►
From YouTube: INSEAD GitLab Case Study: Interview with GitLab CEO
Description
Marco Minervini and Phanish Puranam from INSEAD Interview GitLab CEO, Sid Sijbrandij on remote work.
A
Okay,
so
I
just
will
just
introduce
what
is
the
reason
why
we
are
doing
this
interview.
So
what's
the
main
port
hood
main
purpose
is
to
build
a
teaching
case
for
the
course
with
each
organization
2.0,
and
in
this
course
we
teach
two
elements.
One
one
part
of
the
course
is.
We
teach
MBA
students,
some
core
principles
of
organizational
design
and
this
core
principle
of
organizational
design.
A
We
especially
focus
on
innovative
forms
of
organization,
and
clearly
the
develop
is
with,
is
it's
all
remote
solution
and
other
type
of
solution
they
found
to
the
typical
problem
of
organizing
is
innovative
form
of
organization,
and
the
other
part
is
that
we
teach
some
parts
of
unalloyed
dicks
for
organizational
designs.
We
teach
how
to
use
experiment,
progression,
design
and
machine
learning
and
simulation
to
better
design.
A
A
B
To
big
considerations,
so
first
of
all,
is
startups
versus
the
rest
of
the
world.
If
your
cash
flow
company
is
sort
of
a
start-up,
then
sorry
stock
options
might
not
be
necessary.
So
stock
options
is
an
instrument
for
startups
where
the
value
where
you
create
a
lot
of
value
over
short
amount
of
time.
Your
your
cash
poor,
but
future
value
rich,
that's
probably
obvious
to
you.
B
Then
I
think
there's
a
big
Geographic
divide
between
how
people
value
stock
options.
When
we
try
to
recruit
senior
leaders
in
the
valley,
we
got
laughed
out
of
the
room
and
we
didn't
offer
stock
options.
When
you
recruit
people
in
Europe,
they
they
almost
value
the
stock
options
at
less
than
10%
of
what
the
value
you're
giving
people,
because
they've
never
seen
someone
else
get
breach
of
them.
So
the
how
many
people
are
near
you
who
have
gotten
rich
of
stock
options
determines
how
you
value
them
and
I.
B
Can
totally
imagine
startups
that
give
stock
options
in
the
US
and
now
outside
of
that
at
a
certain
point
we
started
allowing
people
to
choose,
and
predictably
the
people
in
Europe
didn't
get
any
stock
options.
We
discussed
with
our
board
and
our
board
said:
look
we've
seen
this
movie
before
those
people
will
be
very
disappointed
by
the
time
you
IPO
and
everyone
else
gets
rich
of
their
stock
options
and
they
don't
and
it
will.
They
know
it's
their
fault,
but
they
still
will
be
disgruntled.
B
Just
don't
do
it
just
give
everyone
the
same
and
keep
it
simple.
Also,
it's
kind
of
hard
to
keep
track
of
what
option
people
choose.
So
you
run
into
problems
where
you
promote
people
and
you
have
to
like.
We
react
them
and
figure
out
what
what
what
setting
they
choose
and
everything
else.
So
we
opted
to
keep
it
simple
for
prevented
disgruntlement,
it's
already
bad
enough.
The
people
in
the
beginning
get
more
than
people
later
on.
It's
it's
fair,
but
it
is.
B
It
is
an
intuitive
to
some
people,
so
don't
don't
complicate
it
further,
but
I
do
think.
There's
a
big
divide
and
I
continue
to
be
said
that,
for
example,
people
in
Europe
sometimes
talk
about
their
compensation
and
they
don't
even
include
stock
options
as
part
of
that
package,
even
though
it's
an
enormous
expense
to
the
company.
A
B
B
B
B
So
sorry,
like
I,
don't
want
to
I,
don't
want
to
shut
the
conversation
down
like
I
just
I'd
be
happy
to
elaborate
if
you're
like.
Oh,
that's,
we
remote
is
social.
