►
From YouTube: Release Management Pirate Metrics/Funnel
Description
Checkout this discussion on how the Release Management group is building out their North Star Metrics
https://gitlab.com/gitlab-com/Product/-/issues/880#note_311975580
A
So
Gila
what
I
wanted
to
walk
through
was
the
Northstar
metric
process
for
release
management
and
kind
of
think
about
using
the
pirate
metrics
and
funnel
and
really
dive
into
what
I'm
accustomed
to
is
having
a
Northstar
metric
at
the
company
level,
understanding
your
product
lines
or
P&L
contribution
to
that
norick
yeah.
But
it
sounds
like
we're
doing.
North
Star
at
the
group
level
and
we're
allowing
each
stage
to
kind
of
then
have
their
core
North
Star
metric
further
area,
which
is
a
little
different
than
I'm
used
to
seeing.
B
It
yeah
I
think
I
think
what
do
you
describe
absolutely
makes
sense,
I
think
here,
because
my
assumption
is
because
we
have
so
many
stages
and
products
right
and
features
and
groups.
So
we
want.
Each
group
has
your
own
North
Star,
but
I
agree
100%.
We
should
assemble
them
some
way
in
the
bigger
model
to
the
company
North
Star.
Otherwise,
like
it's
still
hard
to
understand
the
real
impact
of
the
work
and
also.
A
B
A
A
CV
is
really
great,
but
then
that
underestimates
the
value
and
attribution
of
our
horizon
three
investments
making
right
now,
obviously
completely
agree
with
you
on
that.
Okay,
so
I
have
this
agenda.
I'll
go
ahead
and
share
my
screen
and
we
have
walk
through
my
thoughts.
We
can
talk
about
what
you
would
recommend
about,
like
our
the
approach
of
getting
this
kind
of
off
the
ground.
Can
you
see
yeah.
B
A
A
We
only
introduce
releases
about
a
year
ago,
okay,
fairly,
immature
offering,
but
I
would
still
make
this
metric,
that
the
current
email,
metric
and
activation
metric
or
really
an
acquisition
metric
like
people,
are
becoming
aware
of
it,
and
then
once
we
realize
that
customers
have
started
creating
releases,
that's
like
the
population
of
conversion
from
they
realize
that
paid
that
the
release
page
exists.
I
want
a
release
page.
This
is
a
metric
that
I'm
unable
to
calculate
today
is
an
unknown
metric.
B
A
Me
pull
up
like
the
gitlab
DevOps
diagram,
because
I
think
it
kind
of
helps
to
see
where,
where
we're
at
in
the
cycle
mm-hmm.
So
release
is
right
here.
If
you
look
at
this
diagram,
so
it's
not
it's
it's
the
right.
When
you
convert
your
binary
or
image
and
you
package
with
all
of
your
merger
quests,
and
then
you
have
a
release
or
a
deployment
to
something.
A
B
A
This
is
kind
of
like
the
all
the
stuff
that
goes
into
making
a
production
a
production
deployment
right.
So
that's
what
gets
me
down
to
since
we
are
so
new
people
were
deploying
with
get
lab
prior
to
their
releases
functionality,
but
in
population
of
the
customers
that
were
deploying
with
get
lab
was
very
small
because
they
were
unable
to
track
that
deployment
over
time.
So
that's
where
like
release
management
and
the
release
stage,
is
supposed
to
be
enabling
ultimately
people's
ability
to
deploy
with
with
gitlab.
A
So
when
I
really
right
now
in
the
issues
that
I've
been
talking
with
my
director,
Kenny
I
am
a
count
of
deployments
as
being
the
metric.
That's
going
to
be
tied
to
eventually
like
iacv
in
revenue
like
the
more
customers
deploy
with
get
lab,
the
more
likely
they
are
to
leverage
upstream
functionality
of
the
CI
CD
stage
or
section
op
section
and
they're
more
likely
to
adopt
monitoring
the
downstream
of
a
deployment.
