1

Business Model and Customer Development

Business Model
One of the interesting thing about a startup is how is your company going to be organized? What we now
know is the most efficient way to think about all the pieces just all the parts is by a business model. And
so the next question is okay, Steve just told me think about a business model but what is a business model?
What are all the pieces? Well let’s take a look. A business model is how a company creates values for itself
while delivering products or services for its customers. Now if you think about it, in the old days we think
about how to organize a company around functional organizations. We think no, a company is about its
sales department or its engineering department and you would draw a work chart but now where going to
draw a very different diagram.

Customer Development Process

One of the really interesting developments about this class is this whole customer development process. It
says you start with your business model canvas hypotheses, and in fact, what you really do is you blow up the
canvas, and you actually post it to the wall, and you use yellow stickies, no pens or pencils allowed, because
you are going to get most of them wrong. But you’re going to make it visible, and you will actually begin
to construct your hypotheses, and the next thing you’ll do is look at them and go ”Hey, there aren’t any
facts in this room.” ”Let’s get out of the building and talk to customers and partners, inventors, and we’ll
learn how to do this with some rigor, with a process.” Not just randomly getting out, but actually design
experiments, run tests, get data, and more importantly, get some insight, and the customer development
process is kind of interesting. The customer development process is actually a 4-step process. The first step
is customer discovery. This is where you construct your hypotheses, and you get out of the building and start
testing your assumption about whether other people have the same problem or need you think they have.
And then you’re going to do customer validation and actually see if your proposed solution actually matches
what you think the customer problem was. This test between problem and solution and your features and
customers is actually sometimes called product market fit. That’s what you’re out testing, and this is what
we call the search for the business model. But now instead of randomly doing this by hiring and firing sales
execs and trying to make numbers that really are just random guesses we’re actually going to have you get
out as early as possible and test some of these primary assumptions. One of the interesting things on the
bottom of this diagram that we’ll talk about is something called a pivot, and the pivot is what will save
your job. Once you find this repeatable and scalable business model then you go into the execution phase
of customer development, and that’s about creating end user demand and scale called customer creation
and then building the organizations to actually build your company for scale by transitioning from customer
development into a functional organization that’s oriented for constant and rapid execution.

Business Model Canvas Value Prop.

Let’s take a look at the first piece called the value proposition. The value proposition answers the question
what are you building and for who? The value proposition says, ”Hey, it’s not about your ideal product,
it’s about solving a problem or a need for a customer.” That is what pain are you solving, what gain are
you creating, and more importantly, who are your customers? Now value proposition is a fancy word for
what product or service are you building. This is where you normally would list all your features and here’s
all the speeds and feeds and benefits and whatever but we’re really going to be asking a different question
than might have been used to. It’s not all about your technology. Your technology is just part of the value
proposition. Customers really don’t care about your technology. The customers are trying to solve a problem
or fulfill a need. By the way, we’ll be talking about this for multiple lectures. The difference between a
problem and a need is–a problem is I have an accounting problem or I want to use the word processor, and
those type of products solve a problem, but they are other things that human beings do–like I want to be
entertained or I want to have a da There’s just some basic hard wired social need or I want to communicate
with my friends. Like Facebook or Twitter. Those are needs. Needs are different than problems. And by
1 Generated from UDACITY transcripts of Course ”How to Build a Startup”

For use at FEUP in Course Unit: PRODEF029 1

the way, if you could find products that solve needs, your total available market as you’ll see later is huge
compared to I solve specific problems.

Bus. Model Canvas Customer Segments

The next thing is, who are my customers? Who are they and why would they buy? And as you’ll hear a
number of times, your customers do not exist to buy, you exist for them. And what you’re going to do by
getting out of the building is figuring out all their geographic, social characteristics, demographics such that
you actually could draw and put up a picture on your wall of who the archetype is or who the persona is of
your customer, but it turns out that in most start ups you might have more than one or two or three types
of customer archetypes than personas, but you need to understand them in detail, and there is no possible
way you could have anything but a hypothesis on day one of who they are.

Four Phases Of Customer Development Solution

The order is customer discovery, customer validation, customer creation and company building.

