Free Preview

This is a preview of some of our exclusive, member only content. If you enjoy this article, please consider becoming a member.

Here’s something we typically hear from frustrated underwriting managers:

• “We bind around 12 percent of our submissions, but our underwriters review 100 percent of submissions.”

• “A 1-2 percent increase in our submission-to-bind ratio would be a game changer.”

Could a business case be made for deploying artificial intelligence (AI) to crack these chronic issues? We think it can.

Executive Summary

Artificial intelligence can affordably help carriers write more and better business right now, asserts Dan Epstein, the CEO of ReSource Pro. Here, Epstein makes the business case for getting on board the AI train and tackles the obstacles of fear, uncertainty and doubt that lead carriers to resist changes that could enhance the success of their underwriters in writing accounts with the right submissions without disturbing the underwriter autonomy that pumps the lifeblood through the system.

Technological Change

The thing about changes in technology is they seem to come slowly at first, then all at once.

It’s tempting to be intoxicated by technology. Its promises are attractive, seemingly on the cusp of fulfilment yet also just beyond reach. Then at some point, you look around and see that the world as you once knew it has changed.

All around us, the population of digital natives grows apace; toddlers are technocrats. Thought leaders tell us that the digital-savvy buyers of insurance are looking for a seamless and intuitive insurance buying experience. The kind of AI that underlies the decision algorithms of Netflix, Amazon and Facebook are assumed to be applicable to the insurance customer and risk management experience.

The dilemma for commercial carriers is not to doubt technology’s relevance but rather to determine where specific uses of it sit strategically on the spectrum of “slowly” or “all at once.” Carriers struggle with the tradeoff between investing in systems today that may soon be obsolete or being left behind. Adding another wrinkle of complexity is figuring out how to integrate new systems and/or replace existing systems that may be suboptimal but still work.

Let’s take one example of a new technology that promises to transform the insurance industry and demystify it: artificial intelligence. There’s a business case for using AI to benefit your business, your people, your customers today. In short, AI can affordably help you write more and better business, right now.

Efficiency Is King

In an environment of seemingly relentless cost and pricing pressures, efficiencies matter. Releasing unnecessary costs can aid competitiveness. Optimizing processes lubricates workflows, increases speed for the customer, and improves quality and risk management practices. This benefits employees, agents and insureds, and ultimately delivers better returns through improved expense ratios and combined ratios.

At a recent workshop, the CEO of a successful regional insurer declared, “Submissions are our lifeblood. Without them, we don’t exist.” Not a controversial point for her to make. Nonetheless, some submissions are more equal than others. The quality of the submission flow immediately informs the production and attractiveness of quotes, which in turn leads to how effectively business is bound.

Standard stuff, sure. Yet at ReSource Pro, our research shows that many businesses lack actionable insights into the submission-quote-bind journey. The data are huddled around somewhere but are not easily susceptible to inquiry. Analyses, often incomplete, of what was planned in submission-to-bind ratio targets compared to what was delivered in a given period must often be composed and interpreted retrospectively.

The great news is hot off the press. Without radical change, AI and machine learning can solve problems and improve results. How? By training a neural network to learn what kinds of business fit the carrier’s appetite, which are most likely to bind and which accounts represent an untapped product opportunity. Algorithms can be built to prioritize business at the underwriter desk level. Efficiency and insight can be applied to the commonly flawed submission-quote-bind process in your operation—and it can start right now.

Consider the possibilities: integrating an algorithm into the concerted efforts of the design and iterations of executive management’s strategy; making augmented intelligence a component of underwriters’ decisions and prioritizations; generating data that can also be useful to you at the point of claims and when soliciting the business you want from your producers, positively helping them understand your appetite.

The obvious question jumps out: If all that really can be done, why wouldn’t you?

Understand the Resistance

This question merits a fair answer. There are barriers to adoption despite the compelling benefit propositions. There may be the visceral concern about whether we trust a thing over a human and whether it’s countercultural to our relationship- and judgment-based underwriting traditions. There may even be the blunt feeling that here is just another add-on bit of technology flaunting itself as a silver bullet.

In this context, as in many others, all of us are vulnerable to that classic change adoption-killer trio of fear, uncertainty and doubt—the fear of complexity, uncertainty as to efficacy and doubts about return on investment.

Any of these objections can become self-fulfilling if the path to the desired outcome is poorly mapped or executed. Or they can be overcome when collaboration and clarity of intent contrive to build the right environment for success. Approached in this way, metrics rapidly render benefits measurable, repeatable and sustainable.

Make AI a Partner, Not a Replacement

We believe passionately that deploying technology to improve returns isn’t about changing what you do entirely but rather about doing what you already do better. AI need not be a monstrous creature supplanting your people. We subscribe to Jim Collins’ philosophy that leadership is about “getting the right people in the right seats on the bus.”

As leaders, our goal is to accelerate their success. Effective algorithms, supported by good processes and engaged people, provide indicators that refine themselves through machine learning over time in a constant self-improvement trajectory. Underwriters use these indicators as tools, quintessentially as guides, learning to trust them to improve results. AI isn’t an isolated solution, nor is it an attempt to rein in underwriter autonomy and accountability that pumps the lifeblood through the system.

To be effective, the introduction of AI into the submission-quote-bind cycle is done in partnership with stakeholders throughout the business: senior executives, underwriting leaders, heads of operations and other relevant participants. When introduced and supported by good change management practices, there is no reason why AI cannot seamlessly fit into the underwriting ecosystem.

The elegance of the AI role in this core cycle of the underwriting process is that it’s both simple (using what you already possess and generate) and simplifying (as data matures, it reduces out-of-appetite submissions). Trawling through submissions and separating wheat from chaff goes with the underwriter’s territory. The territory is easier to traverse when those submissions have already been sorted into helpful “red,” “yellow” and “green” categories. Similarly, as intelligence derived from the analytics develops, quoting can be streamlined and prioritized to maximize those most likely to successfully bind. The financial benefits in expense ratio and in building a robust book over time—with expected co-extensive improved loss ratios—are clear. In a sense, it is “Moneyball” for the flow of business.

Focus on Meaningful Outcomes

Keep in mind that this isn’t about developing a piece of technology, the superior machine learning algorithm; it’s about building a sustainable, integrated mechanism that underwriters will utilize to enhance their success. Integrated effectively into the process, the accretion of data and the use of visualization techniques can enhance communication with brokers, speed response times, improve service delivery to insureds and quite simply ease business flow throughout the value chain.

Underwriters who see positive results from their efforts by hitting target ratios productively will outperform in other ways—such as relationship building and formulating more considered analyses—because they do not have the demoralizing leakage of time and effort of not winning business. As informed patterns of indicators and predictability build through use of the AI, brokers also can benefit from more frequently receiving the right quote for the right risk.

The future is here. Make sure it is distributed in your favor. Right here, right now, write with increased confidence of success.