By Sanil Basutkar, Co-founder at HealthySure
Insurers currently have a lot of data, but they have been slow to monetize this. This data can be used for creating new lines of businesses. As more customers go online and transact digitally, the number of available data points keep on growing. New analytics technology like big data, AI/ML enables insurers to exploit that data in ways they had not previously contemplated. However, many insurers encounter organisational hurdles in becoming data-driven businesses. Others are waiting for commercial applications to emerge before investing adopting these applications. As a result, insurance industry has lagged others in terms of investment in and use of analytics. However, there are some exciting things brewing.
What is Big Data and Machine Learning and how can it be used in insurance?
Data is continuously generated every time we open an app, search Google or simply travel with our devices. Organizations now have a massive collection of valuable information that they need to manage, store, visualize and analyze. Traditional data tools aren’t equipped to handle this kind of complexity and volume, which has led to a slew of emergence of big data software and architecture solutions designed to manage this. Big data generally refers to these massive and complex data sets that are rapidly generated and transferred from different sources.
Machine learning on the other hand are algorithms to simply find patterns in data and use them to make predictions. As new data points get added to these algorithms, the model starts becoming smarter and as a result, more the data points, smarter is the model. So, any machine learning algorithm will predict more accurately with time.
When insurance providers tap into the vast repositories of Big Data that is available to them and combine this data with machine learning and AI capabilities, they can develop new products that can unlock new channels of growth and customers.
How Tesla is enabling custom motor insurance for their cars using data
With Tesla’s approach, every human interaction with the vehicle (turning, braking, parking accuracy) generates a data point that is analysed and used to improve or create new auto pilot algorithms which are sent back down to the vehicle via software updates. Tesla owners are not just driving a car, they are simultaneously training the Tesla AI/ML engines as they go. As a result, Telsa has created one of the most effective crowd-sourced AI/ML training initiatives around today. Tesla has not only used the data points to improve their self-driving capabilities, but it has also launched its own motor insurance for their vehicles. By analysing the driving patterns of their drivers, they can effectively predict risk of vehicle damages. This has resulted in cheaper insurance for their car owners, as their cars have resulted in lower crashes compared to others. This has not gone unnoticed among car manufacturers, and we can slowly see more such customized data products based on driving data points.
How digital health records and genetic data can transform the life and health insurance industry
As genetic testing becomes more mainstream, technology and data analytics can help insurance companies appropriately price their risks particularly in health and life insurance. It is now possible to predict the likeliness of cardiovascular, diabetic and other diseases based on an individual’s genetic data. While there are ethics and regulations that need to be overcome, it seems inevitable that this information will be used to price risks in the future. As medical records become digital, it is also becoming very easy to share these records across the ecosystem. Insurance companies are sure to evaluate the possibilities of risk pricing based on these records. Once data points are captured, it becomes easier to run machine learning algorithms to understand the ideal pricing. India’s digital health ID can also potentially have these use cases.
The scope for other insurance sectors to leverage data
Property, marine, travel or any other form of insurance underwriting happen in a similar way. Products are created based on underwriting rules that are arrived through careful actuarial models. Data analytics can enhance these capabilities multifold and give models that can potentially drive down the price of insurance. The challenge here may not necessarily be implementing data tools to underwrite, but more on creatively building products that can leverage this data.
Insurers that are tapping into data are becoming competitive and those that are lagging will slowly start losing their competitiveness. The insurtech ecosystem is evolving rapidly and are helping traditional insurers adopt the latest technologies. The scope for disruption is very high in insurance because of the emergence of data analytics.