![](https://edge.sitecorecloud.io/guidewiresodb06-guidewire0f2e-productioncd91-5186/media/project/guidewire/guidewire/blog/thumbnails/blog-20201029-analytics-vision-560x320.jpeg?h=320&iar=0&w=560)
The Big Picture
It has been difficult for the insurance industry to formulate a proper vision for their use of predictive analytics. Executives who set their company’s direction tend not to understand the scope of what is possible and lack direct experience with the benefits and challenges. Conversely, those who do have more experience and understanding tend not to be in a position to establish a corporate-wide vision.
The result has been that most insurers have aspirations rather than visions with respect to predictive analytics. They want to do more with their analytics teams. They want to make it easier to get to their data. They want to shorten the time from idea to implementation. In other words, most of the focus has been on current pain points, and the direction of movement has been controlled by these tactical considerations.
In this final installment, we consider what it means to have a vision around predictive analytics.
The Place of Predictive Analytics in Insurance
First and foremost, insurers are in the business of insurance, not predictive analytics. Insurance executives have a vision for their business – what they hope to provide to customers and who they want to be in the marketplace. Any use of predictive analytics should be in service to that corporate vision, just as a claims department or accounting group does the same. Executives should focus on the overall vision with predictive analytics as the means to an end.
The corporate vision often includes, explicitly or implicitly, a certain level of efficiency in their operations. Insurers want to be the best, or cheapest, provider, or to provide the most value to their customers. This can’t be achieved without smooth and efficient operations, and predictive analytics can be a major enabler in this effort.
For example, if an insurer decides that property inspections and loss control is a key offering to reduce costs for all their insureds and provide value to the inspected customers, predictive analytics can help determine those customers who would benefit most from this service. Accurate pricing and competitive rates are key parts of being able to serve the entire market, and predictive analytics plays an essential role here as well. The goal supports the vision and predictive analytics helps make it a reality.
Second, and perhaps more intriguing, the capabilities that predictive analytics can provide may change who insurers want to be and how they serve the market. In this way, predictive analytics can impact and change the corporate vision. The explosion of insurtechs has demonstrated how new ways of interacting with customers and managing business are made possible through predictive analytics and the allied developments in technology.
Guidewire’s Approach to Predictive Analytics
Guidewire believes that predictive analytics is an essential and inseparable component of a modern core system, not that it stands apart. The results from predictive models work in tandem with other business rules, transactions, and workflows to enable a smarter insurance operation. And who is being made smarter? It is, in reality, the end user – the underwriter or adjuster who has a task before them – who is made smarter by being provided with good information about likely outcomes. In this way, those people can become brilliant in the moment that a decision is required.
This blog series has talked about many of the steps required for an end-to-end approach to analytics and has argued that a platform approach that utilizes a seamless integration from first mile to last mile can increase the efficiency and return on investment of predictive analytics efforts. But in some ways the previous blogs have only discussed the tactical components that allow you to achieve this vision. Their importance is really in what they make possible.
Be sure to check out the whole series. Part one covered the need for a comprehensive approach to predictive analytics. Part two discussed data, while part three considered how models are built. Part four and part five discussed how predictive models are implemented and monitored.