Is it controversial to say that insurance has always been data-driven? You can’t price and underwrite a policy or manage claims for customers without having data about the risk, whatever that risk might be. Actuaries are certainly not new, and data is fundamental to their work. It’s probably less controversial to say, though, that insurers have often not been great at managing their data and keeping pace with broader social and technological changes: what was good enough 10 years ago is not good enough now. What was good enough last year might not be good enough now.
WHAT IT MEANS TO BE DATA-DRIVEN
For me, ‘data-driven’ means embracing the fact that data is the real enabler of insurance. It is the blood that flows through the body of an insurance company and brings things to life. And when the circulatory system can’t keep up with the demands of the company and its customers, performance is impacted across every single aspect of the company’s operation. Practically, this means that being data-driven is about switching from the position where data quality, availability and capabilities are the limiting factors to one where they are the fuel for serving customers and for growth.
A data-driven approach opens up the opportunity to perform complex, large-scale and real-time analysis without the need for lots of manual data handling. It means having confidence in your data quality and a very clear understanding of what your data means and how it can be used do drive customer and business value both now and in the future.
THE BENEFITS OF BEING DATA-DRIVEN
The benefits of such an approach will be felt across the whole organisation. Having confidence in your data and knowing that you have a solid foundation on which to build is liberating. Increased visibility of business performance, increased sales, improved pricing accuracy, improved claims control all the way through to strong regulatory and capital management positions are enabled – all without the need for numerous manual processing steps.
With solid data foundations in place, once cutting-edge data-centric tools and techniques are rapidly becoming business as usual, for example:
- Telematics insurance products driven by real-time mobile data
- AI-based fraud monitoring that knows who you associate with and scores your social media interactions
- Claims damage assessments made quickly from photos and real-time accident data
- Machine-Learning-based pricing driven by significant volumes of internal and external data
- The ability to predict flood risks within a postcode and warn when things might get wet
- The ability to identify the young driver speeding in front of a school in real-time
All these examples are data-driven use cases, and all of them require insurers to deploy both new technologies and new ways of thinking. Whilst it can be easy to focus on things like Artificial Intelligence (AI) and Machine Learning (ML), they also absolutely rely on the foundation of great data quality.
HOW TO BECOME DATA-DRIVEN
Whilst AI and ML are appealing, there are often significant steps required to get the organisational data to the level where this becomes effective, with many opportunities that come before implementing these technologies.
Step 1: Understand your ‘as is’
Getting a clear understanding of your business challenges and opportunities is key. Reflect on the data exploitation steps that happen within your organisation and get clear on how you collect, store, govern, share and use your data. There will be many opportunities to improve things: siloed data, duplications, inaccuracies, incomplete data sets, lack of data definitions, lack of clear data ownership, error handling, over-reliance on manual processing and reconciliations as well as both organisational and process-related problems may all figure into your analysis.
Step 2: Build a common vision
With a strong focus on both the long-term and the immediate problems to solve, build a common vision of the future. Business buy-in across the organisation with a strong benefits case behind it will become paramount for your data strategy. Prioritisation is a key part of this. It is also important to remember that the emerging data strategy needs to be part of the wider business and technology strategy; they must work together.
Step 3: Turn your vision into a data strategy
With a clear view of the problems that need to be solved, a long-term direction and business buy-in, it is time to focus on delivery. A data strategy is the combination of your vision with a delivery plan to get there. That plan very much needs to include the people, skills and funding needed to implement it. Technology change is a key part here, and so are changes in culture, ways of working and organisational design that might be needed to make your strategy work.
Step 4: Execute your plan, review and re-plan!
Breaking down the work into bite-size chunks is crucial to mobilising benefits quickly and building both belief and long-term momentum. You should also ensure that an agile mindset and the willingness to adapt and re-plan are embedded as part of your approach, so that you are ready to respond to ever-changing priorities and external influences, like new regulatory requirements, deal opportunities and changing partner demands. As improved data exploitation fosters better control, visibility and understanding, new opportunities will appear!
CHALLENGES TO BECOMING DATA DRIVEN
It’s easy to set out a step-by-step process, but the reality isn’t quite so simple. Expect to cycle through and between the steps many times. From oversubscribed change portfolios and competition from other strategic initiatives, complex and meandering legacy architectures born of acquisitions and compromise, people resisting change to tricky technology problems and choices, all of these can impede the perfectly laid-out plan. Furthermore, the challenge (and the cost) can seem very daunting.
Whilst the long-term view is important, in my experience the best approach is to build momentum through a series of smaller wins. Get started, build collaboration across the business and demonstrate the business case for further investment by delivering tangible benefits.
The topic of data-driven insurance is both so broad and deep. There are so many more angles to consider of this short introduction. However, I did save one until the end because it is massively important in driving delivery – and that’s the excitement, energy and sense of achievement that ‘data-driven’ enables!