image

Although they’re not quite the same as a crystal ball, data science and predictive modeling offer a glimpse into the future in their own right.

“There has been an exponential increase in data over the past several years,” notes William Drake, AVP, Advanced Analytics at Nationwide. “We are getting more and more data with the right context to make informed predictions.”

Instead of guessing what tomorrow will bring, data scientists are tapping into the predictive power of data to gain an edge in the market.

Understanding data science and forecasting

Data science involves, quite literally, using scientific methods to analyze data to gain insights. These insights can be used to help businesses make decisions, develop strategies and, ultimately, drive profit.

Traditionally, data science has been used to take a deep dive into the past to gain a historical view of a given subject. It accomplished this by looking at data linearly. For example, data may demonstrate that younger drivers are considered to be a bigger risk than their older, more seasoned counterparts. 

Today, data science has become much more sophisticated and is being harnessed to develop a more forward-looking model. Looking at this same driver scenario through a predictive modeling lens reveals a much more in-depth look at each individual risk. For example, algorithms can pull further information on whether the younger driver has additional applicable experience, such as working on a farm or serving in the military. It also pulls in other related information, such as a driver’s credit score, years employed and more.

“This approach takes traditional information, such as a driver’s record and years of experience, and puts it together with third-party data to see if there are other factors to take into account when examining a risk,” explains Carole McIntyre, Senior Director of Underwriting at Nationwide. “This is a cutting-edge model that allows us to use data much more efficiently than we have previously.”

Industry trends taking shape

Predictive modeling is already having a major impact on the trucking industry. One prime example is employing data to increase the safety of customers and their fleets, which is paramount.

“Ultimately, we share insights we gain from data with our customers to help them operate more safely,” says Drake. “Telematics is huge, and it’s one of the more obvious applications.”

However, gathering relevant safety data for the trucking industry from a myriad of sources is complicated and involves a large commitment on the part of the organization to ensure it’s done thoroughly and properly.

“It requires an investment, and Nationwide is making that investment,” Drake notes. “We’re here to do more than just help our customers when they have a claim. We also provide support services powered by data that add value for our customers.”

Leveraging insights in Commercial Auto

To gain relevant insights from data, Drake and his team begin by considering what decision needs to be made and then working backwards. They examine data from Nationwide, the U.S. Department of Transportation and other entities, curate it to isolate the data that is useful and then uncover insights into patterns that impact the trucking industry.

When evaluating data, often the analysis will trigger additional follow-up questions. Following this trail of breadcrumbs can provide additional information the business needs to make informed decisions about mission-critical areas. On example is using data to inform risk selection.

“We use volatility insights to understand our risk selection proposition,” he says. “These insights have a direct impact on how we determine tiering for a fleet. We use this information to stack the portfolio in a certain way.”

The development of a risk selection strategy goes hand-in-hand with creating a pricing strategy to ensure a strong portfolio.

“Data modeling helps us determine where the markets are flexible,” explains Drake. “It benefits everyone if we are pricing appropriately and are mutually profitable.”

This is a collaborative effort that will help Nationwide as it evolves to an agency management model. Underwriters work closely with general agents and data analysts to uncover opportunities, work more efficiently and gain a competitive advantage in the market.

McIntyre agrees that state-of-the-art machine learning is having a positive impact on both Nationwide and its partners. “We’re taking our data in real time and coming up with new and better ways to price and segment our business. That’s putting us ahead of our competitors,” she says.

The sky’s the limit

McIntyre believes predictive modeling can support the business well into the future by positioning Nationwide to insure a broader range of accounts that will fit its program at the appropriate price.

“As we continue to collect more and more data, our algorithms learn based on what we are doing and expand upon what we already know,” she says. “Pricing will become less stagnant and more dynamic, and we’ll have greater flexibility.”

For his part, Drake sees a future where insights from data help streamline operations so the business can operate at maximum efficiency—something that will benefit both Nationwide and its partners.

“All our investments in data science and predictive analytics lead to an optimized experience and the right risks at the right prices,” he says. “Everybody wins.”