Building a Finance Model that scales as you grow
Kirk shares his journey in handling forecasting, budgeting, and resource allocation in dynamic business environments.
You Might Also Like
I'm always wondering what's the new SaaS metric that's going to come out.
You know, we've got Rule of 40, we've got the magic number,
we've got semi-new ones like the Rose Formula,
which is just essentially ARR overhead count,
but includes contractors as well as contractors and consultants become more
important.
And what I've realized is those metrics are important.
What's more important is how you evaluate them.
[MUSIC]
Hello everyone, welcome to yet another episode of the Optimized Show,
brought to you by Spamflow.
I'm Adi, the Chief of Staff at Spamflow, and on today's episode,
we are going to be talking about building a financial model that scales as you
grow.
And who better to talk about it than Kirk Kapilal,
a seasoned professor with a strong track record in business modeling and aud
iting
at Deloitte KPMG and EY.
Today, he's the Director of Strategic Finance and Drive Train,
a financial planning and decision-making platform.
Welcome, Kirk, it's my pleasure to host you today.
Adi, absolutely pleasure to be here.
Thank you so much for having me.
So, part, to start off with, why don't you tell us a little bit about yourself
and your experience in finance?
Definitely, yeah, finance always piqued my interest.
And going out of college, I got a job at Deloitte in their auditing team.
And auditing is obviously very important and focuses a lot on the historicals.
And the historicals are really important from a accounting perspective,
but what really grabbed my eye was forecasting and looking ahead.
How can we forecast out the model to make predictions, let's say, about revenue
or expenses and those sorts of things.
And so I found that opportunity at EY.
And so I went over to EY and really got into modeling,
creating complex models, and then found a very specific team at KPMG
that did financial modeling itself.
So it was specifically focused on the three statement models,
the income statement, the balance sheet statement of cash flows,
and then anything that might feed those in terms of schedules.
And I had some really good mentors there.
They helped me learn how to create models.
But then, of course, every consultant eventually leads to the private world
and drivetrain, which, by all intents and purposes, is the future of modeling.
Had an opportunity to be the director of strategic finance there.
And I interviewed, met with the team, saw the product.
And from an Excel perspective, I just couldn't imagine anything more efficient.
And so that's what made me make that jump there.
And so I'm still a lover of Excel, but at every point in a company's life cycle
they need to move to a tool.
And I just couldn't think of a better one than drivetrain, hence my jumping
there.
So, and I'm pretty sure you're the manager to talk about today's topic,
which is like building financial models at Skia, Ashigro.
And I'll talk typically what I'd like to do is, like, sort of dissect the topic
that we're talking about before we head into the different questions.
From your perspective, why is scalability important when you're building
financial models?
Yeah, no, that's a really good question.
So, I mean, there's a variety of ways to think about modeling.
I like to break it out into three sort of major categories.
One is people, the second is process, and the third is practice.
And so the people question is, how big is your team?
How much time can they commit to these processes?
Process would be, you know, how currently are you updating your model?
Actual is rolling the forecast, reforcast, and then practice, which is also
known as industry.
And this will obviously determine your kind of model.
When we're thinking about scaling it, we're trying to make the people process
and practice questions as easy to answer as possible, meaning how,
what's the fewest amount of people I can use to maximize my efforts in the
process part?
What's the fewest amount of processes I need to complete in order to create an
accurate forecast?
And then from a practice perspective, this changes from industry.
If you're in the SaaS industry versus, you know, the biochemistry industry,
your model is going to look a little bit different.
But the goal is, at a scalable perspective, how do I simplify it
for each of those different levels there?
Sometimes in the beginning, that's Excel.
For more, you know, seasoned companies, that's going to be something like a
tool.
So when we're thinking about scalability, we want to ease the burden on the
financial team
and finance and accounting in general.
And so those are the three kind of major categories that I think about
and why it's so important to scale it in the right amount of time.
I mean, that's a very interesting perspective.
I mean, like, you did touch upon how the journey of the company goes from Excel
to something more advanced and complicated, right?