Business Model Canvas Channels

The next is channels. How does your product over here get to your customers over here? We use distribution
channels to do that. Now what’s really interesting is pre-1990s, the only channels to get to a customer was
at physical channel. That is you went to a store, you have sales people, there was physical distribution, but
since the mid 1990s in the last couple of decades, we now have virtual channels. The web, mobile, cloud
and so at distribution channel, the first question you want to ask is, how will I be selling and how will I
be distributing my products. Are they through physical channels or the web mobile or given today almost
every physical channel also has web presence. What is the relationship of how your product gets from your
company to the customers?

Customer Development Done By Founders

Now, one of the interesting things about customer development that makes startups very different from
large companies is that customer development is done by the founders. Why the founders? It turns out that
there’s a couple of interesting things, and this is about human nature. There’s no technology involved. In a
traditional startup in the old days, you would hire a VP of sales who would go out because you were smart,
this was your idea, and you’d say, ”Go try to sell this. Go talk to customers. I’ll just hire somebody to do
this.” But remember, an employee doesn’t have the vision. They’re just executing what you told them. And
guess what happens if they go out and talk to people who say, ”This is the worst idea we ever heard,” or
”No, we won’t buy it.” What happens is they’ll come back and tell you in the first time they do this, and
you’ll say, ”Well, you’re just not describing it right.” And you send them back out for another couple of days
or weeks or months, and they’ll come back again. And if you’re like any passionate founder, you’ll go, ”I
hired the wrong executive. You’re fired.” Now just imagine we run the exercise this time not with a proxy
head of sales or marketing, but we force you to get out, you the founder to start talking to customers. And
if you got that same exact feedback, it might take you 3 customers or 5 or 30, but eventually, smoke will
start coming out of your ears because cognitive dissonance is now coming into effect. It’s that you might
realize that your story or vision isn’t right. But unlike a proxy, a VP of sales or marketing, you have the
power to change the company’s strategy. You have the power to change the entire value proposition, to say,
”What if we had these features?” And customers might say, ”Nah, nah, nah, still not.” But then if you said,
”Well, what if we had this?” they might say, ”Oh, if you could do that, you could have my check right now.”
If you’re a smart founder you’d say, ”Let me get right back to you,” because what you’ll do is you’ll go test
that new feature with 5 or 10 other customers. And all of a sudden you realize that you had the wrong
feature set and just by adding this small, little change, you can now actually get a whole series of paying
customers. Only founders could do that. And what we did in the past is we would wait till first customer
ship, we’d wait till sales didn’t match the revenue plan, and we’d actually make these changes by firing
executives instead of actually having the founders engage in Day 1. So it’s the founder who could change

For use at FEUP in Course Unit: PRODEF029 2

the product, make pivots, and hear customer feedback firsthand. And that’s the idea of getting the founder
outside the building.

BMC Customer Relationships

Customer relationships is kind of a fourth piece, and customer relationship has a really interesting interaction
with these other three pieces. It basically says how do I get customers, how do I keep them, and how do I
grow them, and just like thinking about distribution channels, these are very different for web mobile than
they are for physical channels. But visually, they kind of look like this double-sided funnel. Let’s just take a
look at quickly a web example–in getting customers, you’re going to be o worrying about how do I acquire
them that is how do I get them even to my website, how do I activate them that is how do I make them
do something, and then later on we’ll see after I got them how do I keep them around that is how do I not
lose them through attrition and churn and then what can I do once I have customers to make them spend
more money or use my product even more. One of the things we’ll be thinking about is how do I get, keep,
and grow customers. And just like every other step, you might have hypothesis on day one, but you’re only
going to figure this out when you’re out of the building.

Hypothesis Testing
The next piece about customer development to understand is what is it you’re actually doing outside the
building? And what I think about is you’re really testing on the highest possible level your understanding
of the customer’s problem or need. You implicitly had a hypothesis. I want to extract from implicit to
have you make it explicit. Here’s what the pain and the gain is we’re actually doing for these hypothetical
customer segments that I’ve written down on my canvas. Great, so how are we going to go do that? Literally
we’re going to get out of the building, take our hypotheses, and we’re not just getting out of the building
and randomly talking to customers because if that was the case we could have just sent them a letter or
sent an email. What we’re looking for is not just data, it’s insights, and how this process will work is we’ll
get out of the building, and then as we find those new insights we’ll actually change the canvas. Big idea.
We’ll change the canvas by marking it up and saying, you know, we thought the customers were these kinds
of people. Holy cow, they’re actually these kinds of people, and we thought the features we need, well,
they’re kind of different. They’re actually these features, so what we are doing outside the building is you
start with these hypotheses. In this case let’s just take an example of the customers will be male, 24-35,
live in urban areas, and then we’re going to design some experiments. Let’s go figure out maybe a Google
AdWords campaign. Or if it was a physical product, go out and meet them personally, and then run some
tests and take a look and analyze the data. But it’s not just the data. We’re trying to understand did the
results match the hypotheses? And if not, just don’t give up and say, ”Well, it didn’t.” ”Let’s try another
segment.” Understand why your initial hypotheses were wrong because it’s this why not that might give you
some insight. What you might find out in this case is, oops, we kept getting teen girls in suburbia, and you
can either keep deciding no, no, no, I want men, or you might go ”Well, wait a minute.” ”The teen girls are
actually enthusiastic, in fact, trying to figure out how to buy our product right now.”

