This Opinion piece is by Erik Fogg, Co-Founder and Chief Operating Officer at ProdPerfect, which is an autonomous E2E regression testing solution that leverages data from live user behaviour data.
The insurance industry presents exciting opportunities for AI-driven technology to completely change how companies compete in the marketplace. Disruptive innovators are changing the way pricing is calculated, fraud is detected, claims are processed, and plenty more. As the depth and scope of data accumulated on a per-person basis continue to grow, insurance companies should be looking forward and identifying how artificial intelligence can help make use of this to cut costs and improve the customer experience.
How AI is Disrupting the Insurance Industry
Consumers today looking to purchase insurance expect a speedy and frictionless process. Insurance is a crowded marketplace, with many companies competing to offer the lowest prices and utilize the newest technology to undercut the competition. Balancing the need to provide an efficient customer experience while remaining both competitive and profitable is challenging. The increasing digitalization of the insurance process, from new customer acquisition to fraud detection, risk management, and beyond, creates new challenges and opportunities for AI and insurtech (insurance technology).
An excellent example of digitalization in insurtech is black box car insurance. With this, drivers fit a small telematics device in their car, which collects and transmits valuable data (things like speed, location, time of day, etc.) to insurers to help adjust policy premiums. Safe drivers are rewarded with lower premiums, and risky drivers can be identified.
Some telematics devices also warn users when they detect risky behavior, such as speeding. This data is invaluable for determining risk to insurers, but it comes at a cost. These devices can cause problems for the user if they regularly give out false positives. There is a lot of nuance to calculating risk – is a user sharply swerving engaging in dangerous driving, or are they trying to avoid an accident? Is a user accelerating to legally overtake a slow vehicle, or are they speeding?
Another use case is in pricing. Insurance comparison sites are commonplace, and users can easily contrast and compare plans and prices. Price carries a lot of weight with many consumers, so having competitive prices is very important. Companies should consider not only their internal pricing structure but also the prices of their competitors. These rates require constant adjustment, so companies do not have the luxury of time. All this data must be processed quickly and seamlessly. Defects in this process can have very real financial penalties.
Developing and maintaining comprehensive insurtech test suites is therefore challenging. Teams need to iterate rapidly and innovate constantly. New releases are being pushed several times a week, all of which need to be comprehensively tested without creating a bottleneck in the process. Manual testing that takes days to run is simply not suitable for many insurtech companies.
Current AI Application Trends in Insurtech
A relatively recent approach to pricing, known as Behavioural Policy Pricing (BPP), uses potential new customer behavior to determine the price of a policy or product. This does not replace the traditional approach to determining pricing, but it rather compliments it. Instead of assuming customers have all information available to them and making rational decisions based on this data, BPP instead focuses on asking the customer questions: how do customers perceive and price product information? How do they react to quotes and use them to make purchasing decisions?
BPP is driven by experience and learning. The more information about the customer that is available, the better able you are to deliver a price that a customer is likely to agree to. These factors make it particularly well suited to AI-based automation. Models of customers can be developed that break down available data into categories weighted based on each customer’s value perception to arrive at a personalized price.
All of this requires a lot of testing, and the constant pace of iterating means that tests are constantly breaking. Test automation helps insurtech companies iterate rapidly without losing velocity, but this alone is not enough to help keep pace with rapid development. Continuous testing tools automatically run PRs against full (or partial) test suites and return results to developers in very short time frames. This allows companies to catch and fix bugs before they make it to production, but without disrupting or stalling developers.
The Impact of AI Solutions applied in InsurTech on Customer Experience
AI solutions in insurtech enable companies to offer customers a convenient, personalized service. From customized policies and pricing to detecting fraud and dealing with claims, companies that innovate can offer a faster and smoother customer experience.
Test coverage is therefore critical in insurtech. Defects in the process that do not get caught can have significant consequences, so companies need to know that they are testing the right things. Instead of relying on manual testing to cover all potential use cases, insurtech companies turn to AI tools that analyze the routes users are taking through an app and use that information to generate tests where they are most needed for the most valuable use cases.
This cuts down significantly on both test development and test runtime. By selectively developing tests based on data-driven customer behavior, you don’t waste time developing and running unnecessary tests. You also have confidence that the tests that do exist provide value; this can even mean removing older tests automatically if they are no longer valuable. All of this allows test suites to be kept lean and test runtime to a minimum to accommodate the very rapid nature of insurtech development.
Test maintenance is another timesink for many insurtech companies. Each iteration in a rapid development cycle requires new tests to be developed and may also break existing tests. Requiring developers to take time out from development to fix broken tests slows down the development process. Instead, by utilizing self-healing tests, developers are free to write code without worrying about breaking tests. The less time developers must spend writing and maintaining test code, the faster companies can iterate and innovate on new developments.
These developments drive the innovation that powers the most competitive insurtech companies. Customers are increasingly savvy and price-conscious, but there is plenty of room in the marketplace for those that can innovate. Building lean and utilizing tools like serverless architecture let small teams leverage the power of the cloud to crunch large amounts of data without requiring significant overhead. This is only part of the process; delivering an improved customer experience also requires rigorous testing. The speed and coverage that AI brings to testing processes let teams focus on their product first and foremost.
The low barrier to entry and fast-paced nature of the insurance industry allows disruptive insurtech companies to compete with the large, institutional players. Smart use of AI and technology enables smaller firms to be nimble and offer competitive services that deliver low prices and an improved customer experience