In the traditional model of an insurance company, the role of actuary was relatively easy to understand. They were the people who were trusted to get the pricing of risk right.
There would have been a lot of judgement involved in the pricing exercise – but there would be a lot of data too. Perhaps mortality tables for those actuaries pricing life products – perhaps traffic report data for those pricing motor and so on.
Some of the early questions that cropped up when the insurance industry started to get its teeth into things like big data, machine learning and AI were: What will happen to the role of the actuary? Will actuaries be replaced by algorithms that are quicker and cheaper?
Naturally enough actuaries said ‘no’ while data scientists said ‘yes’ and they continued to exist side by side – but the latest trend of ‘digital-only insurers’ seems in danger of turning such questions on their head.
It may soon no longer be a question of whether data scientists will replace actuaries but rather one of whether all senior staff in insurance companies will have to have a background in data science.
The simple fact is that technology makes it easier for insurers to gather more and more data – most of it from their customers. It provides insurers with the ability to interrogate the data to draw conclusions about far more than pricing. It provides insights on how products can be packaged best to maximise sales. Marketing, branding, distribution, renewals, claims – all of these areas can usefully use data insights that are arrived at with some precision.
In other words, the whole of insurance as a business turns on the expert interrogation of data. Gut feeling and odd bits of customer feedback have been superseded by large amounts of information derived from even larger data sets.
What this means for traditional insurers is already plain to see. They will have to follow the example that has already been set by digital-only insurers where no one gets hired unless they have a solid grounding in data science.
A typical morning for the C-suite within many of these digital-only insurers involves trawling through the data that came in overnight – how many website visits, from which geographies, how long did the visitor spend, did they commit to a product or were they merely comparison shopping, where did they go after they left the website, what messages did they leave, what questions did they ask, were they first-time visitors or repeat visitors, was traffic up or down on the previous day, was traffic up or down on the previous week or month.
And there is also a host of other data to be digested that is derived from ‘social listening’. How is the insurance brand being discussed on social media and what does it mean?
There are many ramifications of this constant trawling of data – but perhaps the most obvious is in launching new products. Some digital-only insurers claim that they can envisage a new product and launch it within a week – including everything.
Insurance is fast becoming a scary but exhilarating business to be in – and much of it starts with data science – and probably ends with it too.
Paul McNamara
Editorial director
Middle East Insurance Review