Other parts of this series:
Analytics has always been relevant for insurers. Now it’s essential
This is the first blog in a series looking at why analytics combined with machine learning is crucial for the insurance industry.
The level of competition in insurance markets is fierce: pressure to reduce costs and improve processes is surging, new entrants are challenging business models, and technology’s advancing at an exponential rate. Together, it adds up to a serious challenge. 86% of insurers now believe they must innovate at an increasingly rapid pace simply to retain – let alone expand – competitive edge.
I believe analytics-driven innovation can make a defining contribution here. It’s got the potential not just to improve profitability, but also to transform the customer experience. In this series of blogs (and on our website) I’ll be introducing and exploring the part analytics can play in different stages of the insurance customer journey.
Insurers have led the way in using data for pricing and risk management. And they’ve invested heavily in technologies and people to support these functions. But it’s also fair to say that the focus on these traditional, actuarial capabilities has led to a partial failure to exploit the maturity of advanced analytics.
Other industries (like banking, retail and technology) are now far ahead in their analytics capabilities. They see data-driven business decision-making as vital. They’re re-engineering their end-to-end value chains around machine learning. And they’re taking a holistic view of new digital technologies – a view that is as concerned with operational processes as it is with front-office capabilities.
Insurers aren’t so good at taking this holistic view. In fact, recent Accenture research shows their use of technology was farthest behind disruptive industries (including players like Google and Uber) when it came to running the operational aspects of their businesses. So why should they bother trying to keep up? Why’s analytics so important, and how is it changing?
At a strategic level, it can help insurers (and brokers) develop new routes to growth and profitability. Obviously, it can dramatically improve core competencies like risk-pricing and claims management. It can also improve end-customer segmentation and understanding – delivering the kinds of personalised products and experiences that customers now demand.
But the biggest trend I see coming in analytics is the evolution from predictive to prescriptive models. Whereas predictive models evaluate the probability of an event occurring, prescriptive analytics identify the action (or set of actions) that are most likely to optimise a result.
Take this example: a predictive model for customer retention evaluates the probability of renewal, while a prescriptive model identifies ‘next best actions’ (like changing price, out-bounding the customer, or offering an incentive) that will improve the probability of renewal.
Machine learning can enhance prescriptive models and estimate the change in KPIs like Customer Lifetime Value or customer satisfaction scores. And these new models are much more closely aligned to what insurers are trying to achieve – improvements in value through making the right customer decisions.
Happily, the shift from predictive to prescriptive models supports a key trend in insurance – the desire to create better real-world outcomes for customers and underwriters alike. Prevention is far better than cure, and analytics is driving insurers’ evolution from being mere financial compensators to being risk-mitigating partners.
Done properly, this will also have a hugely beneficial effect on how the industry’s perceived by the public as a whole. But this evolution from reaction to prevention requires two key ingredients to make it work: the Internet of Things to generate datafeeds, and the right analytics to make sense of it all, increasingly in near-real time.
It’s why we passionately believe now’s the right time for insurers to be redefining their approach to, and use of, analytics. And in the follow-on blogs in this series I’ll be proving my point: demonstrating what analytics can do at every stage of the customer journey.