►
From YouTube: Pricing page AB testing feature comparison display 10620
Description
The Digital Experience team is discussing the results of a recent AB test on the pricing page as well as what we're planning to do next.
A
Hello,
we're
gonna
go
over
the
pricing
test
on
the
about.gitlab
page
for
my
keyboard.
Just
stopped
working.
Sorry
about
that,
and
so
far
the
results
are
pretty
much
an
even
split
kind
of
50
50.,
for
you
know
overall
did
people
click
the
buy
now
purchase
button
or
not
it's
been
more
than
two
weeks
and
we're
not
seeing
any
deviation
from
this
pattern.
A
So,
even
though
there's
not
a
large
statistical
significance,
we
feel
pretty
confident
in
the
results
that
we're
seeing
so
far
and
it's
pretty
even
split
across
the
board
between
free
premium
ultimate
did.
They
click
contact
sales.
It's
there's
not
a
huge
change
between
the
two
and
so
that's
kind
of
the
the
first
thing
we've
learned
and
yeah
on
to
jess.
B
Yeah,
can
you
just
scroll
down
to
the
bottom
of
the?
Not
that
page
of
that
one.
B
So
one
thing
we've
also
noticed
is
that
just
less
than
half
of
the
people
48.3
are
actually
clicking
through
the
test
where
you
get
the
see
more
option
on
the
tiers,
so
we're
thinking
that
maybe
we
can
do
something
there
and
we
want
to
test
a
variant
where
those
additional
benefits
are
just
always
open
and
they
don't
have
to
actually
click
through
and
want
to
see
if
that
will
actually
increase
conversion
by
just
not
having
to
take
that
extra
step
of
clicking
the
view
more
button.
C
Yeah
and
based
on
our
data
right
now,
you
know
we
we
actually
started
really
liking
this
test
option
and
and
we're
really
rooting
for
it.
The
numbers
don't
lie,
though,
so
our
preference,
if
we
had
to
make
a
decision
right
now,
would
be
to
stick
with
the
control,
which
is
the
variant
that
you're
seeing
on
the
left
here
with.
C
It
is
more
of
a
familiar
and
standard
presentation
style,
but
we
think
that
there's
a
lot
of
opportunity
in
that
it's
aligned
with
our
boring
solution,
sub
value
and
while
persecutive
customers
are
evaluating
other
options
for
their
devops
needs,
they
could
be
doing
tab
docking
where
they
have
one
tab
of
gitlab,
one
tab,
github
another
one
etc,
and
it
does
make
it
easy
for
them
to
compare
which
could
lead
and
we're
seeing
slightly
in
the
the
data
to
higher
conversion
rate
and
it'll
also
be
easy
for
easier
for
us
to.
C
In
fact,
iterate
from
we
believe
having
things
more
isolated
instead
of
baked
into
one
component,
will
be
it'll,
be
easier
to
understand
what
changes
are
having
positive
impact
instead
of
having
everything
kind
of
bundled
into
one,
and
if
we
do
one
thing
we
don't
actually
know,
you
know
that
half
the
people
didn't
open
the
benefits
and
the
thing
we
did
there
could
take
us
longer
to
get
to
statistical
significance
speaking
to
a
following
test,
the
one
that
just
talked
about
where
we're
going
to
open
it
on
the
right
just
leave
it
open
see
if
that
has
any
positive
impact.
C
We'd
also
like,
similarly
to
our
campuses,
page
we'd
like
to
make
premium
a
little
bit
visually
different.
So
maybe
it's
a
little
bit
higher
on
the
page,
maybe
a
little
bit
bigger
like
you're,
seeing
here
to
kind
of
nudge
and
encourage
people
to
select
the
paid
option
so
I'll
kick
it
over.
You
did
yes.
B
Yeah,
so
just
one
other
call
out
here
is
that
we
don't
really
know
where
people
are
coming
from
on
their
journey
to
this
pricing
page.
That
could
be
the
first
stop.
You
know.
Maybe
they
heard
about
gitlab.
They
want
to
see
what
the
prices
are
or
it
could
be,
their
last
stop,
maybe
they're
already
a
user,
and
they
are
on
the
free
version
and
now
they
know
they
need
to
upgrade
to
premium.
B
So
there's
a
lot
of
unknown
variable
there
that
we
need
to
just
be
mindful
of
again
just
to
kind
of
understand
where
they
are
in
that
journey
yeah.
We
just
don't
know
how
how
much
they've
seen
before
they
get
to
that
page
in
order
to
convert.
So
it
might
not
even
be
that
the
information
on
this
page
is
as
important
as
we
think
it
is
because
they
might
have
been
on.
You
know
15
other
pages
before
they
got
to
that
spot.
So
it's
just
a
call
out
to
think
about
that.
A
little
bit.
C
A
Yeah
and
so
kind
of
the
next
steps,
we're
thinking
is
to
pull
the
work
into
our
sprint
march,
15th,
with
the
targeted
release
date
of
the
24th,
and
that
iteration
will
run
for
a
couple
weeks.
A
It
seems
like
two
weeks
so
far
has
been
the
point
at
which
either
it's
become
clear,
that
there's
a
winner
or
that
there's
not
a
huge
amount
of
difference
or
not
necessarily
statistical
significance,
but
we
can
see
something
trending
in
a
direction
and
make
a
decision
at
that
point
that
we
can
build
from,
and
so
after
that,
we'll
be
building
the
test
where
our
plan
is
to
build
a
test
where
we
try
and
nudge
people
towards
that
premium.
As
was
talked
about
earlier
with
the
campus
example.