Unleashing the insights locked into Policy Admin Systems and Bordereaux data

The Issues

Our client is a large MGA in personal lines with around 250 professionals, directly or through partnerships, serving the product life cycle from underwriting, compliance, claim management and policy administration services.

The GIROUX.ai consulting team performed a discovery phase and identified that about 80% of the tasks consisted of establishing and maintaining controls when it comes to data quality and resilience. More strikingly, the underwriters would wait a good two to three months to get a clean set of data to perform their work, at times causing friction with partner and insurer relationships. PowerBI reports were providing different results depending on who wrote the SQL query behind the reports. It was a real challenge.

For example, claims data was incomplete, with errors such as missing claims or wrong allocation against the wrong policy. Incorrect policy types led to incorrect rates being applied. Exceptions were not spotted in a timely manner. The Net Earned Loss Ratios were not always correct as a result, especially when providing a market update to the insurer.

Time to Act

Without a rapid turnaround, the insurer ran the risk of alienating its partnerships with the market and some of their insurers were threatening to pull their capacity unless their reporting improved, especially regarding forecasting accuracy on binder performance.

Over a period of three to six months, GIROUX.ai performed a rapid consultancy using its well-rehearsed and proven method looking at translating the MGA strategy into a robust conceptual and physical data architecture. The approach was wholly participative using a series of interviews and workshops over eight weeks. The joint client-GIROUX.ai team then profiled the data sources and started loading historical data up to 10 years and also leveraged on the IT team’s work. Rather than replacing what IT was doing, it was about augmenting the team’s work to bridge the gap between what the actuaries and underwriters needed and what had been built.

Deployment and Results

As illustrated below, the system was successfully deployed within a few weeks and delivered the base requirements to the underwriters and actuaries, as well as other professionals such as claim managers and policy administration staff.

Each given user had the ability to see what they needed to perform their respective tasks. The definition of terms such as metrics, attributes and hierarchies were leveraging what the business was used to so training needs were minimal. A key benefit was that the process of implementation encouraged the organisation to adopt greater discipline when it came to definition of terms for metrics and attributes. Additionally, no SQL or Python skills were required to create reports or query data – all users could easily self-serve. Importantly, the MGA in question managed to secure an additional three-year capacity deal following the rapid turnaround.