Data analytics in selling insurance
Insurers can leverage data analytics, “particularly predictive analytics”, to optimise growth in several areas, according to Kim Global Consultancy founder and director Kim Yan Lim, when she spoke with Asia Insurance Review.
Optimising growth
Ms Lim highlighted some areas where insurers could leverage predictive analysis to optimise growth and efficiency:
• Risk assessment and pricing:
Historical data and machine learning (ML) models can be used to predict risk levels for individual customers or segments, allowing for the development of dynamic pricing models that adjust premiums based on real-time risk factors.
• Customer segmentation and targeting: Analysing customer data to identify high-value segments and tailor products and services to needs, as well as predict lifetime value to focus acquisition efforts on profitable prospects.
• Cross-selling and upselling: Efforts can be improved by identifying patterns in customer behaviour to predict which additional products they are likely to need and using propensity models to determine the best timing and approach.
• Fraud detection: It can be strengthened through the implementation of anomaly detection algorithms to flag potentially fraudulent claims and the use of network analysis to uncover organised rings.
• Customer retention: Strategies can be refined by developing churn prediction models to identify at-risk customers and creating personalised strategies based on individual preferences and behaviours.
• Claims management: It can be optimised by predicting claim severity and complexity to better allocate resources, and by using text analytics on claim descriptions to fast-track simple claims and identify potential fraud.
Success stories
There are several companies that have reported success in using data analytics to drive growth and efficiency, Ms Lim said. She said, “For example, a major insurer in China implemented a big data analytics platform to improve customer segmentation and personalised marketing. This initiative reportedly led to a 15% increase in sales conversion rates and a 10% boost in customer retention within the first year of implementation.”
On the other hand, she said that “a multi-line insurer in India” leveraged predictive analytics for cross-selling and upselling. According to Ms Lim, the insurer “reported a 12% increase in premium per customer and a 5% overall growth in annual premium revenue within 18 months of implementation”.
She also touched on AI, saying that a “large insurance group in Japan” which adopted an AI-powered claims processing system realised cost savings.
She said, “This system reduced the average claims processing time by 30% and decreased associated operational costs by approximately 20%.”
The regulatory landscape
The regulatory landscape “The regulatory landscape for data analytics in the insurance sector across Asia is evolving, with varying degrees of support and caution depending on the specific country,” Ms Lim said. She also noted that
landscape is trending towards supportive regulation, though with careful consideration of consumer protection and data privacy.
For instance, she pointed out that the Monetary Authority of Singapore (MAS) has been “proactive in encouraging innovation while balancing it with consumer protection”.
“[The MAS] introduced guidelines on the ethical use of AI and data analytics in the financial sector, providing a clear framework for insurers to follow,” she said, also noting that the regulatory sandbox approach allows companies to test new data-driven solutions in a controlled environment.
Meanwhile, Hong Kong’s Insurance Authority implemented a Fast Track pilot scheme for digital insurers in 2017. The Hong Kong Monetary Authority also published guidelines regarding the use of big data analytics and AI in 2019.
The human touch
While data analytics can be deployed to handle routine underwriting tasks, Ms Lim warned that human expertise was still vital for assessing complex or unusual risks.
She said, “Experienced underwriters can interpret nuanced information, understand unique business contexts and make informed decisions on risks that fall outside standard parameters.
“This is particularly important in commercial insurance, specialised coverage, or high-value policies where risks are multifaceted and require detailed analysis,” she said.
For all that data analytics can “provide insights into customer behaviour and preferences”, she also said that human interactions are also essential for building and maintaining strong customer relationships.
For instance, she noted that agents and customer service representatives are able to offer personalised advice,
explain complex policy details and provide empathetic support during claims processes.
Ms Lim said, “This human touch is essential for customer retention, especially in handling sensitive situations or addressing unique customer needs that automated systems might not fully comprehend.”
Lastly, she believes that although data analytics can flag potential fraud and streamline routine claims, human investigators are essential for complex claim situations.
“[Humans] can conduct in-person assessments, interview witnesses and use their judgement to evaluate ambiguous circumstances.
“In cases of dispute or where claims involve intricate legal or medical issues, human expertise is irreplaceable for
fair and accurate resolution,” she said.
Bridging the skills gap for data analytics adoption
When asked how insurers could upskill employees in preparation for the adoption of data analytics, Ms Lim first suggested that insurers identify existing skill gaps by conducting a thorough needs assessment.
“Once identified, training efforts should be prioritised based on the urgency and impact of the skill gaps on business objectives,” she said.
From there, she said that “tailored training programmes should be developed to address the identified skill gaps and cater to the specific needs of different roles within the insurance organisation”.
She said, “These programmes can be delivered through a blended learning approach, combining classroom training, online courses and hands-on workshops to accommodate diverse learning styles and preferences.”
Finally, she believes that investment in data analytics tools and platforms will be essential to provide employees with the necessary hands-on experience.
“By providing access to industry standard tools like Python, R, SQL and data visualisation software, insurers can equip employees with the practical skills needed to work with data effectively,” she said.
She also believes insurers should encourage employees to work on real-world data analytics projects, as it would help them apply newly acquired skills and gain valuable practical experience.
The future
Ms Lim said, “One prominent trend will be the increased integration of AI and ML into various aspects of insurance operations.
“This will enable insurers to improve risk assessment, pricing and customer segmentation through more sophisticated predictive models and fraud detection algorithms,” she said.
She also expects natural language processing to continue to gain traction and allow insurers to provide more personalised customer service through chatbots and virtual assistants, as well as automating document processing tasks.
“Finally, data governance and privacy will remain a top priority as insurers navigate evolving regulations and ensure the security and integrity of data,” she said.