Fraud Detection analysis is a modern method of protecting telecommunication companies from a range of frauds and other illegal actions aimed against them. The most frequently encountered hazards are false identities of customers attempting to sign contracts. Other types of hazards encountered in the telecommunication sector are break-ins into cellular network, security violations of computer systems, illegal telemarketing, identity thefts, and abuses in the pre-paid segment.

StatConsulting offers models for early detection of hazards to support the operations of fraud detection divisions. Our models can be routinely applied to analyze new customers' applications and select the ones characterized by elevated fraud risk. The selected applications can then be transfered to experts for individual verification.

Fraud Detection models are also used for monitoring the account balance of post-paid subscribers during billing periods. By using a scoring model it is possible to determine for each subscriber a threshold below which the risk of amount due is acceptable. The exceeding of this threshold may indicate that an unauthorized use of the service by a third party has taken place. Early warning about excessive growth of amount due allows the operator to take some sort of preventative action, such as restricting the access to services or contacting the subscriber.

The employment of Fraud Detection models increases the effectiveness of customer acquisition by streamlining and shortening the process of verifying customer data. The application of the model allows to significantly limit customer acquisition budget while maintaining high standards of protection against abuses. The solutions offered by StatConsulting enable online monitoring of the accounts of activated subscribers for illegal operations. This is of special importance in the case of cellular networks, which may have millions of subscribers, as it significantly lowers operational costs.