National Programs’ Data Science & Analytics team (separate from our Actuarial team) works with company leaders to comb through data, looking for ways to increase profitability. The team focuses on four main areas: marketing, underwriting, pricing and claims, all with the goal of adding to the bottom line. The result? Data insights help create a proprietary model to generate revenue.
“Our process is to understand the business’s KPIs (key performance indicators),” explained Daren Eiri, director of Data Science. “How can we optimize those KPIs? So, we begin to hypothesize, ‘If we do this, how will it increase our hit ratio?’ Then the next question is, ‘Do we have the data we need to test the various hypotheses?’
“Here’s an example,” he added. “Business A wants to increase revenue – let’s say, generate an additional half million of revenue. Our team looks at their KPIs that influence revenue: What’s their total submission count? How many agents do they have appointed? How can they get more business in the door? And then we look at the submissions – are they the ‘right’ ones? How many are they quoting vs. how many are they winning or binding – their hit ratio? We also look at the decline ratio.
“Reviewing them all, we ask: Are there ways we can lift each of those KPIs? For instance, if we can raise their hit ratio from 20% to 25%, that would generate another half million of revenue.”
Moving from data collection to creating a submission grader
Once the Data Science team determines with company leaders what the team’s focus will be, they work together to create the if/then hypotheses. As an example, the data might show that producers in the Midwest send in more bindable business, or accounts that fit this profile have twice the likelihood to bind.
Armed with the data and tested hypotheses, they build a model that ultimately takes each submission and scores it. The team creates a custom submission grader, based on the criteria discovered. It’s a different model every time, Eiri explained, because one size does not fit all.
The team has built them for several of National Programs companies and is in the process of building more. The result? “We’re able to identify the accounts the company is more likely to win up front. And bingo! That’s where their sales team will focus. Plus, they can market to their producers, “Here’s our sweet spot: These 8 areas or channels or class codes, etc.”
As an example, let’s say the company writes BOP coverage, and they already know that restaurant BOPs are very successful for them. If the team digs further, they may find that restaurants that have alcohol sales of less than X% total sales are the sweet spot. They can then share with their producers where the company has most success. These are data-driven sweet spots, and they can focus their agents on who to go after.
Here’s another example: Carrier ABC, a big competitor, suddenly leaves the market. Or has a huge rate increase. With accurate data, the response should be, “Quick! Run a report on every account we know Carrier ABC has, or accounts that we’ve lost to them, so we can market to them now!”
Snapshot Case Study: How the Data Science & Analytics team increased wins for Sigma Underwriting Managers
Catastrophe Wind specialists Sigma Underwriting Managers receive roughly 200 email submissions daily from producers across the country. With an average manual review time of 30 minutes per submission, Sigma recognized a critical bottleneck in the workflow, creating a huge workload for the underwriting team.
Amy D’Angelis, director of financial projects, explained, “There was an overwhelming number of submissions that we received daily, and all of those submissions required manual review. We desperately needed to find a way to improve the process and automate it.”
Sigma partnered with National Programs’ Data Science team to tackle workflow issues and streamline the time-consuming task of manually reviewing email submissions. Eiri and his colleagues quickly assessed Sigma’s situation and created an automated data solution that they named Sigma Submission Express. Eiri describes the custom-built automation in simple tech terms, “It applies computer logic to output decisions based on rules that we program.”
First, the team set up a single inbox for producers submissions. From there, Sigma Submission Express would receive the emails, review the attachments, assess the risk and decide if the submission is in-appetite, declined or needs further review. Once a decision is made, the program sends an email to both the producer and assigned underwriter with the results. For submissions that require additional review, Sigma Submission Express automatically generates a summarized report, expediting the review process once it lands on the underwriter’s desk.
From start to finish, successful submissions are reviewed by Sigma Submission Express and completed in 2-3 minutes, creating a 20-30% daily time savings for the underwriting team. Reflecting on the workflow transformation, D’Angelis remarks, “We now have additional time to service our customers and focus on the accounts we believe will be successful binds that will grow our business.”