►
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
A
So
the
genesis
of
this
is
that
we
in
growth
had
an
okr
around
navigation,
specifically
to
propose
designs
for
two
improvements
to
navigation
and
identify
ten
others
for
the
backlog.
Basically,
we
have
quite
a
large
backlog
of
issues
related
to
navigation
going
back
years
in
some
cases,
and
we
didn't
quite
know
exactly
how
to
prioritize
that
or
what
was
sort
of
the
most
pressing
of
those
issues,
and
so
some
of
this
research
was
done
in
order
to
just
understand
of
all
sort
of
the
known
problems
that
we
have.
A
And
so,
given
that
we
had
this
okay,
the
show
care
actually
came
a
little
bit
into
the
quarter,
and
so
we
didn't
have
a
tremendous
amount
of
time.
And
so
we
had
to
sort
of
be
scrappy
and
think
really
tactically
of
what
we
could
execute
in
a
given
time
frame.
A
So
why
navigation
navigation
is
a
really
persistent
issue
for
us
at
gitlab.
It's
been
something
that
comes
up
very
often
in
sentiment,
surveying
that
we
do
mps
and
sus.
You
can
see
some
of
the
examples
here
on
the
right.
A
These
are
verbatims
from
either
mps
or
sus
speaking
to
issues
that
are
related
to
navigation,
and
then
the
thing
I
always
sort
of
come
back
to
is
navigation
is,
is
really
one
of
the
only
aspects
of
gitlab
that
literally
every
user
touches,
and
so,
if
you
think
about
making
changes
and
improvements,
even
small
changes
and
improvements
when
you
aggregate
those
across
the
entirety
of
the
user
base,
those
gains
and
efficiency
or
gains
and
sentiment
in
terms
of
making
things
easier
or
more
pleasurable
to
use
are
just
massive.
A
A
And
so
like
I
said,
we
didn't
have
a
tremendous
amount
of
time.
We
had
approximately
six
weeks
before
the
end
of
the
quarter
and
thus
when
this
okr
would
need
to
be
wrapped
up,
and
so,
given
that
amount
of
time,
we
really
chose
to
sort
of
look
at
the
ends
of
the
spectrum
when
it
comes
to
usage
of
navigation
and
the
people
that
use
navigation.
A
And
so,
if
you
think
about
that
as
a
spectrum,
that's
basically
you
know
people
who
are
brand
new
to
gitlab
what
you
would
call
net
new
users
who
have
never
used
gitlab
before
and
are
not
currently
sort
of
involved
in
an
organization
or
group,
that's
using
git
lab
and
then,
on
the
other
end
of
the
spectrum.
You
have
basically
highly
active
users
of
gitlab
people
using
gitlab
more
than
three
days
a
week,
people
who
ideally
have
sort
of
established,
routines
and
workflows
that
we
wanted
to
understand
better.
A
So
for
the
new
users,
we
really
wanted
to
understand
when
you're
brand
new
to
gitlab,
and
you
have
really
no
mental
model.
You
really
have
no
background
to
go
off
of.
How
does
navigation
affect
your
understanding
of
the
product
and
how
are
you
able
to
comprehend
that
navigation
and
complete
what
we
considered
introductory
tasks
tasks
that
would
be
appropriate
for
someone
who
is
brand
new
to
gitlab?
So
you
can
think
of
this
as
things
like
creating
a
new
issue,
creating
issue
board
modifying
your
profile,
adding
someone
to
a
group.
A
These
are
not
kind
of
detail-focused
tasks
like
you
know,
creating
and
editing
a
ci
pipeline
or
going
deep
into
you
know.
Maybe
road
maps,
things
that
you
know
are
gonna
happen,
certainly,
but
are
going
to
happen
once
you're
a
bit
more
established
in
in
gitlab
and
then
for
the
active
users
and
routines
or
excuse
me,
active
user
routines
and
workflows.
This
was
much
more
open-ended.
We
really
wanted
to
understand
for
these
folks.
A
What
are
you
doing
on
a
day-to-day
basis
kind
of
how
we
framed
it
was
like
when
you
first
entered
gitlab
for
the
day
like
what
are
the
first
things
you're
doing?
What
information
are
you
looking
for?
How
are
you
accessing
that
information
and
sort
of
what
is
your
goal
and
what
are
the
kind
of
workflows
that
you've
established
to
do
these
sort
of
common
tasks
that
you're
doing
every
day?
