The relationship between the insurer and the insured is drastically changing. Besides transparent markets, in which information is publicly accessible and customers can easily compare policies and premiums from different insurance providers, technological advancements in data analytics are developing at a rapid pace. Therefore, we will discuss if the traditional business models are sustainable and highlight a new business model around claim prevention as an innovative way to differentiate from competition and stay relevant in the insurance landscape.
In order to survive in this drastically changing insurance landscape it is necessary to embrace new technologies and harvest the benefits of data available.
Data Science & Machine Learning are enabling insurers to make deep predictive customer analyses. By deploying smart algorithms that analyze the data and translate these to tangible actions, insurance products can be personalized and tailored to the behavior and needs of each individual customer. This approach will put the customer at the center of organizational decision making. Applying innovative Data Science & Machine Learning solutions will positively influence conversions and portfolio volume on targeted profiles and give the portfolio a profitable boost.
Predictive models have become more and more accurate and are able to capture complex data structures. It is expected that this trend will intensify in the near future. According to Moore’s law the processing power of computers will double every year. This law has been proven for decades and as Intel CEO Brian Krzanich stated: 'Moore’s Law is alive and well and flourishing'. We can only speculate when the limit will be reached.
So which new business models will arise from these developments? The advancements in analytics and the increasing levels of available data will provide insurers with insights in what claims will be claimed in the future with higher probabilities. The list of data that can be used for these predictions is endless and still increasing (e.g., demographics, claim history, sensory data provided by cars and health apps, weather data etc.). Predictions can be made on a corporate level as well as on an individual level. The latter is interesting as we look at the current situation of the insurance landscape. Competition is mainly premium focused due to transparent markets and the perception that the insurance product is homogenous in the mind of the customer. As consequence, insurers have to optimally set their individual premiums based on revenue versus predicted costs while taking into account the volume of portfolio to control for overhead costs. Data Science & Machine Learning can optimize this.
Given the rapid developments in data analytics as well as new data sources that give more real-time insight (e.g. connected cars, smart houses, health apps etc.), insures will soon have a better understanding of the patterns in data that lead to riskful events. It will eventually be possible to translate these insights to predict individual risk to such an extent that every individual is perfectly insured for their individual risk. Premiums will be set based on that prediction. In this scenario, it is important the consumer remains in control of which data is shared with the insurance company. But, the more data will be shared, the more accurate the predictions will be, which will eventually result in a more accurate premium.
From a competitive point of view these accurate predictions seem very valuable. However, in this model every individual will only cover his or her individual predicted risk plus overhead costs. This will undermine the collective character build in the traditional insurance landscape and transforms an insurance provider into a financial institution. In which you as policyholder pay a premium every month to save for an upcoming predicted negative event. So, if individual risk can be perfectly predicted and premiums are aligned with these predictions, how can you differentiate yourself as an insurance provider?
Before we will answer this question, you probably asked yourself, how are insurance companies able to predict risk on an individual level? Large institutions have an information advantage over individuals, their access to data provides them with a better understanding of ‘the big picture,’ which couldn’t be retrieved on an individual level. This enables insurance providers to better understand what high-risk factors are and which behavioral cues lead to an negative event.
Consequently, insurance providers can take actions to prevent these high risk events from happening based on their observations in the data. As we all know, prevention is better than the cure, besides it will produce a substantial decrease in costs. Take for example sensory data from a car, the algoritm is able to recognise that the brake blocks are soon worn out, which is a high-risk factor for damaged brake disk. For the latter, repair costs are much higher than for renewing the blocks only. Therefore, the insurer can recommend getting new brake blocks which are (partly) financed by the insurer. The insured will be happy because the negative event has been prevented from happening and his car is safe again. And the probability the insurance provider has to pay-out a severe claim as result of damaged brake disks has been decreased.
By integrating these mechanisms, the insurance provider can transform from an institution that is handling claims to an institution that is mapping out risks and even able to prevent claims from happening. In this way, the insurer becomes a proactive service provider and can be of real value to their customers. In this relationship policy holders are providing data which enables insurers to help preventing negative events from happening, like an angel watching over their shoulders.
The scenario illustrated above seems futuristic, although, this proactive approach is exactly what the modern demanding consumer is expecting from their insurance provider. According to the World Insurance Report 2018 conducted by Capgemini, customers are foremost interested in an excellent digital service. Besides, tech driven corporations such as Amazon and Google are already heavily investing in and implementing data driven insurance solutions. They already have the technical infrastructures to provide this excellent digital service, and secondly, they generate huge amounts of data themselves through all kinds of consumer products and services they've put on the market the past years. So, to stay competitive to these new competitors, embracing Data Science en Machine Learning techniques will be the solution.
For insurers it is a dynamic time with lots of challenges surrounding them. On the other hand, it is a time of new opportunities to differentiate and to revolutionize the industry. Data Science and Machine Learning are at the basis of this transformation.