Pretotype it (first pretotype edition) - ProductCamp Nuremberg 2014
Pretotype it (first pretotype edition) Make sure you are building the right it before you build it right. ProductCamp Nuremberg 2014
Published on: Mar 4, 2016
Transcripts - Pretotype it (first pretotype edition) - ProductCamp Nuremberg 2014
Make sure you are building the right it
before you build it right
First Pretotype Edition
Copyright (c) Alberto Savoia 2011
This book is dedicated to my family:
To my father, who believed in me and made great sacrifices to invest in me
when I first started and I had nothing but some crazy ideas. Thanks dad,
you were my first VC.
To my mother, who let me leave Italy when I was seventeen to follow my
dreams in Silicon Valley. Thanks Mom, I know that must have been hard.
To my wife, who has always supported my entrepreneurial risk-taking,
makes our house a home and makes parenting seem easy.
To my children, who make me proud every day and who make the tough
job of being a father almost too easy.
This Is Embarrassing
This is not a “proper book.”
Writing and editing a proper book on the subject of pretotyping would
take months. I would love to write that book, but at this time I have no
indication that such a book would be worth writing. Most books fail in
the market, and most of them fail not because they are poorly written or
edited, but because there aren’t enough people interested in them. They
are not the right it.
What you are reading now is a first pretotype edition of the book. I
wrote and “edited” it in days instead of months, just to test the level of
interest in such a book. I had a few friends and colleagues review it, but
don’t be surprised if you find typos, misspellings, bad grammar, awk-
ward formatting and all sorts of misteaks.
Releasing it in its present state is not easy for me.
The toughest thing about pretotyping is not developing pretotypes, that’s
the fun part. The tough part is getting over our compulsion for prema-
ture perfectionism and our desire to add more features, or content, before
releasing the first version. The tough part is getting our pretotypes in
front of people, where they will be judged, criticized and – possibly – re-
Reid Hoffman, founder of LinkedIn once said: “If you are not embar-
rassed by the first version of your product, you’ve launched too late.”
I am plenty embarrassed. I must be on the right track.
At this very moment, millions of people across the world are pouring
their hearts, souls, hopes, dreams, time, money, and energy to develop
new ideas that once launched will flop miserably.
At this very same moment, a much smaller number of people are devel-
oping new ideas that will turn out to be successful – and some of them
will be extremely successful: the next iPod, the next Google, the next
What group are you in?
Most people believe that they are working on a winning product, but we
know that can’t be true.
Most new ideas fail, and predicting the actual market success of a new
idea with any degree of confidence is next to impossible. Some brilliant
“can’t fail” ideas turn out to be gigantic flops, while some crazy “who’d
want that?” ideas turn out to be spectacular successes.
Some people and organizations may be better at predicting success than
others, but even the best venture capitalists, investors or entrepreneurs
regularly invest way too much on the wrong ideas and frequently choose
to invest nothing on the right ideas.
If all we have is an idea for some new product (or service, or book, etc.),
the best thing we can do with that idea is collect opinions about its use-
fulness or market potential. Ideas are fuzzy and abstract; opinions are
subjective and even more abstract; when you combine the two you get a
big fuzzy ball of abstractions and opinions. Not much to go on.
Traditional prototypes can help to test and validate the market potential
of new ideas more concretely and objectively than ideas and opinions.
In many cases however, the development of a “proper prototype” is too
difficult, expensive and time consuming. It’s normal to invest weeks,
months or years, and hundreds of thousands or millions of dollars to de-
velop prototypes. Furthermore, most prototypes are built to answer
questions such as, “Can we build it?” or “Will it work as expected?” in-
stead of focusing on questions such as “Should we build it at all?” or “If
we build it, will people buy it and use it?” Unless you can answer the
latter questions positively, the former questions are of little importance.
Prototypes can help you fail faster, but often not fast enough or cheaply
enough. The more you invest in something the harder it is to let it go
and admit it was the wrong thing. Once you have a “proper prototype”
working, it’s tempting to work on it a little longer and invest in it a little
more: “If we add this one feature I am sure that people will finally use
it.” Prototypes often turn into productypes – a prototype gone too far
– and you can kiss fail-fast goodbye.
Between abstract ideas and “proper prototypes” there are pretotypes.
