►
From YouTube: 2023-09-20 Product Analytics UX/PM Sync
Description
Today we talked through:
1. Upcoming work for Unique Users toggle and how it applies or doesn't to none Product Analytics Visualizations - https://gitlab.com/gitlab-org/gitlab/-/issues/421328
2. Recent learnings about Pain Points from a user survey
3. What's next for the Visualization Designer - https://gitlab.com/groups/gitlab-org/-/epics/9386
A
B
A
I
wanted
to
touch
on
I
had
shared
in
a
video
walking
through
the
opportunity
solution
treat
last
week,
some
of
the
data
from
our
willingness
to
pay
survey
and
the
pain
points
that
we
had
laid
out.
So
we
had
identified
pain
points
based
on
the
previous
research
that
had
been
done,
conversations
that
we've
had
with
customers
and
knew
that
they
aligned
to
things
that
were
already
built
or
things
on
the
roadmap.
I
tried
to
structure
it
in
a
way
that
here
is
a
pain
point
that
you
might
have.
A
We
add
customers
to
give
points,
basically
to
those
items
to
invest
in
them
or
you
would
buy
a
bit
for
a
product
analytics
solution,
and
if
you
give
something
10
points,
it's
twice
as
important
as
something
you
would
get.
Five
points
to
or
20
points
because
of
four
times
as
important
kind
of
give
them
that
scale
to
work
off
of
so
I
shared
those
findings
in
a
video
added.
The
pain
points
the
top
five
anyway
to
our
opportunity
solution
tree
and
we
can
start
working
on
those
opportunities.
A
The
night
part
is
it
aligned
very
well
with
what
we're
already
building.
So
it's
happy
to
see
that
that
we
prioritized
correctly
with
local
market
field,
but
wanted
to
check
in
with
the
team.
Is
there
another
way
that
I
can
share
that
data?
Is
there
more
ways
that
we
want
to
talk
about
this
I?
Don't
think
having
it
in
one
video
and
one
opportunity
solution
tree
is
the
best
way
to
go
so
I
wanted
to
open
it
up
to
what
are
other
modes
of
communication
so
that
we
can
share
this
data.
A
C
B
No,
the
block
is
because
of
we're
waiting
for
backend
to
make
this
happen
from
a
friendly
point
of
view,
it's
a
very
straightforward
process
to
add
the
toggle.
Once
we
have
that
back-end
implementation,
there's
also
worth
noting
that,
in
the
implementation
plan
itself,
we've
added
a
tracking
event.
So
we
can
monitor
how
often
this
feature
is
used
and
get
an
idea
of
the
impact
that
it's
having
with
our
users.
A
D
D
B
B
D
Yeah,
we'll
have
to
be
smart;
well,
initially
we
can
be
dumb
and
what
I'm
thinking
of
is.
We
have
the
value
analytics
dashboard
all
the
value
stream
dashboard,
so
we
just
need
to
be
sure.
That's
this
is
part
of
all
the
filters
that
we
exclude
on
that
dashboard.
I
believe
it
should
be
easy
to
do
for
a
start,
but
in
the
future,
we'll
probably
want
to
be
smarter
about
it
in
terms
of
like
dynamically,
adding
this
filter
based
on
the
type
of
visualizations
that
a
user
is
added
to
their
dashboard
yeah.
A
Okay,
good
good
to
think
about
it,
so
I
added
a
note,
put
it
on
all
of
the
product
analytics
dashboards
built
in
and
custom
and
just
make
sure
it's
not
on
the
value
stream.
Dashboard
cool,
excellent
refinement,
I'm
going
to
skip
over
the
upcoming
design
issues.
I
think
the
majority
of
what
is
already
scheduled
is
the
usage
quota
stuff.
A
Then
I
know
the
team
took
a
hard
look
at
that's
coming
up
in
16-6,
so
I'm
going
to
reserve
time
for
us
to
talk
through
the
visualization
designer
dashboard
designer
conversation
today,
because
it's
fresh
and
we'll
pick
up
this
issue
next
week.
If
that
works
for
everyone
all
right,
all
right,
I'm
going
to
open
it
up
to
our
front
end.
B
B
So
this
morning
I
mean
Jan.
We
sat
down
to
discuss
where
the
dashboard
design
and
visualization
designer
were
at.
We
felt
like
the
dashboard
designer
was
in
a
good
enough
place
that
we
could
release
it
from
its
feature
flag.
Obviously,
we
still
have
the
more
brighter
product
analytics
feature
Flags,
so
we're
still
limiting
in
this
current
phase
to
very
specific
user
customers,
but
this
then
opens
up.