Why
is
that
I'd
love
to
elaborate
it
on
that
from
I'm,
giving
you
strong
opinions
I
feel
free
to
dive
in
more.
If
you,
if
you,
if
you
have
questions,
if
you
ask
me
a
yes-or-no
question,
I'll
answer
it
yes
or
no,
but
to
give
you
a
chance
to
ask
further
questions,
know.
A
A
B
I've
not
seen
a
learning,
but
that
is
practical.
So
right
now
we
don't
have
any
plans
for
that.
We
do
try
to
automate
anywhere.
We
can
an
example
is
right
now,
if
you
create
a
new
job
family,
it
checks
automatically
that
you
put
requirements
in
and
performance
indicators,
any
any
rule
that
you
can
defer
that
you
can
have
be
checked
by
a
computer.
It's
superior
to
a
rule
that
humans
need
to
check
so
anywhere.
We
can
not
bother
humans
with
checking
things
we
try
to.
We
try
to
do
that
and
automate
the
test
in.
A
B
Saving
time
it's
consistency,
it's
mostly
the
tediousness
of
it.
It
get
lab
the
software.
Has
it's
over
a
hundred
thousand
rules,
a
hundred
thousand
tests.
Imagine
that
and
we
have
over
a
hundred
thousand
commits.
So
imagine
someone
had
to
check
those
hundred
thousand
things
a
hundred
thousand
times.
That's
a
lot
of
work
and
we
say
people
that
work.
A
B
So
I'm
not
sure
I
would
want
to
compete
with
an
old,
remote
organization
versus
a
co-located
organization
if
they're
in
the
same
room.
Now,
if
there
are
multiple
rooms,
it
gets
easier.
There
are
multiple
floors,
make
it
easier.
Still
there,
multiple
buildings
get
easier.
Still,
multiple
cities
get
easier.
Still,
multiple
states
or
provinces
get
easier,
still,
multiple
countries,
multiple
continents,
so
the
more
widespread
an
organization
is
the
less
they
benefit
from
being
co-located.
B
While
the
old
remote
organization
will
have
to
adopt
the
processes
you
use
to
control
an
organizational
skill,
namely
writing
things
down.
Recording
things
working
asynchronously.
We
have
to
master
those
things
early,
so
they
become
more
prevalent
in
the
organization
and
hence
they
are
easier
to
scale.
For
example,
we
have
the
best
onboarding
process
of
any
company
in
the
world.
It's
not
perfect.
I
can
see
lots
of
improvements,
but
nobody
else
has
250
tasks.
B
You
need
to
complete
when
onboarding
and
that
works
better,
like
that's
more
useful
at
scale,
because
we
just
hired
600
people
this
year.
They
all
got
to
benefit
from
that.
So
that's
why
I
think
a
remote
organization
skilled
better
and
it
was
counterintuitive
to
our
investors.
They
thought
it
would
break
at
a
certain
point.
Instead,
they
saw
it
working
better
and
better.
C
You
my
question
simply
was
in
these
examples
that
said,
give
we're
looking
at
large
companies
which
are
in
multiple
geographies.
Isn't
it
the
case
that
they
can
occasionally
bring
together
the
people
that
need
to
be
together
in
Peru,
even
if
they
are
across
multiple
geographies
or
buildings
or
groups?
Yeah.
B
They
can
so
can
we,
we
don't
do
it
because
it's
very
inefficient.
Imagine
the
cost
of
this
meeting.
If
you,
if
I,
had
to
fly
to
your
location
or
your
to
mine,
so
they
fall
back
on
extremely
inefficient
things
like
flying
people
halfway
around
the
world.
For
for
our
meeting,
we
don't
fall
into
that
trap.
We
have
meetings
that
are
more
efficient.
B
C
B
Of
course,
there's
benefits
to
physically
co-located
meetings,
it's
easier
to
interrupt
each
other,
it's
easier
to
see
what
everyone
around
the
room
is
thinking.
You
don't
have
the
wireless
issues,
it's
easier
to
look
each
other
in
the
eyes.