Pretotypes make it possible to collect valuable usage and market data to
make a go/no-go decision on a new idea at a fraction of the cost of pro-
totypes: hours or days instead of weeks or months, and pennies instead
of dollars. Pretotyping helps you fail fast, recover fast and leaves you
plenty of time, money, energy and enthusiasm to explore new tweaks or
ideas until you hit on something that people seem to want – the rare and
wonderful right it!
A lot of what you will read in this book will seem obvious to you. But
before dismissing it too quickly, look around at all the products, serv-
ices, apps, books, etc., that are launched every day only to fail soon after.
Most of these new products don’t fail because the people who worked
on them were stupid, lazy or incompetent, nor because they were built or
marketed poorly, but because they were not the right product to start
Chances are that, unless you are just starting in your business or indus-
try, you can look back at some of the products you have worked on and
identify quite a few that in retrospect were not the right product. That’s
certainly the case for me; I’ve been lucky enough to work on products
that turned a few months of work into millions (and eventually billions)
as well as products that turned years of work and tens of millions into
Even though this version of the book is itself a pretotype edition (I’ll
elaborate on that later), it should have enough “meat” to be worth your
while. I sincerely appreciate the fact that you are reading it: please send
me your feedback – I need data to decide if I should invest the time re-
quired to turn this pretobook it into a proper book.
Alberto Savoia (email@example.com)
Mountain View, CA
The concept of pretotyping as well as this book, would not have been
possible without the encouragement and support of Patrick Copeland
– my manager and mentor at Google. Patrick not only helped me to de-
velop and refine these ideas, he makes sure that I practice what I preach
– and launch early and often. He has also helped me to spread the word;
he has given several extremely well-received keynote presentation on
pretotyping at major conferences throughout the world.
I have been very fortunate to have two great innovators working with me
at Google: Stephen Uhler and Bob Evans. Stephen is a born pretotyper
who has helped me immensely by inventing and developing PretoGen, a
tool that makes it possible to generate working pretotypes in minutes. I
use PretoGen to end most presentations by creating custom live preto-
types on demand. Bob, one of the smartest guys I know, has been a con-
stant source of inspiration and lively debates during the earliest days of
pretotyping. As a matter of fact, the idea for pretotyping was born from
our discussions while we were sharing an office.
The other key person in the development, refinement and popularization
of pretotyping is Jeremy Clark, a leading thinker and practitioner in the
area of innovation and the founder FXX Inc. Jeremy and I continue to
work together on pretotyping and we often give joint conference presen-
tations on the subject.
Carlo Alberto Pratesi, professor of Marketing at Roma Tre University
and founder of InnovAction Labs in Italy, has not only been a source of
inspiration and examples but he has been very active in putting pretotyp-
ing in action in Europe.
Finally, I want to thank the hundreds of Googlers (and Google customers
and visitors) who have come to my presentations and workshops. Their
positive reaction to pretotyping, their own experiments with it and their
ongoing suggestions and enthusiasm convinced me that pretotyping is
the right it.
The Right It
The title of this book is “Pretotype It” and the subtitle is “Make sure you
are building the right it before you built it right.”
I will explain and define pretotyping very soon; before I do, however,
we need to address the following question:
What is this it I speak of, and why is it so important to have ‘the right
What is this it I speak of?
In the context of this book, it can be a new product, a service, a book, a
startup, a charitable organization, a video game, an innovative type of
boat, a new musical instrument, a revolutionary genetically engineered
hypoallergenic hamster, …
It is something that does not exist yet, but you have been thinking about
it and would like to – or have to – create it and bring it to life.
It is something important to you, and creating it will require a non-trivial
combination of your time, effort and money, as well as a considerable
amount of your energy, drive, enthusiasm and commitment.
Ideally, it is something that you are deeply passionate about, but it’s OK
if it is just something you have to do as part of your job.
What’s so important about having ‘the right it’?
The odds are heavily stacked against the success of your it. Hopefully
this is not news to you. I’m sure you’ve heard statistics similar to the
following many times:
• 90% of all mobile apps don’t make any money
• Four startups out of five lose money for the investors
• 80% of new restaurants close within one year
Most new its fail. Unless you’ve been given some form of divine dispen-
sation, you have the same odds as everybody else. Chances are the it you
are currently thinking of will not succeed – unless your it happens to be
the rare right it.
If you don’t have the right it then, by definition, you must have the
wrong it. One of the most wasteful and costly things you can do is to
continue working on the wrong it hoping you will be able to make it a
success through sheer willpower and effort. Unfortunately, this happens
very rarely: generally speaking, there is no amount of time, effort or
money that can make the wrong it succeed.
