In the past weeks, I have met with mobile carriers in Latin America, Europe, Africa, India and the US. The conversations have taken us in a very new direction – how to address competition from OTT applications and deal with the revenue loss from enthusiastic uptake of these OTT applications.
In the competitive mobile market, more than 100% penetration is becoming the norm: many customers have more than one phone and several different types of service. Mobile carriers now have to worry far more about customer churn and revenue leakage of any kind.
Yes, operators are worried about fraud. The rapidly morphing cocktails of subscription fraud, account takeover, SIM swap, never pay, etc. are taking their toll on profitability. Without advanced solutions, the communications industry will continue to lose billions of dollars a year – $38 billion by the CFCA’s latest reckoning – some of which they can’t even account for since it slips past the capabilities of rules-based anti-fraud systems.
Fraud, however, is just one part of the conversation. We are called upon because we have the industry’s first big data/machine learning application, and our performance levels at 350% improvement are off the charts in comparison with other fraud detection methods. But it’s not fraud that sends the analysts running to bring their colleagues into our meetings.
The data plane has taken over from voice as the major traffic hub on mobile networks. When we developed our application it was in recognition of the fact that more and more fraud will be perpetrated on data services, but also that there is a broader and increasing need, beyond fraud, for real-time, big data-based network analytics in the mobile operator data center. We are the only company providing deep analytics on the data plane – and it is here that the next big challenge resides.
OTT (over-the-top) IP-based applications pose a very serious threat to operator profitability. Call termination revenues are dropping sharply. Providers worldwide recognize that they need a way to manage OTT adoption without banning it altogether and alienating the entire generation of millenials. But to monetize OTT, carriers need a means of detecting OTT traffic. This is where we have been able to prove some unique capabilities.
At the request of operators trying to understand the scale of Skype, GoogleVoice and WhatsApp usage, we have used our DPI and machine learning engine to run multiple trials to detect OTT traffic on the data plane. The results were eye opening. Our tests found interesting insights into the traffic. It’s not just the well-known, popular apps that operators need to worry about. New applications such as QQ, Viber, WeChat and others have already gathered millions of users.
In order to solve the OTT dilemma, communications providers have no option but to find out what traffic is running through the data plane. Through our OTTExplore feature, we are able to identify types of OTT service used, volumes of usage by subscribers, and even call times. This information allows operators to develop competitive and appealing offerings for subscribers while stemming the OTT revenue bleed.