Business Model Canvas Revenue Streams

The next thing is revenue streams. How do you actually make money from your product and service being
sold to customer segments? You know revenue streams basically ask the question what value is the customer
paying for and then actually as you think about what’s the strategy of how long I’m going to capture that
value Is it I’m going to just have a direct sale or it’s a complete transaction based on price? Is it a free me
model where I’m going to give away the product for free and hope that some portion convert later. Is it a
license or subscription model? That revenue model is different than the pricing tactics. That is what is the
dollar or pound amount or euro amount that I’m going to be charged. Again the only way to figure this out
is being able to interact with tens or hundreds or thousands of customers so you finally understand what
the right revenue streams and revenue model is.

For use at FEUP in Course Unit: PRODEF029 3

Minimum Viable Product
The interesting thing about the customer development process ties back into agile engineering and agile
development hand and glove. Basically it’s this notion of the minimum viable product. Back in the old days
what we used to do is specify the entire feature set of the product from beginning to end. Now, this makes
sense when you’re in a large company releasing version 2.0, 3.0 and 4.0 because you kind of have a feeling of
who the customers are and what they need, so a product manager can be pretty accurate about I’ve been
interacting with customers for the last year and a half, and I think I know what they need. But in a startup
you’re really kind of guessing, and the odds are you’re going to be guessing wrong, so rather than waste
a whole ton of time and money why don’t we actually get outside the building before we build something
and waste a lot of engineering time, and more importantly, cash, because that’s what puts startups out of
business is running out of money. We want to make sure that we actually listen to the people who eventually
will buy this product. We want to make sure we satisfy their wants and needs. Why don’t we just figure
out how to build the minimum viable product? Build the minimum features in order to get feedback. Now,
feedback could take the form of input verbally, or they gave you early orders, or they gave you anything
that was valuable in helping you come to closure on what should we be building in what order? And by the
way, an MVP could be something as simple on the Web as a wireframe or a PowerPoint slide, or for physical
products it could be a physical mockup, or it could be a working part of the system. But as you get more
feedback you could start adding more features. One caveat is a comment I always get is ”Steve Jobs didn’t
build the iPhone by asking customers.” And we really doubt Henry Ford asked customers do they want a
car before one existed. In fact, in his case if you would have asked people about what they wanted they
would have said a faster horse or one with 6 legs. And so the immediate response is ”Well, therefore for new
products you just don’t get out of the building at all,” and that’s just a fallacy. There is a type of startup in
what we call a new market, and we’ll be describing new markets in the Customer Segments lecture. But just
understand that in new markets of course you don’t get out and ask people what features they need. But
you do want to understand how is their day in a life different today versus the day after you give them your
new product? How does their world change? And there’s no possible way sitting locked in your conference
room or your office you would know that without talking to customers.

Business Model Canvas Key Resources

Next piece is you want to think about is what in the key resources? What do you need to make the business
model work? What assets are important? And what’s an example of an asset in a key resource, well finance.
Do you need capital? Do you need a line of credit? Some assets are in resources or physical. Do you need
physical plant like a manufacturing line? Do you need specialized machines? Do you need vans for delivery?
Do you need cars? Is there’s something else you need? Is there intellectual property you need? Is there
patents you need to acquire or protect? Do you need to acquire customer list? Or is it just that you need
to get great people, great shop or programmers in the specific area or great hardware designers or great
manufacturing people. And then finally again at the interaction between intellectual and human capital is
that’s another key resource. What specifically do you need to do to keep this people and who are they.

