Let’s be blunt, the UK motor insurance market, driven by its dependence on aggregators, is broken. There we said it, just like that. You only have to click a drop-down occupation menu on a well-known comparison site and see `Parachute Packer’ come up as a possible 2022 job option to know it’s time for something new.
Someone else who believes things have to change is Adiona CEO Paul Harvey, a British-American who is passionate about developing consumer-focused 100% data-driven insurance products. Also on the Teams call today was Eugene Van Biert Jr from Duck Creek, a company that will need no introduction to insurtech enthusiasts.
But let’s do it anyway; Duck Creek has worked with LexisNexis, Capgemini, Accenture and many more, so they know how to make data work for companies. Now Duck Creek is working with Adiona on their new car insurance product and the timing could not be better, as the FCA ruling kicks in and price no longer becomes the primary benchmark for consumers.
As we move into a new era of loyalty, added value, ADAS equipped vehicles, better customer service and ever faster claims settlements, let’s learn more about the future of car insurance.
IE; Tell us more about how this partnership came about and why you both see such potential in the car insurance sector.
PH; OK, so a little bit of history, the motor market has been making money for decades by price-walking customers, selling claims, exorbitant premium financing, and administration fees. That had to change and thanks to the FCA, it is a genuinely new era and starting to change.
The other factor here is that an outdated set of data was underpinning the market as regards quotes – wealth, credit, age, gender, job, even ethnicity. You do see that bias in the UK and other countries. Insuring motor risks should be about driver behaviour, vehicle type and when and how a vehicle is used. I came to the conclusion long ago, when I was working at Octo Telematics in fact, that data-driven car insurance was going to be the only fair way to assess risk. At one point Octo had six million vehicles connected and that generated a huge amount of data. You begin to see patterns and that era of telematics or connected insurance really started the journey towards the algorithms that allow AI to calculate and adjust premiums – and so much more – today.
IE; But there was always that slow take-up on telematics devices outside of the fleet market and then the notion that the entire insurance industry would agree on one driver score as being the indicator of a safe driver was never really going to work?
PH; That’s true, also I learned from my past experience with Bank of America and Octo. Telematics are nothing more than cheap phones or data devices. I believe telematics is going to be like the ISDN, which was important for a short period of time. We will talk about telematics for a few more years but everything is going to a world of connected cars.
At Bank of America, I learned two lessons: first, big financial services companies have huge technical debt; and second, it is hard to change and despite the size of their budgets, they have very little money for innovation. When there was something new they could take on board, it had to be highly scalable and secure.
So, we started Adiona and knew we needed something flexible, and truly scalable. We felt Duck Creek gave us that type of platform. Our IP was going to be our AI models and digital underwriting. Utilising data for underwriting in real time and at scale is what matters. Adiona uses AI to price cover based on the car, plus where, when and how it’s driven. We will also use digital processes to settle low-value claims and to connect to lots of drivers via smartphone and a connected device, which has other benefits we will get into later.
So, we took the bench-tested product to Insurtech Connect (ITC) Las Vegas last May and tested it live with real data coming from the UK. Our real-time underwriting engine just worked. No drama, just results, a bit like an iPhone demo – kind of underwhelming because it does the job but that’s the secret sauce, that’s what makes it good. Here’s one little example; you can get real-time pricing based on data, so if a driver suddenly does something dangerous you can pause cover, or adjust the price. You proactively manage the risk. Because good drivers want the rewards, and everyone agrees the bad drivers need to pay more – that’s fair. It’s not about postcodes anymore.
IE; Eugene, let me ask you why Duck Creek saw an opportunity with Adiona.
EVB; Adiona is a very interesting insurtech for us at Duck Creek. Usually we work with quite big Tier One companies which is excellent, but there are always opportunities to do something good with smaller start-ups. The thing about Duck Creek’s SaaS (software -as-a-service) systems is that they can work on any scale, that’s the beauty of it. It is that challenge of building something new that also has real scalability which is exciting for us.
Our technology is very open, very API driven and because we are always looking for differentiation in the marketplace we wanted to work closely with Adiona.