Movies are a good example of how next-to-impossible it is to turn the
wrong it into a success at the box office. If the movie concept (the it in
this case) is not right, no star directors, actors or $100M+ budget is go-
ing to turn the movie into a success (e.g., “Ishtar”, “Heaven’s Gate”,
“Howard the Duck”.)
On the other hand, if you have the right it, everything becomes much
easier and the odds of success swing in your favor. In the case of mov-
ies, this would be a movie with little or no budget, an inexperienced di-
rector, no name actors and no expectations, that turns into a blockbuster
(e.g., “The Blair Witch Project”, “El Mariachi”, “Paranormal Activity”.)
Having the right it is essential. Most people or organizations don’t have
unlimited time, energy or money to sustain a long string of slow and ex-
pensive failures caused by chasing wrong its. The goal of pretotyping is
to weed out wrong its and find the elusive right it with the minimum in-
vestment of time, money and effort.
Why do I keep writing ‘it’ in bold and italic?
The concept of pretotyping is applicable to a wide range of ideas for
products or services – software, hardware, websites, games, soft drinks,
hard drinks, books, movies, etc. Since it’s cumbersome to write (and
read) things like “If your product or service is …”, I decided to simply
refer to whatever your idea is as it.
Throughout this book, I write it in bold and italic to differentiate it (your
idea) from the pronoun ‘it’. Since this book is – at least at this time – a
pretotype itself, I might have missed a few its here and there. Hopefully
it will be clear from the context when I am referring to your it.
As you read on, some of good acronyms and mnemonics for it are:
• idea on the table
• idea to test
• innovation to try
What Is Pretotyping?
Now that you have a rough idea of what I mean by the right it, we can
give pretotyping a proper introduction. The best way to do that is by
sharing with you the two stories that started me thinking about this
whole thing: the IBM speech-to-text “experiment” and the the Palm Pi-
The IBM Speech-to-Text Experiment
I first heard this story during a presentation at a software conference a
few years ago. I am not sure how accurate my description of the events
is; I probably got a few details wrong, but in this case the moral of the
story is much more important than the details. With that caveat out of the
way, here’s the story as I remember it.
A few decades ago, well before the age of the Internet and before the
dawn of ubiquitous personal computing, IBM was best known for its
mainframe computers and typewriters. In those days, typing was some-
thing that a small minority of people were good at – mostly secretaries,
writers and some computer programmers. Most people typed with one
finger – slowly and inefficiently.
IBM was ideally positioned to leverage its computer technology and
typewriter business to develop a speech-to-text machine. This device
would allow people to speak into a microphone and their words would
“magically” appear on the screen with no need for typing. It had the po-
tential for making a lot of money for IBM, and it made sense for the
company to make a big bet on it.
However, there were a couple of major problems. Computers in those
days were much less powerful and more expensive than today, and
speech-to-text requires a lot of computing power. Furthermore, even
with adequate processing power, speech-to-text translation was (and still
is) a very difficult computer science problem. Tackling it would have re-
quired a massive investment – even for IBM – and many years of re-
search. But everyone would have wanted such a device. It would be a
sure-fire hit. Or would it?
Some folks at IBM were not convinced that all the people and compa-
nies who had said they “wanted and would definitely buy and use”
speech-to-text machines would actually end up buying them. They
feared the company would end-up spending years in research and lots of
money developing something that very few would actually buy: a busi-
ness disaster. In pretotyping lingo: they were not sure that speech-to-
text was the right it. After all, people had never used a speech-to-text
system, so how could they know for sure they would want one? IBM
wanted to test the business viability of such a device, but since even a
basic prototype was years away, they devised an ingenious experiment
They put potential customers of the speech-to-text system, people who
said they’d definitely buy it, in a room with a computer box, a screen
and a microphone – but no keyboard. They told them they had built a
working speech-to-text machine and wanted to test it to see if people
liked using it. When the test subjects started to speak into the micro-
phone their words appeared on the screen: almost immediately and with
no mistakes! The users were impressed: it was too good to be true
– which, as it turns out, it was.
What was actually happening, and what makes this such a clever ex-
periment, is that there was no speech-to-text machine, not even a proto-
type. The computer box in the room was a dummy. In the room next
door was a skilled typist listening to the user’s voice from the micro-
phone and typing the spoken words and commands using a keyboard:
the old-fashioned-way. Whatever the typist entered on the keyboard
showed up on the user’s screen; the setup convinced the user that what
was appearing on the screen was the output of the speech-to-text ma-
So, what did IBM learn from this experiment?