So to zoom in a little bit onto that, so how do you think financial models
of early stage companies differ from those of late stage companies?
Yeah, yeah, really good question.
And it is a drastic change because, of course, in the beginning,
when you're starting a company, you likely do not have any revenue
or much revenue at all, build models for several companies
where they were forecasting their revenues two to three years out,
but had expenses today.
So when we're thinking about a financial model from their perspective,
they're likely thinking from a market perspective
or a total addressable market perspective,
meaning what's the total market out there that I could address?
How much of that market might I be able to capture?
And then how much of that addressable market that I could even capture
is going to sign up to my product over all of the other products there.
And that's a very high level answer.
But, of course, eventually every company,
if they're doing the right thing, they're going to start making revenue.
And about six to 12 months in to them making revenue,
their model has to change, right?
And this will probably be the first change that they make
from a modeling perspective or forecasting perspective,
because now they have a base that they can work off,
meaning they have a customer list.
Those customers are paying the money.
So it could be as simple as a P times Q model,
where we're essentially taking the price of their product,
the quantity of the product sold or the service given,
and then forecasting it forward.
I've seen other methodologies where they use sales reps,
assuming sales reps should be able to make $40,000 of sales per month or per
year.
And then you just add sales reps as you want your revenue to grow,
hoping that they meet their quotas
and using that as sort of a forecast methodology.
Expenses seem to find their place in the world based on revenue.
This is why I typically talk about it based on revenue.
Financial modeling, by far the most difficult part of it,
is getting the revenue piece right.
If you get the revenue piece right,
your expenses, they matter less,
because you can control for your expenses.
Revenue you can't control for, of course,
otherwise you would likely have billions of dollars from month one.
But expenses themselves, they land into place when you get your revenue right.
That's great in Saikou.
Just to take a step back and discuss financial modeling in general.
So how do you think financial modeling has changed over the past few years
in the context of new technology and, of course, AI?
So I would say, conceptually, financial modeling hasn't changed that much.
But methodically, it has changed substantially.
What I mean by that is this,
the way people are thinking about forecasting likely hasn't changed.
It's going to fall into one of those buckets I mentioned before
based on industry, P times Q, total address for market,
if you're a SaaS model, you're going to use the ARR, MRR corkscrew,
and we can talk about that a little bit later.
If you're an airplane company,
you're going to use a number of seats or utilization or rent,
or these sorts of things, those pieces are likely going to stay the same,
because those are very industry specific.
What has changed a lot is obviously the technology has increased.
And there's a lot of non-value ad processes in Excel modeling.
Even though it's very quick and you have a lot of flexibility,
what's not quick and what is really non-value ad is the process of going into
NetSuite
or your other ERP, downloading your actuals,
formatting them and then putting them into your Excel model
to then compare budget versus actual.
That process is something that 10-year-olds can do, right?
Everyone knows how to download Excel documents,
and without much practice, you can start formatting them.
What we want to do is eliminate those non-value ad processes
by adding in a system that's going to do it for you.
And this is the induction of FP&A tools that have come in
to try and centralize a lot of your systems, your CRMs, your ERPs, your HRIS
tools.
And so one thing the Drivetrain does is it centralizes those data,
allows you to create metrics off of that data,
and then all updates these items.
So essentially that data that you would have to manually pull yourself
now is just automatically already flowing in at the hourly level,
and you don't need to think about it anymore.
Now you can either rest, which is much needed in the financing accounting world
or spend time on real insights, drilling down into the data,
finding correlations, finding perhaps causations,
and making your forecast that much more accurate.
That's very interesting.
I mean, since we work with a lot of finance leaders here that's been for as
well,
and whenever we get to meet CFOs during the events that we host and these like
that,
one of the major topics that comes up at least in between the discussions
between the CFOs is like,
hey, what are the different financial modeling approaches that are out there,
and like which one to choose considering the complexity, accuracy,
strategic planning, and a bunch of other things.
So like, how can financial leaders determine the most suitable strategy for
their companies,
these and goals?
Totally.
Yeah, and this is where that people process practice item that I was mentioned
before,
super important because you can make one level of detail model.
