The world witnessed something very novel in 1996 as Garry Kasparov began to play chess in the Pennsylvania Convention Center. There was nothing unusual about Kasparov’s game of chess, but what was unusual was his unknown opponent, the ‘Deep Blue’, playing from the IBM center in New York. The media even hyped this event as “the future of humanity is on the line”. For every move by Kasparov, the Deep Blue, powered by 256 processors in parallel, could evaluate more than 100 million positions per second and make its move. Kasparov won the match 4-2, losing 1 game to Deep Blue, winning 3 games and 2 ending in a draw. But Deep Blue returned one year later with more enhanced logic and defeated Kasparov in 3½ – 2½. This historic match paved the way to show that computers can mimic or even beat the best experts, when programed with combinatorial analysis allowing them to evaluate multiple parameters simultaneously.
The insurance industry has two things to take from this historic match:
- Deluge of data that is associated in decision making and the combinatorial patterns and outcomes
- Expert knowledge available with the agents in deciding about the products each prospect would need and patterns of success
Gartner predicts that big data infrastructure will be $232 billion across industries by 2016. With that said, the insurance industry has leveraged its inherent statistical expertise and has started making computers do the sedulous work to guide insurers on the next best action to take in business scenarios. This prescriptive analysis guides insurers across all the value chains – sales and distribution, policy management, claims management, customer service, and so on.
For instance in sales in distribution, there are various ways in which predictive analytics can be used across lines of businesses and users. The insurance industry witnesses a significant attrition in agents, sometimes around 80% with life insurance. Since agents are able to create more business with insurers who are easy to work with, insurers working with independent agents cannot achieve their objective of profitable growth without empowering their agents with tools to guide them in the business. 80% of insurers acknowledge this and are providing agent portals to equip their agents with faster and accurate decisions and tools.
For an insurer, it may be interesting to know the behavior patterns that might indicate that an agent is moving out, in order to prompt preventive actions. For an agent, it could be handy to know which products will best suit a certain type of prospect, given the needs, behavioral analysis, past history of conversion, life time value of the customer, etc. It may also be interesting for both the insurer and the agents to know customer retention rates and how customers can be retained, since it is more expensive to acquire new customers than to retain existing ones. One insurer saw the customer retention rate at 85% across multiple years. This retention rate can be misleading. The insurer and the agency might take this to mean that all is going well, however a deeper analysis revealed more information on which policies were cancelled and during what year. This has a lot of input on whether long term customer retention is the issue or new customer follow up needs improvement. This, in addition to the behavioral patterns, profitable products across customers, customer life time value and effectiveness of various retention initiatives, can help guide agents as to where they need to spend the most time. The insurance landscape is transforming to get the “Deep Blue” kind of algorithms packaged within a simple application to navigate the unknown waters.