馃攳
SaaS metrics: the ULTIMATE guide to Software as a Service KPIs - YouTube
Channel: Slidebean
[0]
when you live and breathe software as a
[1]
service terms like ar mrr
[3]
ltv cac become just part of your daily
[6]
slang gross versus net mr churn and how
[8]
it's different from customer churn
[10]
it's a real mess so much so that for new
[13]
company employees i sometimes do like a
[15]
group class on
[16]
understanding sas to make sure that they
[18]
understand the language
[19]
i speak after six years running a sas
[22]
company i figured that instead of doing
[23]
that every single time we could make a
[25]
thorough video so
[26]
if you're just joining the sliding team
[27]
welcome to the team if you're a stranger
[29]
watching this on youtube well
[30]
i think you're going to find it useful
[32]
as well in this video we'll go over
[34]
the basics mrr and mrr we'll go over
[37]
expansion contraction and reactivation
[39]
mrr
[40]
we'll go over churn rate and the
[42]
difference between customer churn
[44]
grows revenue churn and net revenue
[46]
churn and then we'll go over
[47]
understanding what
[48]
negative turn rate is and why we love it
[51]
then we'll do
[51]
uh estimating lifetime value cost of
[54]
acquisition and lifetime value
[55]
to cac ratio and then finally we'll do a
[58]
little bit on understanding what cohorts
[60]
are
[60]
how to read them so let's start with the
[62]
basics mrr and
[64]
ar mrr stands for monthly recurring
[68]
revenue that means the number of
[70]
subscriptions
[71]
that you bill each month the number of
[73]
people who are subscribed to your
[74]
platform
[75]
and who will be billed automatically
[77]
unless they take action
[79]
and cancel their accounts mrr is the
[82]
reason
[82]
sas companies are exciting to investors
[85]
predictable revenue subscription
[86]
businesses can accurately predict how
[88]
much revenue they are going to make
[90]
in a given month if you multiply mrr
[92]
times 12
[93]
you get arr or annualized run rate same
[96]
as before this is a great metric to
[98]
estimate how much revenue the company is
[99]
able to
[100]
generate on a yearly basis now assuming
[103]
that you don't add any customers but you
[105]
also don't lose any customers
[107]
the arr number paints a very good
[109]
picture of the annual revenue
[110]
company will generate when a customer
[113]
subscribes to an
[114]
annual plan say at ninety six dollars a
[116]
year that's just a random number
[118]
unrelated to slight bean
[119]
the opposite happens since ninety six
[121]
dollars is the annual revenue
[123]
the customer will bring the impact of
[125]
ninety six dollars is reflected directly
[127]
on the annual run rate the arr
[129]
that means that to estimate the monthly
[131]
impact of those 96 dollars you have to
[133]
divide them
[134]
by 12 in other words a 96 annual plan
[137]
means
[137]
96 dollars in ar but only eight dollars
[140]
in
[140]
mrr now arr would be beautiful if it was
[143]
guaranteed but there are always going to
[145]
be cancellations luckily
[147]
there's a simple way to estimate how
[148]
much time a customer will stay
[150]
based on historical customer behavior
[153]
and that's why the churn metric
[154]
is so critical for sas and we'll spend
[156]
some time
[157]
talking about it now let's assume that
[159]
this platform is called
[161]
slight bean it's a super cool pitch deck
[163]
builder powered by ai
[164]
and it subscribed 100 new users on a 29
[168]
a month plan in january that means that
[170]
the new mr metric for january is
[172]
2900 that's easy let's go to
[176]
customer churn rate in february this
[179]
company adds another 100 users that
[180]
means that they added
[182]
another 2900 in mrr however
[185]
they lost 10 customers from the january
[187]
cohort
[188]
a cohort is a group of users who share
[190]
characteristics such as
[191]
the month in which they subscribed or
[193]
maybe the campaign that brought them
[195]
in so just to keep track this is how the
[197]
mrr movements table looks like
[199]
the company added 2 900 an mrr but lost
[202]
290 or 10 customers by the end of
[205]
february they have 190 customers
[208]
and 5510 dollars in mrr alright so let's
[211]
estimate
[212]
the churn rate out of those hundred
[214]
users we lost 10.
