Mitigating VoIP Fraud Using Real-Time Data Analytics

Providers already have prohibited these types of activities in the terms and conditions presented in most contracts. Vonage Holdings Corp., for example, gives itself the right to terminate any customer at any time for what it deems a misuse of service. Many other operators make similar statements. Yet relatively few have been able to enforce them, primarily because they have not had the resources in place to identify fraudulent behavior amid millions of calls per day.

It’s even more unfortunate that some VoIP providers choose not to track individual calls at all, seeing it as another expense that cuts into their already thin margins. Ultimately, the lack of record keeping may cause VoIP providers serious legal problems down the road, particularly if the United States Justice Department or Department of Homeland Security, or in the case of publicly held companies, the Securities and Exchange Commission, requests information on specific callers or incidents. In fact, federal and EU regulations require that all communications providers — even international services like Skype — maintain strict financial and operational records. Failure to comply presents bleak implications for the VoIP provider, such as stiff fines or even jail time for executives. Compliance is serious business, and fraudulent activity jeopardizes a provider’s ability to comply — pure and simple.

Fraud Mitigation Strategies

While reviewing call detail records to identify fraud may seem like a cumbersome and expensive process, this task can be accomplished relatively cost effectively for most providers through the proper implementation of data analytics solutions.

For instance, there are several advanced analytics applications in the market place that give operators visibility into call activity and behaviors. These solutions develop algorithms for determining “typical” customer behavior. The algorithms measure several details, such as number of calls, length of calls, time of day, and how many numbers customers typically call. Conducted in near real time, these applications provide a snapshot of how particular services are being used by customers.

Many of these analytics solutions can be integrated readily into existing core CRM systems, so there’s no need to overhaul or forklift the infrastructure. In addition, some providers offer these tools as a managed service, relieving the burden of integrating new technology.

As part of the evaluation process, the analytics solutions will measure typical usage, and determine what average user behavior looks like. Operators then can set their own threshold markers to help define acceptable and reasonable usage, as well as parameters that could resemble fraudulent activity.