A
Whether
that's
looking
into
mrs
dealing
with
issues
whatever
that
may
be,
and
so
that
was
really
led
by
the
participants
themselves
and
the
work
they
were.
They
were
doing
in
their
their
gitlab
account,
and
so
it
varied
much
more
compared
to
the
new
user
experience,
which
was
tasks
that
we
formulated
for
folks
just
to
see
how
easily
they
would
be
able
to
complete
those
tasks
or
potentially
not
easily.
A
So
for
the
new
user
component
of
this
testing,
given
that
we
wanted
to
get
a
decent
sample
size,
you
know
we
wanted
approximately.
You
know
anywhere
from
10
to
15
people,
and
we
didn't
want
to
put
people
through
all
of
the
tasks,
but
we
still
wanted
to
have
a
large
number
of
tasks.
You
know
10
tasks,
I
created
a
unmoderated
testing
setup
and
I
know
a
lot
of
folks
on
the
ux
team
are
very
interested
in
moderated
testing.
A
It's
something
we're
talking
a
lot
about,
so
I
wanted
to
spend
a
moment
talking
about
the
setup
that
I
used,
because
we
did
have
some
hiccups
in
the
beginning
and
some
kinks
to
work
out,
but
it
actually
did
end
up
working
out
pretty
well
for
this
purpose,
and
so
what
I
did
was
created
a
qualtric
survey
that
basically
started
with
the
instructions
for
how
people
should
go
about.
You
know
completing
the
test
and
basically
instructed
them
to
join
a
zoom
meeting,
which
they
were
the
only
member
of
that
zoo
meeting.
A
The
purpose
of
that
zoom
meeting
was
just
to
record
them,
interacting
with
both
the
survey
and
the
gitlab
instance,
and
so
the
survey
would
sort
of
instruct
them
as
how
to
do
that.
It
would
gather
the
information
we
needed.
A
It
would
tell
them
how
to
you
know,
download
share
their
screen,
share
their
video
log
into
the
gitlab
instance,
and
then
it
had
a
randomized
set
of
tasks
for
them
to
complete-
and
this
was
a
nice
usage
of
the
qualtric
survey
logic,
because
you
can
basically
use
what's
called
a
randomizer
and
out
of
the
10
tasks
that
we
created.
A
Each
participant
was
presented
with
a
set
of
five
randomized
tasks,
and
this
is
nice
because
with
task-based
testing,
especially
when
you're
using
sort
of
a
single
interface
where
people
keep
going
back
to
the
same
interface,
to
complete
new
tasks,
there's
going
to
be
a
learning
effect.
So
the
last
task
the
person
completes,
is
likely
going
to
be
easier
for
them
than
the
first
task.
A
And
so
it's
really
important
that
for
each
person
you
randomize
those
tasks
such
that.
The
third
task
on
your
list,
for
instance,
maybe
for
some
people,
will
be
the
first
task
they
complete
for
other
people
will
be
the
last
task
they
complete,
but
once
you
have
enough
folks
that
learning
effect
sort
of
gets
distributed
across
each
of
them-
and
you
don't
have
this
issue
where
oh
the
last
task
on
your
list
is
always
the
last
task.
A
People
are
going
to
perform
as
part
of
their
study
and
so
they're
always
going
to
perform
better
on
it
than
they
might
have
with
the
first
task,
and
so
this
randomizer
that
qualtrics
has
was
really
nice
for
this
purpose,
and
so
basically
the
survey
would
present
them
with
a
task
and
see
here.
This
was
a
really
basic
one.
Changing
the
name
on
your
account.
They
were
instructed
to
sort
of
complete
this
task,
narrate
their
thought
process
as
they
went
through
it
completed
it
as
best
they
could.
A
If
they
felt
like
they
couldn't,
they
could
abandon
it
and
they
could
come
back.
And
then
this
survey
asked
them
a
couple
questions
once
they
either
completed
the
task
or
felt
they
were
unable
to
just
about
you
know.
Yes,
I
was
able
to
no,
I
wasn't
ask
about
their
confidence
and
then
talk
about
how
confident
or
how
their
confidence
was
compared
to
when
they
started
the
task.
