In recent years, the insurance industry has finally started to embrace digital technology, as insurtech startups have forced the traditional players to adapt if they want to survive. Wefox, a digital insurance marketplace, is at the forefront of the changes. Having successfully moved what has historically been a face-to-face business online, the company is looking into the future, a world where the Internet of Things, data science and artificial intelligence will play major roles in our lives, including how we insure what is valuable to us. We spoke with Julian Teicke, wefox’s Founder and CEO, about innovation in insurance, the future of AI, and the evolution in the way we treat our personal data.
What made you decide to implement AI in your product?
I think every business should use AI. The idea behind wefox is to use data to change how insurance works, and of course AI plays a huge role in that.
Insurance provides financial protection in case of unpredictable events. Typically, insurance companies use historical data to calculate the risk of certain types of events, and use that to calculate an average premium customers would need to pay that would allow the company to make a profit, even though in individual cases they pay out significantly more than the customer paid in.
Our idea is to take a proactive approach to insurance. We want to use data from people’s IoT devices to better understand the risk that an individual is in and make more accurate predictions about the likelihood of certain events happening. Then we could use that information to warn the individual that they might be in danger, instead of just paying them after the fact.
If you could actually protect people from something bad happening, that would be incredible. But at the same time, you would need to have a huge amount of personal data to achieve this. How do you make sure that your customers feel that their data is safe and they are not being spied on?
We definitely need to ensure that our customers’ data is secure and is not being misused.
There needs to be a shift from companies owning their customers’ data to individuals owning their own data. I believe this is something we’ll be seeing in the next ten years. The consumer will be able to see what data is shared with which company, as well as the value they are getting in return for sharing it, and if they feel that the value is not high enough, they can stop their relationship with that company.
Obviously we don’t have the infrastructure to make this possible yet, so how do we get there? There isn’t a lot of demand for it yet, because the level of awareness people have about their personal data is not yet very high. Most people don’t seem to care about who they share their data with. It will take a few years for the awareness level to rise.
In the meantime, we try to be very responsible with our customers’ data. Our rule is to never do anything with customer data that we wouldn’t want done with our own. We have many policies in place that cover how we handle customer data. For example, the data that we use in real time to calculate risks is anonymized and can’t be connected to the individuals. We only use individual data when it comes to handling claims.
How did the implementation of AI to your product look like?
We’re actually still far away from where we want to be when it comes to AI. In fact, I don’t think what anyone is doing today is real AI. What we currently have are more sophisticated ways of collecting data and learning from it, but we’ll only get to real AI in about 10-15 years.
Our company collects a lot of data and we have a lab where we work to figure out how to use geolocation data or any other data to engage in risk prevention. But for us, innovation is secondary to having a revenue-generating core business.
We started with something very basic: to help insurance advisers sell more products and help customers have the most convenient experience possible. 90% of all insurance policies are still sold without technology. We set out to make that part digital. This is what generates revenue for us, and then we can use that revenue to innovate.
Many companies focus on innovation from the start, but then never actually make any money and in the end don’t achieve anything. We decided to do it the other way around and start with the business part first.
The idea of AI is often connected with futuristic visions which bring both hopes and fears. What is your take on that?
At this point, the risk of AI replacing humans and making them irrelevant outweighs its potential benefits to society.
In order to use the advantages of AI, we have to think about how we make individuals the sovereigns of their own data, and based on that data have what we call an ‘AI you’ - your own AI, on your own device, that helps you overcome your individual weaknesses and be a better person. The AI must be owned by you, not by any company.
Let’s imagine it is 2005, the beginning of the internet as we know it, before big data, machine learning and AI. If you had to create your product back then, what would it look like?
In 2005 I would not have chosen to do anything with insurance. The internet first disrupted industries with a lot of customer engagement and fast moving consumer goods, such as shopping and travel. Insurance was one of the last industries on that list.
When we started around 2015, the insurance industry had hardly changed since 2005, or even 1995. Insurance companies were not under much pressure to improve customer experience, since the customer is not engaging with the product very often. They didn’t have a lot of urgent operational business. They could stop selling insurance and still do really well for the next 20 years. Nobody would realize that they're not generating any new business because the premiums would still be coming in. There was no urgency to change anything.
What they did in 2005 was put up websites. They made them look really nice because they had the money to hire good agencies, but on the backend they connected to really inefficient legacy systems. And insurance companies have so many legacy systems that some don’t even know how many customers they have, because they have dozens of systems that have been growing for 50 years.
Back in 2005, we would not have been able to do what we do. We need proper IT infrastructure to be able to collect data and analyze it using AI, which did not exist back then. Only now are we able to build the IT systems necessary to disrupt the industry.
If we had started in 2005, today we probably would’ve had the same problems as other insurance companies.
You mentioned that it’s still a few years until you’ll be able to use AI in the way you really want to. What needs to happen to make it be possible?
The first thing is happening already, which is an increase in the number of IoT devices. In a few years everyone will own a number of IoT devices, which will produce significantly more contextual data. Although it’s important to note that these devices should implement privacy by design and protect users’ personal data by default.
Another thing is that we need to work on infrastructure that allows AI to be used in a more responsible way, where data is owned by the consumer. I’m working on this with one of my initiatives, creating a data wallet based on decentralized technology that allows you to own your own data.
Finally, we need to collectively understand how to use all this data and the infrastructure to benefit humanity as a whole and not just the companies and shareholders. We need to find ways to build a responsible ecosystem that doesn’t just exploit technology in the interests of the few, because in the long term that approach doesn’t make sense for anyone. If we end up with something that makes 95 percent of the population feel irrelevant, they will go out into the streets and the remaining few won’t be able to protect themselves anyway. So we need to come together and create something that is fair.
What is your advice for new companies thinking about entering the industry?
In the insurance industry, there is a big gap between the old world of experience and the new world of technology and innovation. These two worlds hate each other, and it is extremely difficult to bridge the gap between them. People who’ve been part of the insurance world for a long time feel like they know everything about it and are often not receptive to new ideas and technology. That’s the reason why many insurtech companies fail. In order to succeed, it is not enough to create a product based on your tech knowledge. You need to be a very good mediator between the two worlds, which is mainly a leadership quality.
I would also recommend that, instead of looking for the biggest possible innovation in the beginning, you start by focusing on adding the most value in the easiest way. Many companies start out focusing on technology and gadgets that sound really cool, but then fail when it comes to creating a viable business. What we did was focus on building a hundred million dollar business in terms of revenue, and now we can easily check out new things. I think that’s the way to do it, and not the other way around.