Unsere Internetseite verwendet Cookies. Indem Sie weiter auf unserer Website navigieren, stimmen Sie der Verwendung von Cookies zu. Mehr Informationen finden Sie in unserer Datenschutzerklärung.

Perspective

P&C Insurance - Analytics Maturity Evolution

December 13, 2017| Von Upendra Belhe | P/C General Industry | English

It is true that P&C insurance carriers have been embarking on the exciting journey of building predictive models. Many small to mid-size P&C carriers have admitted to the trend and are increasingly building their capabilities to nurture and use predictive models for pricing purposes. Consulting and service companies are coming to the rescue and enjoying increased revenue by helping carriers increase their use of predictive modeling within the actuarial function.

Executive leadership at many carriers, especially small to mid-size, is wondering how they should expand their analytics and what direction to take. In this blog, I present the situation of P&C carriers at various stages of the data science maturity evolution.

Stage 1. Predictive models prove to be valuable

At this first stage, P&C carriers, which built sophisticated models for pricing (based on frequency and severity prediction), have seen tremendous benefits over the last decade. Respectively, there are increased expectations from a top-line/bottom-line perspective. CEO/COOs want to glean the benefits of predictive modeling by expanding analytics programs. They start looking to add talent. Some rely on consulting services, but understand their dependence will only increase. Costs may accrue rapidly to run the existing/newly built models. They assume this challenge is common and faced by many. They think areas such as machine learning are too advanced and technical and have very little relevance to their business. They also think the use of metrics, dashboards and reporting for business purposes has very little to do with analytics.

Stage 2. Expansion of predictive analytics is limited by data and talent

In Stage 2, the actuarial staff is excited about their prior success and is adding actuarial talent as part of the CEO/COO’s analytics expansion efforts. However, they wish more data was available, and perhaps external data sources as well, that could be used to increase their success and surpass their competition. They are encouraged to leverage the power of predictive analytics for claims purposes (e.g., large loss, fraud and subrogation modeling). They may realize the majority of insurance data is unstructured and start focusing on how such data could be leveraged for predictive modeling purposes. Unfortunately, the benefits seen over the last decade have left the impression that they need to get increasingly better in building predictive models; build more and build faster. However, they are realizing that it’s hard to find talent. Dependence on consulting service expertise seems inevitable.

Stage 3. Realizing the worth of existing data assets

At this point in the evolution, carriers may have change agents who are looking at the worth of their data assets beyond predictive models. Leadership begins to doubt their reliance on the power of predictive modeling and if it will bring the necessary competitive advantage that they had initially thought. They start to observe trends in the data that were ignored until now. The idea of building predictive models without validated business insights acquired through exploration and adequate data maturity is a failing proposal. They realize that Data Science maturity is characterized by stages, each stage a necessary building block for the next. The ability to predict evolves with extreme data familiarity and such ability has a low correlation with the impact on the business. The realization that analytics has a much broader meaning is growing rapidly and creating the necessary discomfort. Such opportunities to leverage data are no longer limited to pricing or severity analysis. They are coming from claims, marketing, and distribution.

Stage 4. Focusing on business goals

Once carriers start looking at trends in the data, conference room conversations begin to cast an objective/data driven look at the assumptions that drive business. Awareness about strengths in the existing data is noticeable. BI tools and technologies advancing towards data visualization help to demonstrate the strengths in the available data. There is an increasing trend to ask “so what?” instead of asking “what’s new?” The curiosity, questions and data driven discussions are not being limited by the availability of analytics talent resources. These recent experiences are telling them the data is only valuable when you do something with it - analyze it, visualize it. Very large sets of internal, as well as external, data housed in a structure without data schema concerns, allow leaders to think about potential business value as the new area of focus. The recent technological advancements, such as Data Lakes leveraging the Hadoop platform, are becoming a vital element in their data ecosystem.

Stage 5. Exemplary use of data and analytics is behind business transformation

At stage 5, data visualizations are revealing intricate data structures that cannot be absorbed in any other way. Storytelling with data visualizations displaying numerical evidence is drawing an impactful response from underwriters and others across the company. There is a growing emphasis on the “insights economy” where more and more insights can be drawn from the internal and/or external data which cost marginally less as you gather more within a given business focus area. Data Governance is taken far more seriously at this point in the evolution. Most importantly the oversight is chaired by the business and not IT. The enterprise data dictionary is populated and accessible to all parts of the business to contribute to. The metrics and dashboards used by the executive team to monitor and manage the business are well defined, transparent and follow the organizational hierarchy. Reports are streamlined from a resource consumption perspective and use efficient, non-redundant data sources. These reports are actively used to highlight analytics opportunities by the business segments.

Conclusion:

P&C carriers may follow different paths towards their analytics evolution. The characteristics stated above are not a unique set and only serve as an example of the environment. Presence or absence of one or more characteristics stated within the stages above shouldn’t be interpreted as the carrier’s belonging to that stage. That may only mean that the carrier is on the right path to be identified at that stage, if they persist in their effort to inculcate more of those characteristics. Many P&C carriers are still not there yet, but Data and Analytics has potential to become a cultural component through their conscious effort over the coming decade.

If you are wondering what stage of analytics evolution your company aligns with, or, if you would like to explore some questions you could ask yourself, please reach out to Gen Re’s Data and Analytics team for further discussion.

×

 

Stay Up to Date. Subscribe Today.

Autoren

Lernen Sie unsere Experten kennen

View Contributors