Helping insurers compete in the age of disruption
Insurance has always been about risk transfer driven by data and analytics. Insurers were conducting in-depth analyses long before computers or big data. And, much to the same degree, insurance has also always been about trust, risk mitigation and protection. But over the past decade, the relationship between data, analytics and trust has changed, posing significant opportunities and challenges for the incumbents in the insurance sector.
The rise of smart, connected devices, the internet of things (IoT), well-funded InsurTech start-ups and sophisticated data and analytics is disrupting the industry. The flood of new data is allowing insurers to track how insured ‘things’ move through the risk universe. With this trove of fresh information, insurers now have a sizable opportunity to deliver new benefits to their customers and to society as a whole. In fact, prevention over protection is quickly becoming the new norm. It is quite clear that analytics will be a mainstay for many of the decisions that insurers make in the future and as this trend continues, more focus will need to be placed on ethical questions and the power of trusted analytics.
Developing and managing trust in the algorithms
By embracing advanced analytics and black box algorithms, insurers have the ability to hand over many of their business decisions to machines. Machine learning and cognitive computing are being harnessed to perform more complex yet routine tasks and, increasingly, to manage some underwriting capabilities and claims processing. Many insurers are already applying predictive and prescriptive analytics to optimize operations, reduce losses and manage risk. While these uses of data and analytics certainly can deliver significant benefits, they also require the organization, from executives down to the call center agents, to trust that the algorithms (and the insights they deliver) can truly help reduce costs, improve efficiency and make better risk-based decisions. On the flip side, employees may be worried that new predictive analytics tools and InsurTechs will make their specialized skills redundant. They may be skeptical that a robo-advisor-based InsurTech can provide advice and make decisions as well as trained professionals. Executives and managers often view the introduction of outside data as having a diluting effect on their proprietary risk profiles and claims data. Establishing the proper controls and methodologies to manage trust in the algorithms will also require executives and analytics leaders to think carefully about how they develop and manage their analytics across the organization. There is also a need to embrace and learn valuable lessons about the long-term management of trusted analytics from the InsurTechs.
Fundamentally, it may not be possible to understand exactly why a model based on machine learning made a given decision, which could create concerns for regulators. For example, some InsurTechs use historical data to train a machine to make decisions in claims processing based on long-term patterns and thereby embed historic ethical precepts automatically. As the ethical environment changes, these precepts may suffer from a lag, which ultimately could erode the customer’s and the regulator’s trust in the insurer unless it periodically tests and reviews the ethical framework embedded in its analytics. While the wider integration and application of data and analytics in the insurance sector is already prompting provocative questions, it will take some time for insurers to understand the full impact of InsurTechs and data analytics on trust with both internal and external stakeholders. What is already clear, however, is that incumbent insurers must start rethinking the relationship between InsurTech, customers, their customers’ data and trusted analytics as they hasten to embrace new business models.
Download the third article in the Trusted Analytics series: Helping Insurers Compete in the Age of Disruption >>