Pivot
One of the other interesting observations about customer development is this notion of the pivot. The
pivot was a term that my best student ever Eric Reese coined when he noticed the arrow between customer
validation and customer discovery, and he actually gave it a name, which I think is incredibly accurate. A
pivot says what do you do when your hypotheses don’t meet reality? And this is such a neat observation
about startups and why what we now know is much different than before. What we now know instead of
firing executives when our business model doesn’t match what’s going on outside in the real world we fire the
model, and we simply say, ”Hey, our hypotheses were wrong,” so because we’ve been building the product
iteratively and incrementally and keeping our burn rate incredibly low a pivot is a substantive change to
one or more of the business model components. It just simply says, ”Hey, this isn’t our customer segment.”
”Our customer segment is really here,” or ”Wait a minute.” ”Our revenue model shouldn’t be freemium.”
”We should be charging for it from day 1,” or ”Wait a minute.” ”We’ve been using the wrong distribution
channel.” ”We need a direct sales force,” or ”Gee, we have the wrong partners.” By the way, an iteration is

For use at FEUP in Course Unit: PRODEF029 4

a minor change to one or more of the business model components. For example, an iteration would be going
from charging from $9.99 to $6.99. A pivot would be a change of ”Gee, our pricing is going from freemium
to subscription.” That’s a substantive change. The key idea here is a pivot allows you to get out and make
changes. Remember, typically only the founders could do pivots, but it’s actually the heart of what makes
customer development radically different than what’s come before. The other thing to notice about pivots is
that you want to keep them up at a constant speed and a constant tempo, and you want your entire company
operating with speed and tempo in decision making just like a metronome. It’s constant, it’s consistent, and
it’s relentless.

Business Model Canvas Key Partners

The next piece is who are your key partners and suppliers. Partnerships are kind of interesting is we need to
ask ourselves before what’s the deal is what exactly are we acquiring from partners, and also what activities
are they going to perform and when. And this is where a start up sometimes make a mistake of thinking while
large companies do partnerships, I guess I need those two on day 1. Change all the types of partnerships
you need in year 1 are certainly not the ones you’re going to need in year 3 or 5 or 10, and the types of
partnerships could be strategic alliance joint ventures, just regular suppliers and buyers, and you need to be
thinking through who they are and after your getting out of the building and tested them.

Business Model Canvas Key Activities

Next, our key activities. What’s the most important things you need to do for the business to make
the business model work. Are you in the production business, are you making something, or you’re in the
problem-solving business like you’re doing consulting or engineering or you’re managing supply chains. What
are the key activities you need to become expert at.

Phases Of Customer Discovery

Earlier we looked at the 4 phases of the entire customer development process, but now let’s take a look at
the customer discovery process itself and order its phases.

Business Model Canvas Costs

Finally all these adds up on the left hand side over here to cost. What are the cost and expenses to operate
the business model? One of the interest thing things about cost is it’s not just the obvious ones like people
or buildings or materials. What you are going to be asking is what are the entire cost to operate a business
model. You want to think about our what are the most important cost you need to worry about? What
are the most expensive resources you’re going to need to pay for? And what key activities are the most
expensive? And then you want to ask the typical accounting things. What are fix cost? What are variable
cost? Are their economies of scale? And you want to start getting a good handle on what it is that will end
up costing you money to run your business.

Phases Of Customer Discovery Solution

And so the phases of customer discovery start with stating your hypotheses, testing the problem, testing the
solution, or verifying or pivoting.

Market Opportunity Analysis

One of the things we keep asking our startups to think about is how big is this opportunity? That’s just
a fancy word for saying when you’re all done and you’re spending the next couple years in the startup are
you going to make a million, or are you going to make a billion? That is, how potentially large is the
opportunity? And so what we really want to do is do what we call a market and opportunity analysis. What
you’ve already put together is a business model canvas with your hypothesis that says, ”Look, we’re really
passionate about this product or service, and we think there’s a set of customers out there, and we’ve put

For use at FEUP in Course Unit: PRODEF029 5

together a revenue model, our first hypotheses and saying we’re going to make a lot of money,” and that’s
great, but there’s really kind of a method to this madness, and there’s not only one way to do this, but this
is the way we kind of think of it. You want to identify who your customers and markets are, and you have,
but you now want to size the market, look at competitors, and figure out whether this market could grow.