If you have five people on your FBA team, you can make a much less detailed
model
if you only have one person or let's just say half a person,
commonly, you know, FBA teams are built out of the accounting team by somebody
just wants to take on the role.
And so really it's a half role in the beginning.
So from a people side, how many people do you have to really commit to this?
A lot of tools are aiming to relinquish the need for adding FBA people so you
can keep your team
small, let's say two or three people.
But even that's becoming difficult as, you know, sort of your capacity is being
lost
and models are becoming more cumbersome.
And so really what you're trying to answer is this question is,
how complex do I want my model to be and therefore more accurate?
And how simple do I want my model to be but not too simple that no one's going
to believe in it.
The extreme example would be just slapping on 10% at the end of each year and
saying,
I'm going to grow 10%.
People are immediately going to look at that and say, why are you growing 10%?
I need justification behind those numbers.
The alternative is go down to the most deep level that you can by region, by
industry,
by segment, by product line, by price.
And sometimes that can be more accurate but not so much more accurate that it's
worth
all the extra effort that you're going to put in.
Right? So that's the process piece.
The people, how big's my team, the process, how far am I going to make this
thing accurate?
Where's the balance?
Can I do, you know, 60% of the work to get 90% of the accuracy?
And the extra 40% of the work's only going to give me that extra 10% of
accuracy.
And it's obviously a lot more effort throughout my month and I'm going to burn
out my team.
That's the balance that you want to find.
And the practice piece we obviously spoke about when it comes to the industry.
This is really going to determine, you know, especially in the SaaS industry,
how you're going to model things out.
Most, most industries have a pretty sound, let's just say, baseline of
forecasting.
For example, in the SaaS industry, it's the ARR corkscrew model or the ARR
waterfall,
where you're going to go from, you know, beginning ARR, contraction, expansion,
churn,
new, and then ending ARR. That allows you insight into, you know, how are we,
you know,
forecasting SaaS at a more granular level, but not so granular that I'm losing
efficiency.
That's for SaaS. But then a total addressable market model might be used for
something in the
biochemistry sphere when you're talking about new pharmaceuticals or something
to that effect
coming up. And so they're very different, but they're also important to get
right.
So that, like you said, we can have a scalable model that is effective for my
company,
but not too much where I'm burning out the people that's working for me.
So that was very insightful, Gug. You did touch upon people's process and
practice, right?
And basically how you select the strategy depends on the complexity and how far
you want to take
it in terms of the financial modeling. Now, when it comes to execution, so as
per the complexity
or like as per the requirements, what sort of tools or like strategies should
financial leaders
take in order to actually execute it and practice, right? So at what stage is
an excel enough?
And what stage do you really need like a complex tool that works with like a
bunch of other tools
that are using as well? Totally. Yeah. Well, I have a pretty strong opinion on
this. And I actually
just did a podcast with the SaaS CFO Ben Murray, where we actually aim to
answer this exact question.
When does a company move from Excel to a stronger Fpna system? And the question
, let me just say
off the bat is not based off of dollar of earning. A lot of people will say,
you know, something to
the effect of if you have a million of ARR or 10 million of ARR, now you're
ready to, you know,
move to a tool that doesn't work. And I can tell you why. I've worked with some
companies where
their products or services cost $100,000 just for that one product or service.
I've worked with
other companies where that same, you know, their equivalent of the product or
service is $2.99.
So if we're talking about $10 million of ARR, for one company, that's going to
be a lot more
transactions than another, obviously. And so the question isn't, I move based
on dollar amount.
The question is I move based on transaction amount. Once I reach a certain
number of transactions,
which is essentially too cumbersome for my Fpna team to handle within Excel,
you know,
Excel has only got what is it? 1.04 million lines, right? There are some
companies that will
have that many transactions. It's time for you to move over even before that in
anticipation of this.
And I say this to a lot of companies. It's not usually too soon to move to a
tool. If you believe
in your growth trajectory and you've reached a certain point where even with
one person,
the Excel modelings become too too cumbersome, it could be a good time to move
over to a tool.