B
B
It's
not
been
as
high
as
a
priority,
so
we
need
we
wanted
to
figure
out.
You
know
without
too
much
polish.
Where
could
we
get
to
I'm
in
a
reasonable
amount
of
time?
That
would
allow
us
to
start
getting
that
user
feedback,
because
realistically
that's
where
the
you
know
where
we
really
want
to
be
heading
from
a
direction
and
targeting
those
pain
points
for
users.
B
So
we
sort
of
came
up
with
a
multi-state
plan
that
extends
beyond
the
release
of
the
flag.
So
the
first
stage
is
more
or
less
just
the
biggest
pain
Point,
really
with
the
visualization
designer
right
now
behind
beyond
the
ux
issues
that
are
more
broader
and
need
more
ux
designer
input.
Once
we
hire
someone,
it's
that
when
you
go
on
a
Visa
validation
designer
someone
who's
never
used
a
product
before
it
doesn't
explain
to
you
what
it
is
or
how
it
works.
B
B
You
know
aligning
it
slightly
more
with
the
designer
moving
the
CTA
to
the
bottom
and
generally
bringing
the
the
step
process
of
creating
a
measure,
selecting
visualization
type
and
seeing
what
it's
going
to
look
like
before
you
save
from
what
is
currently
a
132
process.
Where
you
look
at
the
left,
look
at
the
right
look
at
the
middle
to
a
look
at
the
left.
Look
at
the
left.
Look
at
you
know
go
through
without
using
a
step
up
just
making
it
more
obvious.
B
But
this
is
not
a
requirement
to
release
as
I
say
it's
not
a
phase
functionally
working,
it's
just
not
great!
So
we
can
do
this
in
tandem.
It's
not
gonna.
Releasing
these
tweet
UI
tweaks
won't
start,
won't
cause
any
breaking
changes
for
the
user
or
just
improve
their
experience.
So
I'm
just
going
to
stop
there
to
see
if
anyone
has
any
comments
or
questions
before
I
discuss
sort
of
what
would
happen
after
the
release.
A
I
think
generally,
the
stage
one
stage
two
sounds
good
I
think
the
biggest
Gap
that
we
have
is
that
our
MVP
is
barely
viable
with
the
change
from
Jitsu
to
snowplow.
We
lost
all
of
that
structured
event.
Data,
like
click
events,
where
that's
a
big
part
of
the
UI
today,
really
all
we're
giving
the
users
is
the
ability
to
change
time
frame
on
existing
graphs,
I
mean
in
essence
that's
all.
A
They
can
really
do
and
then
put
them
into
an
order
that
they
really
want
so
I'd
love
to
see
before
we
release
it
and
turn
it
on
flip
the
feature
flag,
work
with
back
end
and
figure
out.
How
do
we
get
structure
back
and
at
least
get
click
of
them
so
that
you
can
filter
by
click,
events
Group
by
click
event,
and
that's
the
feedback
that
we're
getting
from?
A
But
from
our
our
friends
in
developer,
experience
is
hey,
I
wanted
to
end
of
Ben
and
query
on
those
and
they're
abusing
page
views
today
and
trying
to
jam
data
into
page
views,
because
then
you
can
see
the
data
that
they
want.
Otherwise
it's
just
unstructured
events
that
are
being
sent
back
and
so
being
able
to
see
at
least
all
of
the
events,
and
then
we
can
work
on
now.
I
want
to
filter
events.
I
want
to
query
for
specific
events.
A
What
other
directions
that
might
take
us
would
be
then,
a
later
step,
as
we
really
really
iterate
on
this
and
maybe
tear
it
all
down
to
the
studs
and
build
it
back
up
with
a
new
designer.
But
for
me
that
would
be
before
we
flip
the
feature
flag.
What
I'd
want
to
see
otherwise
you'd
get
a
I
want
to
see
all
the
click
events
Group
by
click
event,
and
it's
all
just
unstructured
data
like
that
just
seems
like
a
broken
experience
and
not
a
valuable
experience
to
me.
B
A
Something
you
can
change
your
measure,
so
eliminate
page
views,
get
rid
of
that
right.
I
want
a
quick
event,
quick
events,
all
quick
events
and
then
start
to
see
The
Click
event,
so
that
I
get
zero
data.
So
in
projects
where
I
know
I
have
click
event,
data
flowing
through
it
shows
up
as
unstructured
data
which,
if
I
have
more
than
one
click
events
I,
have
now
zero
idea
of
what
is
happening
and
where
it's
happening,
because
we
don't
allow
like
we're
not
sending
that
data
along
or
it's
not
queryable
out
of
Click
house.
