Don’t Listen To Your Customers – Do This Instead | Kristen Berman | TEDxBerlin

this is one of my favorite charts in
social science some companies have close
to 50% of their employees saving for
retirement but some have over 90 percent
employees on the right will be sipping
mimosas on the beach in their retirement
communities
well employees on the Left will likely
be hanging out in their cubicles well
into their 70s so what explains the
difference between these two types of
companies that could lead to such a
massive gap raise your hand if you think
it's level of Education raise your hand
if you think it's income raise your hand
if you think it's age it turns out it
has nothing to do with the usual
suspects
let me ensure you - Carla Carla makes
under thirty thousand dollars a year
she works at a retail clothing shop she
has an eight-year-old son and things are
tight but Carla is saving for retirement
she is on the right this secret for
Carla and others like her is the form
that she fills out when she first gets
hired in this chart employees on the
right are automatically enrolled into
retirement savings
it was the default they did nothing and
they start saving people on the Left had
to fill out a form themselves they also
did nothing but instead they didn't
start saving I'm a behavioral scientist
I've worked for the last decade in
Silicon Valley helping hundreds of
companies build products that change
behavior for good when we ask people why
are you saving for retirement over 80
percent say because I care about my
future
it sounds great it's just flat-out wrong
could you imagine answering an interview
question with because it was the default
of course not and only 6% of people do
as a behavioral scientist I study the
gap between what people say they will do
and what they actually do most
organizations today design products by
talking to customers they ask some
questions with interviews they do
surveys they do focus groups and it
sounds logical right the problem is it's
not working financial providers should
help us with overspending but our bank
accounts aren't growing health care
providers should help us with overeating
but our waste lines are definitely
growing social media should help us
connect but we are lonelier than ever
today I'm gonna give you a sneak peek
into my work we'll unpack why we study
what people do and not what they say
we'll go into how to actually understand
behavior and build solutions that help
people be happier healthier and
wealthier traditionally companies ask
three types of questions to understand
you their customer they ask you about
the past what have you done
they ask you about the future what will
you do and they ask you why you do
something but can we trust the answers
to these questions maybe less than you
think let's take each one and figure out
if it could be misleading us so first
the past what have you done
so imagine you work at a hospital and
want to avoid the unnecessary spread of
infection increasing hand-washing is a
proven solution so if you ask people to
report their past hand-washing behavior
can you trust what they say so let's try
this how many of you forgot to wash your
hands the last time you went to the
bathroom please raise your hands and
avoid your neighbors face so are you
telling the truth well we put cameras in
the bathroom to figure out who could be
lying just kidding but it is easy to see
that you may lie when asked this
question in public but even in a private
survey people say they forget to wash
their hands only one in a hundred times
they go to the toilet people tend to
underreport
questions about the past behavior for
two reasons first it's just not socially
acceptable to tell the truth so do you
text and drive of course not
do you eat chocolate in the morning
before work I have never done that do
you forget to wash your hands nope
people often answer what they think
everybody else will answer we could
never fix hand-washing if we built
solutions based on those answers and
second we just forget you've likely been
to the bathroom 2,500 times do you
really remember the 2470 first time we
cannot report accurately if we do not
remember so let's take the second type
of question questions that ask us about
the future what will you do so imagine
you're a healthcare organization that
helps people with high blood pressure or
hypertension patients that fail to take
their medication on a daily basis have
three times higher risk of dying of
stroke compared to patients who take it
properly so as an organization you may
want to ask people if they need help
remembering to take the
life-saving medication in our study 58%
of people said that they don't need any
help to take this medication they think
if taking the medication is important of
course I will do it
sadly hypertension adherence is famously
low 50% of people stop taking it
entirely within the first year how could
people get this so wrong they say they
don't need any help but clearly they do
when asking people about their future
behavior they are imagining an ideal
self in the future I am a perfect person
in the future of course I will take my
life-saving medication they forget that
in the present moment
life is happening you have to take your
kid to work early or there's a work
fire-drill going on people cannot
predict their future accurately because