A
Weren't
super
useful,
but
what
was
really
valuable
here
was
having
them
talk
through
each
of
these
answer
choices
and,
of
course
all
this
was
on
the
recording,
and
so
as
people
would
go
and
answer
these
questions.
They
talk
through
why
they
answered
a
given
response
and
and
what
led
them
to
do
that,
and
they
were
instructed
to
do
that,
and
so
that
was
really
valuable.
Just
hearing,
oh,
I
expected
this
to
be
really
easy,
but
it
was
very
difficult
for
these
reasons,
or
I
expected
this
to
be
very
easy.
A
It
was
very
easy
great,
so
this
actually
worked
out
really
well.
There
were
some
kinks
that
we
still
need
to
work
out.
People
cannot
take
these
unmoderate
moderated
tests
in
parallel.
A
Only
one
person
can
do
it
at
a
time,
unfortunately,
just
due
to
how
zoom
works-
and
I
realized
only
after
interrupting
one
of
my
four
participants
and
knocking
him
out
of
a
session
that
I
cannot
host
a
meeting
while
one
of
these
participants
is
doing
one
of
these
sessions,
because
I
am
technically
the
host
of
these
meetings,
and
so
these
are
all
solvable
problems
I
feel
like
if
we
were
to
use
this
in
the
future.
A
So
I
just
want
to
go
into
some
of
the
findings
that
we
found
from
both
of
these
sessions
and
then
later
on
I'll,
go
into
the
next
steps
that
we're
going
to
do
as
a
result
of
what
we
uncovered.
A
So
one
of
the
big
things
I
think
was
in
terms
of
the
groups
that
we're
showing
people
and
the
groups
that
people
are
members
of,
typically
people
described
being
actively
using,
maybe
one
or
two
groups
at
most,
but
often
they
had.
You
know
three
to
four
times
that
many
groups
in
their
groups
list,
and
especially
if
this
was
their
dashboard.
A
You
know
this
is
something
they
were
coming
to
every
time
they
log
in,
and
this
was
sort
of
their
their
root
homepage
of
gitlab,
and
we
do
have
sorting
mechanisms
for
that
group
to
try
to
you
know,
float
the
the
most
relevant
things
to
the
top,
but
they
weren't
super
useful
for
people,
because
the
sorting
mechanisms
aren't
necessarily
personal
to
the
individual
they're
about
the
group
as
a
whole,
and
so,
for
instance,
someone
showed
me
a
group
that
was
basically
the
second
group
in
their
ranking,
and
they
said
I
haven't
looked
in
this
group
in
a
year
and
a
half,
but
other
people
were
using
that
group
on
a
nearly
daily
basis,
and
so
the
activity
by
those
other
people
kept
popping
that
group
back
up
to
the
top
of
this
person's
group
list,
and
it
basically
gave
it
an
area
of
importance
that,
for
him
did
not
exist,
and
so
I
think
it
was
really
interesting
to
hear
that.
A
Like
oh
yeah,
I
have
all
these
groups
they're
technically
active
groups,
but
for
my
own
purposes,
they're
considered
deprecated
or
archived.
These
are
words
that
people
were
using.
These
are
old
groups,
but
in
terms
of
how
we
showed
them
in
the
interface,
they
were
just
as
important
as
the
one
or
two
groups.
People
were
interacting
with
on
a
nearly
daily
basis,.
A
The
other
thing
when
it
comes
to
navigating
you
know,
groups
and
projects
and
p.
Typically,
the
workflows
that
the
active
users
would
describe
was.
Oh,
I
have
a
handful
of
projects
or
groups
that
I
care
about
that.
I
monitor
typically
every
day
or
multiple
times
a
week,
how
they
were
then
getting
to
those
groups
was
navigating
through
the
tree
structure
of
of
git,
lab
groups
and
projects,
and
so,
if
you
think
about
our
user
interface,
you
start
at
a
high
level
of
a
group.
A
You
then
potentially
navigate
down
to
a
subgroup
and
then
down
to
a
project
and
then
within
a
project.
Whatever
aspect
you
care
about,
whether
that's
mrs
or
pipelines
or
issues
or
whatever,
and
basically
each
time
people
need
to
get
to
a
different
project.
They
have
that
kind
of
ping-pong
up
and
down.
In
order
to
do
this,
so
they
would
start
at
the
dashboard.
They
would
go
down
to
an
individual
project
and
they
would
go
back
up
to
the
dashboard.
