real-time Internet of Things (IoT) data from around the world for customers such as Flytxt.

“This real-time data enables us as an analytics provider to give our telecom customers real-time information on their end customers — such as when a prepaid data balance drops,” said Doshi.

In these instances, the telecom can automatically send customers text messages that notifies them their data allocations are running low. The real-time offers enable customers to renew their data supplies on the spot and at a discount from their mobile devices. The end result is that revenue is boosted for the telecom, thanks to anticipatory selling. The risk of losing a customer also dramatically drops.

“This is an important tool for us,” said Flytxt’s Doshi. “The average telecom mobile subscriber spends one to two dollars per day on voice and data services. Without a real-time means of monitoring and responding to this usage, it would be impossible to deliver offers for service renewals in ways that match up with the customer’s actual demand. Instead, telcos would only first see the drop in data services that a customer has purchased on the following day.” By then, the window of opportunity is lost.

Is it working?

Doshi cites Vodafone, which he says was able to increase revenue by one to two percent, thanks to real-time anticipatory selling propelled by big data analytics.

“The focus is clearly on customer retention and also on understanding the customer experience from where the customer stands,” said Doshi. “Big data analytics are helping telcos because they can see where a customer currently is at in real time with respect to his voice and data usage — but they can also see this information against a backdrop of historical information on the customer, and against information on other customers who have similar buying and usage patterns.”

The ability to combine real-time big data analytics with data that originates from more static database stores within the enterprise is still a work in progress for many companies; this will likely change as more companies architect their transaction-based, stored, and real-time data collection and usage, and then optimize all of it for the benefit of the business — like anticipating customer “buys.”

“One reason we have enjoyed success is because of the unique analytics model that we offer, along with our ability to use sensor-based technology to capture what is going on in real time,” said Doshi. “The other peace of mind that we offer customers is that we assume complete responsibility for the monetization of their data through our methodology.”