Firestarter: skills to lead growth function

Objective & Structure of Firestarter series 2

How to get the best out of this series: treat it like a workbook 3

Chapter 1: Building Growth Model 4

1.1 The Growth Formula 4
1.1.1 Estimating the Number of Potential Customers 5
1.1.2 Calculating Net Value per Customer 9
1.2 Turning Growth Formula Into Growth Model 12
1.3 Turning Growth Model Into Valuation Model 15
Summary of Chapter 1 17

Chapter 2: Building Growth Roadmap 18

2.1 Input metrics versus output metrics 18
2.2 Input & Output Metrics in Growth Model 23
2.3 Objectives & Key Results: Connecting Growth Model to Growth Team Structure 27
Summary of Chapter 2 32

Chapter 3: Building Growth Measurement System 33

3.1 Why & how to set up a growth measurement system 33
3.2 How to improve the growth model with the growth measurement system 40
3.3 Common Root Cause Analysis methods 44
Summary of Chapter 3 48

Growth Planning Workshop 49

Objective & Structure of Firestarter Series
Firestarter is a series of ten essays. After absorbing all of them, you will
be equipped with a minimum but comprehensive set of skills to lead the
Growth function for a new project.

The series is structured in 3 sections:

● The first section, covered in this document, is going to be focussed
on how to plan growth,
● the middle section, with 4 essays, on functional execution, and
● the last section, with 3 essays, on managing different aspects of it.

Why start with planning growth, instead of, say, jumping directly into
101 ways to hack growth?

Growing a business is akin to a journey that you are about to undertake.

The first section then is about the preparation you do for the journey –
preparing the itineraries, bookings you make, packing the essentials, and
so on. After all, paraphrasing Lewis Caroll, if you don’t know where you
are going, any road will get you there.

Further, the first section on planning is divided into three

chapters/essays:
 ● First one, i.e. the article below, is focussed on how to put together a
long-term model for the project,
● the second one on how to prepare a short-term roadmap, and
● the third one on how to put in place a continuous measurement
system to keep track of the project.

Back to the travel metaphor to understand the breakdown: the first essay
is about putting together a map of the territory you will be heading into.
The second essay is then about highlighting the specific route that you
will be taking within the map. The third essay is learning how to use a
compass to know if you are on the right path or not.

How to get the best out of this series: treat it like a workbook
One final thing before diving into the article itself: a tip on getting the
best out of this series. While going through this essay, have a specific
business in mind. And, instead of passively reading, keep applying the
concepts to that business to really internalize them. The articles will
have regular prompts to remind you to do so.

The business can be an idea of yours that you intend to start up. Or it can
be a project you are working on professionally. Or it can be some famous
company that you want to analyze as a case study (but hopefully the first
two, since you will have some skin in the game to go deeper).
Alright! Without further ado, let’s get started.
Chapter 1: Building Growth Model

1.1 The Growth Formula

Wait, did you think of a project? If yes, answer this question: is that
business a woolly mammoth hunter or a bee farmer? That is to say, does
it have a few but very big-ticket customers, or does it have a lot of
small-ticket customers? Or is it somewhere in between?

Consider the table below:

Your project could be on the woolly mammoth hunter end of spectrum:

having just a hundred customers but all of them potentially worth
hundreds of thousands of dollars. Or, it could be on the bee farmer end
of spectrum, having millions and millions of customers and all of them
contributing in cents. Or, somewhere in between? What matters is the
product of the two.

Quantitatively, this idea can be put thus:

Potential value of business = Number of potential customers * Net
value per customer

This forms the bedrock of everything that is to follow in this article, and
we will keep referring to it as ‘the growth formula’. (Note that I am using
the term value for the output of the formula to keep things simple but, to be
clear, the formula’s output is not revenue or valuation. Having said that, the
three concepts are closely linked and I’ll come back to link them all later in the
essay.)

Now, let’s use this formula to calculate the potential value for the project
you have in mind. Starting with the first part: number of potential
customers.

1.1.1 Estimating the Number of Potential Customers

To estimate the number of all potential customers for your business, you
should first know one of them in detail. Or, in marketing terminology,
you need to know the “ideal customer profile” for your business.
A concrete profile can be in terms of:
● Basic Demographic attributes: age, gender, location
● Advanced Demographic attributes: Education level, family size,
occupation, income, etc.
● General behavioral attributes: habits, interests, attitudes
● Behavioral attributes specific to the category your product/service
is operating in: e.g. attitude towards sustainability if that’s a core
proposition of your product.

Once you are clear on that one ideal customer, you can then proceed to
estimate how many similar potential customers exist. But how does one
do that?

There are different ways to go about it. One is a top-down approach (or
guesstimate) where you start with the overall population in the market
that you intend to service and keep narrowing down to the customer
persona. For example, if you are building a sneaker brand for Gen-Z
customers living in metro cities in India who wear sneakers, you start
with the population of India, then narrow it down to how many live in
metro cities, then to how many of them are in the Gen-Z cohort, and
then to how many of them wear sneakers, to arrive at the number.

The other approach would be to look at it from a market category-lens,

especially if you are building a company in an existing category. That is
to say, in the previous example, you can start with the overall sneaker
market size from secondary market research reports, estimate how much
of it serves Gen-Z customers in metro cities, then estimate how much of
that market you can capture (by looking at the pace at which similar
brands historically have gained market share).

