Blog

Breaking the bottleneck of Fraud Management

Breaking the bottleneck of Fraud Management

This article, written by Joseph George (Senior Vice President, Fraud & Security - Mobileum) first appeared on Telecoms.com in August 2017 (Original article here).

Telecoms.com periodically invites expert third parties to share their views on the industry’s most pressing issues. In this piece Joseph George, Senior Vice President of Fraud & Security at Mobileum, argues operators need to revise their approach to countering fraud.

We live in a golden age of data. For operators looking to counter fraud, there has never been more actionable information available at their fingertips, than there is now.

In theory, this is an amazing advantage for communications service providers (CSP) looking to stay one step ahead of increasingly sophisticated ‘fraudsters’, as well as gain insights that can help their business thrive in new service areas. A wealth of useful data, an increased ability to fight fraud, and a way to add to the bottom line all are all wonderful things for operators. However, this explosion of data has also created unforeseen challenges too.

Operators are reaching a tipping point, as the telecoms sector significantly expands its global services with faster broadband, 5G roll-out and more connected devices. Combined with the overall trend of traffic moving from voice to data networks, CSPs are facing huge challenges as many fraud management systems are buckling under the weight of trying to detect and act with the speed and accuracy needed to prevent potential revenue losses.

Simply put, many traditional fraud management systems can’t keep up with the sheer volume of data out there. It’s leaving operators staring at a mountain of overlooked (and underutilized) data, too much of a pain and inconvenience to be analysed thoroughly.

The root problem with many systems is that they can only handle limited datasets, not accounting for volume, variety and velocity of critical data. Also, modern capabilities and features are missing in older systems, including mobility, machine-learning, self-service analytics, and more visual and intuitive interfaces. In fact, some legacy systems still in use today by CSPs monitor fraud by only analysing aggregate records of calls.

There is a real, looming threat that fraud management is becoming a bottleneck, impeding CSPs’ ability to offer and expand services until fraud data can be interpreted and managed. Like a clogged kitchen sink, fraud management is creating a backup. Initially it might just cause a small pipe leak, but if operators aren’t proactive, they could have a full-blown burst on their hands.

That burst may be caused from the pressure of a mounting catalogue of services (and data) CSPs are involved in the delivery of, which they also need to analyse. Although outside their control, CSPs are often best positioned to identify instances of fraud occurring over carrier traffic on their networks. Examples of this include data fraud, international revenue share fraud and bypass fraud, among others. Along with the risk of IoT and sensor networks having fraudulent apps installed, the result is that the blind spots of many current CSP systems are being exposed by emerging sources of fraud.

So how can operators get ahead of the problem? How can they break this ballooning bottleneck, take advantage of that fact that they have access to vast amounts of data, and expand their services? The first step is to go beyond merely detecting fraud. CSPs should look inward, circling back and advancing their fraud protection tactics.

CSPs should have integrated, actionable and prescriptive control of fraud and abuse, based upon a combination of dynamically auto-configured business rules and policy control. By obtaining a high degree of detection accuracy, operators can get a clear understanding of the fraud data they are being presented with, and what it is telling them. With IoT for example, it means having an ability to uncover fraud outside of rule-based detection.

The implementation of predictive, big data technologies and machine-learning is a way to keep up with new frauds in real time, stopping it in its tracks. It also offers the added benefit of creating more parameters and making greater volumes of data available for analysis. All of this can be accomplished by employing a comprehensive multi-protocol solution that is nimble, fast and adds to an operator’s current system capabilities.

It’s no secret that the telecoms sector is significantly expanding its services and capabilities. But it’s the savvy operators who realize that breaking through the bottleneck of fraud data saves time and money in the long-term, and facilitates investment in new opportunities and services that otherwise would have been missed.