It's
easy
to
kind
of
wait
grant.
Before
and
after
the
meeting,
it's
easy
to
take
a
walk
together
or
do
something
fun
together.
There
are
advantages.
B
B
We
have
a
very
high
velocity,
a
meeting,
augmented
with
Google
Docs,
while
allowing
some
people
to
attend
remote
if
they
couldn't
attend,
but
there's
definitely
benefits
I,
don't
think
they're
as
big
as
people
make
them
yes,
I
much
rather
see
someone's
face
than
not.
So
if,
if
this
was
a
hybrid
code,
where
some
people
were
sitting
behind
the
table
and
I
see
like
three
pixels
of
people's
face,
that's
pretty
bad
I
see
your
face.
Pretty
good.
Our
latency
is
pretty
good.
Okay,
we
have
the
disruption
earlier.
There's
a
waste
of
two
minutes.
B
Travel
would
have
been
a
waste
of
two
hours
or
two
days,
so
I
think
I
think
we're
pretty
good.
It's
it's.
The
the
biggest
thing
is
that
sometimes
we
need
to
take
the
initiative
to
interrupt
each
other
because
that's
kind
of
harder
in
a
remote
setting
get
that
we
solved
that
by
making
sure
that
we
put
your
questions
in
a
dock.
So
it's
it's
clear
who
has
a
question
and
who
to
hand
it
off
to.
A
Well,
there
is
that
I
mean
I.
Can
imagine
and
I
completely
agree
with
your
app.
There
may
be
a
lot
of
savings
and
a
lot
of
efficiency
gains,
but
still
I,
guess
that
you
are
facing
some
challenges
in
scaling
up.
What
are
the
type
of
challenges
that
you
are
mostly
facing
and
which
of
them
I'm,
sorry,
which
of
them
are
related,
and
you
guess
that
are
related
when
to
to
the
fact
of
being
horrible.
B
B
However,
it
is
counter
intuitive
to
communicate
changes
to
the
handbook.
The
default
of
people
is
when
they
want
exchange
to
send
a
slack
message
to
send
an
email
to
give
a
presentation
to
tell
people
in
a
meeting
anything
but
making
a
change
to
the
handbook,
because
it
is
slower
for
them.
It
is
quicker
to
do
use
any
other
form,
because
if
they
make
it
in
the
handbook,
they
have
to
find
the
relevant
part
of
the
handbook.
They
sometimes
have
to
adjust
the
handbook
to
make
sure
that
your
change
will
fit
in
afterwards.
B
They
have
to
go
to
a
technical
process
and
a
reviewer
to
kind
of
get
it
and
act
it.
They
have
to
wait
a
bit
for
before
it's
deployed.
It
is
slower
than
any
other
option.
However,
it
allows
people
that
come
with
a
change
after
that
to
build
on
their
change,
so
it
is
when
they,
when
they
take
that
extra
time,
it's
a
foundation
for
the
next
thing,
I
think
of
it
as
Brick
Lane.
So
every
piece
of
information
is
a
brick
with
gitlab.
B
There
is
a
well-structured
house
and
everyone
adds
to
that
one
house
and
we
build.
We
build
a
house
of
how
gitlab
works
and
because
we're
pretty
particular
on
how
we
build
it.
It
has
a
strong
foundation
and
we
can
build
it
very
high
in
every
other
company.
They
sent
the
brick
into
the
heads
of
people.
We
have
now
a
thousand
people
at
gate
lab.
Everyone
is
receiving
five
bricks
a
day
that
they
have
to
add
to
the
house
they're
building
internally.
B
They
have
to
spend
a
lot
of
time
how
to
kind
of
guide
the
different
speaks
with
each
other.
They
forget
things.
Things
are
unclear.
A
lot
of
context
has
to
be
created
because
there's
no
context
around
where
to
place
a
barracks
or
every
big
concert,
the
large
packet
of
instructions
where
to
place
it.
People
interpret
things
very
differently.