IE; AI-driven software is transforming every aspect of insurance, have you seen that process speed up during the pandemic?
EVB; Definitely and it isn’t just because of more remote working, different locations and so on. What we see is insurers and brokers realising that something like quote and claims is a complex business and let’s say there are 100 points in there. If you can use technology to shave off one or two of those points, then you are potentially making huge savings on something high volume like the car sector.
What Adiona wants to achieve is install a new system, at scale, which really reflects the modern, data-driven way of doing insurance, so we can help to build that toolkit which gets the job done.
IE; It is a new foundation isn’t it? For decades car insurance has really worked on a model that was paper based, it just computerised the old dataset.
PH; Exactly. You know I was lucky enough to meet Bruce Carnegie-Brown and visit the Lloyd’s of London building. You see the model of the ships that Lloyd’s was underwriting going back to a time when they met in coffee houses and all they knew was the vessel, the cargo and the Captain. They had to quote on that data!
But we have a modern-day version of that. We have driving behaviour data and know how a vehicle is being used. Thatcham Research is brilliant at testing vehicles for safety and rating them, plus they’re listing the ADAS systems in each car by make and model, classifying them for the future. As an industry we really need that data as the starting point, we have to understand the vessel first.
IE; Yes, I can see that feeding that ADAS info into your data funnel would also help calculate the potential salvage value, or the potential repair costs. So you could make that calculation very quickly after an incident, again that saves time and money for insurers.
PH; Absolutely. You could add real time vehicle resale values too, so in the future I can actually see insurance loss adjustors becoming data and asset traders.
IE; That is an amazing transformation for adjustors, from a clipboard report and some smartphone photos of the damage to data brokers. Let me ask you this; do you think that telematics was something of a side turn on the road?
PH; In some ways maybe so. But perhaps what Adiona is building is a starting point for something that is more sophisticated in the future? We had to start doing something with data as an industry in the past and we learned valuable lessons from it. One of the key elements is how we connect the car data to the driver data in the future. So we are looking at ways to cross-reference the data in the Adiona product, so that you are using layers of information to construct a model that functions in real time.
EVB; One of the great things about 5G is that it frees up insurance brands to obtain data from multiple sources and process it without having to build a data centre somewhere.
If I think back just a few years Duck Creek used to get a fair percentage of enquiries about building an IT facility for a company, somewhere where they could collate and analyse datasets. Now, we hardly ever get those enquiries. Every discussion is about how brands can tap into Cloud-based data flows and create useful channels, or funnels, of data. Because each insurance product needs a different set of data points.
IE; I’m curious to know how the car and driver data can interact within that pipeline, where are the reference points if you know what I mean?
PH; The advantage a new brand has is clean data, they start with no history. In some ways that is good because you have no geographical, or job bias etc. But whether it’s a new start-up or legacy system, you still have to process the data from a baseline. So you match the vessel to the cargo (driver plus passengers) and the voyage – the regular journeys we all make. As time goes by you gather more of that from the driver’s insurance app and the AI keeps accumulating and sifting that information.
If there’s a claim, you will have a huge advantage with a system like Adiona because you have that golden hour to look at repairs, recovery, legal etc. All the stuff that can become very admin heavy later on can be analysed in low value incidents – potentially settled within that golden hour. So instead of selling on complex claims insurance brands can settle them, put them aside and concentrate better on customer service, extra products and the business of pricing more accurately.
Insurance is a social good and one thing we want to do at Adiona is bring affordable insurance to communities who have been unfairly priced out of the marketplace. When you price or renew on real-time risks then you take away that ancient postcode, job, income bias and so on. It really is time for insurers and brokers to drop that mindset and price cover on personalised risk factors, not broad brush strokes.
IE; That has to be good news for thousands of drivers who feel the current system discriminates against them.
PH; Without doubt the way to end that discrimination – and you do see it in cities, or certain areas – is to insure on individual data, not jobs, age, gender etc. That level baseline is an integral part of what we are doing at Adiona.
IE; It’s fascinating to look ahead to a new era in car cover, thank you both for your time.
Be the first to comment