Welcome to Part Two of our conversation with Professor Andy Pardoe, one of the world’s leading voices on the subject of Artificial Intelligence (AI).
In Part One of our interview, The Future of Insurance is Artificial Intelligence, Professor Pardoe discussed how AI is changing the insurance industry from the outside and why it’s so important for insurance companies to act now. In this second half, he explains…
Professor Andy Pardoe is a world leader on the subject of Artificial Intelligence and is in constant demand as a consultant across industry sectors, including financial services, retail, media and, of course, insurance.
An expert advisor to the British Government’s All Party Parliamentary Group on AI, Professor Pardoe is also listed by IBM Watson as one of the Top 30 AI Influencers globally. He holds three degrees from Warwick University (BEng, PhD, MBA), an Honorary Professorship and is Chair of the Warwick Technology Professional Network.
Professor Pardoe, thanks very much for joining us once again! In our earlier discussion you talked about how AI is making rapid inroads into the way the insurance industry operates. Just how significant are these changes?
Oh, very significant. It’s like AI has been nibbling around the edges of the insurance industry for a while and then suddenly it’s biting off huge chunks. There are so many insurance-related technologies entering the market that depend on AI and machine learning, it’s hard to keep up. And the level of innovation is extraordinary.
On the one hand you have very practical AI technologies that save time and therefore cost savings. Tractable, for instance, has tech that lets people take photographs of a damaged vehicle and then their software estimates the cost for the repair. This doesn’t completely remove the human from the equation but it can speed up the work of an adjuster by 100-200%.
And then, on the other hand, you have these really big swings where companies are building AI tech that aims to predict life events, such as getting married, moving house, having children…even the likelihood of health problems. These days everyone’s wearing a smartwatch that can collect data on your activity (or perhaps inactivity!) – your resting heart rate, how much time you spend outdoors.
AI is the perfect tool for analysing all of the data and making assessments about health, and how likely you are to experience problems in the future.
We’re not yet at the point where we can predict every detail of an individual’s life – and we probably never will be – but when it comes to groups of people, we’re already seeing companies make predictions with astonishing levels of accuracy.
It’s not hard to see how fundamentally this will change insurance. The industry relies so much on calculating probabilities to produce accurate, profitable quotes. And AI is already allowing insurers to improve that accuracy and therefore their profitability.
How do you think consumers will react to seeing their premiums change based on their activity and behaviours?
If it means paying less, I think consumers will react very well! Ultimately it’s about fair value both for consumers and for insurers. But then that assumes that individuals will be directly comparing their premiums with other people which I don’t think will happen. In practice, someone might mention in conversation that their annual motor insurance renewal has gone up and complain about it to their friends. But they don’t tend to sit down as a group and all compare each other’s numbers.
What it does allow insurers to do, if they’re using AI-driven software when making their quotes, is offer a better deal than their competition to win more business.
Do you think this will change consumer behaviour at all, or are all of these AI-led changes just happening behind the scenes?
If it changes consumer behaviour at all it will come down to how the insurance company markets what they’re doing. It’s hard to imagine a consumer looking at their life insurance quote and deciding to do some extra exercise so that next year they might be able to get a slightly cheaper deal. But they will change their behaviour if they’re given a direct incentive to do so that makes them feel as if they are in control.
That’s the critical difference. More and more consumers want and expect to be in control of their relationship with service providers. If you can offer them that and reward them at the same time, it is actually possible to change people’s behaviour.
Health insurance is a good example of this, with some providers offering their customers incentives to make healthier lifestyle choices. You can draw a pretty straight line between encouraging a customer to do more exercise and a decreased likelihood of needing to make a claim. But what’s interesting is that the incentives are often unrelated to the cost of the premiums themselves.
Vitality, for instance, gives customers points for exercising. And although one of the rewards is a reduction in the annual renewal price, they also offer things like free cinema tickets, free Caffé Nero drinks and Amazon Prime upgrades. Even the wearable itself can be at least partially paid over several years for if you do enough exercise.
This is significant because it moves the whole thing away from price-led marketing and, from Vitality’s perspective, keeps the customer locked in for much longer.
Ironically, the customer could wind up paying more overall for their health insurance this way, but because they have the ability to impact the price they pay and the rewards they receive, they feel like they’re in control. And ultimately, for many, that’s a more important factor in the decision they make.
AI and machine learning are at the forefront of the pressure on the insurance industry to adapt, but there are a whole host of other factors, like these reward programs, in the market that are stirring things up.
The only thing that we can safely predict about the insurance industry right now is that it’s going to keep changing and it’s going to do so quickly. Insurance companies need to watch what’s happening closely and be ready to respond. If possible, be proactive rather than reactive.
This seems to be the big challenge for insurers. They recognise that AI is going to be important to their business, but when and how do they get started? No one wants to be left behind. But equally, no one wants to jump in too early and make a costly mistake.
That’s a valid concern, but there are a couple of things insurers can do now that will let them explore AI with minimal risk and also future-proof their activities.
First and foremost, insurers need to be capturing as much data as possible from their activities and either using AI-powered software to analyse it, or working with a partner who has this expertise. Whatever you do with AI now and in the future, your success and your profitability in this area is mostly going to boil down to the data that you collect.
I can’t stress enough how important this is.
If you’re an insurer and you’re not properly collecting and managing your data, you are losing ground. It’s akin to keeping all of your money in a bank account and collecting the minimal 0.5% interest when you could be investing it in the wider market. The data you collect now is going to be your currency in the future.
Secondly, it isn’t necessary to develop your own AI tech from scratch. There are plenty of tech firms who have already invested the money to create the innovations insurers can use. Look for insurtech firms with a proven team and product and you can test it out without having to risk time and investment building your own solution.
That’s a good segue into Nuon AI. How do you see Nuon fitting into this picture?
When working its way into an industry, AI tends to go through specific phases. Phase 1 is automation – largely working alongside the human element to speed it up – to reduce costs. Phase 2 directly works on the product or service itself.
Nuon AI is Phase 2 technology because it allows underwriters to adjust pricing in real time, far quicker than they could achieve making manual adjustments, with the end result of maximising revenue.
It’s about working alongside underwriters to enhance their capabilities and data insights.
And as mentioned before, Nuon has a sandbox mode so insurers can run tests using real numbers and activity to see what the outcome would have been if Nuon’s software had been running live. This allows for a gentle introduction to AI so you can build confidence and trust in the tech before putting it to work.
This is a good approach for using any AI-powered tech. If you can test it in a simulation first, you’re taking a proactive approach, but removing most, if not all, of the risk from the equation.
Thank you to Professor Andy Pardoe for taking the time to talk with us for this series! Remember to check out Part One, all about the future of insurance through AI here.