If your business will be creating a new category altogether – the way

Meesho was creating social commerce as a new category, or WhiteHat Jr
was creating a new category of teaching kids to code – one invariably has
to follow the top-down route but has to take extra care while defining
the customer profile. Defined loosely, often out of optimism, it can make
the potential market size look much bigger than it might be. That is, if
Meesho defined the core customer persona of resellers as homemakers,
but with a background in fashion either as part of educational experience
or family business, it would make the potential market much smaller.
However, if we had optimistically defined it as any homemaker, or
anyone for that matter, with intent to earn money by reselling, the
potential market will look much bigger, but the subsequent calculations
will be on increasingly shaky ground.

The key point is to keep the ideal customer profile as sharp as possible,
and thus the potential number of customers on the conservative side.
Without defining the customer persona and your right to win, and just
using the method of capturing some share of a large existing market can
lead to an irrationally large number. To put it simply, yes, taking away 1%
of Nike’s market will make yours a multi-million dollar brand, but the
question is: will you really be able to do that? This is what Bill Aulet says
in his book ‘Disciplined Entrepreneurship’ about this temptation:

The common pitfall is “The China Syndrome,” also known to my students as

“fun with spreadsheets.” Rather than create a new market, the thinking goes,
one could choose a huge existing market, get a fraction of the market share, and
reap the rewards. After all, if you could get even a tenth of a percent of the
toothbrush market in China (population 1.3 billion), wouldn’t you make a lot of
money?

I call such high-level market analysis “fun with spreadsheets,” because you have
not demonstrated in a compelling manner why people would buy your product
or why your market share would increase over time. You also have not validated
any of your assumptions by learning directly from customers—you probably
haven’t even been to China. After all, if entrepreneurship were this easy,
wouldn’t everyone sell toothbrushes to China?

Big companies with lots of resources can afford to work hard to gain
incremental market share, but entrepreneurs don’t have the luxury of resources.
Don’t get ensnared by “The China Syndrome.” Take your resources and apply
them to a narrow, carefully defined new market that you can dominate.

On the other hand, if you are yet to find your narrow market that you can
dominate, in order to do a bottom-up estimation, you might have to
estimate what % of a sufficiently large, existing category will you be able
to capture, to, say, rank possible business ideas. A good rule of thumb
could be that market share % should not be more than the number of
years it would take to gain that market share (e.g. the new sneaker brand
I am building will capture 5% of the sneaker market in 5 years and 10% in
10 years). A larger number would run the risk of making the foundational
number, on which the rest of the calculations, and subsequently the
growth efforts, are going to be based, too optimistic.

If you wish to read more about this topic, especially how investors and
consultants approach it, you can read up on the terms TAM (Total
Addressable Market), SAM (Serviceable Available Market), SOM
(Serviceable Obtainable Market).

1.1.2 Calculating Net Value per Customer

Now, coming to the other factor in the growth formula: net value per
customer. While, the first factor, i.e. number of potential customers, was
dependent majorly on the larger market or category, or the niche within
a category, that your business is going to operate in, the second factor is
more specific to your business model. Let me explain.

Suppose you are launching a new premium clothing label in India. You
pay ₹100k to a lifestyle influencer to give your brand a shoutout. From
that campaign, you get 100 customers. That’s ₹1,000 spent in acquiring
each customer on average. That’s your Customer Acquisition Cost
(CAC).
After the first 100 orders, you get an additional 110 orders from these 100
customers over the course of time. So, each of these customers, on an
average, has given you 2.1 orders. Now, for each order, suppose you make
₹2,000 in profit, after paying all the operational costs (cost of making the
clothing and delivering it being the two major cost heads generally for
an online brand, apart from customer acquisition costs). So, you make 2.1
orders per customer * 2000 per order = ₹4,200 per customer. That’s your
Customer Lifetime Value (CLTV or just LTV).

Net value from customer then is the value your business got from the
customer i.e. LTV minus what you paid for the customer i.e. CAC.
So, for this example,
the net value of each customer = LTV minus CAC
= ₹4200 - ₹1000 = ₹3200.

And, if you had estimated 100k customers as the potential customer base
for this brand, the growth formula becomes:
Potential value of business
= number of potential customers * net value per customer
= 100,000 * ₹3200 = ₹320 million

You can now see why I said that while the estimated number of potential
customers is mostly a function of the market/category size, the net value
per customer will differ significantly between competing businesses in
the same market and is more specific to your business model.
Specifically three things within that:
● How efficient is the business in acquiring each customer i.e. CAC
● And, within LTV,
○ Margin you have in every order which is dependent on cost of
goods and cost of operations
○ How many orders does the business get per customer i.e.
Retention

In the same example, instead of spending ₹1000 on acquiring every

customer on an average, if the business was spending ₹2000, the net
value of every customer drops to ₹2200. And the potential value of the
business then drops from ₹320 million to ₹220 million.

Or, if instead of getting 2.1 orders per customer, with the same CAC of
₹1000, if the business was getting just 1.1 orders per customer, the value
of every customer will drop to ₹1200. And, the potential value of the
business, with the same number of unique customers, drops from ₹320
million to ₹120 million.