A
So
when
we
look
at
rapid
expansion
deployments
that
turning
point,
so
what
I
the
bridge
between
when
somebody
views
the
release
page,
these
are
reading
releases
and
then
they
decide
to
start
leveraging
deployments
is
where
I
noticed
the
drop
off
and
where
I
want
to
like
understand
how
to
target
and
prioritize
my
product
features.
So
that's
exactly
how
I
learned
pirate
so
metric
yeah,
so
just
wanted
to
kind
of
bounce.
B
So
you
think
this
high
level
make
a
lot
of
sense.
Maybe
I
just
have
some
questions.
I
think
we're
asking
a
min
map
might
help
you
as
well,
so
the
release
pageviews
under
which
circumstance
people
will
go
to
release
page
who
will
be
there
is
there
only
people
will
proactively
go
there
or
there's
a
case.
People
will
somehow
stumble
a
pound
release,
page
yeah,
so
people.
A
Typically
view
the
release
page,
we
noticed
in
periscope
around
our
the
release
cycle.
So
it's
usually
a
point
where
people
know
we're
going
to
release
on
the
22nd
of
the
month.
So
then
they
go
to
their
release
page
and
either
people
are
creating
their
release
artifact.
So
you
start
seeing
a
growth,
because
people
are
using
the
release
page
as
a
developer
as
a
release
manager
and
then
that
22nd
comes
and
then
we
have
a
steep
drop-off.
A
So
it's
almost
like
an
s-curve,
which
is
what
we
want
to
see
with
growth
metrics,
it's
a
trending
up
into
the
right
s-curve
and
that's
why
it's
a
good!
That's!
Why
I
think
it's
a
good
metric
for
for
the
release
activation,
because
we're
noticing
that,
as
we
offer
more
features
in
the
release
area,
people
are
viewing
the
page,
the
release
page
more
they're,
finding
it
more
and
more
relevant.
B
A
A
And
we
can
see
that
when
we
look
over
time
there
are
customers
that
are
looking
at
leveraging
the
release
the
release
functionality
for
showing
milestone
planning,
so
they're
gonna
package,
multiple
iterations,
within
a
single
release
object.
So
this
is
going
to
be
their
quarterly
release,
but
they'll
have
a
bunch
of
sub
releases
inside
that
quarterly
release.
So
we
can
see
that,
as
we
start
to
support
that,
like
group
level
functionality
of
releases
that
enables
that
nesting
of
releases
we'll
be
able
to
enable
those
users
that
are
interested
in
adopting
okay.
B
B
B
B
A
No
prerequisites
because
today
it's
pretty
uncoupled,
they
come
exclusively
in
the
planning
category
of
our
product
and
leverage
the
releases
functionality.
They
could
create
at
least
via
a
tag,
and
it
really
is
a
container
or
a
representation
of
the
code
at
any
given
time,
and
then
they
leverage
asset
links
and
uploads
to
that
release,
object
to
create
the
picture
of
that
deployment.
So
it's
really
like
representation
that
this
is
a
moment
in
time
for
this
code.
Okay,.
B
So
they
don't
have
to
UCI
if
they
use,
maybe
even
you,
they
use
source
code
management
like
most
people
use
that
they
can't
jump
to
our
clay
to
release
exactly
interesting,
yeah
I
think
I
think
that's
that's.
Maybe
that's
something
I
think
that
that
may
be
something
you
want
to
dive
into
later
as
some
dimensions
you
want
to
understand,
because
the
the
funnel
is
awesome.
One
way
you
can
get
more
inside
out
out
of
the
funnel
is
kind
of
think
about
possible
steps
before
or
some
some
dimensions
you
can
slice
the
funnel.
B
For
example,
people
visit
release
page
who
are
who
who
are
the
who
are
them
and
are
there
different
personas
or
are
there
different
people
in
that
and
when
you
break
down
that
release
page
view,
do
you
see?
