Big data and Insurance: the important question

There is a lot of ways that Big Data can improve people lives, business performance, systems performance, product and so many other things.

One of the controversial areas is insurance. What is insurance? Let’s ask the Oxford Dictionary:
” An arrangement by which a company or the state undertakes to provide a guarantee of compensation for specified loss, damage, illness, or death in return for payment of a specified premium.”

At the core of the insurance world is the concept of solidarity. Insurance in its original form started in the ancient ages, and it entailed the agreement of mutual aid in its very simple form. It then evolved over time with different methods, but always retained the connotation of mutual support. It then developed in its more modern form in England, in the 17th century, particularly to support maritime trade (Lloyd’s). Life insurance also modernized around those times, where forms of mutual support started emerging and communities would put in contribution that would then be given to the unlucky families with a missing parent. Solidarity has then always been at the core of insurance.

Imagine now that you could use Big Data to better estimate personal risks. Imagine you could use twitter history, or facebook posts, or your wristband or step-counter, to better understand what type of risk you are. And what type of premium you merit. We are not miles away from that. Add IoT on top of that. In a few year time, my refrigerator will probably host my full grocery list, and you could easily work out the family consumption of sat fat, non-sat fat, protein, sugars, alcohol. Do you eat chia seeds? That’s good for Omega 3 and 6. Do you eat bacon? Not so good, we have an increased risk of coronary disease here. How much you drive? How much you walk? How long you sleep? How well you sleep… From a data standpoint, this is the Eldorado. You can get to calculate risk factors with a level of accuracy impossible just a few years ago (and you don’t need all those confidential data. Trust me, a subset of less confidential data is enough to get a good enough profile of you). From a solidarity standpoint, not so good.

What does it mean for insurance as an industry? Which way will the insurance company go? If you have pinpoint accuracy at personal level, what you are really insuring is utter pure chance. You are run-over by a bus. If you leave a perfect life, your premium is minimum. If you leave a less balanced life, your premium is higher. If you live a reckless life, chances are the premium will be so high, you won’t be able to afford it. So insurance won’t be for everybody. It will only be for the wealthy, and the balanced. Which is definitely a smaller market than it is today – and I won’t touch fairness.

Can this be avoided? Not so sure, unless the industry as a whole take a position. Without an industry-wide regulation, the benefits of tailoring premium to specific individual risk is so high that a challenger will just go for it.

Leave a Reply