Company: Finansbank AS
Company background: Finansbank A.Ş was founded on October 26, 1987 in Turkey. In June 2016, it was acquired by Qatar National Bank S.A.Q (QNB Group), the largest bank in Qatar, the leading financial institution in the Middle East and Africa region. The Bank's name has been changed to QNB Finansbank in October 2016. QNB Finansbank is headquartered in İstanbul, Turkey. With its founder and the leading team, it was differentiated from other banks established at the same time in terms of both its main philosophy and its view of banking, and has become a success story brought to life by professional bankers where smart banking is performed. It has achieved many firsts in the Turkish banking sector. Finansbank A.Ş is the first private bank that went public, First installment credit card, The "skip an installment" and "miles through points" campaigns, thus contributing a vital know-how to the sector. In 2009, Finansbank A.Ş offered CardFinans Nakit, "the first debit card worldwide that offers installments" Finansbank A.Ş continued pioneering in retail banking as well. Enpara.com, the first bank with no branches in Turkey that offers services exclusively on digital channels, was established in October 2012. Enpara.com, more than 4 million happy customers were introduced to digital banking at no cost. Enpara.com grows on a daily basis, mainly through user recommendations via word of mouth. QNB Finansbank Financial Services Group includes QNB Finansleasing, QNB Finansportfoy, QNB Finansinvest, QNB Finansfaktoring, QNB eFinans, Cigna Sağlık Hayat ve Emeklilik A.Ş., the technology firm IBTech. QNB Finansbank became a case study at Harvard Business School for four times. The first was about its rapid healthy growth in 1994. Secondly, the sale of Finansbank to National Bank of Greece by Mr. Hüsnü Özyeğin, Third, resale of the bank in 2015 to Qatar National Bank, Fourth, with Enpara.com in 2017.
Contact: Gökhan Dumrul
Title: Expert Specialist / Team Leader
Country: Turkey
Award Category: Innovative Problem Solver
Tell us about the business problem you were trying to solve?
Today on financial sector, corporations need to attract costomer’s attention by announcing new and innovative banking products or making user experience better and easier to lead the market. While allowing customers instant access to products and services, this digital transformation has
increased their vulnerability to fraud.
Especially social engineering fraud attacks are very common in Turkey. In such fraud cases, Many of victims send the money to fraudster using mobile Application that in their own mobile phone, or customers share their personel data like ‘’online banking passwords’’ or ‘’sms including one time pasword’’ with fraudster.
Customer and bank relation is base on trust. In such fraud cases, Even if it is the customer's responsibility to protect their personal information, it is desired to be detected by counter-fraud teams and to prevent customer damage and protect bank reputation. Detect and respond to suspected fraud is a part of cusmoter happiness.
Before SAS Fraud Management Integration, QNB Finansbank had near real time Digital Fraud Tool and adaptation to new products was not fast enough.
- To achieve an appropriate maturity level of fraud controls, usually we had needed developments on our core banking system. We had implemented most of online action like otp on our core banking system with hard coded rules. The process sometimes has been causing performance issues or needed long It Development process for changing.
- Especially recent years, new products for evolving customer needs on Digital banking like Mobile Contactless payment Transactions, Payment with QRCode, Open Banking System etc, required an online fraud responds. Fraud System Respond time was not fast enough.
- Growth of digital banking logins and financial transactions, cause the growth of fraud alarms and needed fraud analysts were very high. Rule creating process was not easy however we had needed to reducing our false positive rates without increasing our losses.
- High false positive rates could lead to late calls and customer complaints.
- QNB Finansbank fraud tool had been working based on transaction level, so It had been creating duplicate alerts for different transactions of the same customer.
- During the alert monitoring process, analysts had must to use core banking System for customer demographic.
How did you use SAS to solve that business problem? What products did you use and how did you use them?
Sas Fraud Management is a real-time fraud tool. In the heart of fraud management process, we used the Sas Fraud management tool.
- We have written smart rules in SFM and we have removed our hard coded rules in our mobile banking frontend and core banking.
- We currently have new real-time and backend blockers / action types on SFM.
- Decline customer transaction,
- Send One time password with SMS or with mobile notification,
- Customer verification process using İnteractive voice notification,
- Automatic customer block, card block, drop mobile activation, drop mobile or internet banking session etc.
- Send information SMS
- We have written the smart rules which kept customer profile. This new fraud system can behaves by itself, detects suspicious transaction, takes appropriate actions, follows the answer of customer himself if needed, updates customers profiles real-time
- We have integrated the customer real-time demographic data to SFM. This way, analysts were able to view customer demographics in the SFM user interface.
We continue integration of new innovative fraud verification process with power of Analytics and machine learning technologies powered by SAS tools too. In the near future, we plan to integrate 2 different fraud process.
- QNB Finansbank have an Digital intelligence assistant which called Q. Q has the ability to chat with customers via QNB Finansbank mobile app, understand their requests, and direct them to the core anking system. Previously, Q used to get to know customers by examining their past transactions and then offer different products to do cross-sell. We have integrated SFM and Q to fill the fraud detection need by combining smart asistant with SAS Fraud Management (SFM).
- SFM's fraud rules works on customer transaction and if in case of a fraud risk in a transaction, SFM notifies the Q.
- Before the customer completes the transaction, QNB Finansbank Digital intelligence 'Q' becomes active on mobile banking user interface and It starts to ask verification questions like a real fraud analyst to customer.
- To complete the transaction, customer must answer questions.The process is completly online. In order to prevent fraudsters different answers are requested from each customer.
- If the customer gives an answer indicating that the transaction is fraudulent during the conversation, the answer is forwarded to SFM.
- SFM rules works again and send prevention actions to core banking system.
- If the customer confirms the transaction, the transaction complete successful whitout any human intervention.
- Today most of smartphone has NFC skills and Most of our customers have an Identification cards with chip. When the customer made Money transfer, transaction details send to SFM.
- SFM rules checks, customers information like device Id. And If there is an suspicious action on customer transactions, SFM send information to our mobile application.
- Customers should use their ID card to complete transaction process.
- Even customers share their online banking passwords with fraudster, fraudster can not belog to customers ID card by social engineering.
What were the results or outcomes?
After Sas Fraud Management Integration;
Sum Up, By preventing fraud with less false positive rates, we contributed to increased customer happiness and bank reputation. At the same time, we increased the customer information and confirmation stages without increasing the manual workload
- The need for it development has been reduced,
- Improved customer experience by preventing transaction delays,
- Number of monitored banking channels and products increase,
- We reduce more than % 60 manuel alerts,
- Customer level fraud process created,
- The number of otp sent to customers has been reduced by 30-50%. Cost of OTP message and customer experience have been improved.
- It was difficult to measure fraud analyst performance; now easy and automated
- Needed time to create a complex, smart rule was almost a few days; now only a few hour.
Why is this approach innovative?
This initiative represents the one and only kind integration by coupling AI driven digital assistance to SAS SFM, positioning the AI technology such a real fraud analyts operating in mobile channel among the banking industry in the region.
Human factor has been reduced with the help of end-to-end automated structure where customers can provide verification self service by interactions with digital assistance in various verification methods designated based on their risk levels.
Rapid financial transaction verifications that can not be handled by human operations on time due to customer experience challenges.
