ork.AI
Organizations are drowning in sanctions alerts. Sanctions screening tools generate a significant number of alerts— of which over 99% are false positives. This can create a huge workload for compliance teams.
That’s why WorkFusion has automated the sanctions screening alert review process with AI-enabled sanctions screening alert review analysts. They are technology controls that intelligently automate tasks like identifying entities from payment messages and comparing them against sanctions lists, drastically reducing the manual effort required to sift through alerts.
Banks and other financial institutions (FIs) leverage these AI Agents to improve their sanctions alert review programs in many ways, such as improving alert review quality and consistency, reducing risk by minimizing errors, scaling to meet surges in alert volumes, and more. But what are the metrics customers measure to prove the success of their sanctions alert review programs?
At a high level, we see across our customer base of FIs that they measure success in four areas:
- Risk Mitigation / Quality
- Cost
- Volume / Scale / Speed
- Revenue / Customer Impact
In total, there are 26 key metrics across the four areas. However, it is rare for an FI to measure all 26 of them. This is due to each FI having its own unique set of circumstances and reasons for hiring a WorkFusion AI Agent. Some may be looking to reduce risk, while others may be interested in reducing costs or scaling for alert volume surges.
Following are the metrics, broken down by the four major areas measured.
Risk mitigation and quality metrics
AI-driven sanctions alert review enables compliance teams to increase alert review quality and consistency, reduce material and immaterial errors, handle alert surges, and avoid missed escalations and missed true positives. All of these are crucial aspects of risk mitigation. As such, this is the area most measured by customers and has the most associated metrics. Customers frequently measure the following metrics:
- Alert adjudication error rate
- Percent of alerts requiring re-work
- Percent of alerts with insufficient narrative or audit trail
- Ratio of analyst time spent on false positive clearing vs. true positive analysis
- Ratio of unproductive escalations to Level 2
- Time from ‘Alert Received’ to ‘Start Alert Review’
- Week-on-week staff level
- Employee satisfaction: attrition / retention rate, ongoing engagement, etc.
Cost metrics
The number of sanctioned individuals and entities has grown rapidly in recent years, especially in the wake of Russia’s invasion of Ukraine in 2022. This has caused a surge in sanctions alert volumes at nearly every FI, and the cost to respond has become astronomical. As a result, many customers have been eager to deploy AI Agents in an effort to control costly outsourcing and / or hiring a large number of additional L1 analysts. This area is covered by six simple and straightforward metrics:
- Total monthly manual hours spent on sanctions alert reviews
- Average manual hours per alert reviewed
- Total monthly alert review outsourcing cost
- Average outsourcing cost per alert reviewed
- Total monthly labor cost for alert reviews
- Average labor cost per alert reviewed.
Volume / Scale / Speed metrics
It is not uncommon to encounter a sanctions compliance team that has incurred a backlog of tens of thousands of unreviewed alerts. This exposes the FI to substantial risks, both in terms of compliance as well as allowing bad actors to take advantage of the financial system. Alert backlogs have only increased in recent years as many FIs have instituted hiring freezes, found it increasingly difficult to find high-quality outsourcing services specialized in sanctions alert review, and faced surges in sanctions alert volumes. So, when FIs in this situation hire an AI Agent, they are seeking to scale their sanctions alert review program. Here are the corresponding metrics which they measure for success:
- Alert-level automation rate: percent of false positive alerts that require NO manual review
- Hit-level automation rate: same as #1, except regarding hits instead of alerts
- Number of alerts processed
- Number of alerts added to backlog
- Change in backlog volume
- Monthly average backlog volume
- Average time alerts are in queue
- Average throughput time
- Average account opening time.
Revenue / Customer Impact metrics
When customers want to open a new account or begin doing business with an FI, speed often matters. When an FI cannot open accounts or provide other services in a timely manner, customer satisfaction (CSAT) suffers. FIs traditionally had to strike a balance between keeping CSAT scores high and taking on risky customers or losing some new customers in an effort to minimize risk. AI Agents enable FIs to avoid having to make those tough decisions because AI Agents perform L1 sanctions alert reviews exponentially faster than humans can, and simultaneously, reduce risk by assessing far more data points and making decisions based on that additional data. To demonstrate the positive impact on revenue and CSAT, FIs measure the following metrics:
- Monthly CSAT score
- Number of new customers (account openings)
- Number of new risky customers (approved customers previously deemed too risky).
What gets measured…improves
By measuring the success of different aspects of their sanctions alert review program, an FI sets the stage to always improve. Depending on the areas initially targeted for improvement, a compliance team may start by measuring just 3-5 key metrics. However, as they gain comfort with the power of AI Agents, compliance leaders expand their usage, and thus, expand success in more areas of sanctions alert review. This enables them to keep pace with bad actors by continuously innovating.
To learn more about AI-driven sanctions screening alert reviews, discover the power of Evelyn and Tara, WorkFusion’s AI Agents built specifically for managing the large volume of Name Sanction Screening and Payment Screening alert reviews.