Why? You're going to get there eventually and it's better to learn it upfront.
And what we've
seen a lot of the time is people in the industry are adding these tools like
drivetrain. And they're
seeing, oh, hold on a second. Now I don't need to hire another Fpna analyst
because my modeling
has become so simple that I don't now need to go in to add another person
because all of this
can be managed by one person alone or two people alone, whatever the case may
be. So it's based
on a transaction amount, not on a dollar amount. And not only are you going to
save time on the
modeling piece, but it's going to be simpler as well. It's a double win there.
Of course,
you mentioned that one of the metrics that's going to be super important is to
transaction
about or rather than the volume amount. So what are some other key financial
metrics,
at least in the SaaS world, that finance leaders should consider? And how can
they be incorporated
into a financial model? Yeah. Well, at this point, I was talking with a
colleague about this as well.
I'm always wondering what's the new SaaS metric that's going to come out? You
know,
we've got rule of 40. We've got the magic number. We've got semi new ones like
the Rose formula,
which is just essentially ARR overhead count, but includes contractors as well
as contractors
and consultants become more important. And what I've realized is those metrics
are important.
What's more important is how you evaluate them. So you look at something like
ARR or MRR. I
mentioned before the ARR corkscrew model or the ARR waterfall. This is so
important to get right,
because if you get this right, you have a very predictable forecast and a very
predictable plan.
If you get it wrong, then you can throw away all your predictability. And as a
result,
you're probably going to overspend or underspend based on what your company can
do. So what that
really means is this, you have a beginning ARR or MRR amount. And you take a
look at that at a
monthly level in January, let's just say it's 10,000. And then you've got two
expansion features.
One's code new ARR, one's code expansion ARR, new is truly new customers
expansion is maybe we have
customers upgrading to a premium service that we have. And then you have the
opposite of those,
which is contraction ARR, the opposite of expansion, or just churn, which is a
customer that's left
us. Now, the reason I highlight it's why it's so important to get this right is
because, you know,
a drivetrain, we've got a lot of customers and I don't know if two customers
actually calculate
this the same way. And these are very common metrics, but everyone does it a
little bit differently.
I've seen some customers that any customer that enters the door for them is new
ARR or new MRR.
And if they leave, it's churn. But then I've got other customers where they
look at it like this,
since they give a three month free trial or a freemium sort of plan, those
sorts of things,
they do not count a customer that has entered within three months and left
within that same
three month period as either new or churn. They don't count them until the
fourth month, right?
And so this is obviously very business specific. You want to make sure that it
makes sense for
your business to get it right. Every product and service is a little bit
different and you
want to make sure you account for that. If you can get those pieces correct,
new churn,
expansion, contraction, you will have a very predictable SaaS model. Once you
've focused on those,
then worry about the other thing. Worry about your rule of 40, worry about your
magic numbers,
worry about your rose formulas. Those things are important and you can
definitely calculate those,
but you're not going to get any of the complicated sort of fun metrics, right?
Unless you get the
foundational ARR and MRR views correct. Undisturbed. Of course, you talked
about the four major
metrics that the financial leaders need to keep in mind before they start
building out the model.
But typically what I've seen, at least happened with a lot of companies, is
that when they're
starting out, at least in the early stages, they try to be a little bit more
scrappy, right? Like
rather than worry about scalability, they start up like, "Hey, you know what,
something's better
than nothing. So I'm just going to get started with this." And then sooner than
later, they actually
find out that, "Hey, you know what, it's not working." And then they try to
model and things
like that, right? So when you're starting out, or ask your building out your
financial model,
so how can financial leaders ensure that their models remain scalable as their
companies grow
and evolve? Yeah, well, I would say this. You're going to make your models more
scalable.
If you're doing things more or less right from the beginning. So how can you do
that? There's a lot
of great resources out there, right? On LinkedIn, you're going to find a
variety of items on how to,
you know, track your SaaS metrics properly, how to build a financial model. At
DriveTrain,
we have an endless blog full of methodologies and financial models that people
can download.
