►
Description
Tim from the Expansion team hosts a review session of 30 ideas to generate small easy experiments focused on inviting members to GitLab and upgrading customers to higher-paid tiers.
Issue - https://gitlab.com/gitlab-org/gitlab/-/issues/239009
A
All
right-
and
I'm
also
going
to
share
my
screen
just
so
that
way
we
can
kind
of
work
through
this
together
and
get
things
going
if
I
can
figure
out
how
to
get
rid
of
this
cool
all
right.
So
so
today,
and
first
of
all,
just
thank
you
for
taking
the
time
out
of
your
day
to
sit
down
and
and
look
at
a
bunch
of
different
experiments
with
me.
I
think
it'll
be
pretty
fun,
looks
like
jeff's
joined
it
too.
A
The
phil
and
I
were
talking
to
one-on-one
and
he
was
like
what,
if
we
just
come
up
with
the
smallest
little
ideas
on
the
back
of
you
know
that
ask
me
anything
with
lin
lincha
from
airbnb,
and
we
thought
like
instead
of
working
on
these
bigger
workflows,
which
we
love
doing
and
we've
seen
some
pretty
good
results.
But
there's
like
there's
like
a
bunch
of
little
things
that
we
could
probably
learn
a
bunch
of
stuff
really
really
fast.
If
we
were
to
to
kind
of
go
down.
A
What
are
the
ideas
that
we
could
ship
within
one
to
two
weeks
very
quickly
and
then
instrument
and
learn
a
bunch
of
stuff,
so
that
way
we
can
use
that
to
kind
of
drive
additional
work
in
the
future,
and
so
this
is
kind
of
the
first
step
we
we
generated
a
list
we've
got
over,
looks
like
we've
got
just
about
30
ideas.
My
goal
is
to
get
the
30
and
we
we
did
that
we
got
a
lot
of
input
from
the
team.
So
I'm
grateful
for
that.
I
know
jackie
and
mate.
A
You
all
had
a
pretty
good
made,
some
pretty
good
contributions
here
and
the
list
isn't
done
yet,
but
what
I
want
to
do
today
is
actually
just
go
through
what
we
have
and
and
basically
just
do,
a
feasibility
review
and
talk
more
about
okay.
Does
this
idea
make
sense
for
us
as
a
group
and
then,
if
it
does,
is
it
something
that's
small
enough
that
we
can
work
on
quickly
and
what
I'm
gonna
do
is
drop
all
of
these
experiments
right
into
this
this
sheet
and
then
I'll
get
them
organized.
A
A
We're
gonna
focus
on
column
c,
which
is
feasibility
and
we'll
worry
about
the
other
stuff
later.
If
we
can
we'll
talk
about
g,
just
with
some
of
the
I
feel
like,
we
have
a
good,
a
good
like
a
good
group
on
the
phone
with
doug
and
mate,
to
also
talk
about
effort
so
like
if
we
were
to
do
a
handful
of
these.
What
is
the
actual
effort
associated
to
it?
So
I
figured
we
could
just
kick
it
off.
A
B
A
Is
so
the
purpose
of
this
one
is
just
for
us
to
identify
the
ideas
that
we
want
to
then
go
further
explore
what
I'm
going
to
do
after
this
jeff
is
take
all
of
these
ideas
and
then
go
to
work
on
is
the
day.
Is
the
data
available?
If
it's
not,
how
easily
can
we
get
it,
and
if
we
can,
what
do
those
success?
A
And
then
we
can
prioritize
a
list
for
13
5,
13,
6.,
okay,
makes
sense,
yeah
cool,
all
right,
so
row,
two
we've
got
an
idea.
It
says
basically
link,
but
this
one.
This
is
an
idea
that
I
jotted
down.
It's
it's
basically
around
link
placement
for
invite
members.
I'm
thinking
that
we
could
just
simply
add
an
invite
members
link
throughout
the
ui
and
see
which
one
gets
the
most
clicks.
D
D
B
A
A
D
I
think
this
one
makes
sense
if
combined
with
the
first
one,
particularly,
we
could
be
running
how
many
experiments
can
we
run
at
the
same
time,
if
we
can
do
five
ten
of
them,
we
could
run
a
couple
of
these
combinations,
different
combinations,
different
combinations
of
copy
and
placement
and
see
which
one
works.