This point is worth repeating once more: while you can still get away
with ‘fun with spreadsheets’ assumptions on the first factor of growth
formula, if the assumptions are conservative enough, it is really
important to calculate the second factor specifically for your business.
How? With real data being generated by your business. That is: How
much is it actually taking for your business to acquire a customer? How
many times is a customer coming back to buy from your business? How
much does your business make per order?

Therefore, the first two orders of business for a growth lead have to be to
put together: 1. an ideal customer profile, and 2. net value every such
customer is bringing in. If the project is in ideation stage, and the latter
number has been calculated using industry benchmarks, it has to be
validated as soon as the business becomes operational and first few
customers come in.

1.2 Turning Growth Formula Into Growth Model

Alright, so you have got a ballpark number for potential customers for
your business, and the value every customer is generating. And thus, you
have both factors for the growth formula. Now it is time to turn the
two-dimensional growth formula into a three-dimensional growth
model. How? By adding the dimension of time.

For example, let’s say the business you are trying to model is a fitness
app with a potential global market of 1 million customers providing a net
value of $10 each, thus giving it a potential value of $10 million. But wait.
1 million customers to be acquired over how much time? All of them to
be acquired tomorrow? Or over the next 10 years? And, over what period
of time are they providing a net value of $10? In the first month? In the
first year? Over five years? (The dreaded ‘over the course of time', or
‘lifetime’ part of lifetime value.)

Adding the pace at which the business will acquire these customers and
the pace at which the customers pay the business back (directly or
indirectly) is how this simple growth formula will convert into a slightly
more complex growth model. Let’s take a look at it, but with a woolly
mammoth hunter kind of business instead to keep things simple.

Suppose you have built a really sophisticated piece of software that’s

relevant only to 5 companies in the world, but each of whom will pay you
1 million. The growth formula then is a cakewalk: 5 accounts * 1 million
per account = 5 million. Now, let’s add the dimension of time to it.

Let’s say that we will acquire only 1 account every year. Acquiring them
will cost $1 million each and the annual revenue from each of the
customers will be $500k and the revenue will continue for 4 years, after
which they will stop using the software. Putting it all together, we get
this table:
Now, let’s consider all the accounts. Remember we are acquiring 1
account every year and the contracts expire after 4 years, so account #2
will start in year 2 and end in year 5; account #3 will begin in year 3 and
end in year 6, and so on.

Adding them all up:

If you add the net value across all the years, you still get 5,000k or 5
million: the exact same number that we had got from the growth formula
(5 accounts * 1 million per account).
However, after adding dimension of time, a new picture has emerged in
the growth model. The company is going to be unprofitable in the first
year of its operations, it breaks even in the second year, and it therefore
would need an external capital infusion to get through the first couple of
years. This is where investors enter the picture. But the
person/institution giving money to keep the company going needs to
know how much the company is worth. Is it 5 million dollars? Or is it
some other number? How do we calculate that?

1.3 Turning Growth Model Into Valuation Model

A bird in the hand is worth two in the bush. That’s kind of what turning
a growth model into a valuation model is all about. It is obvious that a
promise of a cupcake today is a better offer than the promise of a
cupcake a week later. But how much better? Is 1 cupcake today
equivalent to the promise of 2 cupcakes a week later, or is it equivalent to
the promise of 3 cupcakes, or to that of a dozen. You get the picture (and
hungry).

Without getting into the concepts of Net Present Value (NPV) and
Discounted Cash Flow (DCF), which you can read about if interested,
the idea is to calculate and know the current value of any future money
before adding them.
So, in the previous growth model table, the 500k in year 8 won’t be the
same as 500k today, or 1000k in year 7 won’t be the same as 1000k today,
and so they can’t be added together. Hence, to value the business, all the
future values need to be converted to their corresponding value today,
before adding them up to arrive at the valuation of the business.
Classically, this would be put as: the value of a firm is the net present
value of all its future cash flows.

Now, as a growth lead, while you will probably never be asked to value
the business, and it will be left to the functional experts, it is important
to at least understand how you go from a business idea, to thinking about
who will buy the product, to thinking about potential value of the
business, to putting together a growth model for revenue and costs, to
valuing the business. And, how all these concepts connect.

This is also a point that will recur through this series: having a
cross-functional job like growth lead means knowing just enough in
varied fields (analytics, marketing, sales, product management, finance,
people management, etc.) to get the core job of growing the business
done. And the point of this series is to deliver that minimum knowledge
required across such different fields.
Summary of Chapter 1
So, let’s recap all the ground we have covered so far:
● We picked a business idea
● We fleshed out its ideal customer persona
● Based on the persona, and/or the market category, we estimated the
potential number of customers for the business
● We looked at our customer acquisition cost data, our margins in
every order, and how often our customers come back to buy from
us, to arrive at net value per customer
● We added the dimensions of time to turn it into a growth model
● And, finally, we got an idea of how growth model can be used to
value the business itself

In the next chapter, we will start looking at how we convert this

bird’s-eye view of a plan to a step-by-step roadmap.
Chapter 2: Building Growth Roadmap

2.1 Input metrics versus output metrics

“You do not rise to the level of your goals. You fall to the level of your
systems.” Goes the famous James Clear quote. For example, reading a
book every month is a goal. Reading 10 pages of a book, just before
bedtime everyday, is a system. We don’t fail to read a book a month; we
fail to read 10 pages every night. And this idea extends to matters beyond
individual productivity too.