Actually
there
are
different
type
of
people
and
maybe
their
conversion
rate
of
the
funnel
later
is
different
as
well.
That's
one
one
one
thing
and
then
also
maybe
about
because
release
is
a
stage
that
or
or
a
feature,
that's
not
in
the
very
top
of
the
feature
adoption
right.
So
what
are
some
different
scenarios?
People
adopt
other
features.
A
B
B
Yeah
I
think
the
when
you
try
to
improve
a
drop-off
in
funnel.
One
very
effective
approach
is
to
segment
basically
think
about
all
the
possible
dimensions.
You
can
segment
a
funnel,
because
it's
very
often
that
maybe
one
type
when
you
segment
for
a
certain
environment,
the
drop-off
is
not
bad,
but
certain
other
environment
or
certain
other
dimension.
B
The
drop-off
is
very
bad
and
then
what
you
can
do
basically
either
try
to
move
more
people
to
the
better,
funnel
meaning
try
to
push
more
people
to
use
the
environment
that
naturally
has
a
smaller
drop-off
or
get
get
more
personal
or
target
customer
that
naturally
have
a
better
conversion
right.
That's
one
thing:
the
other
thing
you
can
do
is
to
try
to
fix
whatever
that
doesn't
work
in
the
you
know
funnel
that
has
a
bad
conversion.
B
Maybe
there's
something
you
can
do
to
help
that
group
of
customer
or
to
support
that
environment,
to
make
that
conversion
rate
a
little
bit
better.
I
think
your
thinking
is
actually
pretty
advanced.
You
already
have
high
level
Northstar.
You
already
have
funnel.
You
are
already
thinking
about
how
to
make
the
conversion
rate
better.
So
that's
kind
of
one
direction
you
can.
You
can
explore.
That's.
A
Perfect,
this
is
exactly
what
I
need
to
like
iterate
on,
because
I
feel
like
right.
Now,
our
leadership
isn't
as
familiar
with
this
method
of
targeting
a
P&L
and
I've,
been
personally
like
I'm,
struggling
because
I
don't
have
a
contribution
margin
that
I'm
able
to
like
look
at
and
be
like.
Okay,
well,
I
know
that
I'm
driving
this
much
revenue.
How
can
I
activate
better
revenue,
whether
that
yeah
I'm
gonna
target
enterprise-class
features
to
make
sure
that
I
get
more
1
million
dollar
deals
firm
or
I'm?
A
B
We
shouldn't
work
together
on
that.
Actually,
that's
something
I
want
to
work
out
like
I
feel,
like
the
the
current
methodology
of
trying
to
look
at
small.
Only
is
hard
right
like
more
you.
People
using
it
doesn't
mean
anything
like
some
features,
naturally
have
a
lot
of
people
using
it,
so
how
we
can
tie
that
to
iacv
or
arr.
A
B
Think
I
think
we
obviously
have
like
the
the
challenge
I
run
into
is
we
have
so
many
features
and
I'm
not
super
familiar
with,
like
each
features.
Role
in
customer
journey,
like
which
feature
play
a
bigger
role
in
initial
purchase,
which
feature
is
more
like
a
retention
feature
which
feature
is
maybe
more
of
a
great
feature.
B
A
I
would
say
that
that
is
that's
awesome.
I
would
love
to
partner
with
you
on
on
that
front,
mainly
because
I
did
a
couple
of
opportunity.
Canvasses
and
I
started
like
looking
at
just
the
top
25%
of
our
renewing
deals,
the
next
couple
of
words
and
then
like
manually,
reaching
out
to
each
of
these
sales
people
and
being
like
hey.
How
important
is
release
governance
or
release
orchestration
to
these
guys
yeah,
and
what
tools
are
they
using
today
and
then
the
market
share
of
ZB
labs
and
cloud
me
are
really
high.
