Increasing Importance of Data to insurance industry

Increasing Importance of Data to the insurance industry

Insurance has always been a data-driven industry. However, the bar for gaining a competitive advantage using data as an asset keeps getting raised. Following are some industry trends that have made data and analytics ever more important to the insurance industry:

  • Explosion of third-party data availability from vendors and industry consortiums
  • The exponential growth of structured and unstructured data due to digitalization, social media usage, connected devices, geospatial data, call recordings, etc.
  • Technologies such as cloud, analytics, AI/ML are making it easier to monetize data assets


We strongly believe carriers must look at how to enhance their data management and analytics capabilities to achieve the following business goals.

 

data management



Following are some illustrative results achieved by a leading life & annuity insurer who enhanced business insights and decision-making capability through data modernization.

  • $4 Million savings through favorable reinsurance outcome
  • $2 Million incremental earning by optimizing product mix and market positioning
  • $1.7 Million margin preserved through enhanced policyholder retention strategy
     

Use Cases

The following are just a few selected use cases out of many that illustrate how data and analytics can play a key role in improving business performance of insurers.

Accelerated underwriting:

Various surveys have shown that medical tests and slow underwriting are major factors for poor customer experience while buying life insurance. Use of real-time third-party data and advanced analytics enable insurers to quickly underwrite cases without requiring medical tests.

Customer experience:

It is increasingly clear that customers expect insurers to ‘know’ the customer and deliver a personalized experience. Predictive analytics and machine learning models using internal and external data help insurers understand the customers better, provide targeted experience, improve customer retention, and increase customer lifetime value. 

Fraud mitigation:

According to surveys, the top 3 areas where insurers have experienced fraud include policy application, claims, and agent fraud. Use of embedded, real-time analytics at the data intake stage, predictive analytics for risk scoring, voice and facial analytics can be used for mitigating impact from fraud. 

Improving persistency:

Persistency rate plays a major role in profitability. Some products become profitable in 2nd or 3rd year. With continued low-interest-rate environment applying stress on the balance sheet, there is renewed focus among insurers on improving persistency. This applies to individual as well as group insurance business. Advanced analytics and machine learning models are used to forecast lapses and develop targeted retention strategies.

Financial wellness:

Plan sponsors continue to be concerned about their employee's financial wellness and retirement readiness. The retirement industry has taken several steps to address this issue, such as automatic enrollment, auto-increment of contribution, managed services. However, participant behavior plays a major factor in the success of these measures. Retirement plan providers have found behavioral analytics as a key tool to determine offerings that increase the rate of success.

Product portfolio management:

A protracted low-interest market environment has put significant strain on the insurance industry. While it is a problem in general impacting investment returns, some products that offer long term guarantees put additional strain on the balance sheet. Exposure analysis, predictive analytics with the simulation of future economic trends can help insurers fine-tune their product strategy, closed book strategy, etc.

How do we build an effective data management strategy?

Insurers can follow this 6 step approach to build an effective data management strategy.

  1. Start with business objectives. Without having a clear understanding of business objectives, you could waste valuable time and resources collecting and analyzing the wrong data. We recommend focusing on three or four use cases to start with and build it further based on success and learnings. The use cases must be associated with compelling and material financial impact (ROI)

  2. Identify business stakeholders. This should include internal (e.g. marketing, sales & distribution, customer service, finance) and external (e.g. agent, consumer, employer/plan sponsor). Create a persona for each stakeholder.

  3. Identify what key decisions need to be made, and what insights and information required for decision making to meet the business objectives.

  4. Establish robust data processes. Once we know what data is required and how we will use it, we need to establish a process for collecting, preparing, cleansing, storing and distributing the data. We need to answer questions such as:

  • What data sources to use; do we need internal and external data?

  • Do we need both structured or unstructured data or both?

  • How do we ensure completeness and accuracy?

  • How do we store data, and how do we keep it secure

5. Identify the architecture and technology required to deliver the value proposition. Several components are required to be architected together:

  • Data ingestion tools
  • Data lake
  • Data quality
  • Data security & privacy
  • Data science tools – exploration, analytics, visualization, streaming analytics
  • Data governance tools
  • Metadata management
  • Infrastructure, cloud adoption
     

6. Establish data governance which includes

  • Data collection and storage policies, processes, data stewardship
  • Data quality, profiling, lineage
  • Data security and privacy, regulatory compliance