Companies generally have clarity on their overarching goal – that one

lofty number to rally the troops around. Such as, getting to 1 million
active users, or 1 million lifetime customers, or 100 million transactions
per month. And, rightly so, since just having one clear goal gives a
simplifying clarity of that one thing that really matters. Magnitude of the
number energizes the team.

But goals are outcomes of things people do, not something people work
on as tasks. You read 10 pages before bedtime, and the book gets
finished as an outcome. Similarly, the goals that businesses take up are
not something they work on, on an hourly basis; the goals get achieved
as a result of the right input tasks/levers being worked on.
This is all to say: yes, a business needs to be clear on what the ultimate
goal or their north star metric is. But, they also need to realize it is but
an output/dependent metric. They need to know the input/independent
metrics that make up the output metric. And what are the levers that
need to be worked on to move the needle on the input metrics. Let’s
understand these with an example.

Suppose we have launched a new online-only restaurant, which has

become a runaway success. Every morning we check the sales numbers,
and, sure enough, the numbers are always going up and up.

Except one day when it is not. And then the trend continues for a week.
The number is now consistently trending down. Why is it happening, we
wonder? But we don’t know the reason for why it is decreasing because
maybe we never really understood why it was increasing either.

The sales numbers are an output metric, yes. So, what are the input
metrics? That is, if we had to write a formula for it, what would be on the
right hand side of this equation: Sales = ?

We think about it, for a bit, and the formula that has all the independent
input metrics that make up sales is:

Now, suddenly we have a headway: if sales numbers as output metric are

trending down, then one (or more) of the input factors on the right have
to be going down too.

Suppose we find out that traffic is the same as before, and conversion
seems to be fine too. Which means it must be the average order value
that is trending down. And so, if we just fix that number, the output
metric will be back on track too. But wait, how do we fix it?
We can go about fixing a number only if we know what makes up that
number. And so, once again, we have to think of the formula that makes
it up:
Average Order Value = Average items bought per purchase * Average item price

So, we dig in once again to figure what’s happening? Are people buying
fewer items? Or are they buying the same number of items but the less
costly ones? We find out it’s the former and in order to fix this input
metric, we pick up a lever. Say, prepare a better cross-sell plan to induce
customers to buy more items together in a purchase. While picking up
the right levers is, in itself, a combination of scientific and creative
thinking, and one entire section in this series (chapters #4-7) is dedicated
to it, the point of this example is to demonstrate how we went from the
output metric of sales numbers to its input metrics, and then one-level
deeper further, to finally arrive at a lever i.e. an activity that can be
actioned on.
But, wait! Did we not decide that average order value was an input
metric which influences the output metric of sales? Now we are saying it
is, in fact, an output metric itself, influenced by other input metrics.
Welcome to the concept of hierarchy of metrics.

Going back to the reading example: we defined the number of pages read
per day as the input to the number of books finished as an output metric.
The same concept extends there too. Once we move the magnifying glass
over “number of pages read per day”, it can be broken down into three
components: duration of time the book was open, % of time we were
focussed, reading speed of words per minute of focussed time. So, the
number of books read in a year might be the ultimate output metric (the
north star metric, that is; should it be the NSM is an altogether different
matter), then comes a bunch of input metrics which build to that number,
and which might be output metrics themselves.

We will expand upon this hierarchy of metrics again in this essay as well
as the subsequent essay on setting up a continuous measurement system.
For now, the key takeaway is this: a business needs to know its output
metrics, the respective input metrics, and the levers to move the
independent metrics.
2.2 Input & Output Metrics in Growth Model
Now, let’s combine the ideas from the previous section with ideas from
the previous essay. But, with a different example.

Suppose we are starting a mobile game app business, which will make
money by serving ad impressions to the user. We write down its growth
formula. We think of the ideal customer profile and estimate the number
of potential users. We put down an estimate of net value per user which
we have to validate with real data once the business is operational.

Let’s say the estimates are:

Potential users = 10 million
Net value per user = 10 dollars = 12 dollars lifetime value (LTV) minus 2
dollars of customer acquisition cost (CAC)

How did we arrive at an estimate of LTV? Well, it’s again all about
breaking down the output/dependent metric into its input/independent
parts. In this case:

Lifetime Value per customer =

Avg. value we got per customer in their 1st month of using the app +
Avg. value we got per customer in their 2nd month of using the app +
Avg. value we got per customer in their 3rd month of using the app + …
And, how do we calculate the average value per customer in, say, the 3rd
month of using the app? Let’s say we started with 100 users, and by
month-3, only 20 of them are opening the app (i.e % of active users –
retention). But the ones who are active, open it 10 times per month on
average (i.e. number of sessions or app opens per active user –
engagement), and every time they open the app they get served 4 ads (i.e.
number of ad impressions per session – monetisation).

So, total ads served in 3rd month = 20 * 10 * 4 = 800

Thus, average ads served in 3rd month per user = 800 / 100 = 8
And, if ad networks pay us $0.1 per ad serve, net value per user in
month-3 = $0.8

Putting it all together, a hierarchical breakdown emerges:

Now, we have to turn this into a growth model by adding the dimension
of time. In the previous essay, we had taken an example of an enterprise
software business with very few customers and very high LTV, and so we
had prepared the growth model at customer level. In this case, however,
since the calculation will run into millions of rows, we will combine
customers acquired in the same month into a cohort. Also, we will track
the value per customer monthly, instead of annually, considering the
lifecycle of a user might not extend beyond a year and might differ
significantly between months (weeks even).

