Getting employee buy-in may be the biggest key in getting a company to successfully adopt predictive analytics and other cutting-edge technologies, said Martin Welch, CEO of Hawaii Employers Mutual Insurance Co.
Welch has spent 35 years in the property/casualty industry, primarily with a workers’ compensation focus.
As head of HEMIC, considered the largest writer of workers’ comp in Hawaii with $72 million in annual written premium, he’s overseen the organization’s implementation of predictive analytics into its business.
Welch said it isn’t just about whether the technology will make a business better, but whether the company culture can actually survive it. This means organizational buy-in from the top-down.
Welch spoke with Insurance Journal about challenges of getting his company, and his customers, to embrace predictive analytics and new technologies. This has been edited for brevity. Listen to the podcast to hear the full conversation.
Insurance Journal: Can you give me a few words about HEMIC, and any recent changes you’ve gone through that you care to talk about?
Welch: Well, I think we’re pretty used to change here at HEMIC. In fact, it just might be in our DNA. To start with, we’re only a 19-year-old company, and that’s still a teenager by insurance company standards especially for a line like worker’s comp.
I’ve always said this really gives us a great advantage from a technological standpoint because we entered this century with a clean slate. That enabled us to establish a paperless process from the beginning, and really build an organization that I think is very data driven.
That’s unlike most of our competitor’s I think who probably had to pull their organizations away from clunky legacy systems of the past if you will. The thing about technology though is it’s continuously improving, and it’s updating. No matter what state-of-the-art systems we may think we have, or you may think you have, pretty soon there’s going to be something better available.
This year, our big change is we made the decision to convert to a newer even more modern operating system than the one we’ve utilized for the past 19 years. We’re in the middle of a 12-month implementation plan.
It’s testing our ability to manage a huge integration project, and at the same time maintain the service levels, and the financial results of our day to day business. That’s a huge change, but so far, I think we’re managing it pretty well.
IJ: You have personal experience implementing predictive analytics into your business. Can you talk about the challenges?
Welch: Well, starting with the technology itself, I think the real success of predictive analytics is dependent on the quality of your data. As an organization, you need to be able to capture, store, manage, access a lot of different data points about your customers, perhaps many more than you did in the past.
Frankly, this ability to capture data I think really needs to become an obsession in an organization that wants to use predictive analytics. Getting your organization to put a greater priority on that data, collection primarily is the first challenge.
As good as we think our predictive analytics partner is, Valen Technologies, they can’t create that data obsession inside our organization. We have to do that, and that really needs to start even before the analytics.
I don’t think that’s the overriding challenge. As I like to say, technology is the easy part. It’s the people and the culture change that I think is always the most challenging. Trying to get your underwriters to trust the analytical model in a larger percentage of their book is always going to be a struggle.
As our analytics get more and more sophisticated, we can get a much better result through the portfolio pricing that predictive analytics can bring to smaller straightforward risks.
It’s this letting go on the part of the underwriter that’s important, and it really only serves to free them up so they can focus on the larger more complex risks that can benefit from their human underwriter judgment.
This underwriter judgment adjustment is never easy, and it must continue to include more, and more risks into that predictive analytics model. I think the key to this is to get your best underwriters involved in building that predictive analytics box.
IJ: If you can only pick one piece of advice to a business about to embark on a similar journey, what would that be?
Welch: Well, I’ve heard a number of conversations with my colleagues in the industry, and other individuals kind of internally debating this whole predictive analytics decision. I think my biggest piece of advice would be to get started.
I think “journey” is a great word to describe the process of incorporating predictive analytics into your business, but because it’s a journey, you can’t expect to build a perfect and-all predictive analytics process right out of the gate. It’s going to take a number of iterations, it’s going to take some adjustments, recalibrations if you will along the way.
You have to be patient, but you have to get started on it. I know it’s a huge investment, and you want to be sure you do it right, but waiting for the next better upgrade before pulling the trigger I think is a fool’s errand. It’s like waiting for the next iPhone release, or the next model year to buy a new car.
You can always wait, but look at what your business is missing in the meantime. As good as I think our process is today, I know it can and will be better next year, and the year after that.
IJ: Can you name a few advantages of having predictive analytics be a part of one’s business?
Welch: I think first and foremost, it’s discipline. As an underwriter, and as a corporate leader, I’ve always believed that the key to underwriting success is discipline, and that’s effectively assessing the risk, putting an accurate price on it.
Predictive analytics helps us to codify our organization’s underwriting discipline. People understand what it means.
I think another advantage would be actionable business information. You’ll be able to see trends, patterns, correlations in your results that will enable you to make adjustments in your selection pricing models. Even loss prevention or claims attentions on a real time basis because of the predictive analytics feedback.
Maybe a more subtle advantage is providing clarity for your underwriters that in a business that’s always been some blend of art and science, it’s both intuitive and analytical, and how we practice underwriting. Predictive analytics tells us that the science, the analytical really needs to dominate the process. Not our gut.
I think today, across our book anyway, this blend is probably 70 percent, 30 percent science over art. When I started in this business, it was probably closer to 50/50, if that. I suspect that today’s 70 will become 80 percent and higher as our industry’s use of analytics evolves even further.
Was this article valuable?
Here are more articles you may enjoy.