Insight

Prepare and prevent, don’t repair and repent: Churn management with data-based solutions

Published April 10, 2023

  • Data & AI
  • Insurance

The competitive and cost pressures in the insurance industry are continuously on the rise. This compels companies to focus not only on acquiring new customers but also on retaining existing ones. This must be done more efficiently than ever before. But how can insurance companies determine which customers are worth retaining? What specific measures can be implemented? Advances in data analysis and artificial intelligence are opening up new possibilities for churn management, enabling companies to address and continually enhance customer retention based on data. 

Traditional vs. data-based churn management

First of all, it is crucial to understand the distinction between traditional and data-based churn management. Conventional churn management operates with relatively stable and internally homogeneous customer categories, typically categorized in a four-field matrix and managed as part of retention strategies. In contrast, data-based churn management allows for a detailed analysis and individual assessment of each customer. Leveraging advanced data analysis methods and technologies, churn scores can be determined for each individual customer, providing insights into their individual churn risk and the most influential factors. This enables customer advisors to intervene in a targeted and effective manner, tailoring approaches to individual needs rather than relying on generic measures. Data-based churn management introduces new dimensions of customer loyalty, representing a significant step forward compared to traditional methods.

Multiple data sources for effective churn management

To achieve effective churn management, it is essential to leverage data from diverse sources, incorporating both internal and external data. Internal data, encompassing policy and claims records, provides insights into contract-related details and reported claims. Pre-contractual information, such as requested coverage and risk data, also proves valuable. Additionally, external sources like PSD2 data provide insights into banking transactions and purchasing behavior. Environmental data, relevant environmental events and external data from third parties also play an important role. The combination of internal and external data allows for a comprehensive view of the customer, covering both their interactions with the company and external activities like exploring alternatives. This approach opens avenues for creating well-founded and individual customer profiles, crucial for effectively identifying and addressing churn risks. Relying solely on internal data for analysis often results in incomplete insights, overlooking potential churn risks.

Figure: Various data spheres provide a well-founded customer churn risk (churn score) and offer automated and data-supported recommendations for action

Explanation of the data types:

  • Policy data: Contract-related data, including terms and coverages
  • Claims data: Data related to reported claims such as claim amount and processing time
  • Inquiry data: Pre-contractual data covering details like requested cover and risk data
  • PSD2 data: Banking data that complies with the PSD2 directive, including purchase data
  • Environmental events: Environmental events relevant to the customer’s needs, for example, flooding
  • Aggregator data: External data from third-party sources, often obtained from data marketplaces

Emphasis on utilizing external data sources

In addition to your own data warehouse, leveraging external data sources is becoming increasingly important in customer churn management. These sources enable a broader perspective of the customer and their relationship with the company. A customer-based analysis requires the consideration of all platforms and interaction channels through which the customer engages with the company or its competitors. Simultaneously, it is equally crucial to conduct a product-based analysis that encompasses not only the company’s own products that might prompt termination, but also alternative products offered by competitors.

 

Valuable insights from contextual interaction data

Harnessing contextual interaction data marks a pivotal step in refining customer churn prediction within customer churn management. This data enriches the comprehension of temporal patterns by supplying insights into the frequency and reasoning behind customer interactions. It is widely acknowledged that the nature of customer interactions has a significant influence on the assessment of churn risk, for example in instances of complaints. Beyond analyzing the type of interaction, data-driven processes empower us to derive emotions and sentiments from conversations and correspondence. This enables companies to incorporate this dimension into the assessment of termination risk, contributing to a more accurate and holistic forecast.

Leveraging unpleasant touchpoints

In churn management, the emphasis often lies on evaluating positive customer interactions, such as inquiries and additional coverage requests, while neglecting the often unavoidable experience of filing claims in the insurance industry. By introducing data-driven churn management, companies can quantitatively assess the effectiveness of relevant touchpoints and interventions, taking into account not only premiums but also the actual payouts and, with a certain probability, expected payouts in the event of a claim. However, poor claims handling or a lack of support during these interactions can also have an impact on the individual churn risk. Therefore, incorporating this form of customer interaction into the data model is crucial for a personalized perspective, enabling individual churn management under the responsibility of the sales department rather than a compartmentalized approach. Churn management, as a result, becomes an invaluable and integral part of proactively shaping customer relationships.

From trigger event to sales intervention

Analyzing and predicting trigger events throughout the customer relationship is another aspect that enhances data-driven customer churn management. Events such as completing education or relocating can significantly impact customer behavior, serving as early indicators of potential churn. The integration of PSD2 data, provides insights into transactions and financial activities, further reinforces these predictions. For instance, a direct debit at a bridal store may signal changing life circumstances that could prompt a shift in insurance needs. Leveraging this information not only enables more precise churn forecasts but also contributes to the data-driven calculation of customer lifetime value. It is essential to adhere to all regulatory requirements, including data protection regulations. These are not categorically in conflict with the use of data-driven approaches but rather safeguards the integrity and security of all parties involved.

Sales opportunities

Sales constitute one of the major cost factors in insurance companies, emphasizing the importance of efficient resource utilization. Data-based churn management is a useful tool here. By predicting customer churn, companies can implement targeted and personalized interventions with their customers that go beyond traditional “spray and pray” or relying solely on the agent’s intuition. Cost-effective post-intervention analyses are also feasible, allowing for a continuous enhancement of efficiency in after-sales.

What stands in the way of data integration?

Integrating external data sources to enhance data-driven churn management poses several challenges for companies. One significant obstacle is the lack of specialized expertise and technical skills to effectively combine and analyze disparate data. The heterogeneous nature of this data with its often diverse formats further complicates the integration process, especially in organizations with an already complex data infrastructure. Overcoming these challenges requires a strategic commitment from top management to foster the necessary capabilities. Additionally, engaging external expertise can prove helpful. Moreover, centralized control and enterprise-wide data governance are crucial to ensure seamless and efficient data integration. In a fragmented data landscape, responsibilities are unclear or those involved are not in continuous communication.

Data fuels hyper-personalization

The future success of insurers hinges on an individual, data-driven approach to customer churn management. Customers demand personalized solutions and prompt service. The challenge for insurers is to integrate internal and external data sources to create a holistic and personalized customer profile. This is crucial as customer churn can manifest through various interactions and activities. A comprehensive database is therefore essential for success. Harnessing innovative technologies from data science and machine learning enables targeted interventions and effective success monitoring. In essence, data-driven churn management is emerging as the cornerstone of long-term insurer success.

Authors

  • Dennis Glüsenkamp

    Lead Data Strategist – Germany, Frankfurt

    qdive

    LinkedIn
  • Simon Rodler

    Senior Consultant – Germany, Munich

    Wavestone

    LinkedIn