D
D
There
was
no
reason
why
in
a
very
concise
place-
and
you
can
so
we
really
frustrating
and
weeds,
but
unless
everybody
in
the
company
discipline
processes
and
publish
when
to
make
changes-
and
you
know
see
history
involved
from
a
process
like
function
standpoint-
it's
really
really
really
really
hard
to
be
remote
at
all.
Much
less.
A
B
B
Hard,
basically
forbid
people
from
making
changes
in
any
other
way.
So
if
you
send
the
email
where
they
changed,
the
the
the
reply
will
be.
No,
you
didn't
just
make
a
change
link
to
the
link
to
the
change
in
the
handbook.
That
is
the
only
that's
a
source
of
truth.
It
isn't
change
there.
There's
no
change.
B
B
A
C
A
B
Yeah
I
think
it
helps
if
you
make
the
tools,
because
you
can
make
your
own
tools
that
help
with
all
remote
and
people
at
the
company
tend
to
be
more
familiar
with
the
tools.
It
is
no
coincidence
that
tool
makers
frequently
discover
new
concepts,
not
comparing
ourselves
in
any
way,
but
like
the
Wright
brothers
were
manufacturing
bicycles,
so
they
are
access
to
a
lot
of
tooling
to
make
that
and
they
make
it
make
like
lightweight
mechanical
structures.
B
B
Think
that
the
tooling
is
getting
the
tooling
we're
making
is
available
to
anyone
we're
not
using
anything.
That's
not
open
sourced,
so
I
think
that
that
necessity
is
dissipating
the
necessity
that
everyone
in
the
company
is
familiar
with
the
tooling
is
still
there
and
actually
we're
kind
of
under
investing
there.
A
B
I
think
all
remote
is
fit
for
all
knowledge
work.
If
you're
dealing
with
hardware,
it
is
not
a
good
fit.
So
if
you're,
Tesla
or
if
you've
run
a
hospital
or
you
have
patients,
it
is
not
a
good
fit,
but
anything
not
Hardware,
related
I
think
is
a
fit
I,
think
you're,
seeing
it
as
software
first
first,
because
it
is
you
it's
much
easier
to
adopt
or
remote
if
you
do
it
from
the
beginning
and
the
fastest-growing
companies
tend
to
be
software
companies.
Second,
software
companies,
people
work
at
software.
B
Companies
tend
to
be
good
with
software,
and
so
there
they
have
an
easier
time
embracing
the
tools,
but
I
have
no
doubt
that
there
will
be
journalists
and
other
things
we'll
start
embracing
this
as
well.
All
the
good
talent
agencies,
anything
anything
knowledge
work
is
digital
and
there's
no
reason
to
be
in
the
same
location
in
the
open,
Open
Office
plan
and
put
your
headphones
on,
like
you
can
just
work
from
home
or
something
near
your
your
place
and.
A
B
I
missed
the
last
part
of
your
question
that
put
instead
struggle
in
going
on
remote,
there's
there's
some
things
like.
If
you
need
high
trust
environments,
we
all
have
reptile
brains.
If
we
meet
someone
in
person,
it's
a
different
take
away
time,
meeting
someone
in
assume
call
so
what
about?
For
example,
banking,
banking
is
a
high
trust
industry.
It
deals
with
money.
Maybe
we
want
to
see
someone
in
a
suit.
I
want
to
see
that
they're
taller
taller
people
tend
to
meet
you
more
successful
in
banking.
We
want
that
power.
B
B
B
Your
audio
wasn't
great,
but
I
understood
the
question
to
be
about
flat
on
hierarchical
organization,
and
that
is
frequently
something
that
technology
CEOs
do
I
think
it
is
totally
unrelated
to
the
concept
of
over
mode
I'll
give
my
quick
opinion
about
it.
I
don't
think
it
makes
any
sense
and
to
run
a
business
in
that
way.
I
think
every
significant
business
that
tried
that
and
many
have
tried,
has
not
had
good
experiences
with
it.
I
think
it's
something
people
would
like
to
believe.