A
They
would
navigate
back
down
into
an
individual
project
and
back
and
forth
and
back
and
forth,
and
they
really
described
this
as
as
kind
of
labor
intensive
one
phrase
someone
used
was
everything
in
gitlab
feels
very
detached
from
everything
else,
in
that
it
feels
like
those
objects
or
those
entities
within
gitlab
are
far
apart
and
that
there's
a
lot
of
clicking
or
effort
that
has
to
happen
to
go
from
one
to
the
next
and
kind
of
what
we
saw
and
what
even
some
people
were
able
to
articulate
was
they
want
to
be
able
to
move
horizontally
if
you
will
across
projects,
even
if
those
projects
maybe
span
different
groups
or
aren't
not
are
not
necessarily
closely
related,
and
this
was
difficult
in
our
ui
and
people
kind
of
just
had
to
put
in
the
effort
to
do
this
manually.
A
So
another
big
finding
we
have
these
frequently
visited
units
in
the
top
now
for
both
projects
and
groups
for
almost
everyone
we
spoke
to.
These
were
just
empty
and
again
remember.
These
are
very
active
users
of
gitlab.
These
are
people
using
gitlab
more
than
three
days
a
week
and
also
keep
in
mind
that,
as
we
spoke
to
people,
they
were
navigating
all
through
their
gitlab
incidents
and
clicking
on
groups
and
projects.
A
As
we
were
talking,
and
so
even
by
the
end
of
these
sessions,
you
would
expect
had
they
never
used
gitlab
before
just
that,
navigating
during
the
session
would
have
populated
these
units
and
in
almost
every
case
it
wasn't,
and
so
we've
been
talking
with
engineering
about
why.
Potentially,
this
is
happening
from
what
I
understand
has
to
do
with
the
fact
that
we
store
this
data
in
local
storage
in
the
browser,
and
that
has
some
issues.
A
For
the
the
few
folks
where
we
did
see
content
being
populated
in
this,
it
really
wasn't
what
they
felt
was
recently
visited
and
it
really
didn't
provide
value
to
them
in
that,
like
it,
wasn't
the
things
they
needed
to
access
and
that
may
have
been
just
due
to
the
limited
sort
of
real
estate
here,
I
think,
typically
between.
A
There
were
very
very
few
instances
where
these
brand
new
users
to
get
that
board
just
unable
to
complete
a
task,
and
I
think
that
was
really
positive
and
actually
a
really
great
learning
for
us,
because
it
actually
helps
sort
of
direct
our
effort
into
where
we
spend
our
time
and
where
maybe
larger
problems
do
exist
and
the
problems
that
we
did
uncover
as
part
of
that
study
were
pretty
minor
and
actionable.
A
So
one
of
the
things
that
we
discovered
this
unit
here-
and
I
don't
know
what
this
unit
is
called
in
pajamas
terms-
I
I
tried
to
look
it
up
and
I
couldn't
I
couldn't
find
it.
But
one
of
the
major
problems
that
new
users
had
with
this
button
is
that
this
is
both
a
button
and
a
drop
down,
and
people
expected
that
when
they
were
to
click
one
of
the
items
in
this
drop
down,
that
would
trigger
the
action
that
it's
describing
and
that's
not
actually
the
case.
There's
a
second
click
required.
A
So
if
I
were
to
click,
for
instance,
new
subgroup,
the
button
state
would
change,
it
would
become
a
new
subgroup
button.
I
would
then
have
to
click
new
subgroup
in
the
button
itself
to
execute
that
action,
and
so
multiple
users
had
this
thing
where
they
said:
okay,
new
subgroup,
they
clicked
it,
they
kind
of
waited
a
beat
expecting
there
to
be
a
transition,
nothing
would
happen
and
then
they'd
say
oh
and
they'd,
maybe
click
the
drop
down
again.
A
They'd,
look
at
it
and
then
eventually
they'd
say:
oh
okay
and
they
click
the
button
and
they'd
be
on
their
way.
But
there
was
this
sort
of
moment
of
confusion
for
folks
where
they
really
did
expect
the
click
on
the
drop
down
to
be
the
click
that
would
execute
that
action.
But
again,
this
is
this
is
something
actionable.
This
is
something
that
we
can
address.