they cannot forecast all of the other
things going on in their environment and
how they will react let's take another
example so a large company in Los
Angeles wanted to encourage their
employees to eat healthier in surveys 60
percent of people say that they would
take an apple if it was put out so the
company began putting out free apples in
their cafeteria and what happened nobody
took the apples so the company called us
to unpack what was going on
first we observed their lunch hour so
sure enough apples were prominently
displayed but after two hours no apples
were taken but after one look at their
cafeteria we figured out what was going
on the apples were competing with the
fries
when asking people questions about their
future behavior organizations assume
that our responses represent strong
preferences sure we like apples but if
the environment isn't conducive to
taking apples they're not cut up or the
next to the greasy fries then we won't
take the Apple interview questions are
asked in a vacuum as if the environment
of decision making has no effect on our
behavior so unless we could replicate
the exact environment of decision making
we cannot trust people to tell us what
they will or they won't do what about
the third type of question questions
that ask us why we do something these
types of questions are the most
misleading this is the Carla why are you
saving for retirement question when
asked people say that they are saving
for retirement because they care about
their future and it's true they do care
about their future but that is
completely unrelated to why they are
saving or not saving they choose to save
because it was the easiest option at the
time it was the default we act as if we
are highly rational people and this is
reflected in how we answer questions I
thought about this deeply and I
considered all the options and I made
the best decision for me it's unlikely
we'd say something like I was just too
lazy to uncheck that box or I did
whatever else is doing or I actually
have no idea
and yet these statements are closer to
the truth so if we want to figure out
why somebody does what they do we cannot
ask them but if we can't trust people to
tell us the truth how do organizations
understand behavior and build solutions
that help people
so the answer is behavioral design
behavioral design uses insights from the
field of behavioral science to inform
design decisions behavioral design is a
three-step process first we do a
behavioral diagnosis this is a literal
map of every single step every single
action somebody must take to get to the
desired outcome and then we identify the
psychological biases at play when making
a decision this could be social norms or
loss aversion or optimism bias these
biases affect our decision-making and
third we study what people do and not
what they say this is a real-world
experiment we've done behavioral design
with companies like Google Aetna PayPal
and many others let's take an example so
student loan debt is crushing us college
students however two million students
we're eligible to apply for free
financial aid and did not apply when you
ask students why didn't you apply they
say lack of information about financial
aid eligibility
however when Harvard researchers didn't
experiment they found that giving
students information about financial aid
eligibility had no effect on behavior
change this is a domain in which asking
people questions will surely lead you
astray our team was tasked with
increasing applications and we used
behavioral design the first step we did
a behavioral diagnosis and found out
that it was over 20 steps and close to
an hour for students to complete this
application second we identified that
cognitive overload was the bias most
present in student decision-making
weighing back-and-forth the option to
apply or not apply
created this cognitive overload that
pushed the decision off until tomorrow
tomorrow came the student still have
decided and the deadline passed and
finally we ran an experiment to reduce
cognitive overload for half the student
population we texted them that they
could apply and a deadline for the other
half we texted them letting them know
that filling out the financial aid form
was now a part of the enrollment process
we didn't force students to apply but we
removed the decision deadlock by making
it part of the standard process for
people who did not apply the prior year
our experiment tripled the likelihood
that a student would apply for financial
aid if this is rolled out an additional
two hundred and thirty thousand more
students may apply and get financial aid
this is behavioral design from diagnosis
to identifying bias to experimentation
we create meaningful behavior change
because instead of asking people why
didn't you apply we watch what they do
and shape the environment to lead to the
best outcome the stakes are high so
people want to eat healthy they want to
save for retirement they want to take
their pills but they don't because in
nearly every domain solutions are still
being built based on interviews surveys
and focus groups we need to move from
relying on asking people questions and
move to studying what they actually do
this is behavioral design this is the
future for organizations that want to
make us happier healthier and wealthier
thank you
[Applause]