

Insurance innovation of the month – Ping An’s Automated claims solutions
Ninety’s “Insurance Innovations of the Month” feature contains regular snapshots of real and recent insurance innovations, based on conversations with the innovators behind them. They are drawn from anywhere in the world, and from all categories of insurance. If you’ve got an innovation you’d like to highlight, drop us a line at ninetymarketing@ninety.com.
Insurer(s): Ping An
Market(s): China
Insurance type(s): Motor insurance
Problem:
Ping An insure about 55m cars in China, making them the largest auto insurer in the world. They receive about 15m claims per year. In insurance, claims are the messiest part of the process; they are complex and most often manual.
Solution:
For their policyholders, Ping An have a whole suite of apps that sit on a platform called ‘Good Driver’. It has a very similar branding concept to their ‘Good Doctor’ platform. One of these apps allows you, if you wish to make an auto insurance claim for damage to your car, to start the process by photographing and submitting the VIN plate in the corner of the windscreen to confirm the claim relates to the insured car. Users can then submit photos of the damage to support their claim, which the app measures with GPS. The application uses an AI algorithm, with a database that holds every external part of 98% of cars in the marketplace. The algorithm runs through 23 different damage scenarios – so, multiple ways to assess a claim from something significant to a minor ding to an external part of a car, including image correction for handling reflections in glass, plastic, chrome, etc. Within seconds, the algorithm assesses the damage, and prices both the parts and labour costs for repair. Those numbers are completely different in different parts of the country as labour costs vary. The application then gives the user a quote, which they can either proceed with and the details are then forwarded to a repair shop; or, the user can elect to settle the claim themselves. By pressing the button on that option, the system instantly credits to the e-wallet the quoted amount, which the user accesses to pay for the repair.
Outcomes:
This is a very powerful process. During the COVID-19 pandemic, especially from January to March 2020, Ping An have observed a big increase in the app’s usage. In this period, about 40% of claimants were using the fully digital, automated process, up from just a small percentage of users previously. In normal times, 70% of policyholders partially use the app, although most people rely on a combination of online/offline services, such as the support of claims adjusters. Ping An have just created a new version of this platform. Through Global Voyager Fund, they have invested in the German company, Finleap – the largest fintech company builder in Europe. Finleap has created 15-16 next-generation, mainly B2B, platforms, to support the Financial Services industry. They have created capability for actions like collections, asset management, and Solarisbank (bank in a box, or banking as a service, with all necessary licensing included), etc. What is more, with Finleap, Ping An have created Joonko, a platform for auto claims. Part of the investment deal (Voyager Fund has invested in Joonko separately) was that Joonko would consume some of Ping An’s technology. They have launched an interesting solution called ‘Claim It or Eat It’. It is exactly the same claims platform, interfaced to a third party system for parts and labour for the German market, and providing a visual scan for parts for German car models. If you are a Joonko client, for example through Allianz, and you’ve had a collision, you can scan the damage, and the app will tell you what it will cost to fix, and whether it’s better for you to make a claim, or pay for it yourself. In other words, it determines if the renewal price impact outweighs the claim saving.
This investment and partnership is a great example of solving an interesting but difficult problem at Ping An, with new relevance post-COVID-19, and scaling into new markets through a version that can be an overlay for other insurers.
Social impact:
As a social, for-purpose business ourselves, Ninety is always looking for the social, human or environmental impact of new ideas in insurance. Here, the impact is clear: the process allows users to maintain their financial solvency when experiencing car issues and get back to the pre-accident stage in a stress-free and quick manner.
Tips to other innovators:
- Ping An has different approaches to different innovations; in this case, it was an internal build unlike Good Doctor, which we mentioned in our last Insurance Innovation of the month, which was a new stand-alone business.
- The process of innovation – lots of large companies have a very top-down approach and they ask to see goals upfront and review ideas via committee. Then possibly they decide to assign some funding and revisit the idea in three months to review where the innovation thinking is at. The process is slow but potentially, in a year or so, they will have something they can take to the market. Ping An’s learning has been that the real world doesn’t work like this and innovations require many, many iterations, and hundreds of real-time corrections. The idea of presenting a process in 6-12 months is a complete anathema to them. Even Peter Ma, Chairman of Ping An, completed about 200 iterations to get to the first working prototype of the driver claims application that is described above. That’s a lot of senior management time and very detailed interaction. As you’d expect with any user experience or technology, sometimes it’s the big things that matter, but often it’s the small things that make a difference.
Categorization for insurance innovation portfolio planning:
Ninety Ten Big Ideas: this is an example of Ninety’s “Personalised Products” and “Customer Interactivity” Big Ideas.
Ninety Ten Types of Insurance Innovation: this innovation combines the following Ten Types categories – 1) Product & Proposition, 2) Added-Value Services, 3) Customer Experience, and 4) Channel & Distribution.