[216]
that means that our customer return rate
[217]
was 10 the formula for this
[220]
is the number of users lost divided by
[222]
the number of users
[224]
at the start of the period that means we
[226]
have 10 divided by 100 remember we are
[229]
not counting the 100 users that we added
[231]
in february
[232]
those are new customers who weren't here
[234]
at the beginning
[235]
of that period now churn is an essential
[238]
formula if you divide
[239]
one by a churn rate in this case one
[242]
divided by ten percent or one divided by
[244]
0.1
[245]
you'll get the average lifetime for a
[247]
customer that is the number
[248]
of months so let's talk about customer
[251]
lifetime
[252]
a customer churn rate of 10 means that
[255]
the average the average customer will
[257]
stick around for 10 months
[259]
if you actually break that down you'll
[260]
see that a cohort of 100 customers
[262]
where you lose 10 of those customers
[265]
every month means that by month
[266]
10 you'll have around 38 customers by
[269]
month 24
[270]
two years later you'll still have around
[272]
nine customers even at month
[274]
45 you're still expected to have one
[277]
customer
[277]
so again what the churn does is it just
[279]
gives you the average lifetime of all
[282]
of these customers with a customer churn
[283]
rate of 5 the average lifetime would be
[285]
around 20 months
[286]
a turn rate of 3 means that the average
[289]
lifetime is
[290]
33 months and so on now the lower the
[292]
churn rate the longer the average
[294]
lifetime
[294]
now if you have the average number of
[296]
months that a customer is staying on the
[297]
platform
[298]
and you know on average how much that
[300]
customer pays
[301]
you you have all the information you
[303]
need to come up with a lifetime
[306]
value using this sniping example again
[308]
let's say this fictitious
[310]
but totally awesome company had a 29
[313]
a month plan and a 12 dollar a month
[316]
plan so let's say the company had
[318]
50 000 in mrr and 2000 customers you can
[321]
divide the mrr number
[322]
by the number of customers and get a
[324]
number called the rpoo
[325]
the average revenue per user in this
[328]
case it's
[329]
50 000 divided by 2000 that means
[332]
25 some people also call that number the
[335]
our pa or average revenue per account
[338]
this is mostly to distinguish that the
[340]
rpu or rpa
[341]
metric tracks your accounts not your
[344]
users a company using slack may have
[346]
50 or 1 000 employees and the sas
[349]
metrics team at slack mostly cares about
[351]
the average revenue
[352]
per company or per account not for each
[354]
human
[355]
being using slack so if we know that on
[358]
average an account is worth
[359]
25 a month for us and with a churn rate
[362]
of 5
[363]
we know that the average time with us
[365]
will be 20 months
[366]
we can multiply our poo 25 by months of
[369]
lifetime
[370]
20 to get 500 that is our customer
[373]
lifetime value and we can simplify that
[376]
formula even more
[377]
all right great so the formula for
[379]
lifetime value is
[380]
arpu divided by the customer churn rate
[383]
the most basic business rule is that
[385]
whatever you're paying to bring
[386]
customers in needs to be less than the
[389]
money that you generate from set
[390]
customer
[391]
and the customer left and value is a
[393]
great number to help you estimate that
[395]
if based on your churn rate and your
[397]
average revenue per user
[399]
you can estimate that the lifetime value
[400]
for a customer is a thousand dollars
[403]
you know that you can be aggressive and
[405]
spend 400
[406]
to acquire that customer because you
[408]
know that eventually
[410]
you are gonna make that money back as a
[412]
general
[413]
rule of thumb it's considered ideal that
[415]
the cost of bringing a customer
[417]
is one third of the lifetime value or
[419]
less so
[420]
for the fictitious slight bean with a
[422]
500
[423]
lifetime value we should aim to spend
[425]
around 160 dollars
[427]
to acquire a customer for the longest
[429]
time our cac or cost of acquisition
[432]
was much lower than that actually which
[434]
was a sign we could afford to spend more
[436]
aggressively
[437]
on customer acquisition are you still
[438]
following me after all this
[440]
here's a five second video of a puppy to
[442]
give you a break
[443]
okay so now let's get into the advanced
[445]
stuff
[446]
revenue churn all of this time i've been
[448]
speaking about customer churn
[451]
the number of customers that we lost
[453]
versus the number of customers that we
[455]
had
[455]
but you can also measure churn from the
[457]
revenue side
[458]
the amount of subscription revenue added
[461]
versus the amount of subscription
[462]
revenue
[463]
lost in our most basic example the one
[465]
we used before
[466]
we lost 10 customers from january to
[469]
february
[470]
that's a 10 customer churn now if we
[473]