Well,
I
wouldn't,
I
would
expect
more
from
placement
experimentation
than
copy
experimentation,
but
I'm
I'm
you
know,
that's
allow
ourselves
to
be
surprised.
I
guess.
A
B
I
just
want
to
say,
like
that's
where
our
metrics
come
into
play
like
how
many
experiments
we
can
run
is
entirely
dictated
by
our
conversion
rates
and
our
population
sizes
and
so
like.
If
we
have
something
as
a
30
conversion
rate,
that's
a
lot
easier
to
run
multiple
experiments
compared
to
something
has
a
two
percent
conversion
rate,
so
it
just
it
all
depends
on
like
what
your
metrics
is
and
how
many
people
are
going
through
a
given
experience.
A
So
that
way
we
have
a
good
framework
before
we
kind
of
just
do
a
bunch
of
work,
but
I
I
just
want
to
get
the
list
the
list
going,
so
the
third
one
is
were
there
any
other
ones
on
the
second,
any
other
comments,
no
cool
all
right.
So
the
third
idea
was
that
we
could
just
basically
run
an
experiment
on
links
or
buttons
see
which
one
gets
the
most
clicks.
D
A
D
C
A
Yeah
I
like
kristen
pattos
over
at
trip.
She
and
I
went
back
and
forth
on
it
a
bunch
of
times
because
the
the
book,
the
book
now
button
for
them,
is
just
it's
hideous,
but
it
it's
the
one
that
people
click
the
most
like
they.
They
tried
all
different
kinds
of
colors
and
it
went
back
to
the
bumblebee.
A
I
just
thought
it
was
kind
of
cool,
but
it
could
be
completely
different
for
a
completely
different
reason.
Consumer
buying
behavior
is
probably
different
than
devops
usability,
so
I'm
going
to
just
say
no
we'll
just
we'll
just
call
it.
No
because
it
it
doesn't
sound
like
it's
a
good
fit.
It
would
take
some
some
overhead
to
actually
have
that
conversation
and
and
and
and
convince
our
our
ux
leadership
to
to
allow
us
to
basically
go
against
the
system
that
we've
designed.
C
A
A
D
Yeah-
and
this
is
the
one
that
we
are
currently
working
on-
or
at
least
I
am
right-
I
kind
of
just
wrapped
it
up
today.
I
asked
you
and
phil
to
review
it
if
you
think
it's
ready
and
what
we'll
do
is
we'll
just
show
a
model
so
that
we
don't
have
to
go
and
build
the
whole
thing
of
allow
accepting
invitations
from
that
owner
of
the
group.
We
just
say
hey.
This
is
not
ready
yet,
but
we
we
can
measure
how
many
people
click
it
right
and
you
can
click
that
link.
D
A
D
Yeah
and
it
actually
it's
actually,
it
actually
helps
because
it
makes
it
a
bit
more
useful
right
because
at
least
we
tell
the
new
users
hey.
You
need
someone
else
to
invite
new
members
for
you,
because
you
can't,
at
the
moment,
right
yeah
right
now,
there's
very
little
chance
that
users
will
learn
about
that.
Basically,.
A
D
B
Members
I'd
be
very
wary
of
just
toggling
this
and
not
giving
admins
a
notice,
but
also
an
ability
to
toggle
it
off,
because
I
can
imagine
very
real
situations
where
people
don't
want
reporters
say
to
be
able
to
invite
anyone
to
a
group,
because
reporters
might
be
loose
affiliated
people
non-organizational
people
like
just
people.
You
do
not
want
to
have
that
power.
D
A
Approve
it,
that's
why
I
find
this
idea
so
interesting,
because
I
feel
like
we
could
test
it
with
just
we
could
test
it
very
easily,
a
number
of
different
ways.
One
would
be
to
do
the
very
conservative
way,
which
is
what
the
path
that
mate
is
going
down
and
then
there's
a
more
aggressive
way
to
just
see
if,
like
what
does
like.
What's
the
what
of
the
lower
permissions
that
are
provided,
who
are
the
people
who
invite
the
most
users?
A
B
B
A
Thank
you
for
bringing
me
back
to
reality.
