The big data revolution isn’t coming - it’s already started, and it will change insurance in at least two key ways.
The first big change will see insurers being able to build a much deeper and more accurate profiles of their customers.
The second is more challenging. Insurance will have to embrace a big shift in culture to become much more positive and proactive in its relationship with customers.
‘’Big data will be transformational for the industry,” said Insurance Council chief executive Tim Grafton.
BUYING AND SELLING INSURANCE
As insurance companies begin to mine big data it will help them get a clearer picture of what their customers want, experts say.
Michael Naylor, a senior lecturer in finance and insurance at Massey University, said there was a whole new range of data resources and techniques available for insurance companies to use, including information collected from a person’s social media right down to their seemingly mundane gym membership and grocery shopping purchases.
“To combine all that data together would not be easy, technically speaking, for an insurance company to do, but once it’s done it will be a lot easier for the company to know about each individual customer and then the company will be able to do ‘customer underwriting’,” he said.
“What it means, essentially, is that you won’t have to ask the clients a lot of questions to gather data. It will simply already be there in front of the companies, collected from each clients’ digital footprint. The companies will know more about you than you know about yourself. They will know, for instance, if you are googling wedding dresses and then will be able to offer you, for instance, life insurance, or if you have started taking yoga classes at your local gym and offer you a specific health insurance policy that covers you for injuries sustained during a yoga class.”
The data could be collected by a “digital agent” – a piece of software that monitors and collects a person’s digital footprint, Naylor said.
“Currently people give themselves away to Google and Facebook for free, so why wouldn’t people want to charge a digital agent for learning their online surfing patterns?
“The software would build a profile of an individual and then these digital agents would get the profile, control the profile and filter the profile to the people who want to get it – namely insurance companies.”
“No forms,” Naylor said. “Insurance companies will be able to tailor-make a policy for an individual without that individual having to do anything. The person’s profile and the insurance company’s software would come to an agreement and give the customer the best deal available.”
Insurance companies are already starting to build profiles of clients and provide bespoke insurance packages based on a client’s specific needs and wants, Grafton said.
“This will occur and has occurred,” he says.
“For example, a start-up and relatively new entrant to the UK market, Bought by Many, has specialised through extensive research in a niche market, pet owners; they’ve aggregated key needs and developed products with underwriters. This shows the emergence of more sophisticated and responsive broker models.”
Big data will also mean more interesting and accurate predictors of risk being discovered and developed.
“Insurance is a sector that has always been heavily dependent on information to inform its decisions about risks, so a quantum shift in the ability to source, amass and analyse data almost instantaneously provides a platform to significantly enhance the insurance experience for customers.
“Smart analytics will drive more risk prevention and better customisation to meet individual consumer needs which will also reduce claims costs.”
One big change – and benefit for insurance companies – would be better and more efficient risk assessment for underwriting, Grafton said.
“Underwriting based on historical data may become less relevant in some lines. There will be faster and more insightful processing of claims and automated processes, will make operations more efficient.”
Property insurance premiums would be influenced by the higher volume of data available to insurance companies, Grafton said.
“Big data will provide risk information at a granular level – for instance, for individual houses based on publicly available data that can be aggregated and modelled to deliver a house’s risk profile,” he said.
“Alternatively, real-time telematics can risk-profile drivers, which can provide consumer benefits by encouraging lower risk driving rewards.”
Customers will benefit because insurers would need to ask fewer questions of their customers to be informed about a risk they are insuring.
“Insurers would have the ability to proactively deliver new product offerings for consumers based on better understanding and knowledge of their needs, leveraging the technology to act as a virtual e-broker.”
The increased use of blockchain technology would enable claims settlements to be virtually instantaneous in some situations, Grafton said.
“Take the example of someone with travel insurance heading to the airport who receives a credit payment from their insurers triggered by confirmation from the airline to the insurers that their flight has been cancelled.
“Technology has the potential to build high levels of trust and confidence in the delivery of the insurance promise. And, of course, the use of multiple platforms will enable consumers to purchase their insurance on their device of choice, which will almost certainly be a mobile device, and to have their claims settled rapidly via the same platform.”
ON THE ROAD
It’s not all a question of buying more insurance, however, and in some areas customers may find they need less. Car insurance is one of them.
“Car insurance has probably another decade and then it will be utterly unprofitable,” Naylor said.
And the cause of that death? Driverless cars.
“Once we move to driverless cars the crash rates will drop markedly. The car insurance industry as a profitable field is doomed.”
The reason car insurance on its own would stop being profitable is simple, he says.
“If you don’t already have to drive your car and it crashes it’s not your fault because you were not driving, it’s the fault of the software or something similar. If there is no one in the car you can’t blame the driver and you can’t blame a person.”
And while it may take a decade or more for driverless cars to become common on New Zealand roads, Naylor said there would be a “tipping point” for the car insurance sector.
“What we are arguing about is not a matter of when self-driving comes in, it’s when the (car insurance) cash flow starts to decline,” he says.
That reduction in cash flow will be because of fewer crashes, and therefore fewer claims and ultimately less need and want from customers to take out big car insurance policies. Insurance companies will have to fight for the business of ‘’bad’’ drivers – the ones that are involved in crashes with other drivers – and things will start to become more competitive, Naylor said.
“Already all the high-value cars have a range of in-built crash avoidance technology,” he said.
“So as an insurer’s good customers that own very expensive cars don’t crash, they will be competing for the customers that still do all the crashing.”