B
It
is
true
that
is
not,
unfortunately,
not
true
I'd
get
that
we
try
to
combine
the
best
of
consensus,
driven
and
hierarchical
organizations.
There's
a
good
link
for
that
that
I
see
shadows
will
drop
in
our
leadership
page.
We
talk
about
how
we
make
decisions,
namely
anyone
can
provide
input
getting
the
best
of
consensus,
different
organizations,
and
then
one
person
makes
the
decision
getting
the
speed
of
hierarchical
organizations.
The
counterintuitive
thing
is
that
the
people
providing
input
has
to
be
okay,
with
the
fact
that
the
person
making
the
decision
will
not
acknowledge
their
input.
A
B
The
high
end,
so
if
yeah,
if
there's
a
proposal
like
frequently
it's
the
person
who
has
to
do
the
work
but
sometimes
there's
multiple
people
involved.
So
there's
a
couple
of
people
who
review
we
try
to
keep
decisions
that
are
taken
by
the
high
frequency
to
like
it,
DRI,
oh,
and
to
ensure
that
they're,
coherent.
A
How
do
you
so
given
that
okay,
the
first
step,
is,
disagree
and
then
commit
so
that's
the
DRI
that
takes
the
decision
and
put
further
the.
How
do
you
intend
and
the
that
there's
a
farther
disagreement
so
the
next
stage
you
have
a
disagreement,
even
though
the
perspective
of
the
others
were
not
thinking
in
the
kinds
of
the
first
stage.
A
B
There's
a
difference
between
not
taking
people's
perspective
into
account
and
making
sure
they
feel
hurt.
I
think
there's
a
very
big
difference
between
the
two
and
it's
very
easy
to
just
look
at
the
perspectives
of
other
people
like
they
will
comment
in
the
issue.
You
can
read
that
making
sure
everyone
feels
hurt
is
a
lot
of
work
yeah
after
an
hour
comment
and
say:
oh
I,
understand
your
point,
but
this
and
that
etc.
B
So,
there's
there's
taking
into
account
versus
make
up
people
feel
hurt
big,
big,
big
change,
so
the
the
commit
is
not
just
for
the
DRI
that
the
arrival
the
site,
but
they
want
so
they
already
committed.
That's
why
the
person
has
to
do
the
work.
We
try
to
make
them
the
DRI,
so
they're
motivated
to
do
the
work
yeah
committing
is
for
everyone
else,
excepting
I
can't
the
other
person
is
the
DRI.
They
think
something
else.
B
I,
don't
think
it's
gonna
work,
but
I'm
gonna
try
to
make
them
successful
and
then
the
disagreeing
is
about
the
people
who
had
an
other
opinion
are
free
at
any
time
to
go
back
to
the
DRI
and
keep
advocating
their
cause.
It's
not
like,
because
the
deer
I
made
a
decision
you
have
to
now
shut
up
and
keep
compliant.
You
have
to
comply.
B
You
have
to
make
as
long
as
the
decision
didn't
change,
you
have
to
make
them
successful,
but
you
can
just
keep
nagging
them
until
the
end
of
time,
because,
if
you
care
so
passionately
about
something
that
is
a
signal.
That
is
a
signal
that
there's
that
there's
something
that
might
have
been
overlooked
and
as
a
DRI,
you
will
have
to
accept
that
people
keep
disagreeing
with
the
decision
and
that's
okay
and
you
don't
even
have
to
write
a
big
apology.
I
don't
have
to
write
a
big
answer
back
every
time,
but
it
is
okay.
B
Yeah
so
give
up
used
to
consist
of
gift
my
version
control
and
get
lamp
CI
so
basically
to
create
a
verify
stage,
a
separate
applications
that
was
all
of
get
lab
to
applications.
That
did
that
and
then
Camille
said
we
should
combine
it
to
and
Dmitry
we
proposed
it
to
said.
No,
obviously
not
these
are
integrated
perfectly
together.