Yes,
you're
right,
but
man
is
it
interesting,
okay
on
to
the
next
one?
So
what
if
we
were
to
message
one
member
groups
to
invite
more
members,
and
do
you
think
we
could
try
this
if
we
can't
do
it
with
segmentation
through
our
email
marketing
through
marketing
ops?
B
So
we
can
do
this
with
email.
I
actually
know
exactly
how
we
could
do
this.
I
mean
marketing
probably
has
their
own
way,
but,
like
I
could
literally
do
this
with
a
user
pull
and
qualtrics,
not
that
I'm
saying
we
should,
but
I
know
how
it
would
be
definitely
possible.
This
is
not
going
to
be
possible
within
that
messaging.
We
do
not
have
these
kind
of
targeting
abilities
within
that
messaging.
B
C
C
B
A
B
I
mean
marketing
should
probably
do
it,
because
I'm
sort
of
using
a
tool-
that's
not
meant
to
send
pure
marketing
emails
to
do
so,
but
I
just
I
like.
I
can
connect
the
dots
and
I
know
that,
like
I
know
how
to
pull
this
data
from
the
warehouse,
I
know
how
to
generate
a
list
of
users
and
then
we
could
send
an
email
to
those
list
of
users
so,
like
I
think
we
have
all
the
steps
we
need.
Yeah.
A
D
E
A
Yeah,
I
think
you're
right
though,
because
it
is
inside
of
the
app
if
they're
not
really
like
on
landing
pages,
unless,
of
course
it
was
an
empty
state
that
we
were
looking
at,
but
when
you
think
of
like
the
the
pages
that
get
iterated
on
with
these
huge
buttons
and
they
are,
they
do
tend
to
be
tacky
and
they're.
Typically,
you
know
one
one
to
two
sentences
that
typically
run
on
to
one
another
and
then
there's
just
like
a
big
huge
button
for
people.
It's
the
only
thing
that
they
can
do
on
the
page.
A
A
Yeah
yeah,
it's
pretty
cool
all
right,
so
the
next
one
I've
got
here
is
we're
doing.
Okay
on
time,
we've
got
to
keep
moving
quickly,
so
I'm
proud
of
us
at
an
invite,
add
an
invite
member
cta
on
features
that
add
the
most
value
on
users
when
used,
in
collaborations
with
mr
pipeline
value
stream
management
et
cetera.
A
D
E
A
A
B
This
seems
a
little
like
a
different
situation
because,
like
with
invite
members
like
we
have
an
invite
members
functionality,
they
just
don't
have
access
to
it.
This
is
something
that
doesn't
exist
and
so
we're
populating
the
ui
with
something
that
explicitly
doesn't
exist,
which
I
don't
love
and
from
like
a
user
sentiment
perspective.
I
might
be
worried
about
like
if
we
start
down
this
road
of
like
we
added
a
new
thing
like
surprise,
it's
not
real.
B
C
E
C
B
Yeah
I
mean,
I
think
in
general.
We
could
investigate
because,
like
how
I've
seen
people
do,
it
is
like
you're,
not
necessarily
inviting
users
in
the
conceptual
sense
you
have
some
sort
of
auth
provider,
whether
it's
active
directory
or
ldap,
or
whatever
and
you're
almost
connecting
the
two
in
many
cases.
So
like
there's
that
so
yeah.
B
I
think
this
is
maybe
worthy
of
wider
study
because,
like
maybe
csv,
is
the
right
way
to
do
it
maybe,
like,
I
think
about
like
a
lot
of
survey
tools,
give
you
like
a
bulk
answer:
functionality
where
they
just
give
you
a
huge
text,
input
and
it's
basically
one
answer
per
line,
and
you
just
paste
in,
like
you
know,
eight
lines
of
answers
and
then
they
become
individual
answers
like
that
could
be
something
that
we
could
expect.
I
think
there's
like
a
lot
of
possibilities
here
and
I
think
it's
worth
doing.
A
Okay
or
give
it
a
yes
and
explore
it
a
little
bit,
I'm
going
to
write
up
issues
for
all
the
all
the
ones
that
end
up
with
the
s
and
even
some
of
the
ones
that
just
take
that
are
bigger
than
the
bread
box.
