Recently I worked with a client to analyze and solve some significant underwriting problems. Initially the company’s CEO was concerned because its loss ratios were higher than its peers’ loss ratios. As we often see in these situations, there was a lot of organizational noise about where the problems were. Prior to our involvement, they instituted some quick fixes and waited to see the results. But the CEO had lingering uncertainties: How do I know we’re getting better? Are we doing the right things? Are we doing what we say we’re doing?

Most people rely on their management reporting systems to answer those questions, but those systems are usually looking in the rearview mirror. Others count on reinsurance reviews, but those only happen once a year and tend to focus on items of interest to the reinsurer. Furthermore, most reinsurers won’t be as blunt as an auditor — after all, a carrier is their customer. Managers tend to be too close to the action because they are fighting daily fires. They may form opinions from a limited set of transactions and cannot see the forest for the trees. These traditional approaches tend to fix one symptom at a time and tend to be reactionary or flavor-of-the-month exercises. They don’t get deep enough to determine the underlying root causes.

We recommend a more strategic approach, and even then there are options. You can step back and analyze your processes from end to end, or you can jump to the end of the story and look at the finished product through an audit program. The important thing is to start somewhere — and start before your numbers go south. In a world of seemingly endless data, there is no substitute for looking inside an underwriting file. We’ve found that working backward from an audit or file level can be very informative, if you know what you’re looking for. In one situation, we found that underwriters were ordering additional credit information and spending time duplicating work that their predictive models were already doing. This showed us that the financial underwriting guidelines were never adjusted for life in a predictive modeling world. In another situation we found that the acceptable frequency and loss ratio benchmarks were out of date and were improperly categorizing average and below-average risks to be above average. In that particular project we started down the path of looking at their audit program. What we uncovered, however, was a much broader set of issues.

Sorting Out What Really Matters
A good audit program should identify trends, reveal process and communication breakdowns, and uncover causes (e.g. conflicting memoranda) and shine a light on them. In our experience, errors on the part of individuals do occur, but the bigger problems arise out of system failures in the service chain.

For this project we built our approach on three principles: speed, simplicity, and focus. We asked our client to pick one important product line. We laid out a plan to quickly build an audit program for one line in a manner that could be replicated for the remainder of their product line. Our workshops included management and front-line staff and began with one simple question: When you open up an underwriting file, what do you expect to see that tells you that the underwriter did a good job?

The subject matter experts then put everything on a whiteboard, developing an exhaustive laundry list. We then asked a second question: Can we get this list down to the items that really matter? We went through the list again and came to agreement on 11 items or attributes. Agents believe that underwriters have a bit of Lieutenant Columbo in their DNA — there’s always room for “one more question.” However, those same agents have often told me that we live in a “seven-question world” and people do not have the time to go through a laundry list of questions to write a piece of business. We got close.

Next came the fun part. For each of these 11 attributes, we asked, “What is the standard we expect underwriters to apply?” This is where we saw some red faces. For some attributes there was quick consensus on the standard. On other attributes there were significant differences of opinion. For some, there were no standards at all. Our project now touched on something bigger: Who is responsible for setting the standards and policies for a product line? Although we had started down the path of installing an audit process, we ended up uncovering a broad set of issues regarding ownership of the staff product management function. Working with client management, this organizational weakness was addressed by creating designated leads for each product line.

Armed with our new product line leads, we established the standards for all 11 attributes. Keeping it simple, we assigned equal weight to each attribute. Now we had a methodology to score each file and aggregate the results at the department, product, manager, underwriter, and attribute levels. A new file review tool was then created for data capture.

The Importance of Process
When embarking on an audit, you never know where the road is going to lead you. The following are a couple of examples of process issues uncovered on the journey.

• Electronic work queues are somewhat abstract. In days gone by, we would pose as a quote or a policy and walk around the building through the various steps in the process. Today the circuit exists in cyberspace. Work travels by clicks and keystrokes but still goes through steps, changes hands, and gets stuck in bottlenecks. In our recent project, we conducted desk interviews with front-line staff asking questions such as; “Why do you have to look at that?” or “How often do you reject something because of that?” We found numerous small policies in the underwriters’ work queue where there was no value added by their involvement. Management wasn’t aware that these items were going to underwriters. New procedures were quickly implemented so that these policies would go straight to the processing staff. This created immediate capacity in the underwriting ranks. Upshot: Today’s electronic workloads can look like stick counts, and they don’t tell the whole story.
• Another example is unnecessary referrals to management and an unintended blind spot. On this project we found that underwriters were responsible for performing an in-depth review of policies that fit a certain profile and determining whether corrective actions were needed. If the underwriter wanted to take action, he or she was required to have the manager approve the actions. Yet, managers rarely disagreed with the recommended actions. We suggested eliminating the bottleneck on the front end for those rare cases, and instead relying on their audit process to provide a back-end quality review. This change freed up both front-line and management capacity. Enough capacity was created in just a couple weeks to fund the resources for the entire audit process. The blind spot was that the underwriter could choose to take no action, and no referral was required. Our suggestion was to include a random sampling of “no action” policies in their audit program.