With estimates for all input metrics, let’s say, this is what the forecasted
growth model comes to:

If we map this to calendar months – month-1 for customers acquired in

Jan-24 cohort will be Jan-24, month-2 for this cohort will be Feb-24,
month-3 for this cohort will be Mar-24, for example, and similarly
month-2 for Feb-24 cohort will be Mar-24, and so on – the previous table
transforms to:
[Try replicating this calculation in a sheet. Better still, replicate this calculation
for a different business idea altogether.]

Now, let’s say that the company becomes operational with this growth
model and aims for 1000 new customers in Jan-24, 1500 new customers
in Feb-24, and 2000 new customers in Mar-24, and targets a net revenue
of $15,800 as per this model. However, the revenue for Mar-24 is
different than was projected in the above table. How do we track the root
cause of the difference between the projected and actual numbers?

We can look into what were the input metrics for net revenue (see
hierarchy of input metrics in the image above) and see where the
divergence is. If the difference is in, say, value per customer in month-2
(actual value of $1.7 instead of projected value of $2.0) as input metric, we
can look for the independent metrics that compose this number.
So far, we have tried to understand what input and output metrics are,
how metrics are part of a metric hierarchy (where input metrics are
themselves made of other metrics), and how these concepts fit with the
concepts of growth formula and growth model, to help us put together a
mathematical model of the business.

But we had started this article with the premise that such a quantitative
understanding of a business will translate into an actionable roadmap of
tasks & activities to be done to drive growth of the business. Let’s get to
that.

2.3 Objectives & Key Results: Connecting Growth

Model to Growth Team Structure
There is an anecdote of a guest of Elon visiting SpaceX, stopping by an
employee on the floor impromptu, and asking them what they were
working on. The employee, instead of just telling the task that they were
working on, started with SpaceX’s mission of making interplanetary
travel possible, what one of many technical challenges in achieving that
objective was, and how the specific project they were working on was
going to solve it.

The purpose of any goal-setting framework like OKR (Objectives and

Key Results) is precisely this. To give an edifying clarity of: What is the
long term goal, how will we know if that goal has been achieved, and
what do we need to do to make that happen?

Let’s connect our growth model and the concept of OKRs, with the
example from the previous section. Let’s say that the company has set
the objective of being the most popular mobile game app in the world in
the trivia category. Okay, the why is clear. But it’s not very quantitative,
which will be brought in by the Key Results part. Which is where the
Growth Formula and Growth Model comes in.

In the long run, as per the growth formula, having 10 million customers,
each giving $10 of net value are the key results. In the short run, as per
the growth model, getting to a net revenue of $15,800 by the third month
of the operation is the key result.

To achieve this short-term key result, we will have to achieve different

input metrics:
● number of customers to be acquired,
● customer acquisition cost,
● revenue retention in subsequent months (i.e. revenue coming from
Jan-24 cohort customers in Feb-24 and Mar-24; revenue coming
from Feb-24 cohort in Mar-24).

The metrics hierarchy then maps to the key results hierarchy, which,
in turn, informs the growth team’s structure, goals, and tasks. Let’s
understand this point, since it connects what we've learned so far, and is
the key point of this essay.

The organization has the overarching goal of net revenue, which might
be the CEO’s key result. The input metrics for it are number of
customers, cost, and revenue retention. The CEO will assign one or more
of these metrics to the functional leaders. Let’s say that revenue
retention metrics have been assigned to the Head of Growth. They then
break down this key result, as per the metric hierarchy in the diagram
above, into % of active customers (retention), # of sessions per active user
(engagement), and impressions served per session (monetisation), and
assigns each of them to respective teams. The person in-charge of user
engagement might break it further down into input metrics, and so on.
We can visualize it so:
So, is it input metrics all the way down? Where do we hit the bottom of
the well? And, how do we make the magical leap from key metrics to
levers?

Generally, after 2-3 levels of going down the metric hierarchy, we might
reach a reasonably independent metric. Let’s understand this point with
the online restaurant example. We started with an overarching goal of
sales. Going down the metric hierarchy, at level 1, we get the key metrics
of traffic, conversion rate, and average purchase value, which when
combined give us the sales numbers. Following conversion rate as a
branch, we get, at level 2, % of visitors who view at least one item, %
among them who add to cart, % among them who complete the checkout
process.

Now moving focus on the ‘% of visitors who view at least one item’, we
can further dice it in different ways, but we have perhaps reached that
point where instead of figuring out the mathematical formula for it, we
can put down the business drivers for it. What could be the drivers for
increasing the % of people who reach the online restaurant listing page
and click on at least one item? Or inversely, what could be the drivers for
decreasing the % of people who don’t click on even one item and bounce
off? Perhaps, the page load speed on the technical side. On the
merchandising side, it would perhaps be the popularity of the products
being displayed up front. What other driver can you think of?
And finally, we pick up the levers that will improve these drivers of the
independent metrics. Such as, improving page load speed from x to y; or
adding so-and-so popular dishes to the assortment. This leap from the
hierarchy of key metrics to a list of levers/activities to be done, to grow
the business, thus comes from a deep understanding of the business
drivers for the independent metrics. And this understanding can come
from diverse sources: customer research, domain experience, data, and,
of course, first principles thinking.