We're
trying
to
fill
right
now,
all
right.
So
next,
one
is
add
the
invite
members
option
to
the
quick
menu
which
is
the
plus
icon.
A
D
A
A
So
hypothesis,
we
believe
that
users
will
be
more
likely
to
see
value
in
gitlab
and
eventually
upgrade
they
invite.
Colleagues
to
the
platform.
However,
we
believe
the
new
users
they've
been
kind
of
challenging
to
figure
out
how
we
actually
invite
a
colleague
and
we're
talking
about.
I
wonder
where
we're
talking
about
doing
this.
A
A
D
A
All
right
all
right,
let's
shift
gears
so
now
we're
let's
hit
the
reset
button.
So
we're
not
thinking
about
inviting
members
anymore.
Now
we're
going
to
shift
gears
to
the
focus
area
of
upgrading
customers
so
or
context
or
context
shifting
going
on
here
midstream,
I
think
we're
all
capable
of
doing
it.
The
first
one
for
for
upgrading
customers
is
to
is
the
link
placement
for
upgrading
customers.
A
A
A
It's
but
we
have
to
so
forget
about
that.
Like
we'll,
we
got
to
figure
that
out
and
quite
frankly,
I'm
like
exhausted
thinking
about
it,
because
that's
where
the
majority
of
our
business
is
and
we
need
to
figure
it
out.
So,
even
if
it's
something
that
we
just
test
and
we
look
at
the
numbers
in
aggregate,
then
we
just
have
to
be
okay
with
it
yeah
sorry,
I
don't
want
to
get
all
but
yeah,
so
I
think
I
think
this
one's
feasible
doug
mate.
A
This
one,
so
it's
similar
to
the
one
that
I
wrote
above
it
this
one
being
so
jeff
to
your
point
like
self-managed
and
sas
very
different
from
an
upgrade
flow
perspective,
and
I
feel
like
the
documentation
that
we're
linking
people
to
is
different
so
like
the
destinations
of
those
links
and
of
those
buttons.
I
feel
like
we
could
test,
and
I
bet
you
they're,
going
to
be
different,
we'll
see
higher
conversion
on
on
self
managed
for
for
different
destinations.
A
B
Yeah
this
is
similar
to
the
one
I
was
speaking
to
earlier,
like
as
long
as
this
information's.
In
the
data
warehouse,
we
can
pull
a
list
of
users.
The
one
thing
that
I
kind
of
forgot
last
time
that
we're
going
to
need
to
make
sure
we
consider,
because
normally
for
my
purposes,
I
don't
have
to
pay
attention
to.
It
is
the
marketing
opt-in.
So
we
would
need
to
scrub
this
against
the
marketing
opt-in,
because
this
would
be
marketing
communication,
and
that
is
something
that
involves
marketo,
so
that
adds
a
wrinkle.
B
D
A
D
Yes,
so
in
if
you're
on
bronze
in
the
sidebar,
you
won't
see
any
links
to
the
security
stuff
right
and
if
you
are
on,
I
don't
know,
I
know,
is
it
gold?
I
don't
know
which
one
it
is
where
you
actually
get
security
features,
but
if,
if
you're
on
gold-
and
you
have
the
security
features,
but
you
haven't
set
up
anything
go
there
and
you
have
an
empty
state.
D
D
Yeah,
there's
that
and
there's
always
the
balance
issue
right.
How
can
you
balance
the
amount
in
that
particular
cyborg,
because
we
know
it's
quite
dense
already,
so
we
we
could
run
an
experiment
or
two,
but
if,
if
it
turns
out
to
be
successful,
I
think
we
would
need
to
have
a
logic
behind
it
and
say
these
behaviors
tell
us
that
these
users
or
this
team
of
users
would
probably
benefit
from
this
feature,
and
that's
when
we
would
show
that.
B
Yeah
long
term
I
mean
one
one
way
I
could
see
we
do
it
is
we
give
admins
and
group
owners
the
ability
to
sort
of
toggle
like
joe
features
not
available
on
my
plan,
and
so
if
they,
if
they
don't
want
this,
they
could
the
same
way
they
can
toggle
off
for
a
given
project
or
group.