A
So
if
there's
two
solutions,
I
know
that
I
can
erode
that
and
how
upsell
of
existing
customers
yeah
influencing
net
new
prospects
is
challenging
without
having
the
feature
set
already
there.
So
then
that's
where
I'm
noticing
we're
very
good
at
influencing
expansion
of
current
all
base.
Our
players
are
very
geared
toward
getting
deep
into
enterprise-class
deals
and
we
have
been
able
to
influence
like
two
upsells
into
ultimate
mm-hmm,
the
financial
services
area.
A
A
B
That's
that's
a
good
approach
like
I
think
maybe
I
can
borrow
some
of
your
methodology.
Now
we
are
working
on
the
North
Star
metric
was
every
p.m.
so
I'm
thinking.
I
want
to
ask
that
information
from
the
PM,
at
least
as
a
first
step
seems
like
you
mentioned.
My
sink
is:
what's
your
feature
order
your
products,
roll
in
initial
purchase
versus
retention
or
expansion?
Like
basically
do
you
think
those
three
are
the
main
category?
B
A
I,
like
that
slice
that
segmentation,
because
then
it
kind
of
groups
together
where
get
lab
is
playing
today
and
yeah.
We
really
want
to
play
in
the
future
and
what's
it
gonna
take
us
to
get
there,
because
right
now,
I
see
product
managers
churning
and
crashing
with
this
idea
with
sales
I
want
more
ultimate
features,
but
what
they
mean
is
I
want
to
sell
more
competently
in
the
enterprise
space.
B
A
A
challenge
for
me
as
a
salesperson
trying
to
reach
a
quota
to
go
to
an
enterprise
customer
and
be
like
it
lab,
can
meet
your
needs
so
like,
but
I,
don't
think.
Pms
are
necessarily
thinking
when
they
hear
I
need
more
ultimate
features:
mm-hmm
oublie
as
a
get
loud
package.
They're
thinking,
oh
well,
in
release.
I
need
to
offer
more
of
this,
whereas
when
you're,
a
buyer
persona
you're.
A
B
A
B
Yeah,
so
the
approach
will
be
icing,
CID
and
CAD,
so
the
data
manager
and
telemetry,
so
the
high-level
approach
is,
you
will
create
a
ticket
in
data
board
and,
as
talking
about
here
is
the
selected
Northstar
metric
in
your
case
is
like
number
of
deployments
seems
like
right
or
or
or
like
I
would
suggest,
I
think
the
will
ask
for
alternatives,
because
they
they
will
help
you
to
see
that
what
which
one
is
available
like
maybe
deployments,
is
available.
Then
that's
awesome.
If
something
else,
it's
not
available.
What
are
alternatives?
B
You
are
willing
to
start
with
right
and
then
they
will
help
you
and
they
will
ask
you
some
questions.
Why
do
you
think
this
makes
sense
and
blah
blah
blah,
and
they
will
also
ask
you:
what's
the
current
situation?
Is
this
data
available
or
not?
Do
you
need
to
actually
track
this?
It's
if
it's
not
available,
then
in
it
there
are
some
instrumentation
need
to
be
done.
If
it's
available,
then
there
might
just
be
some
dashboard
data
team
can
help
you
viewed
right
and
and
then
so.
B
The
initial
effort
will
be
really
focused
on
North
Star,
but
you
can
add,
like
some
secondary
metrics,
just
to
have
it
there
so
that
they
can
think
about
it,
but
initially
I
think
because
of
the
bandwidth.
They
will
probably
mainly
focused
on
making
the
Northstar
metric
available
to
you
yeah,
and
do
you
know,
do
you
have
number
of
deployments
tracked
as
of
today?
Okay,
it's.
A
B
B
B
A
Can
for
like
inner
sourcing
companies
there
you
so
there's
short
answer-
is
there's
cool
and
use
cases
where
total
sense
to
derive
value,
but
for
most
get
lab
companies
that
are
looking
at
trying
to
further
their
CD
adoption.