A
This
is
you
know
something
it's
not
a
major
issue
which
was
really
great
to
see,
especially
for
the
brand
new,
so
next
steps
based
on
what
we
found
in
this
research,
so
in
terms
of
product
improvements,
kind
of
the
big
ones
are
fixing
frequently
visited.
Our
first
step
in
terms
of
nvc
is
just
let's
make
the
unit
work
as
it's
intended
to
work.
A
A
The
other
thing
that
we're
doing
is
going
to
be
instrumenting,
the
top
bar
navigation.
So,
unfortunately,
right
now
we
do
not
have
any
instrumentation
on.
As
far
as
I
know,
any
of
our
navigation
elements,
which
makes
it
a
bit
challenging
to
understand
the
relative
frequency
of
usage
of
those
elements
and
understand
just
at
an
aggregate
level.
A
What
aspects
of
the
navigation
folks
are
using,
so
the
expansion
heap
is
going
to
be
putting
in
work
to
to
instrument
at
least
the
top
bar
navigation,
so
we
can
get
a
better
idea
of
usage
and
that'll
help
inform
some
of
the
next
research
that
we
do
as
well
the
things
that
we
want
to
investigate
in
terms
of
research
for
this
quarter.
A
One
aspect
is
dashboard
redesign,
prototype
testing,
something
we
heard
a
lot
was
that
the
homepage
or
the
dashboard
is
not
terribly
useful.
For
folks,
as
I
was
saying,
it's
not
showing
people
the
information,
that's
pertinent
to
them.
It's
showing
people
sort
of
everything
and
whether
that's
all
your
projects,
all
your
groups,
all
your
activity,
it's
not
the
things
I
care
about
as
an
individual.
A
Now,
what
I
care
about
as
an
individual
is
a
very
complex
problem,
that's
very
specific
to
the
individual.
So
it's
it's
not
that
we
have
a
very
easy
solution
that
we
can
put
in
place.
A
That'll
solve
everything
for
everyone,
so
we
need
to
try
to
better
understand
what
is
important
to
an
individual,
and
how
can
we
frame
that
and
how
can
we
show
that
information
in
a
way
that
will
work
for
a
variety
of
people
in
a
variety
of
situations,
so
how
we're
hoping
to
accomplish
that
we're
still
kind
of
early
on
and
how
we're
formulating?
A
This
is
doing
some
rapid
iterative,
prototype
testing
so
using
small
prototypes
of
potentially
even
individual
components
that
could
live
in
that
space
to,
of
course,
validate
those
as
potential
solutions,
but
also
use
those
as
sort
of
a
jumping
off
point
for
larger
conversations
around
what
type
of
information
is
relevant
to
you.
What
would
you
want
to
see
here?
A
A
As
often
as
once
a
week
such
that
we
can
get
to
a
place
very
early
on
where
we
can
feel
reasonably
confident
that
you
know,
we've
got
a
solution
at
a
conceptual
level
that
will
work
and
then
we
can
go
and
put
in
the
design
work
to
really
flesh
it
out
and
then
test
out
that
flushed
out
solution.
A
And
then
the
other
area
that
we
want
to
investigate
for
research
is
new
users
and
established
groups.
And
so,
as
I
said,
net
new
users,
people
who
are
brand
new
and
not
really
associated
with
the
larger
organization.
They
tend
to
do
pretty
well
with
basic
introductory
tasks.
A
And
so
we
want
to
shift
our
focus
from
those
folks
to
people
who
are
new
users,
but
coming
into
an
established
organization
or
group,
because
they're
going
to
have
much
more
existing
content
to
interact
with
they're
going
to
potentially
have
much
larger,
more
complex
tasks
they
need
to
complete,
and
we
think
that's
really
where
there's
probably
a
lot
of
things
we
can
address.
A
It
also
is
nice
because
it
aligns
with
the
goals
of
growth.
In
terms
of
you
know,
expansion
and
increasing
activity
and
increasing
seat
count,
and
so
it's
where
we
want
to
try
to
focus
more
of
our
time
and
make
sure
we're
serving
those
users.
Well
all
right,
that's
it
for
me.
If
anybody
has
any
questions
about
unmoderated
testing
or
this
navigation
work,
please
feel
free
to
reach
out
to
me,
or
you
can
ping
the
navigation
channel
on
slack
and
yeah.
Thanks
for
having
me.