This March, in Indonesia we had a hard wake-up call: debit card skimmers are still at large. We found out that more than 1,400 cards were being skimmed from a single team of perpetrators. And 1,200 of those skimmed cards came from a single bank. According to the news, these fraudsters, which started less than a year ago in July 2017, are working around several big cities in Indonesia.
The media reported that the skimmers stole up to 22 Million IDR, which is around $1,500 USD (rate USD 1 = IDR 13700), per customer. If we extrapolate the numbers, assuming a customer loses $1,000 USD on average (and the bank has to pay that balance back), and around 1,000 customers are impacted each year, that’s approximately $1 Million USD that they stole in just a few months by a single team of perps. Also, it’s quite likely that were more incidents than reported, but the banks did not share the actual number so as not to alarm the public.
From the bank’s point of view, while the monetary loss was not monumental (What’s $1 million to the billions the bank makes a year?), from the overall business point of view, this will impact:
- Customer experience
- Brand Name of the Bank
- Being famous for the wrong reasons as an “easy target”, more will try to find ‘loopholes’
A poor perception of trust, customer experience, brand name and getting a reputation of being an easy target, can have a long-lasting negative impact to the bank. To combat this, banks need to implement adaptable anti-fraud solutions as soon as possible that can help them prevent and reduce the activity of these tricky fraudsters.
Based on my experience as a technologist, I have looked at these acts of fraud from the tech angle to see how they might be prevented.
How to prevent bank fraud using TIBCO solutions
First, to understand fraud/skimming, you need to understand how it happens and when it happens. It all depends on data. While most of your data is probably in a data-warehouse, there might be another data source that you use. You might use data from excel, another database or table, a web-service, or even from another SaaS tool such as Google Analytics.
If you have so many disparate sources of data, you can start the fraud detection process by using TIBCO Data Virtualization. Data virtualization gives you a seamless way to collect data from many disparate sources and make all your data uniform so you can a 360-degree view in a single dashboard. After all, you know what they say, “garbage in – garbage out”. You need to have good data going into your dashboard so you get good data coming out of it.
Then, using analytical tools such as TIBCO Spotfire combined with your historical data, you can start to analyze your data to find which kind of transaction is likely to be fraudulent or not.
After doing the analysis, the next step would be implementing the analytical model in real-time, either as an event rule or as a streaming analytical event. Both can be used separately or in conjunction with one another. Of course, we’re not stopping at that.
While we can monitor and score the transaction in real time, we also need to put in Case Management whenever we find potential fraud. This enables us to catch suspicious activities and also ensures that we don’t block the card unless we’re 100% sure that this is a fraudulent transaction.
The Financial Fraud Accelerator is available from TIBCO Community to help you speed up your time to understanding the data, visualizing the data, building the analytics model, doing machine learning, model scoring, and so on. This accelerator or template is an all-in-one, end-to-end solution for fraud detection.
TIBCO’s fraud detection tools can be implemented either in a modular manner or as an end-to-end solution, depending on your needs. So, what are you waiting for? Fraud can be lethal to banks.