Putting it all together, the templated way to understand the relationship

between objectives, key results, and levers would be: We want to do x
(objective), as measured by y (key result), by doing z (lever).

For example, I want to be well-read, as measured by reading 12 books in

2023, by reading 10 pages every night before bedtime. Or, I want to be
healthy, as measured by improving my body mass index to 20, by eating
healthy on 90% of the days (lever #1) and exercising for 30 minutes on
80% of the days (lever #2). Levers are something you work on. Metrics are
something you measure to understand if the intended outcome has been
achieved.
Summary of Chapter 2
After reading the first chapter, you should be able to put together the
growth formula and growth model for your business idea. After reading
this chapter, you should be able to:
● build a visual metric hierarchy of your growth model,
● identify the independent/input and dependent/output variables in
it,
● map the independent metrics to their respective business drivers,
and thus the levers that will improve them, and finally,
● visualize the growth team structure that corresponds to the
independent metrics & associated levers.

In the next chapter, we will close on the first section of the series, with
steps to set up a continuous measurement system to track if the growth
system is working as intended or not. It will also introduce a few key
analytical concepts that further sections will build on.

The second section of the series, to be published in February, will go

about answering the million dollar question: how do we gain enough
cross-functional knowledge to know which levers will move the growth
needle and which won’t?
Chapter 3: Building Growth
Measurement System

3.1 Why & how to set up a growth measurement system

Startups often oscillate between the two states of not consuming enough
data or trying to consume a lot of data.

You are in the ‘not consuming enough data’ state if you know the
revenue numbers, the marketing costs, the cash in the bank, and, yeah,
that’s about it. Now don’t get me wrong. This is good too. That is to say,
not knowing them would be worse. But the problem with tracking just
these numbers is that these are output metrics.

You are in the ‘trying to consume a lot of data’ state if one day you felt
bad about not tracking enough metrics, and went the opposite extreme:
you opened a spreadsheet and wrote down every metric you could think
of and then asked someone to update the sheet with all those numbers
every morning by eleven. But it has been a few days since you last went
through it.

But why is the second state bad, in theory? Isn’t being on top of numbers
supposed to be a good thing?
The problem is that such a laundry list of metrics will not be effective in
achieving the objective of a measurement system. So, let’s start with the
why. Why set up a growth measurement system?

The core objective of setting up any measurement system is threefold:

● To know the overall health of the business at a glance with the
output metrics
● To be able to do a quick diagnosis, if some output metric is off the
trend, with hierarchy of input metrics
● To be able to make predictions and catch things going off the rails
early, with the lead indicators

Let’s understand them with an example. Suppose we are running a call

center, with 10 agents. Every morning we come in, we want to, first,
track overall health of the operations. How do we do that?

Any system will have two complementary metrics – a volume metric and
a quality metric – that capture the overall performance at the highest
level. You will see this pattern everywhere, once you internalize this
concept. Such as, pages read versus knowledge gained. Calories ingested
versus nutritional quality. GDP growth versus quality of life. In fact,
negative externalities often arise because there is no counter-balancing
quality metric being tracked.
So, for this call center, the top-level volume and quality metrics can be
the number of tickets cleared and the customer satisfaction score
respectively.

But do these numbers mean anything by themselves? Is 720 tickets

cleared the previous day with a 57% satisfaction score good or not? How
do we know that? This context can be provided by adding how it has
trended over time i.e. a time series trend. And/or adding the targets or
projections against these actuals.

Alright. So, we set this up and have a high-level picture of the 2 most
important metrics, as well as their context. We look at this in the
morning, are happy with the numbers, and can move on with our day,
without having to go through a laundry list of every data point the
business has generated.

But what if the metrics are significantly deviating from the past trend, or
from the projections we had made? How do we do a quick diagnosis?

This is where the metric hierarchy, that we covered in the previous essay,
comes into play. Both the volume and quality metrics will need to be
split into their respective input metrics. We can quickly look at these
level-2 metrics and understand the reason.

For example, let’s say we have started a new content app. What should
the top-level volume and quality metrics to track its growth be? It can be
the number of unique people opening the app and average engagement
per user respectively. We are also tracking this over some time instead of
just the previous day’s numbers, so that we are aware of the past trend
and any pattern of minor variation (e.g. higher app opens on weekends
than weekdays).

Now let’s say, the number of unique people opening the app (i.e. number
of active users) is dipping. How do we know what’s happening? From the
ideas of the previous essay, we will have its metric hierarchy. For level-2,
the formula can be the sum of new (first-time) users and returning users.
For level-3, the number of new users can be split into different
marketing channels, and the number of old users is split into different
cohorts.

So, before jumping into investigating myriad possible reasons every time
the top-level metric seems off (did a new competitor launch, is the new
app version crashing, etc.), we can first make a quick diagnosis by going
down the metric hierarchy, and have a clear sense of the specific
direction to go in for identifying the problem. We will come back to
some common ways to do root cause analysis in the last section of this
essay.

So far, we have talked about setting up a top-level tracking of volume and

quality metrics, and their level-2 and level-3 input metrics for quick
diagnosis. Is this all? The measurement system defined above looks at
the past and catches any aberrations early. But can we have a system
which predicts if these metrics are about to deviate from the trend?
Enter the concept of leading indicators.

