A recent report from Gartner has stated that organisations that use predictive analytics will increase their profitability by 20% by 2017. There is a great deal of focus placed on the benefit of analytics for improving profitability. Less commented on, but no less relevant, is the role that analytics can play in preventing loss – specifically, loss through fraud, says William Lawrence, regional practice lead: Fraud at SAS Institute.
Current fraud detection solutions rely on recognising pre-existing fraud modalities and flagging similar events when they occur. And it’s that “when” that’s important. The fraud has already taken place by the time the event is flagged, and while this may create a neat paper trail to the perpetrator, it doesn’t do anything to prevent the fraud from occurring in the first place.
In addition, fraud detection solutions that rely on identifying events that are similar to previous instances of fraud can’t do anything about new methods of committing these kinds of crimes. If there’s one thing that we know about fraudulent individuals and syndicates, it’s that they are constantly looking for new ways to defraud businesses. Given this trait, it’s completely pointless to rely on a solution that only recognises the replication of historical fraud.
We need to know the future
Clearly, what is needed is a solution that can predict the future. While ten years ago, this may have been a laughable notion, we now have at our disposal a tool that makes science out of prediction – predictive analytics.
In much the same way as analytics can make sense of disparate data sources, see patterns and predict future business trends, they can also see patterns in fraudulent behaviour, and predict when fraud is likely to occur before it has happened. This in itself becomes a deterrent.
Let’s use procurement fraud as an example. Procurement fraud is an international problem, but a recent report by PWC has indicated that it’s a serious problem in South Africa specifically. According to the report, 59% of South African companies have experienced procurement fraud compared to 29% of companies globally – and those are only the instances of detected fraud.
The reality is undoubtedly a much higher percentage. And another thing that we know about procurement fraud is that it always grows – perpetrators start with a small fraudulent amount to test the system and then expand their activities, given what they’ve learnt.
How analytics does it
By turning the power of analytics to detecting procurement fraud, we can use a combination of features working in collaboration to recognise not only known patterns of fraud, but also new methods and processes for defrauding organisations that have never been seen before or that are specific to that particular organisation’s processes and employees.
Predictive analytics for fraud detection can use geographic location, similar bank accounts, personal details of employees, unusual amounts processed by the financial systems, unusual activities outside of the financial system (such as regular meetings with suppliers) and text analysis of the data contained in emails, to predict when and how fraud is likely to happen.
This is a much better way of dealing with fraud and reducing financial losses than the old “pay-and-chase” scenario.
So, while businesses are focusing on the profit generation capabilities of analytics solutions – a 20% increase in profitability by 2017 – the reduction in loss due to fraudulent activities that analytics can provide are equally compelling. In addition, King III requirements state that company directors can be held accountable for instances of preventable fraud, and the good news is that with analytics, most fraud is preventable.
Analytics are going to make companies more profitable without a doubt, but – just as important – they are going to make companies more honest, financial controls tighter and fraud a far less prevalent evil in the modern workplace.
