By: Eric Truntz, Senior Principal
It’s a challenge as old as the most primitive and rudimentary markets. How do you create optimal products that fit the needs of active buyers and how do you bring them together? As with many industries, insurance and financial services markets are further complicated because customer-product alignment changes over time while the pivotal life events that trigger active buying follow an uncertain progression. Savvy carriers have successfully met this challenge by watching the competition’s product offerings closely while rationalizing consumer behavior as possible. With those insights, carriers place their bets and mimic what seems to be working for others. While product differentiators can and do exist, they are often small, non-transformative variations. For the most part the playing field was leveled by everyone having access to the same information.
This situation has most certainly changed.
For decades, insurance carriers have invested in systems capabilities to improve service, increase efficiency and reduce operating costs. An extremely valuable by-product of these investments is the customer, activity and product information captured by these systems at each interaction. Within this data hides insights into how to identify the highest value customers, what makes them so, what they buy, when and why. The same can be found for the least valuable and most expensive customers and products. The availability of technologies to integrate, explore and interpret that data into models of experiences for customers and products represents a fundamental change in how future differentiators will be produced. What’s more, since this data comes mostly from within an organization, it isn’t available to the competition. A carrier with the capability to recognize patterns in data and shape products and interactions to capitalize on those patterns has a distinct advantage. For now, the playing field is no longer level.
It likely comes as no surprise that digital strategies around predictive analytics and machine learning have earned “table stakes” status among the business strategies at most carriers. While the tools and techniques used to reveal such high value insights from data are highly specialized, the biggest barrier for most carriers today is assembling and accessing high-quality data for analysis. The most successful algorithms and highly tuned models are a direct reflection of the high-quality data on which they were built.
What may come as a surprise is the position that the key to using data to inform and transform insurance offers isn’t solely in the purchase of an analytics platform and hiring top data science talent. The key is in establishing low-cost, well governed data source(s) to feed the customer, product and behavior data you’ve gathered to the analytics platforms you select. Many analytics strategies have been strangled by the task of navigating the myriad of legacy models and data issues directly. Establishing an analytics and machine learning data repository(s) is a crucial prerequisite. The judicious application of big data technologies in contemporary data architectures enables the ingestion and curation of legacy data sets while minimizing impacts to legacy systems. Such curated data sources and supporting practices can be built incrementally; the most successful approaches often start with a single business theory, and an inventory of the business information required to prove or disprove that theory.
Information follows from implementing the architecture and governance practices to ensure consistency and quality of the data representing the business. From there, when provided with dependable data, the scientists and their toolsets do the rest with shocking efficiency.