And there is also what he refers to as the "tipping point", the moment when vehicle crashes fall to such a low rate that it effectively renders car insurance unprofitable for companies to sell.
"If we say 15% of the cars are self-driving they won’t crash, they will ‘talk’ to other self-driving cars on the road and they will all avoid each other and therefore avoid crashes.”
Once autonomous vehicles were 35% to 40% of cars, people-driven cars would find it very difficult to find another car on the road to crash into, he said.
Jeremy Holmes, principal of actuarial firm Melville Jessup Weaver, says another way big data will influence car insurance is via telematics, which are devices that monitor insured driving behaviour and incorporate this into their premium.
“This sort of thing hasn’t really taken off in New Zealand yet but is very big in the United Kingdom and United States,” Holmes said.
“When you’re recording second-by-second driving information covering numerous metrics (speed, location, direction, G force) for hundreds of thousands of insureds, then it doesn’t take long for the data sets to become so voluminous as to present big challenges to data users. This probably falls well into the territory of what people like to call big data.”
In order to survive the brave new world of big data, insurance companies will have to undergo a radical culture change, Naylor said. In short, they have to start being “more positive”.
“Currently insurance companies are fairly negative, you deal with them when something bad happens. You deal with them when you are trying to claim, you deal with them when they demand renewal money and that’s why people don’t like insurance companies.”
But the good news is insurers could use big data to create positive experiences and a higher level of service for clients.
One example would be using pre-set software in a car to know the moment a client’s car engine started to get problems and thus book a repair or a service on the driver’s behalf - or sending a customer a rental car because they knew the moment the car had broken down.
“And the companies have to give clients information they like receiving,” Naylor said.
“They have to have ongoing contacts with clients through social engagement.”
That may sound hard, but a company like Amazon proves it can be done.
“I mean Amazon was a book warehouse at the start and they moved into being a company that basically shows you things that are nice.
“Insurance companies have been focused on cost control and on risk, they have not been focused on their customers. They really must change their entire culture to become customer-engaged and friendly. They have to become aware that everything they do needs to be based on what the customer will respond to, and that’s a real hard trick.”
But insurers had to change, he warns.
“If insurers do nothing they will gradually shrink, and they will end up being just a piece of software in a box. What they must do is turn around and say ‘well, we have all this data, we have, for example, a real-time link with drivers, what can we use it for? What can we leverage?’.”
The obvious step would be for insurers to offer customers a range of things to buy that the companies could make a profit from, like the health insurance company in Singapore that provided day care, Dr Naylor said.
“When Amazon made the Kindle they lost money on every Kindle sold,” he says.
“But what they did was they then used that product that was in everyone’s hands to sell books and music and things like that.
“Amazon then dominated the market and sold a range of value-added things around that. Think about Facebook, it makes all its
money from advertising. So why would you make your money from insurance when you can make your money in a range of other ways?”
The role of actuaries would change as more insurance companies started to “mine” big data, Holmes said, and actuaries would no longer be sitting in the back office, crunching data.
“I think the role of actuaries is definitely changing,” he said.
“If you look at the pricing teams for large insurers (in NZ and abroad) then they’re generally made up of some actuaries as well as other technical experts. Insurers understand that your pricing team doesn’t need to be all actuaries.
“Actuaries are not simply data scientists or insurance statisticians, although we share some of the same skill sets. The value that an actuary adds is to take the technical results and overlay with commercial acumen, market knowledge, judgment and common sense.
“Actuaries now and going forward will act as a conduit between the data technicians and the decision makers to help them understand what the numbers actually mean in terms of achieving their goals.”
Dr Naylor said the traditional role of actuaries would be replaced by artificial intelligence algorithms and that going forward “most things will be done by software”.
“It will be the software that will be doing the underwriting,” Naylor said.
“And therefore the actuaries will actually have to supervise the software systems. Computers are very stupid and will make mistakes that are obvious to people.”
SET TO PRIVATE?
But, will the growth of big data, should customers be worried about privacy? Maybe, Holmes said.
“It’s reasonably easy to find out what information an entity holds on you (you just ask them) and you can correct it if it’s wrong,”
he said. “It’s no surprise that entities are collecting more and more data on people these days – internet browsing behaviour is an obvious example. Fortunately, individuals have some control over this by you choose how you use the internet.”
And while data collected from your internet browsing and social media habits is not really incorporated into insurance pricing in New Zealand yet, says Holmes, this may change in future.
“But for an insurer to go to the effort of analysing all this data, they need to have some belief that it might actually make a worthwhile difference in the insured’s assessed risk profile.”
On the issue of privacy Grafton said high levels of protection needed to be built into any insurer’s data storage systems.
“Insurers need to operate with high standards of transparency in dealings with customers about data collection and the use of the data (including what it will not be used for),” he said.
“Full compliance with appropriate privacy legislations will be essential, and should any breaches of privacy occur, then customers need to be advised immediately.”
Dr Naylor is blunter. “There are massive concerns about privacy,” he says, “but young people are not that bothered. Older people don’t understand that everything we do online is public.”
Expect to see insurance companies offering discounts to people who are “prepared to expose their lives” to show they are a lower risk, Dr Naylor said.
“The people who don’t will be higher risk and therefore there will be a massive increase in their premiums.
“And I think soon enough customers will start to see the benefits that flow from insurance companies knowing a lot more about them via their online presence. It’s really going to do some amazing things.”
Professional IQ College offers workshops, online courses, webinars and qualifications.
For upcoming events:
Where members can access industry Resources & Media Content