It would appear that the insurance industry is well aware of the benefits offered by technologies such as artificial intelligence (AI) and machine learning (ML). A recent survey by Deloitte revealed that 98% of UK insurance executives believe cognitive computing will play a disruptive role in the industry in the years to come.
Meanwhile, according to a recent study by McKinsey, the use of such cutting-edge technology is likely to become mainstream across all industries and spheres of business. However, the insurance industry will likely be one of the key beneficiaries of the early adoption of AI and ML when it comes to the actual business value.
This is primarily due to the volume of data that the industry has access to, as one of the essential elements of creating predictive AI models is the availability of large datasets. A well-trained AI algorithm is an extremely powerful tool for any insurance organisation and it can be applied across a wide range of insurance products. It helps organisations to better understand everything from the possibility and probability of claims, to assist customers in choosing the right coverage options to fit their needs.
Yet how big will the impact be? A further report from McKinsey recently reported that AI and related technologies will have a “seismic impact” on all aspects of the insurance industry, from distribution, underwriting and pricing, claims, and more.
However, to really take advantage of the power of AI and ML and apply these innovations to day-to-day operations, organisations within the insurance sector have to address several key challenges.
To start with, the insurance industry is highly regulated, meaning that customers’ personally identifiable information (PII) must be protected. This is a major reason as to why organisations in the sector may be behind other industries when it comes to data collaboration, as a huge amount of time is necessary to clean this information before it can be shared, and also to make sure data compliance regulations are met. Due to this heavy burden, AI or data-related projects can take much longer to implement and deliver value than initially expected, compared to other industries.
Secondly, despite being able to accumulate and manage terabytes and terabytes of data, insurance organisations face some limitations with the information they have, as the actual dataset itself may be under-representative. The most valuable data for an organisation can often be hidden in an under-representative customer segment, making any model trained on this biased dataset return biased results. In fact, data quality limitations are the most cited major barriers that prevent organisations from utilising their data assets. Even more concerning is that according to Forrester, over 70% of data actually goes unused for analytics by companies.
The long term effects of such data challenges are potentially harmful to insurance organisations. This is because cognitive models that create false assumptions about customer segmentation lead to higher costs for new customers acquisition (insurers already spend over £600 each year per client), inappropriate offers being made to customers, which ultimately makes them less likely to purchase, or worse.
However, these challenges can be solved by a more sophisticated approach to the development of AI and ML models. Opportunities for organisations to generate synthetic data have emerged as a highly credible method of protecting PII, while also eliminating the limitations that the insurance industry is facing with under-representative datasets. Sophisticated AI and ML technologies are now able to generate new, entirely ‘synthetic’ datasets from the original information that are highly statistically accurate (up to 95%) but importantly, do not reveal customers’ PII.
This approach to data could transform the insurance industry, with financial services organisations like JP Morgan already exploring its potential. Its biggest impact could come in the ability to personalise insurance products to customers.
When it comes to understanding the customer, AI has enormous value for the insurance industry as it can unlock new revenue streams through previously unknown or unavailable insights. This is crucial for the industry’s future as it currently faces huge disruption due to a host of factors, from self-driving cars to the sharing economy, all of which will challenge the traditional profit models for insurers. However, AI and ML technologies are built specifically to extract insights from data which encapsulates consumers’ preferences, interaction, behaviour, lifestyle details and interests. Not only this, but the technology is also developing to such a stage that it can analyse datasets for bias and ‘rebalance’ them to better serve every customer segment.
When synthetic data is accurately implemented, research has found that it can give the same results as real data. Crucially for this highly regulated and competitive industry, the key benefits include full data privacy compliance and a major reduction in the time needed for product development and testing.
Using this approach, insurance companies can offer highly relevant and timely products to their customers, boosting both customer loyalty and profit. This matters hugely as EY found that more than half of insurance consumers (57%) across the world only want to hear from their insurance provider twice a year. This puts a huge premium on the quality and accuracy of outbound communications to customers, as any misstep may have a detrimental impact on an insurance company.
By having an insights-led approach to customers, thanks to AI and ML models, the industry could unlock an entire ecosystem of opportunities to become, in effect, a one-stop-shop for their customers’ needs.
For example, in the area of car insurance, when providing accident coverage, an insurer can create an ecosystem that offers roadside assistance, devices and apps that monitor and reward safe driving habits, as well as providing customised offers for repairs and maintenance. Research has found that more than 75% of insurance customers globally are open to such an ecosystem from their insurers.
While it is clear that more and more insurance organisations are waking up to the benefits of using AI and ML technologies, any approach to maximise data’s value must have a deeply coherent strategy underpinning it. Used in the correct way and with a well thought out strategy in place, the opportunity from these technologies offers limitless potential to the insurance industry.