The
continuous
delivery
adoption
releases
are
only
valuable
in
context
of
being
able
to
monitor
their
own
deployments
to
certain.
So
here's
a
really
good
example
say
that
I'm,
a
customer
and
I
have
300
end
users
that
all
have
15
different
versions
come
on
use
the
releases
page
to
show
that
I'm
deploying
this
version
to
this
AWS
environment.
A
This
is
AWS
environment
and
this
is
AWS
environment
chain
of
custody.
Of.
What's
that,
current
production
environment,
for
that
particular
customer
or
than
the
allure
version,
and
they
would
want
to
show
their
binary
attached
that
they
would
want
to
have
their
release
evidence.
They
would
want
to
be
able
to
show
all
of
that
information
in
one
place
and
that's
where
release
pages
will
eventually
get
to,
but
that's
kind
of
where,
like
I'm,
hesitant
to
to
declare
that
creation
of
release
is
a
defunct
just
because
we
haven't
built
out
the
full
functionality
yet.
A
So
it's
still
in
like
this
growth
mode,
where,
like
page
views,
are
still
kind
of
a
good
proxy
for
our
people.
Finding
in
release
pages
yeah,
we
haven't
offered
the
connecting
point
between
release
functionality
and
deployments
yet,
but
we're
going
to
like
that's
that's
eventually,
where
we're
going.
Okay,.
B
A
B
A
B
A
B
Okay,
okay,
so
it's
kind
of
like
sales
basic
give
you
a
rough
idea
right.
I
mean
that's
better
than
nothing,
but
is
I
think
it
would
be
hard
for
sales
to
attribute
to
a
single
feature,
though
I
would
imagine
right
like
if
you
ask
them
particularly
for
release.
If
you
have
this
feature
for
release
or
not,
does
that
change
yeah.
B
B
A
Do
you
see
that
today
it's
because
releases
and
deployments
are
so
highly
correlated
and
since
I
don't
have
like
count
of
releases
created
versus
count
of
releases
created
and
those
that
are
that
are
deploying
uh-huh
I'm,
not
able
to
ascertain
if
that
sounded
like?
If
we're
Fiverr,
there
is
I'm
retaining
because
I'm
deploying
with
gitlab
rather
entertaining
because
I'm
crated
releases
so
I
need
that,
like
release
is
created
first
start
out
and
then
I
can
show
the
Venn
diagram
of
like.
What's
attributable
okay,.
B
A
Okay,
okay,
but
we
do
have
some
compelling
ultimate
features:
okay
and
like
that's
more
like
expansion
of
base,
typically
customers
that
are
looking
at
us
at
a
whole
they're
like
I,
want
to
know
if
you
have
an
executive
dashboard
for
deployments,
and
that
is
in
my
area
right.
That's
the
environments,
dashboard
right,
okay,.
B
Yeah
cool
I
think
you
are
very
good
track
like
it.
A
lot
of
yeah
like
I,
think
you're
thinking
is
pretty
advanced
in
terms
of
like
the
metrics
and
Northstar
just
begin
to
build
the
funnel,
and
hopefully
we
can
begin
to
help
you,
like
figure
out
the
segments
and
different
ways
to
look
at
the
data
so
that
you
can
get
a
better
opportunities
figure.
It
out.
B
A
Like,
depending
on
who's,
renewing
and
after
I
have
a
conversation
with
that
customer,
it's
coming
up
for
renewal
if
they
want
to
have
more
front-end
work
and
their
deal
is
high
enough.
Then
typically
I'll
prioritize,
building
out
more
front-end
work,
because
the
release
manager
persona
can
be
technical
and
cannot
be
technical.
So
it's
it's
a
hard
like
right
now.
It's
I
I
sometimes
feel
like
I
miss
stepping
because
I
don't
necessarily
have
the
data
100%
to
validate
that
I
should
be
investing
deeper,
an
API
or
deeper
and
UI
functionality.