Turkcell is one of the EMEA region’s leading communication service providers (CSPs), with over 30 million subscribers across its customer base. That creates an extremely complex operational ecosystem for reducing fraud vulnerability and tackling risks to revenue generation.
Fraud prevention is a major challenge in high volume telecommunications environments. A 2019 report by Europol’s European Cybercrime Center and cybersecurity firm Trend Micro estimated that telecom fraud costs EUR29bil (~USD35bil) annually.
Neural Technologies has a long history of delivering anti-fraud solutions that meet these significant operational challenges. That experience enabled us to increase fraud detection at Turkcell by 12%, reducing time between analysis and intervention by 27%, and improving collection amounts by 89%, thanks to the introduction of Neural Technologies’ powerful Optimus platform solutions.
Adapting to the requirements
As a leading national CSP, Turkcell boasted a mobile subscriber base of 33.9 million customers at the time of engagement in 2008. That customer base consisted of 44% prepaid, and 56% postpaid customers.
With a high number of postpaid customers, Turkcell was faced with the challenge of understanding and identifying potential fraud risks across more than 20 million unique postpaid users. The company was keen to introduce measures to reduce risk of bad debtors’ uncollected payments, at the same time implementing a future-proof fraud and credit management system that could adapt and evolve to changing customer needs.
This ability to adapt is critical to successful anti-fraud implementation. A customizable solution can be adjusted to specific operating conditions, avoiding the blunt approach of off-the-shelf fraud reduction solutions.
Turkcell turned to Neural Technologies for a fully customizable and adaptive solution that could be quickly scaled as and when needed. The open nature of the Optimus platform meant it could easily integrate with existing legacy systems, at the same time allowing custom implementation that reflected the operational scenario of the CSP.
The Optimus Fraud and Credit Management system provides an automated solution that is built on extensive machine learning and artificial intelligence technology. This automated fraud alert approach enables operators like Turkcell to eliminate time-consuming manual processes, freeing up department resources to focus on more complex, real-time fraud investigations.
Crucially, this also enables a more customer-focused approach to credit management, speeding up processing and detection of standard events, while enabling better oversight and resourcing commitments for complex cases.
This automated approach further allowed Turkcell to better address the concerns of high-value customers who generated a substantial share of the operator’s revenue, embedding specific procedures to reduce risk while safeguarding customer experience in this vital customer segment.
Implementing a customized solution
The Optimus platform utilizes advanced neural technology to analyze, understand, and identify risk. That enabled Neural Technologies to implement key features that accommodated Turkcell’s particular operating needs.
Dynamic spending limits were introduced which adapt to reflect customers’ changing risk levels. Unlike a fixed credit limit, this enabled Turkcell to adapt its offerings rapidly to meet evolving customer scenarios. This solution not only protects those customers less able to pay, but quickly adjusts to provide greater revenue opportunities for the CSP by rewarding customers with higher credit limits where possible. This delivers greater customer satisfaction, while also providing a responsive solution that encourages further revenue generation for the operator.
Advanced warning systems were also integrated to prevent ‘bill shock’ for customers. This was targeted towards the particular challenges around high data roaming costs or internet browsing shocks for users traveling frequently overseas, a particular pain point for Turkcell customers. Event triggers would automatically send an SMS to alert customers in advance of any bill limit breaches.
This automated approach to customer communication is embedded in the Optimus platform. Predictive analytics enable continuous monitoring of customer usage, predicting and sending alerts of spending limit breaches before the end of a billing period. This allows the operator to predict bad debt in advance, and reduce bad debt exposure. This predictive solution combines with automated customer communications for standardized events, managing potential credit risk in a timely and reliable manner.
Analyzing the benefits
Neural Technologies’ advanced fraud prevention solutions have enabled Turkcell to reduce credit risk while improving customer service satisfaction. Not only was it possible for the CSP to view detailed behavior of existing customers, but the predictive modelling enabled an automated response that tackled credit risk before it emerged.
Thanks to the highly-customizable nature of the platform, Turkcell was also able to introduce new fraud alert and prevention measures on prepaid services, monitoring withdrawals from prepaid balances, and introducing an automated barring process to respond to emerging threats in real-time.
With deployment of the Optimus platform, Turkcell demonstrated clear benefits across the full length of its value chain.
- 12% increase in fraud detected in proportion to total revenue
- 27% decrease in time period between analysis and fraud deactivation
- 24% increase in fraud barring
- 43% decrease in number of subscribers presented for analysis
- 21% increase in number of investigated subscribers
- 38% increase in justified deactivations, not including false positives
- 25% decrease in reactivation
- 89% increase in collection amount
To find out more about how our fraud detection solutions can unlock value for your enterprise, contact Neural Technologies to arrange a free consultation.