Success Metrics For Your Data Program

Summary

In the conversation with Rachel Alt-Simmons from AXA XL, the focus was on the development and evaluation of success metrics for data programs. The discussion brought to light the complexities and strategies in evolving a large insurance company into an organization driven by data. AXA XL's unique difficulties arise from managing complex, custom risks with small yet intricate datasets. Rachel highlighted the significance of fostering a data and analytics culture, which includes aiding underwriters with decision tools rather than replacing them. The conversation also included the use of external data to bridge internal data gaps and the application of technologies like natural language processing to glean valuable insights from document-heavy processes. Besides, Rachel shed light on the significance of developing a communication strategy to encourage data-driven discussions across the organization, and how AXA XL is using platforms like DataCamp to improve data literacy and skills among employees.

Key Takeaways:

• Success metrics for data programs differ across organizational roles and initiatives.
• AXA XL encounters unique difficulties with small, complex datasets in the insurance industry.
• Natural language processing is a vital tool in gleaning insights from document-heavy processes.
• Building a data-driven culture involves aiding existing expertise with decision tools.
• Communication strategies are vital for encouraging a data-driven environment.

Deep Dives

Success Metrics for Data Programs

Success metrics within an organization can differ significantly depending on the role and objectives of the data program. Rachel Alt-Simmons emphasized that while some areas might focus on quantitative metrics like return on investment (ROI), others might prioritize qualitative measures such as user adoption and the decommissioning of legacy systems. In AXA XL's context, success is also measured by the ability to align data initiatives with key business goals, such as profit maximization and risk reduction. Rachel pointed out that the metrics also depend on whether the data team is engaging in experimental work to build skills and explore new technologies or directly addressing business problems. "Success is about providing client value and improving profitability," she stated, highlighting the necessity of tying data science work to business returns.

Challenges of Data Transformation in Insurance

Evolving an insurance company into a data-driven organization presents unique challenges, particularly when dealing with complex, custom risks. AXA XL's data is not large in volume but highly complex, requiring the capture of numerous exposure elements unique to each risk. Rachel explained that their focus is not on automating roles but on enhancing underwriters' decision-making capabilities with data-driven tools. The insurance industry's competitive nature limits the amount of data that can be requested from clients, thus necessitating the use of external data sources to supplement internal datasets. Rachel noted, "The challenge is making sure that whatever data we use can be trusted in the decision process."

Natural Language Processing and Data Extraction

The use of natural language processing (NLP) is a key tool for AXA XL in managing document-heavy processes. Insurance processes are traditionally reliant on paper documents, making data extraction challenging. Rachel discussed the use of NLP to automate the extraction of relevant data from documents, which can often come in various formats and lack standardization. This technology allows AXA XL to extract valuable insights locked within text-heavy documents, facilitating better decision-making and risk assessment. "NLP helps us uncover what we're leaving behind in documents," Rachel remarked, highlighting the potential of NLP to transform data accessibility and usability.

Building a Data-Driven Culture

Building a data-driven culture at AXA XL involves more than just technical advancements; it requires encouraging a mindset shift across the organization. Rachel has been leading initiatives like Analytic DNA to develop data literacy and skills among employees. A key component of this transformation is communication—sharing data stories and successes to encourage collaboration and innovation. Rachel mentioned the importance of identifying early adopters and champions within the organization to drive these changes. Additionally, she highlighted the role of platforms like DataCamp in providing the necessary training and resources to support this cultural shift. "Everyone has a responsibility to be more data-driven," she emphasized, indicating the collective effort required to achieve this vision.