The Importance of Focus
The final phase of our project was to develop a plan to ensure the new audit program would be executed efficiently and reliably. This is where the principle of focus took center stage. Our recommended approach created focus by identifying the subject(s) to be audited, defining when it will be reviewed, and dedicating resources to conduct the audit. To achieve consistent, reliable results, reviews should be conducted in a finite time period by dedicated resources that will not be interrupted or distracted by other demands. We helped our client design a calendar that would enable them to conduct the audit and, very importantly, allow sufficient breathing room for staff to absorb the findings and implement corrective actions. The calendar included developing a heat map that identifies the most critical product lines or geographies to be examined as well as scheduling the three stages of an audit: the prep phase, the execution phase, and the recommendations phase. We recommended an annual calendar with “white space” for the unexpected. Through this planning process our client was able to schedule resources, line up SMEs and management personnel, and communicate the plan to staff up front.

Managing Barriers
One barrier for audits is that they are sometimes viewed as a lack of trust in the front-line staff. At Nolan we believe that a well-designed audit program analyzes the whole process — not just an individual — and is focused on identifying the aggregate performance of the underwriting function. Another barrier to the audit process is time. Everyone’s too busy. For this client we overcame that barrier by keeping it simple, establishing a workable annual schedule of audits, and uncovering things through the initial audit that could be eliminated. We avoided the “fear of gotcha” by involving front-line staff working side by side with management on performing the audit, and in turn, we eliminated the typical “we-they” thinking. 

Results Matter
This project delivered immediate benefits.

• Knowing the “11 Things That Matter” field, UWs are now able to cover gaps in the renewal process and eliminate file handling by a second downstream.
• Small policies now bypass the underwriters’ work queues, speeding cycle times and decreasing underwriter workloads.
• A systematic audit process is in place with quantifiable results showing which attributes require clarification or additional training.
• Focusing on what matters is bringing clarity and consistency, resulting in more capacity and higher quality.

With this recent project example as a backdrop, here are some things to consider for your underwriting process.

• Technology and predictive modeling are transforming our business. The concepts of “straight-through-processing” and “exception underwriting” are increasing the need for a reliable validation process. You want to know that the risks passing through your screens are what you think they are.
• As underwriters, we’re wired to internalize underwriting discipline. However, productivity pressures, automation, and increased analytics can lull some into putting more emphasis on processing than decision making. Do you want someone handling exceptions with a processing mindset?
• Today, with technology and modeling, underwriters are able to carry significantly higher caseloads. However, what has been removed from their work queues are the easy ones. Now they get a steady diet of “exceptions.” What criteria are they using to “clear” exceptions?
• Have your agents figured out how to bypass the system? Is there now a need to think about an audit process or sampling methodology for your distribution partners?
• Underwriting is a dynamic discipline; so is the discipline of managing the underwriting function.

Protect Your Blind Spot
What’s the best way to make sure you have your blind spot covered?

• Stay neutral. Oftentimes we let bias become reality. Don’t be afraid to test an assumption. You know what happens when you make an assumption.
• Be systematic. Develop an annual heat map and plan the calendar. Sit down with your team and ask questions such as, “What haven’t we looked at in a while? What segment is growing rapidly? Where have we had changes in our staff? Where have we had changes in our agency plant? What new underwriting initiatives have we launched recently?” Don’t be afraid to ask questions that may give you answers you don’t want to hear.
• Be transparent. Make the audit process part of everyday routine versus a special event. Communicate the plan, the evaluation criteria, and the purpose of the review. Include staff where possible in the process and the reviews. Give people the chance to see file reviews, and allow them to come back to the auditor with corrections. Reviewers sometimes overlook things. This step helps increase buy-in and credibility.
• Be clear and concise. Audit reports can become tedious and recount every deficiency found. You are looking for trends and system-wide weaknesses, not one-offs. Get to the point and say, “Here’s what we’re doing well, here are the two or three things we must do better, and this is how we plan to fix it.” Quick and direct is easier for people to absorb and adopt.

Everyone has a blind spot. A solid system of checks and balances that supplements your management information systems can protect your blind side. Rolling up your sleeves and diving into the actual files adds a qualitative dimension that information systems cannot reproduce. More importantly you get a first-hand look into what is really going on. The old adage still applies, “Trust, but verify.”

I welcome your questions and comments about this article. Please contact me at ray_mazzotta@renolan.com.