How do you fix a category where the business and customer seem
to have distrusted each other for over 300 years? Through radical
transparency, says Yael Wissner-Levy at insurance challenger
Lemonade, aided by behavioural economics and AI.

Founded in 2015, Lemonade has US$180m of financial backing;
currently in 21 US states, it plans to expand to Europe in the near
future. It might be easy to characterise AI as dehumanising, but
here Lemonade shows how it can be used to help both sides have
a more Real & Human relationship with each other.

What was the opportunity for a challenger in the insurance category?
Both of Lemonade’s founders didn’t know a thing about insurance.
They were both veterans of the tech industry, but it was precisely
their ignorance of insurance that allowed them to think about it
in a truly transformative way, incorporating technology and AI
on the one hand, and behavioural science research on the other.

As the founders started looking at the insurance industry they
found that 25% of Americans feel it’s okay to defraud their insurance
company. When you look at that number and you look at who those
people are, they aren’t criminals. They’re people like you and me.

Then you start thinking to yourself, ‘Wait, have I ever rounded up
when I’m filing a claim?’ or similar white lies that we like to tell
ourselves. We’re all victim to these white lies, and also the
perpetrators of these white lies. It’s a victimless crime.

What happens in traditional insurance is the insurance company
has its eye solely on the profit. Every time they pay you a claim
it’s less money for them. This fuels a very unfortunate series
of events, what we call the vicious cycle, in behavioural
economics speak.

Both the insurance companies and the policyholders are not bad
people. It’s the system that brings out the worst in these people.
And if you change the system, if you create one that will work
to bring the best out in people, their behaviour will change.

What is different about the Lemonade model that is helping change
those behaviours?
Part of building a system that brings out the best in people is
transparency. It’s a buzzword and you see it everywhere. It’s the
new ‘disruption’. Everyone was disrupting a few years ago, and
now everyone’s transparent.

On a communication level we practice real transparency, and try
to use that to the fullest, even if it’s uncomfortable: we talk about
our mistakes, talk about things that we could have done better.
True transparency, in an industry that has been lacking it, is refreshing.
On the product and business model level we’ve built a system that
will bring out the best in people.

It’s transparently shown to you that we treat your money as yours,
not ours. We take a flat fee from your premium. Everything else goes
towards your claims. We call it your money. If there’s money left over,
we ask you to choose a cause to give back your money to.

When you make a claim, you sign an honesty pledge and film a video
of yourself describing the incident. Trust is the key here because
it’s trust between ourselves and our customers, and trust within our
customers themselves, because the Giveback works in a way that
is built on trust. As a policyholder of Lemonade, I’m actually joining
a community of people who are committed to the same cause,
and they’re going to behave in a non-fraudulent way.

Are you seeing that shift in your customers’ attitude to the category?
Here’s a cool example that surprised the insurance executives that
work with us on our team – it started during the first few months when
we began paying claims, and now it’s something that happens all the
time. We started seeing policyholders who had their claim paid and
then found their stolen jacket or laptop or whatever, write back to
us and ask us if they can return the money. They got the US$1,000
that they claimed for, but they’d like to return the money because
they’d found the item. We had overpaid people who had overclaimed
and they’d say, ‘Is it okay if we return it? Who should we write the
cheque back to? We feel bad taking this money.’ The insurance
executives on our team said this is rare behaviour and they had
hardly ever seen it before.

How do you use AI within the company?
The idea for Lemonade was to design insurance for the 21st Century –
the easiest and simplest way for people to get a policy that matches
their needs. When they need to make a claim, we have processes
there that can be automated, without the paperwork and hassle.

This is not to say that humans are not part of the ecosystem or not
needed. Quite the opposite. It just means that there are a lot of things
that humans are doing today that can be done better and faster by AI.

How can a company embrace AI as a part of the customer experience
and still be a Real & Human brand?
We spend a lot of time working on microcopy in chat, for instance,
so it has to feel human.

Our bot Maya is based on a human person on our team, as is Jim;
it’s like the exoskeleton of them. Both are bots, but their tone of voice
and their capabilities are human-like. Maya is able to get you a policy
customised for your specific needs in seconds. Jim can review,
approve, and pay your claim instantly.

But not everything can be dealt with by the bot. And the minute
it cannot, it escalates immediately to humans.

No matter if you speak with a bot or a human, everything – even the
smallest of details – puts the customer at the very centre. People who
have been denied claims after they’ve spoken with a member of our
human claims experience team have said, ’You know what? Thank you.
It was just such a good process. I know my claim wasn’t approved,
but this was amazing.’

And that, for us, is everything.