Reducing Fraud, Risk With Advanced Analytics

With new banking channels and a younger generation of banked customers continually changing the business landscape for financial services and banks comes the added challenge of new risk profiles. This is not just a data management issue, but a fraud issue too, and one that has teams nervous about how to increase their exposure and profitability while minimising their risk.

Innovating with new products and services and building new channels to deliver an omnichannel experience to cater to this shifting business dynamic is an absolute must. In particular, banks have to cater to a very different customer demand served up by the millennial generation. These new products and channels are increasingly digital, and with digital solutions comes a flood of data, which is where the risk and the opportunity lie.

Scattered systems

When it comes to digital sprawl, there is no less is more, only more is more. New applications hooked into legacy systems often create more siloes in your technology, not less, leading to systematic data scattered across the business. Additional products also mean you have multiple systems to review and adjudicate for fraud which makes a case for centralisation.

Fraud detection and risk prevention can only be achieved when you collect the data at hand and assimilate it into the processes required to build the protection you need. This can be an exceptionally manual and arduous task for data and IT teams when there is both system and data scatter. It is not unusual for business analytics teams to build risk profiles from what is deemed a relevant dataset, only to find out later that the models are flawed and don’t factor in pockets of data stored in disparate environments.

When systems aren’t sticky or cohesive, you are also missing a trick in creating a single profile or streamlined user experiences. An experience that will follow the customer through the channels they opt to use to engage with their products.

Data synchronicity

To achieve a centralised view that will help with fraud and risk discovery, you can’t start with an analytics solution. Analytics is the end game; data is the playing field, and your systems are the stadium. Critically the first place to start is ensuring that your systems, processes, and applications are integrated. Making use of integration tools to consolidate events in systems is the crucial starting point, and there is a lot to be said for the middleware that connects processes.

By connecting events and embracing an event-orientated approach, you simplify the risk surface and create an environment where you can apply an analytics solution to data. This helps power reasoning as well as identify and mitigate fraud and risk. Modern analytics leverages artificial intelligence and machine learning algorithms to help with automation, but both are useless unless the data they draw inferences from is sound. It is also good to link to an external fraud identification network to ensure your efforts stay current and feed these models.

Reactive and proactive

The adage refers that prevention is better than a cure. But as financial professionals toying with the daily challenges, fraudsters throw your way you know that you have to be proactive and reactive. Using a rule-based approach in your environment allows you to identify risks and be reactive, where event-based correlations are needed to be proactive in fraud detection.

Both are needed, and both are necessary. When building systems for the future we naturally want to be as proactive as possible, proactively gleaning insights into user patterns and potential fraudulent events will allow you to reduce fraud instead of just reacting to it. This can be achieved when systems are integrated, data collection is centralised, events are continually monitored, and a streaming analytics solution is deployed.

The business loop

The secret sauce in risk assessment and fraud prevention is the consistent collection and analysis of information from all customer channels. This allows teams to continually evaluate the rules associated with risks and then adapt and apply new risk management rules as needed.

Ultimately you want to reduce fraud and reduce the fraud dispute ratio in your business, this is where analytics does its job. It helps bring your data closer to your business decision-making, highlights where customer experience improvements can be made, and turns risk from a negative tick box to an ongoing improvement process.

By Clinton Scott, Managing Director at TechSoft International