A
So
it's
not
a.
It
was
working
before
with
Jitsu
what
I,
what
I'm
trying
to
say
is
we'll
get
back
to
that,
because
we
know
we
need
to
on
the
back
end
anyway.
We'll
get
to
that
point.
Let's
release
it.
I
like
the
ux
weeks.
Those
are
awesome.
That's
a
great
usability
Improvement
for
the
users,
I
think
and
I
glossed
over,
that
you
guys
did
a
great
job
of
iterating
on
the
existing
design.
A
That'll
get
this
into
a
much
better
place
until
we
have
a
ux
designer
in-house
and
we
can
really
rethink
this
experience
and
really
rethink
the
entire
opportunity
of
hey
I.
Don't
want
to
learn
another
query.
Language
Let
me
just
get
my
data
because
we're
going
to
approach
it
kind
of
from
that
really
high
level.
I
think
this
is
one
possible
solution
for
it.
B
Okay,
that
makes
complete
sense.
So
almost
a
stage,
zero,
add
click
events
as
it
were,
it
can
be
done
in
tandem
with
stage
one
but
add
click.
Events
make
sure
click
events
are
actually
working
properly,
which
will
require
obviously
working
with
instrumentation
to
make
sure
we're
getting
that
data
in
the
right
places
and
then
rendering
that
from.
A
Yeah
I
think
we
want
to
give
a
new
designer
really
an
opportunity
to
explore
the
solution
space
and
give
us
all
an
opportunity
to
explore
the
space
before
we
put
too
many
other
tweaks
in
I.
Think
it
made
sense
to
like
the
big
usability
gaps
that
we
have
in
visualization
designer
today,
where
it
yelled
at
you,
because
you
didn't
complete
a
step.
You
didn't
know
about.
Let's
fix
those
like.
Let's
make
this
a
lot
easier
for
users
to
use
and
in
the
current
state,
but
let's
not
over,
invest
there.
A
A
C
B
C
B
The
these
visualization
this
whole
visualization
designer
was
part
of
the
prequel
concept
that
just
got
transitioned
in,
which
is
why
it's
behind
the
flag,
because
it's
not
other
than
just
proving
that
it
can
work
and
Junior
and
starting
I,
think
you
you
Jones
and
Kevin.
You
you've
started
having
some
conversations
interview,
conversations
around
visualization
designer
views
as
well,
so
we
have
got
some
user
feedback,
but.
C
B
A
Yeah
Jackie,
our
hypothesis,
is
that
if
we
put
it
out
with
some
improvement,
we're
going
to
get
a
better
level
of
feedback
like
I
said
from
users
who
can
build.
Visualizations
use
our
sdks
more
broadly,
because
if
they're
just
getting
page
like
page
view,
data
and
rough
user
data
accounts
is
not
as
valuable
as
if
you
can
start
to
instrument
a
vent
and
really
see
like
a
user
selected
this,
and
then
they
did
this.
A
If
I
can
start
to
be
that
event,
data
flowing
through
there's
a
lot
more
value
for
our
platform
teams
that
are
our
first
ideal
user
and
their
product
folks
to
see
what's
really
happening
in
their
applications
like
it's
great.
If
I
can
see
how
many
users
I
have,
but
I
really
want
to
see
if
they're
getting
value,
and
a
lot
of
that
relies
on
send
an
event
that
something
was
accomplished.
And
let
me
count
those.
So
that's
why
I'm
pushing
for?
A
C
That
makes
total
sense.
It's
like
hard
to
get
feedback
from
users
on
a
complex
workflow
if
the
data
is
not
realistic
or
useful
to
them
anyway.
So
that
makes
complete
sense.
I
just
wasn't
sure
if
there
was
something
I
should
go
back
and
look
at
to
see
how
we
got
to
here,
but
it
sounds
like
it
hasn't,
had
any
ux
input
up
until
now,
which
is
fun.
B
C
A
We
had
a
version
of
the
dashboard
designer
that
included
visualization
designer
that
we
ripped
out,
because
we
didn't
want
to
spend
a
ton
of
time
there
and
get
to
a
point
where
you
could
build
a
custom
dashboard.
That
is
basically
rearranging
the
panels
as
a
first
step
and
then
bring
in
hey.
Here's
now
designs
that
or
visualizations
rather
that
you've
created,
because
we
think
that
there's
a
great
path
forward
on
the
customs
of
custom
dashboards
with
other
visualizations
from
within
gitlab
that
we
don't
want
to
wait
on
either.