measure that by the amount of revenue
[475]
lost
[475]
then the formula would be very similar
[478]
in that case since we
[479]
lost 290 we divide that by 2 900
[483]
the revenue we had at the beginning of
[484]
the period and we get
[486]
10 and that is called our gross
[489]
churn and oftentimes the number is quite
[492]
similar to the customer churn number
[494]
so why do we measure those separately
[496]
well
[497]
because you can lose revenue without
[500]
losing
[500]
a customer in our example we had every
[502]
single customer on a 29
[504]
a month plan so each customer lost
[507]
represents a revenue loss
[509]
of 29 however if you have different
[512]
plans you might have people downgrading
[514]
a plan and let's look at that in a
[516]
visual version say in january
[518]
fictitious slide being added 30
[520]
customers to its starter plan
[521]
which is 12 a month and then they added
[524]
70 customers to their premium plan which
[526]
is
[526]
29 a month that's 30 times 12 plus
[529]
70 times 29 total of 2 390 dollars
[533]
great now just like before by february
[536]
we are going to lose 10 customers
[537]
remember
[538]
we might add new customers in february
[539]
but those are not affecting our churn
[541]
metrics
[542]
since they're all new customers we're
[544]
just comparing the customers that were
[546]
already with us
[547]
at the beginning of the period okay so
[549]
let's say that we lost
[551]
five customers from the premium plan and
[553]
five customers from the starter plan
[555]
the amount of lost customers is 10. from
[558]
a batch of 100 customers
[559]
that is that 10 customer return that we
[562]
had seen
[562]
before however the amount of lost mrr is
[566]
205 dollars because there's a mix of
[568]
people from different plans
[569]
that translates into about 8.7 percent
[572]
in gross revenue churn notice how it's
[575]
different
[576]
from that 10 customer churn rate
[587]
now when you are dealing with multiple
[589]
plants you might also have people
[591]
switching between plants following the
[593]
same example
[594]
let's assume five customers from the
[596]
premium plan
[597]
downgraded to the starter plan so seven
[600]
people went from paying 29 a month to
[603]
pay
[604]
12 a month and that is a contraction of
[607]
revenue
[607]
of 119 now that gives us a
[611]
gross mrr churn rate of 13.6
[615]
and as you can see in this particular
[616]
scenario our customer churn rate was a
[618]
bit
[619]
deceiving remember that customers that
[621]
downgraded a plan are not counted as
[623]
lost customers because they are still
[624]
paying for
[625]
a plan it's just a much cheaper plan
[628]
so we might think that churn is 10 but
[630]
not really
[631]
we are really losing 13.6 percent of
[634]
revenue every month
[635]
and that's gross revenue churn and it's
[637]
an ugly number but there is some good
[639]
news
[640]
still customers might upgrade their plan
[643]
while we have contraction revenue
[645]
we might also have expansion revenue
[647]
from customers moving from the starter
[648]
plan
[649]
to the premium plan or for example
[651]
adding new seats to their accounts
[653]
remember we are measuring accounts not
[655]
individual
[656]
seats we might also have reactivation
[659]
revenue those are customers that
[660]
canceled and that came back to
[662]
reactivate their subscription
[664]
technically that's not considered new mr
[667]
because these were our
[668]
customers already so it's classified
[670]
separately it's classified as a
[672]
reactivation mrr now
[674]
the sum of all those variables
[675]
reactivation expansion contraction
[678]
and lost customers gives us a number
[680]
that we call the
[681]
net mrr turn or the net revenue turn
[684]
and at least for us at slide bean it is
[686]
the most important number
[688]
of all so let's look back at our 13.6
[691]
gross churn scenario let's assume that
[694]
while we had
[694]
cancellations and contractions we also
[697]
had some expansion mrr
[699]
15 customers decided to upgrade their
[701]
starter plan
[702]
to a premium plan maybe they wanted to
[704]
track investor views on their pitch
[706]
decks
[706]
maybe they wanted access to our premium
[708]
pitch tech templates who knows why they
[710]
upgraded but they upgraded
[712]
so that translates into 255 an
[715]
extra mrr from the same group of users
[717]
that we had at the beginning of the
[718]
period once again
[720]
these are not new customers these are
[722]
existing customers
[724]
also let's say that we saw seven users
[726]
come back from the debt
[727]
they reactivated their accounts some of
[730]
them on the starter plan
[731]
and some of them on a premium plan and
[732]
they added 169
[735]
in subscriptions so how do we estimate
[737]
the new
[738]
churn rate this is the formula
[748]
so their result here is minus
[751]
4.18 and don't worry this isn't your
[754]
rusty
[754]
high school math skills the result of
[757]
this formula