Let’s say that the content app has good user engagement. However, that
engagement is coming from content created by creators in the near past.
Maybe in the last 48 hours. So, even if the content creators suddenly face
a problem in creating content, there will be enough content liquidity in
the system for some time that the top-level engagement metric on the
consumer side might not show a dip. However, if the problem continues,
the engagement metrics will dip in future. Hence, drop in the number of
new content items is a leading indicator of drop in the consumer
engagement metric.

How do we identify the leading indicators of any metric? It will come

from the deeper context of business drivers for that metric. For example,
employee attrition rate is one of the key metrics for the people system in
an organization. However, it is a lagging indicator of employee
engagement. While that number might seem okay at the moment
(against time series trend, or against projected numbers), it does not
really predict what the attrition rate will be in the next quarter. The
leading indicator for it can be employee satisfaction score. That is, even
if the attrition rate is stable currently, but if the satisfaction score has
started trending downwards, it is likely the attrition rate will increase in
the future.

Another example is retention rate of customers for a platform. While it is

the most important independent metric in the growth model, it is a
lagging indicator of customer satisfaction. In that, a stable retention rate
does not guarantee the rate to be stable in future too. Declining
customer satisfaction scores (or Net Promoter Score - NPS) will mean the
retention rate, and therefore revenue from the existing customers, will
trend downward in future.
So, in summary, a good growth measurement system needs to have:
● Complementary volume and quality metrics, for a quick overview
● Trend of these metrics, and/or comparison against projections, for
context
● An optional view of metric hierarchy, for quick double-click and
diagnosis
● Early indicators of independent metrics, for quick prediction

3.2 How to improve the growth model with the

growth measurement system
Now, let’s say we have a situation on our hands where the actual
numbers are consistently veering away from the projected values. If we
do not change the model being used to make the projections, it makes
the projections meaningless. But if we keep changing projections every
time it’s different from actuals, it makes the act of building projections
futile. So what do we do?

Let’s go back to the growth model we had built for the mobile game app
business, in the previous essay. It had, in its growth formula, 10 million
potential customers and a net value of $10 per customer ($12 LTV minus
$2 CAC). In its growth model, there was 6k of net revenue projected in
the 1st month, 11k in the 2nd month, and 15.8k in the 3rd month.
In this model, the independent metrics that are driving rest of the
calculations were:
● Number of new users per month. Especially the degree of
acceleration in that number i.e. 1k new customers in the first
month, 1.5k customers in the second month, and so on.
● Cost to acquire new users
● % of users returning in successive months i.e. retention
● Revenue per ad impression

And so, once we dig into these input metrics to find the reason for actual
numbers diverging from what was projected in the growth model, we
might find one (or more) among the following things happening:
● Number of new users per month is increasing but not at the pace
that was expected
● Cost is not constant as we are trying to accelerate the number of
new users
● Fewer than projected % of users coming back in successive months
● Revenue per impression being paid at lower rate by the ad
platforms than was projected

At this point, we have to dig into the business context of the original
assumption. For example, why did we assume that the number of new
users per month will keep accelerating? Was it as a result of more
channels to acquire the user being opened up? Was it because of more
supply levers being opened up? Was it based on industry benchmarking
of similar stage apps? Similarly we have to look into the business logic
for quantitative assumptions for other metrics too.

We can reflect on the assumptions and see if they have been proved false.
If yes, we have two options:
1. Modify the long-term growth model by changing the independent
metric to reflect the actual number. That is, if month-1 retention is
actually 12%, use that in the growth model rather than, say, 15%
which was originally used.
2. Modify the short-term growth roadmap by making it a key result
to improve the independent metric by picking the right lever
against it. In this case, increasing month-1 retention from 12% to
15% within a certain time period using certain levers (we will cover
the frameworks to pick the right levers for different independent metrics
in a growth model, in the next section of this series - essays# 4 to 7).

After one of the options has been chosen, we can operate for some time,
collect new data, and re-visit these steps.

This is similar to the build-measure-learn loop described in the book

Lean Startup by Eric Ries. We are building the growth model and
roadmap, with some business assumptions. As the business gets
operational, we are measuring the actuals against the projections. And,
we are accordingly learning from the difference between actuals and
projections and the root cause for it. And subsequently, either changing
the assumptions in the growth model or changing the growth roadmap
to focus on the errant assumptions.

We can run through this loop every quarter (or the short-term planning
horizon being followed at the startup). However, in order to be able to
tweak one or more assumptions in the growth model, without having to
re-do the entire thing, it should be fairly modular.

A well-compartmented growth model will have 3 parts:

● Input: with all independent metrics and business logic for these
assumptions
● Processing: where all calculations take place, but with only
transformations (formula in case of spreadsheets) and no
hard-coded values
● Output: a simplified view of projections for top-level metrics

If there is such a clear separation of concerns in the growth model, one

can look at the input sheets and get a clear picture of what the business
drivers are. And such an abstraction will help non-specialists look at
assumptions in the input and the simplified output views, without
getting bogged down by all the transformations in the processing part.
3.3 Common Root Cause Analysis methods
We saw how divergence between actuals and projections between output
metrics can be looked into, by traversing down the metric hierarchy and
finding the independent metrics causing it. For example, retention %
dropping and causing a drop in revenue against projected revenue. But
how do we investigate a drop in the independent metric itself? Retention
%, in this case.

