By Jon Jacobson, CEO and co-founder of Omnisient
Transactional retail data – such as grocery shopping history – that is used as alternative data provides deeper insights into consumer behaviour and lifestyle choices. For insurers, this data can be incredibly predictive of risk. For instance, grocery shopping data reveals a person’s dietary habits, which can be indicative of their overall health. Regular purchases of fruits and vegetables suggest a health-conscious individual, potentially correlating with lower health risks and, therefore, lower insurance premiums. Conversely, frequent purchases of processed foods and junk food could indicate a higher risk for conditions such as diabetes or heart disease.
In the financial sector, grocery shopping data can offer insights into an individual’s financial stability and responsibility. Consistent spending on premium products might suggest financial well-being, whereas a sudden shift to lower-cost alternatives could signal financial distress. These indicators are invaluable for refining credit risk models, improving the accuracy of lending decisions, and managing loan portfolios.
And this is not a pipe dream – working with Africa’s largest grocery retailer and Africa’s largest banks, we’ve been able to achieve a 41% GINI lift on scoring of thin-file clients using their shopping data. This has meant that 3.2 million individuals now qualify for credit who would have otherwise been declined due to lack of credit history.
PPDC enables multiple parties – such as insurers, banks, and retailers – to securely share and analyse data without data ever exchanging hands or revealing individuals’ identities. This is achieved through privacy-enhancing technologies (PETs), such as differential privacy, anonymisation, and secure multi-party computation (SMPC).
Applications of PPDC in actuarial science
1. Refined risk assessment models: Integrating retail data such as grocery shopping habits into traditional actuarial models, allows insurers to develop more nuanced risk profiles. For example, an actuary could create a ‘healthy eating score’ based on the nutritional quality of a policyholder’s grocery purchases. This score could then be used to adjust health insurance premiums more accurately, rewarding individuals who maintain a healthy diet with lower premiums.
2. Improved predictive accuracy: PPDC allows for the secure development and testing of predictive models using a wider range of data sources. For instance, actuaries can build models that predict health risks, financial stability, or even the likelihood of a policy lapse by analysing patterns in retail data.
3. Enhanced customer segmentation: The ability to securely analyse alternative data sources enables actuaries to segment customers more effectively. For example, grocery shopping data can be used to identify policyholders who exhibit unhealthy lifestyle habits, who could then be offered tailored health and wellness programmes to motivate and facilitate healthy behaviour that can lower risk.
4. Proactive risk management: PPDC enables actuaries to identify early warning signs of risk, such as changes in financial behaviour that might indicate an impending premium lapse. By detecting these patterns early, insurers can intervene proactively – offering payment plans, premium relief, or other support to prevent a lapse. This not only helps maintain the customer relationship, but also reduces the likelihood of costly lapses and cancellations.
Overcoming challenges and ensuring compliance
While PPDC offers significant advantages, there are challenges to its implementation. Actuaries must work closely with data scientists, information security and compliance professionals, and legal teams to ensure that PPDC initiatives are secure, efficient and compliant with all relevant regulations.
Moreover, as regulatory frameworks evolve to address the use of alternative data, insurers must stay ahead of the curve by developing and adhering to best practices in data privacy and security. This includes conducting regular audits, implementing robust data governance frameworks and maintaining transparency with customers about how their data is being used.
As the insurance industry continues to evolve, the ability to integrate and analyse diverse data sources securely will be a key differentiator for insurers seeking to lead in innovation and customer satisfaction. The future of actuarial science lies in the intersection of advanced data analytics and privacy-preserving technologies, offering unparalleled opportunities to refine risk assessment and drive better business outcomes.
|
