Perfection takes time—and time is a valuable commodity. It’s also a priority for the actuarial science research we conduct at LifeScore Labs, where our mathematicians and statisticians are dedicated to perfecting underwriting through machine learning and artificial intelligence.
When we applied our new machine learning-based algorithm to years’ worth of historical policies from MassMutual, the new approach outperformed traditional underwriting by 6% on the basis of claims. Needless to say, that improvement made the underwriting, actuarial and finance executives at MassMutual very happy.
A need for speed
Very happy executives are a very good thing. But we weren’t content to stop there. We were most interested in the performance metrics of the company. And as anyone who sells life insurance can tell you, the most important performance metric is time, as in underwriting speed.
So when our AI and machine learning experts created a risk model that beats traditional approaches by 6%, we still had one burning question: Is it fast? How much time from inquiry to sale? How much time from when a consumer asks about life insurance to when he or she purchases a policy?
In the first two years of operation, the algorithm reduced approval time by 25%, leading to a 30% improvement in acceptance rates.
As MassMutual deployed the new LifeScore Labs algorithm, we found that the new approach led to quicker sales. In the first two years of operation, the algorithm reduced approval time by 25%, leading to a 30% improvement in acceptance rates.
Simply put, the company saw a 30% boost in policy acceptance by consumers in ultra-preferred policies and shaved a week off approval time. Now that’s impressive.
While we can’t divulge actual numbers, the ROI of faster approval time and policy acceptance saved MassMutual millions of dollars in operational efficiency. And the longer LifeScore Labs’ predictive model is in place, the more that ROI may increase.