What if you could help people help themselves through improving their physical, and as a result financial, wellness?
Or detect when insurance or other financial instruments are being used for criminal activity, like money laundering?
Or make life insurance available to people who previously wouldn’t qualify?
MassMutual uses data science to drive all aspects of our customer experience. This ranges from designing algorithms to recommend products and personalize the way in which we interact with our policyholders to speeding up our underwriting processes to gleaning macro-economic and climate information about global longevity trends.
Data science is an inter-disciplinary field that uses scientific methods, processes, and systems to extract knowledge from data. So how does MassMutual approach data science and, perhaps more importantly, how do we capture its value?
The Data Science Team at MassMutual thinks about data science along three primary dimensions: skills, technology, and data. We consider these to be our fundamental building blocks and believe that all questions and problems can be answered or solved with a combination of the three.
Because of the inter-disciplinary nature of the field and the diverse problems facing our industry, the MassMutual data science team has a wide range of backgrounds, all deeply rooted in science and mathematics.
Our team is comprised of mathematicians, computational geneticists, computer scientists, econometricians, biostatisticians and the like. We believe that a diverse set of skills and perspectives leads to effective problem solving and continuous learning and, of course, the ability to take on our industry's biggest challenges.
Technology powers everything we do in data science. We use big data computing systems to power our data warehouses, spark clusters, and machine learning pipelines. We have a passion for open source and leverage it (and contribute) wherever possible (check us out on Github!).
Our team uses that data to extract trends and patterns that are useful for improving the experience of our policyowners, the quality of products and services, and to drive innovation in our industry.
For example, we use customer and marketing data as input into our product recommendation algorithms. These algorithms examine patterns in the preferences and life stages of our customers and then recommend products that are best suited to their current needs. This includes not only identifying particular products but also the way in which they should be purchased, online at massmutual.com, through their workplace, or through one of our advisors.
Real world applications
This approach has created tremendous value for MassMutual across a variety of fundamental problem spaces.
When consumers interact with MassMutual and indicate they are interested in learning more about insurance, data science algorithms use the behavioral patterns generated on our website to predict product preferences and recommend ways in which the consumer would mostly purchase such a product, whether that be through an insurance agent that has been selected based on these data or through a completely digital experience.
Add to the mix what we’ve been doing with our underwriting. In a traditional process, actuaries and underwriters spend time examining the risk factors of life insurance applicants and determining a premium based on the independent evaluation of a limited set of data points. Using machine learning algorithms, we have developed a system that estimates risk more accurately and efficiently and as a result, reduces the time a customer must wait before receiving an offer for a policy and getting coverage.
All this work results in more people getting insurance faster, based on their own needs and preferences.
This is just a sample of the areas data science is improving at MassMutual. It is rewarding and satisfying to know we are helping people find the right type of protection on the right terms, making it easier for our policyowners to secure their future and protect the ones they love.
More from MassMutual...