The Javelin Strategy & Research Identity Theft Report 2013 states that 5.16% of US customers suffered from identity theft amounting to US$20.9 billion. Moreover, Tablet users had the highest probability of fraud at 9.6%. Victims of data breach had a 22.5% likelihood to becoming fraud victims. Hence, it is clear that while organizations are deploying more processes, technology and resources to prevent fraud, the fraudsters are having a ball. One thing fraudsters do, is to think outside the box. So we have to take a leaf out of their book and be innovative in our approach to prevent and detect fraud. Below are some ideas on the same. Share with me your thoughts on what you think about them.
1) Voice Print Analysis
Presently, in most of the banks, a call center agent asks a set of questions to verify the identity of customer for telephone banking. Internal employees, external fraudsters and organized crime groups can easily steal information about date of birth, place of birth, address, secret questions, and card number.
Now voice-printing software is available for authentication of voice. The system automatically verifies the caller voice with the customer’s sample voice to identify fraudulent callers and protect the account.
Secondly, maintain voice records of earlier fraudsters. When system detects a fraudulent caller, it automatically checks against the previous fraudulent call records. Hence, the system will flag if a fraudster has previously conducted a telephone banking fraud. With this, it will be easy to nab the fraudster, if the police had caught him/her in a previous case.
A new voice identity technology is available that captures the tone of the voice and the type of communication. The software can monitor quality of calls and customer satisfaction from call center agents’ conversations with customers. This will cut manual quality control checks significantly and result in savings in quality control department costs.
2) Track through Photographs and Location Mapping
Besides having voice-printing software, use a system similar to WhatsApp to identify of customers. WhatsApp sends text messages, images, video recordings, audio recordings, and the location. If banks invest in a similar application and allow customers to download the application on their mobile phones and tablets, the number of telephone and internet frauds will reduce.
If a fraudulent caller is flagged, then the call center agent can request the customer to send a selfie or video. If it is the wrong person, usually the caller will cut the conversation and drop the attempt to commit a fraud.
If the caller is able to circumvent this control, the application will also track the location. Applications track the frequent places a customer visits or calls from. If the caller is from an unusual place, then s/he can be tracked immediately. For example, if a British customer is tracked to a place in India, the call centre agent can ask the caller to verify their location.
3. Track Spending Behavior
Sometimes high value fraudulent payments are processed resulting in huge losses. A study done by Vivek K. Singh*, Laura Freeman*, Bruno Lepri, Alex (Sandy) Pentland for “Classifying Spending Behaviour using Socio-Mobile Data” determined the spending behavior of customers from the social interaction patterns on mobile phones. For example, it showed that more social couple and couples with diverse business interests tend to spend more.
Using big data, insights on spending behavior of customers can be analysed based on personality traits. Tracking social patterns and payment patterns can flag out anomalies when the payment is not in line with the spending pattern. Moreover, a location map can identify the location of beneficiaries of previous payments . Hence, fraudulent payments can be identified at the time of processing itself.
Another advantage from this technology can be for processing retail loan applications. If prospective customers are willing to give the data of mobile phone transactions, then at the time of processing the application itself, the bank can identify which customers are likely to overspend and default in future. The bank can ask for additional securities and guarantees.
Moreover, if the application is installed in the loan customer’s mobile after loan disbursement, the moment s/he is about to overspend which might result in default of EMI, the bank can send the customer an alert to pay the EMI first.
4. Fraud Risk Conversations
According to psychological studies on emotional intelligence, Negative Emotional Attractor’s activate defense systems and build resistance to change. On the other hand, Positive Emotional Attractors (PEA) activates parasympathetic nervous system and makes a person more conducive to listen and change behavior. An effective team has a 3:1 ratio of PEA:NEA. Another study shows that improving peer-to-peer conversation increases productivity of the team by 30 to 40%.
However, risk management reports are mainly critical hence activate NEA. Moreover, the communication, training material, and code of conduct are all geared towards creating fear and guilt. Hence, it is not surprising that attempts to educate business teams on fraud risks fail.
Fraud risk managers can build a positive interaction model using technology platform. A study conducted by Erez Shmueli_, Vivek Kumar Singh_, Bruno Lepri and Alex ”Sandy” Pentland on “Sensing, Understanding, and Shaping Social Behavior” enables tracking of human behavior through big data analytics. The analytic helps in understanding the behavior, the tone of the conversation and the trust relationships between people.
Using this technology, an organization can use a social networking platform to communicate fraud risks through blogs, videos, and stories. The write-ups and stories should be from the business teams. From the comments section, the application can identify the key influencers and trust holders to bring about change. Thus, change the conversation to change the behavior.
The days of holding a gun to rob a bank are nearly over. Fraudsters use social engineering to obtain sensitive information to conduct account takeover frauds remotely. Hence, organizations need to use socio-physics, social networks, and technology to beat the fraudsters in their own game. Being a leader in adopting the latest technology to prevent and detect frauds has an additional advantage, the fraudsters have not discovered the antidote to it. Hence, fraud risk managers have the right weapons to fight. The right tools can make a hell of a difference.
- Javelin Strategy & Research Identity Theft Report 2013
- Classifying Spending Behavior using Socio-Mobile Data - Vivek K. Singh*, Laura Freeman*, Bruno Lepri, Alex (Sandy) Pentland
- Sensing, Understanding, and Shaping Social Behaviour – Erez Shmueli_, Vivek Kumar Singh_, Bruno Lepri and Alex ”Sandy” Pentland