5 key factors in insurance AI success

Register now

Increasingly, insurance companies are leveraging artificial intelligence (AI) and machine learning to optimize processes, reduce costs, and increase efficiency. For example, Ant Financial’s Dingsunbao app is able to make damage assessment and provide detailed analysis including claim amount, damaged parts and repair plan by leveraging AI technologies such as image recognition. Similarly, a public insurer in Germany employs an algorithm to manage the large amount of email correspondence by detecting keywords, sorting correspondence according to topics, urgencies and departments, and suggesting next best actions.

In addition to unlocking greater efficiencies and lowering costs, AI and machine-learning technologies can also be applied to help insurance companies acquire new customers, cross-sell and grow revenues. For example, AI and machine learning can provide insights to support more effective customer segmentation, automate and personalize product recommendations, and enable more intelligent and customized self-service product research for customers.

However, there are a number of factors that insurance companies must first address if they want to fully leverage AI and machine-learning technologies to support sales and revenue growth.

Low Data Quality
To begin, we have to address the issue of poor data quality, a challenge that is not unique to insurance companies. According to an article in the Harvard Business Review, only 3% of companies data meets basic data quality standards. Common quality issues include data that is incomplete, poorly defined, incorrect, out-of-date, irrelevant or simply difficult to interpret. Some concerted effort will be required at the holistic or organizational level to address the issue of poor data quality. Bad data increases costs and leads to bad decisions. The popular term in the early days of computing, “Garbage in, garbage out,” is still very much apropos today.

Customer-Centricity vs. Company Centricity
Another common issue that threatens the success of AI and machine-learning projects in insurance marketing and sales is embarking on projects without first establishing an accurate and deeper understanding of customers’ needs.

During a recent visit to the Apple App store on our iPhone, we found apps offered by several leading insurance companies with very disappointing reviews. This include apps from recognizable insurance brands. Clearly, these apps were developed hastily and not enough attention have been invested to determine if they can be useful to customers. Without focusing on the customer and finding out what they need and desire, AI projects are doomed to failure.

Security Concerns
Hacking and IT incidents are on the rise and insurers must do more to analyze and address vulnerabilities. According to an article in TechRepublic, there has been more than 3,800 data breaches exposing more than 4.1 billion records in the first half of 2019, representing a 54% increase from the same period in 2018. Alarmingly, finance/insurance are amongst the top five industries most at risk for data breaches, according to the article.

Overlooking Critical Milestones
Insurers tend to overlook critical opportunities to deepen customer engagement and acquire new customers. For example, many insurance companies fail to recognize when a customer’s life situation changes such as when a customer gets married, welcomes their first child or takes out their first mortgage. These life events are often the very moments when we have the opportunity to deepen our relationship with the customer.

AI and machine learning can be used to automate sending out personalized messages or offers, track and measure the success rate of these efforts, and suggest improvements.

Penalizing and Failing to Reward Loyalty
Another issue that needs addressing happens when insurance companies penalize and fail to reward customer loyalty. Insurance companies must begin to take a more sustainable approach. We can start by offering customers a more holistic view of their protection along property, life and health, and point out any gaps in coverage. Then using a combination of predictive analytics and behavioral economics, insurance companies can build loyalty and up-selling campaigns where customers can be rewarded for holding multiple policies or for renewing policies.

Insurance companies have an opportunity to innovate, unlock new streams of revenues and deepen customer relationships by leveraging the tools of AI and machine learning. However, insurers have to first address some basic challenges such as ensuring they are generating high quality data. They must also start to take bolder steps towards adopting a customer centric culture, build a deeper understanding of their customers, and start rewarding customers for their loyalty.

For reprint and licensing requests for this article, click here.
Machine learning Artificial intelligence