May 18, 2015
3 Key Steps for Predictive Analytics
by Dax Craig
The basis of competitive advantage has changed for predictive analytics. Having them used to be an edge, but now you have to do more.
The steady drumbeat about the dire need for data and predictive analytics integration has been there for several years now. Slowly, many carriers have started to wake up to the fact that predictive analytics for underwriting is here to stay. According to Valen Analytics’ 2015 Summit Survey, 45% of insurers who use analytics have started within the past two years, and, of those that don’t currently implement analytics, 56% recognize the urgency and plan to do so within a year. Although it used to be a competitive advantage in the sense that few were using predictive analytics, it can now be viewed as table stakes to protect your business from competitors.
The real competitive advantage, however, now comes from how you implement predictive analytics within your underwriting team and focus its potential on strategic business issues. New competitors and disruptors like Google won’t politely wait around for insurers to innovate. The window to play catch-up with the rest of tech-driven businesses is getting narrower every day, and it’s either do or die for the traditional insurance carrier.
All of this buzz about data and predictive analytics and its importance can be deafening in many ways. The most important starting point continues to center on where to get started. The most pertinent question is: What exactly are you trying to solve?
Using analytics because everyone is doing it will get you nowhere fast. You need to solve important, tangible business problems with data-driven and analytic strategies. Which analytic approach is best, and how is it possible to evaluate the effectiveness? Many insurers grapple with these questions, and it’s high time the issue is addressed head-on with tangible steps that apply to any insurer with any business problem. There are three key steps to follow.
First Step: You need senior-level commitment.
You consume data to gain insights that will solve particular problems and achieve specific objectives. Once you define the problem to solve, make sure that all the relevant stakeholders understand the business goals from the beginning and that you have secured executive commitment/sponsorship.
Next, get agreement up front on the metrics to measure success. Valen’s recent survey showed that loss ratio was the No. 1 one issue for underwriting analytics. Whether it’s loss ratio, pricing competitiveness, premium growth or something else, create a baseline so you can show before and after results with your analytics project.
Remember to start small and build on early wins; don’t boil the ocean right out of the gate. Pick a portion of your policies or a test group of underwriters and run a limited pilot project. That’s the best way to get something started sooner than later, prove you have the right process in place and scale as you see success.
Finally, consider your risk appetite for any particular initiative. What are the assumptions and sensitivities in your predictive model, and how will those affect projected results? Don’t forget to think through how to integrate the model within your existing workflow.
Second Step: Gain organizational buy-in.
It’s important to ask yourself: If you lead, will they follow? Data analytics can only be successful if developed and deployed in the right environment. You have to retool your people so that underwriters don’t feel that data analytics are a threat to their expertise, or actuaries to their tried-and-true pricing models.
Given the choice between leading a large-scale change management initiative and getting a root canal, you may be picking up the phone to call the dentist right now. However, it doesn’t have to be that way. Following a thoughtful and straightforward process that involves all stakeholders early goes a long way. Make sure to prepare the following:
- A solid business case
- Plan for cultural adoption
- Clear, straightforward processes
- A way to be transparent and share results (both good and bad)
- Training and tech support
- Ways to adjust – be open to feedback, evaluate it objectively and make necessary changes.
Third Step: Assess your organization’s capabilities and resources.
A predictive analytics engagement is done in-house or by a consultant or built and hosted by a modeling firm. Regardless of whether the data analytics project will be internally or externally developed, your assessment should be equally rigorous.
Data considerations. Do you have adequate data in-house to build a robust predictive model? If not, which external data sources will help you fill in the gaps?
Modeling best practices. Whether internal or external, do you have a solid approach to data custody, data partitioning, model validation and choosing the right type of model for your specific application?
IT resources. Ensure that scope is accurately defined and know when you will be able to implement the model. If you are swamped by an IT backlog of 18-24-plus months, you will lose competitive ground.
Reporting. If it can be measured, it can be managed. Reporting should include success metrics easily available to all stakeholders, along with real-time insights so that your underwriters can make changes to improve risk selection and pricing decisions.
Boiling this down, what’s critical is that you align a data analytics initiative to a strategic business priority. Once you do that, it will be far easier to garner the time and attention required across the organization. Remember, incorporating predictive analytics isn’t just about technology. Success is heavily dependent on people and process.
Make sure your first steps are doable and measurable; you can’t change an entire organization or even one department overnight. Define a small pilot project, test and learn and create early wins to gain momentum by involving all the relevant stakeholders along the way and find internal champions to share your progress.
Recognize that whether you are building a data analytics solution internally, hiring a solution provider or doing some of both, there are substantial costs involved. Having objective criteria to evaluate your options will help you make the right decisions and arm you with the necessary data to justify the investment down the road.