The new age insurance: Personalisation with AI & data analytics applications

Personalisation: The key to the digital future of insurance

Today’s consumers are open to giving insurers personal data on their behaviour and habits, but only if they benefit from personalised offers and attractive discounts. Shifting expectations call for insurers to issue policies and finetune pricing in a more dynamic, granular way.

Driven by AI and data analytics, two key trends are taking shape:

1. Usage-based insurance aligns customer behaviours with the premiums they pay. Dubbed "pay-as-you-go", this model first gained traction in the auto industry, enabling drivers to pay based on miles clocked. 

But with the rise of telematics, insurers today can tap into data not only on drivers’ mileage but also their travel routes and driving habits. As AI tools receive real-time data from IoT sensors, they can dynamically adjust premiums to, for instance, reward cautious driving behaviours – unlocking a new frontier of personalisation.

2. Customised claims handling draws on automation to streamline the once-lengthy claims process.  

Kristin Warne, Head of Life & Health Claims and Business Management Client Services at Swiss Re, shares that automation leads to reduced administrative burdens and a better experience for both customers and employees. 

“In today’s era of digitisation, customers still deal with paperwork for life and health claims. Now, customers expect to be able to lodge claims digitally and have a personalised experience,” Kristin says. 

“By tapping into internal data sources, customers won’t need to re-enter relevant information in a digital claims form. Questions can also be more tailored to the customer's product and circumstances. Coupled with technology that validates information from other sources, claims assessment becomes more transparent and faster for less complex claims. Companies can then spend an appropriate amount of time engaging with customers for more complex claims.”

In fact, predictive analytics can potentially move insurance toward a proactive rather than reactive approach to manage risk – for example, by pre-emptively alerting customers of potential faults in their home. 

In the event of actual property damage, for instance, smart sensors can trigger claims seamlessly and transmit all necessary data for advanced algorithms to validate claims.

As customer habits shift post-pandemic, insurers who offer AI-driven personalisation will be best placed for success. A 2022 survey by TransUnion and J.D. Power found that consumers are driving less post-pandemic, yet paying the same or more in traditional coverage. This drove 49% of consumers to opt-in for a telematics-based programme. 

The challenge for insurers lies in creating the right incentives for consumers to willingly allow collection of their personal data. To achieve this, insurers must add value by enhancing the customer experience. 

Common pitfalls in deploying advanced analytics and AI

As insurers dive into the uncharted waters of AI, implementation challenges are par for the course. According to Kaushal Silva, Client Business Solutions APAC at Zühlke, insurers face three main pitfalls today:

1. IT/Business misalignment 

AI transformation calls for a company-wide strategic shift, but many insurers today lack the right mindset to move toward a data-driven culture.

“Insufficient end-to-end collaboration between business and IT teams can create a poorly thought-out business case,” Kaushal explains.

“IT may be pushing AI adoption, but their solutions are not aligned with the business team who are closer to the clients and know what they want. On the flip side, the business team may have an amazing idea, but IT is unable to bring out its business value. Companies will need to align between business and IT.

2. Poor data quality and management 

AI solutions are only as good as the data you feed them. Many insurers struggle to unlock the value of their data effectively. Only 22% of insurers say they have at least a consistently accessible set of core data to build models, and up to 75% lack a taxonomy to harmonise diverse types of data.

Incomplete or poorly consolidated customer data threatens insurers’ ability to detect claims fraud. Without well-organised data on customer habits and behaviours, insurers will also face difficulties in providing accurately personalised services.

Effective data flow and governance across organisations and those they work with is essential for shared decision making and efficiency gains. Learn more about data ecosystems and how they can solve your data quality problems.

3. Lack of maintenance 

AI transformation does not stop at deployment – it involves continually evaluating and retraining machine learning models to account for changing customer data. One key mistake that insurers make is launching a great product offering – then neglecting to maintain their ML models after.

Constantly iterating can be a challenge, so Kaushal encourages insurers to plug into ecosystems to get access to entire networks for technical expertise.

“Your best solution for customers today will be irrelevant in a couple of weeks,” says Kaushal. “The whole point of customer centricity is constantly adapting to customer needs. Companies need to see AI not as a one-time investment but an investment through the cycle.”