Total Available Market

The first word we want to think about is something called Total Available Market, and I like to think of
Total Available Market as a pie. It’s the entire pie. Total Available Market says, look, how many people or
companies or whatever your unit of sale is would want or need this product, and how large is the market
in dollars or units if they all bought? That’s just a pretty nice calculation. And the question is, okay,
how would I find out? In different industries, there are analysts that specialize in vertical markets and
enterprise software, it’s Gartner and Forrester; in video games, it’s the MPD Group; in consumer research,
it’s Nielsen; if you want to understand how many mobile startups there are, you go to the Mobile Startup
Genome Project, et cetera. So you need to ask some questions as which industry analysts kind of follow your
specific domain, and then also Wall Street analysts follow competitors in this business. That is, they follow
whoever’s public; in fact, some of them write very nice research reports. If you have access to university
library or a friendly broker, you can get a great free industry analyst report and I would be asking others,
as well, and Google is your best friend here. I truly would be spending time trying to understand, is this
a $1 billion market, and by the way, if you throw out those numbers, the first thing I’m going to ask you
is, well, that’s nice, but break it down for me. Help me understand how many users, who are the players,
what are the competitors, et cetera. You need to understand this total available market pie in some detail
before I’ll let you get to the next step, and the next step is how big is my slice? And we tend to use a fancy
word for that: What’s the served available market? The served available market means okay, well instead of
the theory that there are 7 billion people in the world, how many people really can use a mobile app? Oh,
well a mobile app is kind of dependent on how many people have mobile Smart Phones, and if I’m making
the mobile app for, let’s say an Android platform, then the first question is how many theoretical people are
going to be using Android platforms in this year and in the next 5? Oh, now I can start estimating what’s
my served available market. And so many people have the money to buy the product; that is, are you a
99-cent product or are you a $99 product? Now you’re narrowing the market based on pricing or availability
and you want to do some thought experiments like how large would the market be in dollars if they all
bought– if everyone in the served available market bought, how large would this be, and this is kind of your
first test to say, ”Oh” or ”Wow.” So you want to understand this before dollars and units. How do you find
out? Well, this is one where you’re really out of the building and talking to customers. Then I think of this
again using the pie analogy the first step was try to understand the number of people in the world, but now
we can narrow it down if I have the Android app to number of Android phones–capable phones, and then
down to how many would buy it at my price, but now we really want to get pretty specific. Who example
am I going to sell in years 1, 2, and 3? How many customers is that? How large is the market if they all
bought? That is, what we now are coming up with is the total number of dollars if you had 100% market
share, your revenue isn’t going to exceed this number, and many units would that be? Again, how do I find
out? Boy, this is really about getting out of the building and talking to customers and talking to potential
channel partners, and talking to competitors, et cetera. You really, at the end of this exercise, now have the
first pass hypothesis about is this business model canvas worth executing for the next couple of years?

Market Size Summary

Zero. So in summary, the questions you want to ask are how big can this be, how much of it can we get,
what’s the growth rate of the market, is the market itself declining or is it in flux or is it taking off? and
most importantly here is talk to customers and the sales channel, and next important is you can get some
pretty good market size estimates by competitive approximation if it’s an existing or resegmented market,
or even if we’re going to clone a market outside the US. Next important is Wall Street Analytic Reports
are great. Market research firms like Forrester and Gartner are great, but something to keep in mind about
market research forecasts going forward. Market research data is wonderful on the size of markets in the
past, but if market researchers were great at predicting the future, they’d actually be running hedge funds,

For use at FEUP in Course Unit: PRODEF029 6

so one of the things I don’t accept from my students is, ”and it’s going to be a $40 billion business in the
next 7 years.” Well, what’s its size now? So therefore, I just don’t accept guesses from a market research
firm, unless you’ve actually cross validated that with some really heavy customer input from outside the
building.

Estimate SAM
So how would you estimate the served available market or SAM for an Android Game within an order of
magnitude? Choose from the list below.

Estimate SAM Solution

Would it be the total number of Web users? Well, for an Android game, they’re not on the Web. They might
hear about it on the Web, but that’s not really gonna help you understand the served available market. Is
is the total number of mobile phone users? Well, that might be the total available market, but you’re really
interested in just the Android piece and so number 3 might make sense, the total number of active Android
users. But if we get a little more sophisticated, we’d understand that the real answer is number 4; it’s not
only the total number of active Android users, but they need the right level of the operating system to
support the particular game’s features, and so you really want to get the SAM as tightly defined as you can.

For use at FEUP in Course Unit: PRODEF029 7