Single
sign-on,
same
concept
of
or
the
user
is,
there's
no
benefit
and
he
kept
he
kept
at
it
and
then
to
meet.
B
You
change
his
opinion
and
they
came
to
me
and
I
laughed
them
out
of
the
room.
Obviously
not
everyone
in
the
industry
has
separate
apps
for
this,
but
customers
want
to
mix
and
match
and
he
kept
at
it.
He
convinced
me
and
we
changed
it
and
we
launched
a
revolution
in
DevOps
where
integrated
products
are
now
starting
to.
B
We
started
that
trend
years
ago
now
our
most
important
competitors
following
that.
So
that's
a
good
example
of
for
and
there
wasn't
a
quick,
reasonable
response
from
both
me
in
the
meeting.
That
was
wrong
and
it
takes
someone
that
that
is
that
has
passion
to
keep
at
it
and
nothing
worse
than
that.
You,
what
are
your
right
or
wrong,
not
being
able
to
speak
anymore,
is
the
worst
thing
not
being
able
to
improve.
Your
argument
is
the
worst
thing.
What
we
do
sometimes
say
is
like
okay.
B
Well,
this
is
the
data
on
based
on
which
these
are
the
assumptions
we
may
come
to
this
decision
like
what
part
of
this
is
wrong
and,
if,
like,
even
if
all
the
traders
are
exactly
right
and
there's
nothing
to
learn,
one
person
comes
to
a
and
one
person
comes
to
B
and
there's
nothing
else.
Then
it's
there's
not
a
lot
to
talk
about,
but
frequently
there's
an
assumption.
You
can
challenge
or
there's
a
test.
You
can
do
other
things,
yeah.
B
A
B
With
iteration
there's
a
bit
of
a
risk
of
that,
but
the
people
who
work
at
gate,
lab
or
so
I,
don't
know
they
care.
They
all
have
ideas
about
the
future.
So
I've
never
seen
someone
do
something
kind
of
boneheaded
where
they
were,
they
didn't
have
a
vision
of
the
future.
Everyone
has
a
vision
of
the
future.
The
big
problem
is
trying
to
like
take
that
leap,
and
even
if
you
go
into
a
short-sighted
solution
its
if
you
take
only
a
few
steps,
it's
probably
easy
to
dig
you
out
of
that
hole.
B
The
big
mistakes
are
when
you
spend
a
year
making
something
amazing
and
then
that
pants.
That
is
a
much
bigger
problem
and
I've
used
to
be
very
skeptical,
for
example,
on
driving
on
key
performance
indicators,
because
I
think
I
thought
people
would
take
shortcuts
with
the
clarity
that
they
provide
so
far
far
surpasses
the
risk
of
adverse
behavior
and
perverse
incentives.
A
B
A
B
Of
one-way
and
two-way
doors,
so
if
it's
a
two-way
door,
you
can
just
return
like
don't
even
think
long
about
the
dishes
in
Disco.
This
is
decision
to
integrate
the
two
applications.
Together
we
talked
about
earlier.
It's
a
one-way
door
like
that
is
that
was
many
months
of
engineering
work.
That
was
a
very
big
deal,
so
in
that
case
you
try
to
do
more.
B
Homework
collect
more
data,
get
more
opinions
debated
longer
to
make
to
make
sure
you're
not
doing
that,
so
the
trick
with
iteration
is
key
splitted
up,
but
very
frequently
you
can
split
up
things
more
than
you
think.
I
just
held
my
first
iteration
office
hours
this
week,
three
people
came
with
a
thing
they
fought
was
beyond
splitting
up
and
in
every
case
they
walked
away
a
little
I-car.
A
A
B
For
example,
the
data
analysts
like,
should
he
be
located
centrally
in
the
finance
department,
or
should
every
division
have
a
bit
of
a
few
data
analysts
and
I
think
that
the
answer
we
come
to
is
like
as
long
as
we
can
get
them
a
good
manager
that
really
understands
data
analysis
we'd
be
happy
to
place
them
in
the
different
parts
of
the
company.