I will share with you three frameworks which will serve you well in most
root cause analysis across multiple growth problems. They are:
● Funnel thinking
● Segment thinking
● Cohort thinking

Let’s say the default rate for a bank has gone up massively. How do we
look into it and understand the root cause of the problem?
One way to do this is to dice the data by funnel i.e sequential steps. An
account is considered to have defaulted if, say, 3 successive payments are
missed. So, we can first look at what % of people missed the 1st payment.
Of them, what % missed the 2nd payment. Of which, what % missed the
3rd payment. And try to see which step of the funnel has had a sharp
increase.
If the increase is across all parts of the funnel, it is not the root cause.
This is the crux of root cause analysis, or discovering any data insight in
general: you have an insight only when you have arrived at a point where
the patterns are diverging.

Since it’s not a problem with funnel steps, we can dice the data by
segments. A segment of a population is sufficiently similar to each other
and dissimilar to other segments. We know of common ways to segment
individuals (race, gender, age, etc.). In this context, relevant parameters
to segment would be the size of business. Are the large businesses
defaulting, or the medium ones, or are the small ones? Another
parameter could be sector. Are the primary industries defaulting, or are
manufacturing industries defaulting? Or it could be a combination of
parameters: while other businesses seem to be fine, it’s the service
industries with a turnover more than 100 crore that are defaulting.

And, if there is no pattern emerging either from funnel or segment

(increase in default rate is across all stages of funnel, and across all
segments), we could slice the data by time. Also known as cohort, or
vintage. That is, all loans that were granted in 2020 form one cohort, all
loans that were granted in the 4th quarter of 2021 form another cohort,
and so on. So, if we find that default rates were high only for the loans
which were granted in 2019, and not for the other cohorts, we have
zeroed in on an insight.

To be clear, cohort is not the same as segment. For example, a hotel

which had its loan granted in 2020 will be in a different cohort than a
hotel which got its loan in 2021, but they will be in the same segment (i.e.
hotels) in this analysis. While two companies with different turnover and
operating in different industries, but which got the loan together would
be in the same cohort, albeit in different segments.

Another example of the difference is that generations (e.g. Millennials,

Gen-Z) are cohorts, while income brackets (e.g. HNIs, Middle Class) are
segments.

Let’s go over these techniques, one more time, but in the context of a
growth model. Let’s go back to the example of the lifestyle brand we had
modeled in the first essay. We had arrived at an LTV of 4200 with 2.1
orders per customer and 2000 as margin from every order.

However, after 3 years of operation, we notice that the LTV has

decreased from 4200 to 3200, with the decrease coming from average
orders per customer dipping from 2.1 to 1.6. As we saw above, we have
two courses of action:
● Change the growth model: by changing the underlying assumption
of orders per customer
● Change the growth roadmap: prioritize identifying and executing
on levers to increase orders per customer
But, before that, as the first step, we need to understand why it decreased
from 2.1 to 1.6. To identify the root cause for it, we have to do funnel
thinking, then cohort thinking, and then segment thinking.

In funnel thinking, we have to look at sequential steps for all customers.

So, we will look at what % of customers were coming back to transact in
the 2nd month originally and what the number is now. Then in the 3rd
month, and so on. Is it a dip across all the months? Or in a certain month
of the customer's journey.

In cohort thinking, we have to group the customers based on when they

were acquired. So, we look at LTV of customers who were acquired in
January of 2021, February of 2021, March of 2021, and so on. Is the dip
across all the cohorts? Or for a few specific cohorts.

In segment thinking, we have to group the customers based on their

attributes. So, we can look at LTV of different customer persona, or by
segmenting based on the marketing channels or campaigns they were
acquired from. Is the dip across all segments? Or have we over-acquired
over time from a persona or channel which had low LTV.

Often, the search for a divergent pattern is not so clear cut: it’s not 100%
this or that; but 80% due to segment and 20% due to one funnel step; or
70% in this funnel step and 30% in another. However, in most cases,
following these steps will lead you to the major driver for the change.
Summary of Chapter 3
In this essay, we learnt the following concepts:
● How to set up a growth measurement system for any business,
using
○ Volume and quality metrics as complementary pair
○ Trends and hierarchy of input metrics
○ Leading and lagging indicators
● How to improve a growth model, using
○ Separation of concerns
○ Build-measure-learn loop
● Common root cause analysis techniques, especially
○ Funnel thinking
○ Cohort thinking
○ Segment thinking

With this 3rd essay, we have reached the end of the first of three sections
of the Firestarter series. This section was on growth planning, and by
now you should have the skills to build a growth model for an idea, build
its growth roadmap, and set up a growth measurement system, along
with tweaking the model after finding the root cause of deviations.
Growth Planning Workshop
With the planning base covered, we will now dig into the juicy executional
bits of growth over the next 4 essays (published on Skilletal on
Wednesdays starting February 1st). Answering questions like: How do we
decide which growth levers to pursue? How do we decide the growth
channel strategy for a business? How do we plan a campaign? And many
more concepts and frameworks which form the pareto of executional
aspects of different functions within growth.

Meanwhile, if you are interested in learning the skills of the first section by
applying them through a series of tasks, this workshop might be for you.