Stuart Rose, Market Strategy Director, Predictive Analytics at Guidewire
The concept of “predictive analytics” has been a vital part of the insurance industry for an incredibly long time – long before the IT industry picked up on it as a way to describe the amazing potential in being able to apply predictive powers of big data. Back in the day, predictive analysis was done entirely manually; some would say that actuaries were the first ever data analysts since their inception in the 1700s. Today, of course, predictive analytics is far more sophisticated, using algorithms to try and more consistently pinpoint accurate outcomes.
With such a long history of dealing with predictive analytics, you would expect insurers to be experts. A lot of insurers are. However, insurers still fall into the same trap as other companies that need to maximise their data – making sure it is consistently useful and actionable, rather than simply a number on a page.
As they strive to become fully developed digital businesses, more than ever insurers are keen to understand what happens next and what the best decision is at that moment. Predictive analytics is becoming essential to quickly identify what is relevant and irrelevant in data sources, a task that has become increasingly important as new, interactive variables like telematics and social media have become dominant.
So what can insurers invest in to make sure they are maximising the predictive analytics opportunity?
Machine learning tools constantly search data interactions to find hidden patterns in the data that can use to forecast future behaviour The machine learning that insurers can adopt is not so distant from the methods employed by TV streaming websites that know which film or TV programme to recommend next. The algorithms are applied to data about policies and claims to create models of client retention and loss ratios. These models are invaluable in identifying details that the competition cannot see and picking out the most useful market segments to target with more accurate pricing and underwriting. It is also an opportunity to find the best and most relevant customers, and focus on managing their claims in the most cost-effective way.
In today’s competitive insurance industry speed to market, more frequent rate changes and online real-time customer decisions are essential. Insurance companies can no longer wait days or weeks to look at different what-if scenarios before making a decision. Decisions need to be made in seconds, minutes or hours not days or weeks. Machine learning allows data scientists to prepare, explore and model multiple scenarios using large data volumes never before possible, and it provides much faster processing for complex analytical algorithms – both of which deliver better answers quicker to those who need them for decision-making.
While predictive analytics is in the DNA of insurance, there is clearly still more to be done in terms of reaping the full benefits for the majority in the industry. That said, insurers should be feeling confident that predictive analytics is a step in the right direction. Increasingly, companies are choosing to replace their entire legacy software systems with modern core systems that can integrate with analytics and offer a full transition to digital, enabling them to make real-time data-driven business decisions. As new products are developed and additional market segments are targeted, expect to see the adoption of predictive analytics to increase dramatically.
Becoming a data- and analytics-driven insurance company requires much more than data and really good models alone. It also requires tools that the business user can understand and leverage easily, so that he or she can take full advantage of the data. If an insurer’s claims adjusters, underwriters, and sales managers do not have a way to interact with the output of the predictive model, the initiative is pointless.
Insurers are most successful with their predictive analytics models when the proper predictive models, data and infrastructure converge. But mindset is also key: Adapting and succeeding requires an organization to be primed to use models effectively at the business level. When this occurs, insurers can more easily embrace advanced analytics, see the value in their existing data and commit to the ongoing advanced analytics journey.