Individual
features
like
if
you
don't
want
your
project
to
have
ci
you
can
just
toggle
ci
off
and
it'll,
take
it
out
of
the
nav
so
similar
to
that.
B
We
could
do
a
lot
in
the
ui
to
just
help
them
understand.
Yes,
roadmaps,
for
instance,
are
a
thing
we
have
and
show
something
on
that
empty
state,
page
of
like
and
here's
what
you
would
get
here's,
how
they
work
so
on
and
so
forth,
and
that
to
me
is
is
valuable
like
I
would
want
that
like
if
I
were
actually
needing
road
maps
for
something
other
than
like
my
silly
personal
projects
like,
I
would
probably
upgrade
to
get
that,
but
right
now
we
just
make
that
discovery
really
challenging.
A
You
know
it's
yeah,
so
I
think
it
is.
I
think
it's
feasible
and
I
think
there's
a
couple
different
ways
that
we
could
do
it
and
I'm
jumping
to
problem
solution
in
my
mind
already,
but
even
it
was
just
a
small
little
tool
tip
that
sits
over
the
the
leading
feature
of
a
stage
and
then
it
shows
the
features
that
are
associated
to
that
tier
and
then
the
other
features
that
are
available.
I
mean
there's
so
many
different
ways
that
we
could.
C
A
A
A
B
So
I
think
one
thing
I
just
want
to
call
it.
We
want
to
maybe
like
time
box
this,
because
right
now
we
have
a
nurture
stream
email
that
has
a
video
link
in
it
and
that
video
is
approaching
two
years
old
and
if
you
watch
it,
it
shows
a
version
of
gitlab.
That
does
not
look
like
what
you
see
today,
and
so
I
think
we
just
need
to
be
mindful
that,
like
that
video
or
that
email
or
whatever
is
gonna,
have
a
shelf
life.
A
A
E
A
A
B
A
A
A
A
A
B
A
Exactly
because
that's
point,
when
you
watch
all
of
catherine's
videos
and
the
video
that
jeff
did
about
the
sus
and
usability
from
a
nav
perspective,
it's
people
just
get
lost
in
it.
They
don't
know
where
they
are,
whether
they're
in
groups
or
they're,
on
their
their
name
like
their
personal
name,
space
or
in
their
projects.
It's
yeah.
A
B
B
A
A
We
should
work
on
these
they're
just
bigger
pieces
of
work,
so
I
don't
want
to
lose
the
work
I'll
document
them
I'll,
create
issues
and
I'll
just
put
them
in
the
backlog
for
for
further
work
when
we
can
get
to
them.
I
think
the
next
step
after
I
do
this
is
to
clean
up
the
list.
Get
them
in,
get
them
organized
a
bit
more
and
then
we
can.
A
What
I'll
do
is
I'll
identify
whether
or
not
the
data
exists
for
us
to
to
get
some
insights
about
these
before
we
move
forward
just
so
that
we
can
understand
what
the
reach
and
the
impact
would
be,
and
then
I
also
think
it
would
be
helpful
to
just
have
one
more
session
to
talk
about
effort
once
they're
scoped
a
little
bit
further.
We
could
say
okay,
so
for
this
experiment.
A
I
think,
through
the
through
the
next
couple
iterations
on
this
I'm
going
to
have
to
work
really
closely
with
with
with
the
rux
counterpart,
so
jackie
jeff
mate.
I
I'll
probably
lean
on
you
a
little
bit
doug
there's
only.
I
only
saw
a
handful
in
here
that
are
really
like
straight
engineering
opportunities,
but
if
anything
else
jump
out
I'll
I'll
ping,
you
all
so
just
keep
an
eye
out.
You'll
probably
see
a
follow.
A
couple
follow-on
issues
from
me.
A
E
I
I
told
the
ux
team
this
in
our
slide
channel,
but
I
think
that
the
user
flows
that
that
we
did
have
been
really
helpful
actually
because
I
looked
at
those
user
flows.
While
I
was
doing
this
exercise-
and
it's
really
nice
to
have
it
all
in
one
place
and
not
have
to
worry
about
like
going
into
the
app
and
getting
things
to
look
the
way
you
need
them
to
look,
you
just
look
at
the
user
flows
and-
and
it
drags
lots
of
ideas.