Illustration of Feedzai Genome visually detecting fraud

Listen to: How Genome Saves Time, Stops Crime (10 mins):

A fraud analyst's job sometimes seems like a game of Battleship. Too often it feels like analysts make random efforts to stop fraud that are ultimately unsuccessful. But what if fraud analysts could see fraud even before it happened? 

That was the question our Product team asked itself three years ago when we first launched Genome, a visual link analysis tool that sits in our Case Manager. 

We’ve continued to develop Genome with one goal in mind: to make fraud analysts more efficient at catching fraud. Our latest Analyst Tools release includes a new architecture of Genome that provides users with a 30% increase in navigation, productivity, and data ingestion speeds, self-service rules that offer more flexibility, and customer-level notes to provide a full picture of a customer’s history. 

But what does that actually mean for data analysts who are waging a daily battle in the fight against fraud? Let’s meet three different fraud analyst personas – Larry, Louisa, and Luke – and explore how Genome’s fraud visualization capabilities benefit each of them.

Larry: Level 1 Analyst 

Larry is based out of London and works for a large bank. As a Level 1 analyst, he’s the in-house Case Manager, and by extension, Genome’s largest user group. On the front lines in the fight against fraud, he reviews hundreds if not thousands of transactions a day. He’s under incredible time pressure, unusually having only 60 seconds to make the right risk decision for each case. That’s very little time to make a decision and even less room for error.

Pain Point: Too Many Alerts, Too Little Time

L1 analysts like Larry have too many alerts to review and too little time to review each one individually. If they can’t make a decision on time, fraudulent transactions can get approved. This can result in direct monetary losses to the bank in the form of chargebacks causing friction for the customer and reputational damage to the institution. If they don’t have the right data, or the time crunch leads them to make the wrong decision and decline transactions from legitimate customers, the negative effect doubles.  

How Feedzai Genome Helps Larry

Genome features are designed to follow Larry’s workflow. Genome offers analysts access to high-quality alerts, enabling them to focus more of their efforts on more important issues. Analysts can quickly review layers of information involved in a single transaction, by bringing the relationship between entities to a first-class citizen, something impossible to do in a tabular-like data representation. This information empowers Larry and other L1 analysts to find connections and review multiple cases at once, hence detecting fraud more effectively and taking action faster against activities like bot attacks. The new Case Manager update also offers an always-visible toolbar feature that enables analysts to quickly take action anywhere when reviewing alerts.

Genome helps banks save money through more effective fraud prevention. If Larry is able to detect more fraud in less time, his employer reaps the benefits. One major bank saves roughly $1 million annually for every second on average of alert review time that L1 analysts save.

Louisa: Level 2 Fraud Analyst

Louisa is a Level 2 (L2) analyst or a senior investigator. She delves into more complex fraud patterns and is the best at what she does. She’s concerned about improving the overall risk strategy by both analyzing what fraud cases and patterns were missed and what was wrongly classified as fraud when in fact, they were legitimate customer activity. She often spends her time focused on higher-value transaction alerts and looking for fraud patterns hidden under layers of data.

Pain Point: Detectives Need the Full Fraud Picture

Like Larry, Louisa is pressed for time and under pressure to deliver results. Her decisions need to be accurate, or they could cost her employer considerable losses in chargeback fees. If Louisa can’t do her job effectively, hidden fraud patterns will continue to go overlooked and undetected. Fraudsters will continue to exploit these vulnerabilities for their own ends for as long as they are able to 1) get away with it, and 2) profit from it.

How Genome Helps Louisa

With Genome, investigations that would otherwise take days or weeks to complete can be completed in minutes or hours. Genome helps Louisa identify patterns hidden in layers of transaction data by reviewing subsets of events or reviewing activity. Once the shape and reach of the fraud case are identified, seamless integration with Case Manager Self-Service rules allows Louisa to immediately write and test rules to further prevent similar cases herself, without needing to rely on others to write rules instead. 

For example, say Louisa notices a red flag that a customer is caught up in a romance scam. She reviews an alert that the customer is making a high-value transfer to a recipient at another bank. Genome enables Louisa to quickly review past transactions. With these capabilities, Louisa can quickly identify a history of suspicious transfers between the two accounts. She now sees the recipient’s account is linked to several other, suspicious bank accounts and customers from her institution. From there, she can write a rule for high-level transactions and test this rule against three months’ of transactional data, preventing future damage and protecting other existing bank customers.

The rule can be written to automatically approve or decline transactions based on specific criteria to address patterns that are problematic. If the rule does not bear the anticipated results, analysts can go into the Self-Service tab again to disable it. 

When Louisa discovers more complex fraud, she relays it to Larry. The more effective Louisa is, the more effective Larry and the rest of the team will be. Including Larry’s boss, Luke.

Luke: Analyst Manager

Luke is Larry’s manager. The more effective Larry is at his job, the better it reflects on Luke. Saving his organization money and using available fraud analysis resources effectively is Luke’s main focus.

Pain Point: The Buck Stops Here

If fraud analysts like Larry are unsuccessful at catching fraud or drive up false positives, then financial losses arise caused by chargebacks, or worse, customers can leave frustrated by false positives. All of this reflects poorly on Luke and could ultimately put his job in jeopardy.

How Genome Helps Luke

Genome also provides a top-down view of the top fraud cases that need to be addressed. Luke can take greater ownership of his team of analysts’ workflows using this tool.

Case Manager offers a Queues feature that enables him to make analyst assignments based on channel, event type, and cases. Queues can be combined with Automation Rules (sometimes known as Robotic Process Automation) to automatically sort alerts into different categories. Luke can also access Dashboards. The Dashboards tool provides him with an overview of analyst alert assignments and how they are performing. A List Management feature, meanwhile, helps him keep track of cards traced to problematic individuals or regions. 

The less time Luke, Larry, and Louisa have to spend chasing fraud, the more effectively they can stop it. Genome opens new possibilities to different levels of analysts, enabling them to find fraud more effectively instead of searching through individual data for patterns. Let Genome help your organization save time and money with a more focused approach to fraud.

How can you build a bank from the ground up? In this engaging, on-demand webinar, What will the bank of the future look like?, Feedzai’s Pedro Barata joins SolarisBank’s Jorg Howein at the FinTech Finance Virtual Arena to discuss what it will take to build the bank of the future from scratch.