The Big Bang: Data and Its Explosive Impact on Our Lives
There are estimates that say people create 2.5 quintillion bytes of data daily. That's 18 zeroes. So with each swipe, click, tap, share, repost, purchase, geotag or comment, we're filling up 100 million blu-ray discs – each with the capacity to hold 25GB worth of data. Stacked up, it would measure the height of four Eiffel Towers…. each day.
It's a nice stat. But does it mean anything?
Despite the huge numbers we're looking at, data is also extremely personal and a stark contrast to the extremely impersonal "personal information" we used to cull before.
Insurance is heavily reliant on data to formulate models of the future and put an appropriate price tag on risk. We're always looking for the right bit of data to help us along, but how do we know when we have enough?
Data is like science, and science is constantly evolving. The amount of data is clearly never enough but it's more about how we use the right dose of data to help make society more resilient. And the machines we have these days have the capacity to absorb, learn, direct us. Maybe then it's not a question of how much but how we use it and how we constantly apply it as this science evolves.
Data: Personalising Insurance on a Grand Scale
In the past, we assessed the risk pool and set premiums based on generic and often one-off, and sometimes outdated information. Population census reports, hospital data, or customer claims. The approach tended to be more reactive and generic.
The growth of new, real-time and unconventional data gathered from health and wellness apps like sleep, eating habits, exercise levels, moods and biometrics like glucose levels, blood pressure allow us to gain deeper insights into the behaviour of our customers and make more accurate assessments of a person’s actual risk profile. We are able to design more personalised and dynamic insurance protection to suit the unique and changing needs of our customers, and encourage them to manage their health conditions through incentives.
Better Managing Chronic Conditions with Dynamic Pricing
According to our latest Health Protection Gap Report, households managing chronic conditions account for most of the gap in Asia. This is especially true in emerging Asia, where rapid economic and income growth, as well as urbanization, has led to changes in lifestyles and the prevalence of lifestyle diseases, particularly diabetes, hypertension, and high cholesterol. These largely preventable diseases increase the financial burden of a household and the society at large.
New analysis has been looking at the relationship between certain chronic conditions and behavioral risk factors, which can be modified through behavioural changes – for example, exercising more, dieting, smoking less (or quitting) and changes in alcohol consumption. Dynamic underwriting can be a real game changer in driving behavioural change with the right incentives.
For example, in Thailand, Swiss Re has designed the first dynamic pricing product for diabetics. With digital health management tools and the ongoing underwriting process, the insurers can closely monitor a diabetic’s medical condition and adjust the premium based on periodic health data provided by the customers.
If we're able to change their habits and get clients to eat better, embark on a lifestyle change and be societally productive, it also means that we've succeeded in helping to raise that person's quality of life. On a more macro level, it means the societal burden is reduced and resources that were to be allocated for a diabetic patient can now be reallocated to someone – or something else. The ripple effect is real.
New Risk Pools
The exponential growth of new forms of data in real-time also enables us to better assess new and emerging risks, allowing us to tailor insurance offerings to a specific group of people in need.
For example, in Singapore, we are working with our partner on the region’s first parametric health product that helps pregnant women with gestational diabetes. According to research, one out of five pregnant women in Singapore suffer from gestational diabetes.
The unconventional data also provides insights into a customer’s potential protection gap, allowing insurers to tailor and recommend suitable offerings. For example, a pharmacist can offer cardiovascular health protection insurance to a regular buyer of hypertensive drugs based on outcomes of the predictive model which indicates the person or a family member may be suffering from a heart condition.
We also use data analytics to project when a customer will end his health policy and help insurers to take actions to suggest other protection options as a lapse could again have a devastating personal effect and cause a wider-ranging societal one. In a recent case study, our predictive lapse model and a predictive underwriting model showed 90% and 95% accuracy in lapse and underwriting decision predictions respectively.
There are many more ways to leverage the explosion of traditional and unconventional data. As we integrate machine learning – a computer algorithm which examines large sets of data and identify common patterns to solve assigned tasks - to Magnum, our automated underwriting system, we can systematically classify, store and process new sources of data more effectively and efficiently.
Data science is constantly training our underwriting engines to identify and analyse the unspoken patterns behind this data using machine learning, AI training to federated learning, enabling insurers to incorporate these capabilities to handle even larger volume and more types of data in their future solution design.
The explosion of data and new technology provide insurers the opportunities to transform the insurance customer experience, making them essential health and wellness partners in people’s lives and the bedrock of resilience in society. We're also pretty certain data will have a profound and personal impact on how we live our everyday lives in the future.