[758]
was indeed minus 4.18 so what does that
[761]
mean
[762]
it's negative churn yes it's precisely
[764]
that
[765]
it's sas nirvana really it is sas
[768]
nirvana negative churn is the number
[769]
that gets
[770]
sas companies funded negative churn
[772]
means that
[773]
every month your existing customers are
[776]
growing they're paying you more every
[777]
month you make
[778]
more money from the same batch same
[781]
cohort of customers
[782]
on top of adding new customers so
[785]
negative mri turn is a beautiful thing
[787]
to see
[787]
and one that all sas ceos strive for
[790]
okay so i hope that made
[792]
sense let's look at the very last part
[793]
of this which is cohorts
[795]
for every single metric that we've
[796]
talked about today we are comparing the
[798]
lost
[799]
or earned customers in a period versus
[800]
the number of customers
[802]
on the previous period if we ran these
[804]
metrics for june the formulas would be
[806]
the same we'd be comparing
[808]
lost contraction expansion and
[810]
reactivation customers
[811]
that occurred in june regardless of when
[814]
they signed up
[815]
for the platform and this is great
[816]
because it paints a clear picture
[818]
of your average retention rate and
[820]
that's the number that you're going to
[821]
need to project
[822]
your future revenue however sometimes
[824]
you want to look at a specific cohort of
[826]
users
[827]
the users that convert it on a given
[828]
month and you want to measure their
[830]
behavior against
[831]
users that came before or after and this
[833]
is useful for example if you opened a
[835]
new marketing channel in a given month
[836]
or if you ran a promo
[838]
discounts tend to bring low quality
[840]
customers you can also use cohort tables
[842]
to measure your effectiveness when
[844]
increasing
[844]
onboarding or retention and this is how
[847]
a cohort table
[848]
looks if you look horizontally you're
[851]
looking at the cohort of users from a
[853]
given month and how retention has been
[855]
affected over
[856]
time if you look vertically you're
[858]
looking at
[859]
the evolution how many customers are you
[861]
managing to retain
[862]
by month six or by month nine and how is
[865]
that evolving
[866]
with new cohorts coming in every new
[869]
month
[869]
and you can run cohort tables based on
[871]
your customer turn or your gross churn
[873]
or ideally based on your net mrr churn
[876]
and here you'll see
[877]
the negative turn reflected as a
[878]
retention rate higher
[880]
than 100 that means that you're making
[882]
more than 100
[883]
of revenue from those customers now i
[886]
have seen people try and run these
[888]
metrics manually i have seen
[890]
it but that is just silly if you ask me
[893]
it is a real mess to come up with these
[894]
formulas
[895]
and it is a real mess to make everything
[896]
get tracked correctly
[898]
add a complication if you're billing
[900]
some of your customers
[901]
say via stripe and some of your
[902]
customers be at paypal you have
[904]
two data inputs that look drastically
[906]
different you want to consolidate all of
[908]
this into a spreadsheet
[909]
that actually makes your life easier so
[912]
every single chart that you saw in this
[913]
video comes from church mogul
[915]
who also happened to sponsor today's
[917]
video we've been using chart module
[918]
for four or five years almost since we
[920]
started slide paint
[921]
thermogl is a subscription data platform
[923]
that plugs into your payment systems
[925]
like stripe
[926]
or paypal or chargebee and automatically
[928]
generates all the subscription analytics
[930]
you need to run your business
[932]
like a pro now i met nick the ceo of
[934]
churchmodule way
[935]
way back when our youtube channel sucked
[937]
and we did a small interview that had
[939]
like 500 views
[940]
but when he heard that we were doing
[941]
this video on sas they agreed to sponsor
[943]
it and offer you guys a deal now chart
[945]
module is free for any company with
[947]
under ten thousand dollars in mrr
[948]
i hope that by now you understand very
[950]
well what mrr means
[951]
if you have over ten thousand dollars in
[953]
mrr you can get a 50
[955]
a month credit for 12 months so that's
[957]
600 in credit
[958]
total by using the code slide bean when
[960]
you activate your account
[961]
this code will only be valid for a month
[963]
after this video releases
[965]
so make sure you redeem it as soon as
[966]
possible i hope that this crash course
[968]
was useful
[969]
once again if you're joining slide being
[971]
welcome to the team for the youtube
[972]
people
[973]
if you like this video make sure to hit
[974]
that subscribe button and if you think
[975]
other founders
[976]
around you will find it useful we
[978]
appreciate a share
[980]
we'll see you next week
[989]
[Music]
[996]
you
Most Recent Videos:
You can go back to the homepage right here: Homepage