And then these are all in Google Sheets right now, where people can take them
over,
model out their SaaS business within those sheets, so that when they are
already jumped to a tool
like DriveTrain, they've already been thinking about it the right way. They've
been thinking about
it the way that systems interpret data. They've been thinking about it the way
that their business
and industry needs to interpret that data. So if you start small, let's just
say you downloaded
a DriveTrain Google Sheets model, and you're modeling out in that way. By the
time you get
to DriveTrain or to a tool like DriveTrain, you're already going to have been
doing it
right the whole time, that you know that the lift is not going to be that hard,
right? So get it
right from the beginning. If you're already in a little bit, refresh, right?
There's always a new
template that you can download, rethink it. Eventually, if you're going the
wrong way, the
person that turns back sooner and goes the correct way is the person that's
ahead, right? So just
get it right in the beginning, and then further down the line, you're going to
thank your past self
a lot when you have to then do a full implementation of a brand new system and
rethink how you're doing
your modeling. So, Kirk, we've been talking about what people should be doing
to build out the
scalable financial model, right? But now I just want to touch upon something
that they should not
be doing. So what are some common pitfalls to avoid when people build out their
financial models?
That's a really good question, and you probably saved people hours just by
asking the question.
So the biggest, by far, the largest pitfall that people fall into when they
want to do modeling
is they want to do the classic financial accounting methodology of getting it
right to the scent.
And in historicals, that very much matters. Auditors care about that, account
ants care about it,
they have to care about it. We're dealing with real historical numbers that
actually happen,
and we need to have a level of accuracy, a threshold that is much higher than
forecasting.
You do not need that on the forecasting side of the FP&A side. And if you do
focus on that,
you're going to end up on that spectrum on the right side that I was saying
before,
where you get so complex that your model to update it is two full days of
effort.
All of these systems worth of downloading, you're going to be spending two
hours formatting,
and I've fallen into this trap before, and I'm sure every finance leaders had
this trap
before, where they're forecasting at such a deep level because they want it to
be accurate,
that it ends up costing them two or three financial days a month. That is the
biggest pitfall to
avoid. If you can find the balance between this is concise enough, this is
complex enough,
and it will not charge my team three days across all of them and make them
stress out that they've
missed one sale that's going to break the whole model across a 42-tap model.
That's what you need
to focus on. If you can get that piece right, more or less you're likely going
to have avoided a lot
of the pitfalls because in Excel, especially having a single formula across
every single row is very
important. There's always going to be the need for a hard code somewhere for
some reason.
The smaller the model, the more simple the model, the more likely that's to
occur, but of course,
not too simple. Otherwise, people aren't going to trust your model at all. They
're going to think
it's just something you superfluously made in order to appease the CFO.
Super, thank you so much for that, Kirk. I mean, the thing you mentioned always
but I'm actually thinking you'll see a day for some of our listeners out there.
Of course, finally, I want to touch upon a few emerging brands in financial
modeling. What do
you think are some developments that happened recently that the finance leader
should be aware
about? I think everybody's a keyword or a hot word is AI. Everybody wants to
see how this
AI piece is going to layer on top of financial modeling to essentially improve
the process
and make it easier for FP&A teams to get more accurate results and make a model
quicker.
I do think from what we've seen, there's a lot of room for that. In financial
modeling,
you've got two pieces working together. There's the science piece of it, which
is the accounting
and finance. This is true accounting. You've got your income statement,
your balance sheet statement of CFO. That scientific piece of accounting is not
changing.
That part is where AI can really help because those scientific pieces is what
AI just excels at.
The other piece of it is the artistic piece. This is the creativity of how am I
thinking
about my model? I don't have to think about the SaaS industry in a specific way
. I don't have to
think about the airline industry in a specific way. That piece means that the
user is able to
create a financial model that makes sense for them or for their individual
company.
