(Image credit: Gerd Altmann/Geralt.)
Insurers are focusing on updating their data capabilities, including big data and analytics, in 2019. One significant component of these efforts is an investment in data science teams. New and improved data sources, such as third-party data and data collected via the IoT, are providing new opportunities for insurers. Insurers learning how to better harness the power of their own data and how to incorporate reliable outside information is one of the catalysts launching data science programs.
Data efforts should be part of an organization’s overall strategy and culture if they are to be successful. Discovering an interesting pattern of customer behavior won’t impact the bottom line or improve the customer experience if that information isn’t acted upon. Insights should be tied into carriers’ strategic goals; that’s where the true ROI of data science investment lies. It isn’t always easy to convey the ROI of data science initiatives, but it’s necessary to secure buy-in. To be successful, data science programs need clear support from the top of the organization.
Investments in Data Science
IT executives at carriers of all sizes have successfully secured that support. Almost a quarter of the CIOs interviewed as part of Novarica’s Insurer IT Budgets and Projects study reported that their organizations already rely on big data tools like NoSQL and Hadoop, and many more carriers are piloting these capabilities. Third-party data providers are a crucial part of the growing reliance on analytics, with a small number of insurers using data sets like raw internet consumer data and weather data to inform their strategies.
Both large and midsize insurers are increasing their investments in data science and related initiatives. According to Novarica’s study, 27 percent of large life/annuities insurers and 35 percent of large property/casualty insurers are expanding their data science efforts to some degree, while 13 percent of large life/annuity insurers are piloting an initiative. Midsize insurers are similarly active in the space, with 20 percent of life/annuity carriers and 24 percent of property/casualty carriers looking to expand their data science efforts. Another 20 percent of midsize life/annuities carriers and 4 percent of midsize property/casualty carriers are piloting initiatives this year.
Preparing for Success
One of the best ways to obtain the most value from data science initiatives is by identifying insights that can be directly embedded into the organization’s workflows. Partnering with business unit leaders is crucial to align insights with optimal business processes. On the tech side, integrating analytics into core systems is most often achieved through an API or through a custom interface. In addition to workflow integration, population of data lakes from core systems or other new data sources is an important capability to fuel data science efforts. Nearly 50 percent of property/casualty carriers and almost 40 percent of life/annuities carriers have deployed predictive analytics initiatives that are integrated with core systems. Another 40 percent of life/annuities carriers have a deployment on the horizon.
In addition to integrating with core systems, having a test-and-learn approach to data science initiatives can allow carriers to quickly identify which hypotheses are worth pursuing. Equally useful is learning from failed experiments to provide insights to apply to the next iteration. Testing out small pilots using various internal (such as customer interactions and claims notes) and external (such as IoT sensors or wearables) sources of data can help insurers feel confident in their analytics results.
Investment in the appropriate resources to drive success in a data science program is key. People who understand the possibilities that data can provide, have the skills to organize and mine data, and are able to translate that into business value are critical. These roles are most likely filled by very different people who work collaboratively. No matter how the team is organized, support from the top can ensure that data specialists are empowered to do their best work. Collaboration between data science teams and business units is crucial, especially when it comes to understanding what business problems the data results are solving.
Data science programs are complex and can be costly; identifying the right opportunities to launch such programs can help secure buy-in from the leadership team. Every team is vying for funding and resources, but the ROI that typically compels the C-suite to prioritize a project is harder to quantify for data science initiatives. Using business-focused KPIs can ameliorate concerns over data science ROI.
Acquiring the right talent will also alleviate worries about the efficacy of the data science team. These teams should be designed to include at least one key member with an understanding of the business’s needs and concerns. Not only does this understanding help the business and data sides of the house stay aligned in their goals, but showcasing a commitment to operationalizing data insights can benefit the program during funding and prioritization discussions.
Insurers are well-known for having large stockpiles of consumer data, but operationalizing that data is key. Combining data sources and platforms ensures that the data science team has a reliable data set to work from, and investing in a data science team with business knowledge will ensure that the program’s goals are aligned with those of the organization. Keeping these components in mind can help insurers realize long-term benefits from their data insights.