If,
if
that's
not
the
case,
we
place
them
centrally,
but
we
dedicate
them
to
a
certain
division
to
make
sure
that
every
everyone
gets
their
fair
share
of
time
and
effort
and.
B
I
think
that
happens
in
two
different
ways.
First,
is
our
our
KPIs
and
our
KPI
sirs
there's
some
top
goals
and
they
kind
of
trickle
down
some
of
the
calls
that
are
repeated.
So
these
are
the
top
goals,
and
some
of
these
are
like
sales
related.
Some
of
these
are
people
related.
Some
of
these
are
engineering
related,
and
then
these
are
to
go
the
top
goals
of
sales,
marketing
etc.
So
that's
how
they
kind
of
trickle
down
now
for
engineering.
B
D
B
C
Thank
you
so
I
have
two
related
questions
around
this.
One
is
the
back
turn.
A
lot
of
misinformation
is
transparently
visible
to
external
people
outside
the
company.
Is
that
there's
clearly
a
conscious
decision?
Why
would
you
take
that
decision
and
second,
do
you
think
this
will
change
post,
IPO,
yeah.
B
So,
on
the
bottom
of
our
strategy
page,
it
quotes
Peter
Drucker,
who
I
assume
know
do
you
know
who
that
is,
and
we
try
to
have
fewer
assumptions.
It's
very
hard
and
he
says
strategy
is
a
commodity.
Execution
is
an
art,
so
I
subscribe
to
that
very
well,
it's
not
about
having
the
the
best
strategy.
The
strategy
tends
to
be
obvious.
It's
about
executing
the
best
and
to
execute
the
best.
It
needs
to
be
super
clear,
nothing
is
cleared
and,
like
writing
it
down
explicitly
and
having
the
whole
world
be
able
to
amend
it.
B
People
need
to
be
aligned
to
it.
How
can
you
have
people
on
line
to
your
strategy?
If
you
only
tell
it
after
you
hired
them,
our
strategy
is
very
important.
You
need
to
be
aligned
to
it.
I
will
only
tell
you
after
you
join
the
company,
I,
think
that's
bananas,
so
every
company
in
the
world
does,
if
your
strategy
is
really
important,
you
want
to
make
sure
people
buy
in
before
they
join
and
they
that
people
that
don't
buy
in
don't
actually
join
the
company.
B
B
B
I
think
that's
an
excellent
question.
I,
don't
know
the
answer
to
that.
I.
Think
there
there's
a
couple
of
things.
First
of
all
you
can
make.
You
can
have
a
lot
of
information
you
put
out
as
a
company
as
long
as
you
disclose
the
channels,
the
avenues
you
do
so
upfront
so
we'll
have
what
have
a
lot
of
those
and
then
reg
FD
about
a
material
disclosure.
It's
probably
the
most
important
one.
B
So
we
got
to
figure
out
what
that
means
to
the
cut
for
the
company,
and
we
got
to
get
very
explicit
with
examples
and
training
materials.
Now,
there's
a
lot
of
things
that
are
not
material,
anything
that
has
to
like
as
a
first-order
effect
on
revenue
and
profit,
so
costs,
and
that
will
be
subject
to
that.
But
there's
a
lot
of
things
that
have
indirect
effects
and
although
there
will
be
hesitancy-
and
there
will
be,
people
will
want
to
refer
to
what
is
common
in
other
companies.
B
I
think
in
the
end,
there's
there's
a
you:
want
public
companies
to
disclose,
fix
and
I
believe
that
the
regulator's
in
the
US
are
very
wise
and
I
think
they
will
recognize
that
investors
in
give
up
or
get
more
information
than
ever
before,
allowing
them
to
make
a
better
empowered
decision
and
I
think
that
that's
gonna
be
a
very,
very
tough
thing
to
do.
But
III
would
love
to
work
with
the
regulators
to
to
get
there
together
and
we'll
up
we'll
see
how
far
we
get,
and
these
are.