Big Data, especially powered by with other disruptive developments like driverless cars, the Internet of Things and artificial intelligence (AI), are among six technologies that will make or break both existing players and new entrants in insurance.
Tom Warden
has been leading research and innovation in insurance for
25-plus years, most recently as the Chief Data Officer at AIG
Life and
Retirement. Previously, he was vice president of research at Allstate
Insurance and head of the Allstate Research and Planning Center. I had
the opportunity to interview him on Big Data in the insurance industry,
and how insurers should be thinking big, starting small and learning fast when it comes to Big Data.
Thinking Big
Chunka Mui: Let’s set some context. How are you thinking about Big Data as applied to insurance?
Tom Warden: Big Data is an ecosystem consisting of
expanding types of data, manipulated by massive computing and analyzed
by widening number of data science and AI methods. I am most excited by
the prospect of leveraging Big Data to create new products and new
markets, and to radically change how customers perceive insurance. Big
Data provides the means to transform today’s insurance business models.
Mui: Help us understand the big picture. How do you
see Big Data transforming insurance? Assume that we can start with a
clean sheet of paper.
Warden: Insurance today consists of passive
products. Insurers talk about being their customers’ partners, but they
really are not. Big Data from cars, homes, businesses and government
could enable insurers to better analyze risk and anticipate loss events
before they happen.
Insurers can learn what puts each driver most at risk—and incent him
to change his behavior to avoid accidents. They could automatically
trigger mitigating actions to prevent small incidents in homes, like
leaky or frozen pipes, from turning into catastrophes. They could use
streaming data from commercial plants and equipment performance to
anticipate and prevent major incidents. Thinking broader, insurers can
help multi-national companies manage the huge number of risks they
insure against in a more comprehensive and efficient manner.
By preventing bad things from happening to their customers rather
just restoring things afterward, insurers could provide products that
customers will want to buy, not just have to buy.
Mui: This is a great opportunity for insurers. Does
it also pose a threat to the industry, because others could better
leverage this potential?
Warden: Insurance has operated like a closed shop.
Insurers collect and hold data about those insured and their losses.
Actuaries are like guild craftsman trained to work on that data.
In the new world, many other players can collect and control the most
important data: car manufacturers, smart phone providers, Internet
companies and telecoms, industrial equipment manufacturers, to name a
few. Insurance companies risk having one of the most profitable parts of
their value chain attacked by those who control data flows that enhance
understanding of risk and losses.
Starting Small
Mui: You have been working in this area for several years. What is the best way to get started?
Warden: First, companies need to recognize that Job 1
is leveraging data to optimize the profit machines they already have.
Best practice today is to embed more predictive modeling in all parts of
the value chain. You don’t necessarily need Big Data to do that. You
need smart people, a disciplined approach and a culture of cooperation
to monetize data-driven insights.
Second, they need to focus on the transformational aspects of big
data. Pull together some of the company’s best and brightest from across
all functions. Give them just enough resources to create more than
ideas. Separate them from their day-to-day work. Give them deadlines to
provide discipline. Oh, and don’t forget to include some millennials!
Have the team start with what your customers value today and challenge
them to add what Big Data enables.
On both thrusts, they also need to learn from early-adopters’ mistakes.
Mui: What are some of those mistakes?
Warden: Innovating in a vacuum is the biggest
mistake. Too often, data scientists are clueless about how the business
makes money. Make sure people from the relevant business functions are
part of the team from the beginning. Everyone needs to be broadminded,
creative and collaborative.
Allowing one or two senior voices to kill things too early is another
big mistake. The whole point of the process is to go far beyond the
status quo. In general, senior leaders struggle with suspending their
disbelief long enough to give radical concepts a fair trial.
A third mistake is not placing enough importance on properly managing
data through its lifecycle to assure its validity. The principle of
Garbage In Garbage Out still holds for Big Data.
Mui: How do you keep expectations aligned in such a high profile area? How do you get everyone on the same page?
Warden: Most employees at insurance companies are
focused on optimizing the present while few are assigned to creating the
future. Leadership’s job is to create a vision of the future that
demonstrates to all employees that they play a part in today and
tomorrow’s success. They should be rooting for each other, not against.
Continually talking about this vision, the progress being made to
realize it and adjusting as new things are learned should keep everyone
on the bus.
Learning Fast
Mui: A demo is worth a thousand pages of a business
plan. What are the most dramatic demonstrations of the potential of Big
Data in insurance?
Warden: Usage-based insurance pricing is the most
dramatic example of the power of Big Data in insurance. Not only is the
amount of data massive, there are great opportunities to contextualize
the driving data with external information: weather, road congestion and
more. This leads to smarter risk assessment and smarter pricing. Add
data about the driver’s interests, to-do lists, etc. into the mix, make
everything live and adaptive and voila: you’ve created smart driving from the customer’s perspective.
Identifying fraud in claims and amongst internal parties like
underwriters and sales people is another place where Big Data is a
powerful weapon. Fraud is generally perpetrated by smart and cunning
people. Staying a step ahead of them requires sifting through massive
amounts of seemingly unrelated data to find and validate intricate
patterns.
Risk management is another high potential area. Finding, analyzing
and simulating wide varieties of factors (weather patterns, societal
trends, mortality trends and their causes to name a few) to increase
understanding about what the net effect is on a company’s book requires
very advanced data skills and powerful computing environments.
Mui: Many companies have been exploring Big Data for
several years now, at least. What are the tough questions that Boards
and senior executives should be asking about Big Data at this point?
Warden: Big Data is not “magic,” but it is a powerful
weapon when pointed at the right target. Insurers need to align the
right weapon with the right target. Predicting events on the fly from
continuous streams of granular data is a Big Data problem. Predicting
when a customer may be on the verge of canceling a policy may not be.
Insurance by and large is not a “high-frequency” business when compared
to say, credit cards. Leaders need to make sure they are not funding the
shooting of rabbits with howitzers—or elephants with peashooters.
No comments:
Post a Comment