Where I think AI can help is on the first piece, on the scientific piece,
on the second piece, it can definitely make an attempt and add some value, but
it's going to be
much tougher to replace the time that people are spending on the financial
modeling there in that
case. One thing that it can do is let's just say you have a financial model and
you were to ask
some equivalent of chat GPT. Give me five takeaways about this model. Of course
, in the future,
we're going to be able to see answers such as you missed your budget likely
because or likely driven
by three of your five sales reps who didn't hit their quotas. Had each of your
sales reps
increased their quota by 10%, you would have exceeded budget by 20% or
something to that effect.
So these insights that we can draw from it, those will be extremely valuable
because rather than
spending all this manual effort to get those insights, AI is going to be able
to give it to us
really quickly, which will just essentially make the market substantially more
competitive because
now instead of three months down the line and increasing the quotas for your
sales reps because
you already know that they're at capacity, you can do it three months quicker.
Imagine that
multiplied by all of the other insights that could come out of financial
modeling and you have a
very competitive market accelerating very quickly. I also foresee this as a
result,
having less people in the FP&A space. Even though at drivetrain, we've
typically seen FP&A,
CRO or any sort of revenue operations and HR teams using our tool, I do foresee
in the distant
future, less people using that due to AI. It's not there yet, but I can see
that coming.
Got it. So you just touched upon a few things that you see are going to be
future developments
as well. It's a how can finance leaders stay informed and adapt to these
evolving trends to
you know, at the lake or stay ahead of the curve? I would say the best answer
to that is keep up
to date with the people that are really interested in it. For example, I
mentioned earlier the SASEFO
Ben Murray, he loves it. We've got the FP&A guy. He's posting stuff constantly
about all of these
different types of models. A lot of these guys are very much SASE based as well
. At drivetrain,
we have an unlimited list of blogs. Just go take a look through them. You won't
believe how much
you're going to learn. Even if there's things that you think you already know,
there's this great
quote that says it's the job of great teachers to return us to the simple
truths that we are
so reluctant to accept. These blogs are filled with great information. And if
you take a look at
them, you're going to learn something out of them. If you keep your finger on
the pulse of the
industry, you're not going to get distracted by bad modeling practices or you
're not going to miss
the key insights that come out and you'll very much be a part of it. If you're
a part of it and
you know what's going on in the industry, you're setting yourself up for
success when you do have
to jump to a tool or create a more complex model. Got it. So I think we have
come to the end of the
episode and usually Kirk, what we like doing as funful as ending the episode a
little bit of
liberty. So recently I watched this movie called Dune and the main character
basically,
Paul Atreides. So he gets his power to sort of like see multiple futures all at
once. They call
him the quesad Haryg or Lisa and Haryg, but something like that. So what would
you do if you can get
that power? Like maybe there's a shift in your brain with like AI or something
like that. So
what would you do if you see multiple futures all at once? What would I do if I
could see
the future? Is that what you mean? Yeah, so I mean that we're talking about
financial planning
and analysis. So what if you could see into the future and like see model it
out all at once,
like right in your head? I think what I would do is I would, if I had the
capability of doing this,
is I would model out every single public company on earth in my mind based on
my forecast and
compare it to what actually happens, then evaluate where I went wrong and why.
Because
what normally happens is nobody questions things when they go right. When your
forecast
you're accurate, everyone goes, yeah, you were meant to make it accurate. What
people are really
interested in is finding a pattern of why we get it wrong. And so I would love
to find a unifying
answer to the most common reasons people get it wrong. Are we overestimating?
Are we underestimating?
Why are we doing that? And how can we model just a little bit better? This is
great book called
Tetlock Cole, the University of Pennsylvania called Super Forecasters. And he
tracks people that are
essentially have a higher ability to forecast than others. And so I'm imagining
my
Paul Atreides mind being one of those super forecasters. And I think that would
be the
best way that I could put my powers to good use. Super. Thank you so much for
joining us today,
Kirk. It's been a pleasure to host you and understand what the future of
finance holds for all of us.
What an amazing episode. I'm sure our audience will have learned a lot about
building a financial
model that scales as growth. You can of course connect with Kirk on LinkedIn to
continue the
conversation. Check out the episode description for his LinkedIn and any
additional resources
that you need. Until next time, stay informed, stay inspired and keep thriving
in your national
journey. Thank you for tuning in.