Predictive modelling changes business

Jan Lindeboom |

Predictive modelling is hot. Big Data, Artificial Intelligence and Machine Learning are applications that are going to determine how business is conducted in the future and will overturn current business models. The raison d’etre of the insurance sector is traditionally largely based on predictive modelling. Risks are predicted based on (historic) data and a corresponding premium is calculated. Today, computer power and data are available everywhere in large quantities. That means opportunities and in particular, work to be done in the insurance sector. Further steps have to be taken to safeguard the raison d’etre. Move to insurance 3.0.

The computer knows who you are and what you are going to do

How predictable are you? Human behaviour is determined for 95% of people by habits and incentives that are (un)consciously perceived. The fast-growing computer power and available data makes this human behaviour increasingly easy to predict. A few years ago (with the help of smartphone data) the computer could already predict where we would be in 24 hours with an accuracy of up to 20 metres! The pictures on our smartphone also make us predictable while our behaviour on social media reveals our personality and people are becoming increasingly aware of this. This produces different reactions. Some people go to greater lengths to protect their personal data whilst there are also plenty of people who experience the convenience of personalised services and/or are willing to exchange their data for discounts or other benefits.

Insurance becomes customisation

With more data we are better able to match products to the need of the consumer. At the start of this journey consumers will perceive this as a service, but this will be short-lived. Customisation will soon become a hygiene factor; customisation in product and in lead time. Those who are not providing it will see their market share fall. For insurance companies this means, for example: supporting more variations of a product, continuously developing new products, accepting changes faster and analysing the risks in a much more personal way.

“Protect against” will become “Protector of”

When insurers are able to predict risks, the consumer will want to be warned about them. They will no longer simply be insured against calamity but also be warned about potential danger. In the next stage there will be demand for certainty of outcome. It will no longer be a matter of being “protecting against”, but become a case of being “protector of”. When the consumer makes his or her data available and the insurer is able to use this to avoid damage and perhaps able to guarantee outcomes then a win-win situation is created.

New opportunities, different requirements

Insurance 3.0 places new requirements on the IT infrastructure. Usage, storage and access to data is crucial. It must be possible to adjust and adapt products quickly. And what about scalability and security? IT systems no longer last 10 years. A modern IT infrastructure is essential and is the domain of specialists. That is one thing that we don’t need predictive modelling to foresee.

About the Author

Jan Lindeboom

Jan Lindeboom works as Principal Consultant at Keylane and has been involved at the cutting edge of business and IT in the insurance sector since 1998. As well as working as consultant, he gives lectures and regularly publishes items on new technology and its impact on business and society.