Here’s what I’ve heard: After being initially impressed by the “technol-
ogy”, most of the people who said they would buy and use a speech-to-
text machine changed their mind after using the system for a few hours.
Even with fast and near perfect translation simulated by the human typ-
ist, using speech to enter more than a few lines of text into a computer
had too many problems, among them: People’s throat would get sore by
the end of the day, it created a noisy work environment, and it was not
suitable for confidential material.
Based on the results of this experiment, IBM continued to invest in
speech-to-text technology but on a much smaller scale – they did not bet
the company on it.
As it turned out, that was the right business decision. Keyboards are
proving hard to beat for most text entry tasks. Thirty years ago most
people could not type; but look at any office (or coffee shop) today and
you’ll see people of all ages and professions typing away on their lap-
tops. In devices where a full-size keyboard is not possible, such as mo-
bile phones, speech-to-text can be the right it, but otherwise the key-
board is still the device to beat. The keyboard is definitely the right it.
The IBM approach was ingenious, but what would you call it? The
speech-to-text setup with the typist was not what one would consider a
“proper prototype” – not unless they were planning to actually hide liv-
ing and breathing typists into computers. They didn’t prototype a
speech-to-text system, they pretended to have a speech-to-text prototype,
and used it to test users’ actual reaction to the product. This way they
were able to collect valuable market data based on actual usage instead
of opinions, and they did that with a very small investment of time and
I thought that this was a very interesting and valuable approach, and that
it was different enough from prototyping to deserve its own name (more
about that later) and more study. But first I set out to find similar stories
and uncovered another brilliant example.
The Palm Pilot Experiment
The IBM speech-to-text story started me thinking about the concept of
pretotyping, but this next example is the one that convinced me that it
was worth pursuing further.
Introduced in 1996, the Palm Pilot was a palm-sized digital device with
four basic functions: a calendar, an address book, a to-do list and a sim-
ple note taker. The Pilot was the first successful PDA (Personal Digital
Assistant.) But Jeff Hawkins, Palm’s co-founder and one of the inven-
tors of the Pilot, did not take the eventual success of PDAs for granted.
Quite the contrary. According to a March 1998 story on Time magazine
Hawkins, 40, Palm's chief technologist and Pilot's creator, designed one
of the first handheld computers, the GRiDPad, a decade ago. It was an
engineering marvel but a market failure because, he says, it was still
too big. Determined not to make the same mistake twice, he had a
ready answer when his colleagues asked him how small their new device
should be: "Let's try the shirt pocket."
Retreating to his garage, he cut a block of wood to fit his shirt pocket.
Then he carried it around for months, pretending it was a computer. Was
he free for lunch on Wednesday? Hawkins would haul out the block and
tap on it as if he were checking his schedule. If he needed a phone num-
ber, he would pretend to look it up on the wood. Occasionally he would
try out different design faces with various button configurations, using
paper printouts glued to the block.
Here’s a photo of the pretotype Jeff built (you can see this artifact at the
Computer History Museum in Mountain View, CA):
I can only imagine people’s reaction when Hawkins pulled out a block
of wood out of his pocket and tapped on it, pretending it was a working
device. They must have thought he was crazy. Yes, crazy like a fox!
That piece of wood with paper printouts convinced Hawkins that he was
on the right track. He had answered the first, and most important, ques-
tion: “If I had a Pilot, would I actually carry it with me and use it?” And
his answer was a definite “yes!” He knew he had the right it; now he
could focus on the next set of questions, such as: Can we build it this
small? How much would it cost to build? How long would the batteries
last? It was time to invest in building a “proper prototype.”
The Palm Pilot was not just successful, it was hugely successful and had
a tremendous impact. The Pilot was the predecessor to today’s smart
phones, and it all started with a little piece of wood – just like Pinocchio.
Fake It Before You Make It
The speech-to-text and Palm Pilot stories have several things in com-
Both teams had doubts about the eventual usefulness and adoption of their
innovation. It was a cool idea. It made sense. It solved a problem. But
was it the right it? Would people actually use it? Jeff Hawkins, in particu-
lar, had just been burned by investing years to develop a product, the
GridPad, that was “an engineering marvel but a market failure” (i.e., the
wrong it) and was “determined not to make the same mistake twice.”
Because of their doubts, both teams wanted to test the usefulness of their
idea with a prototype and collect feedback from actual usage of the prod-
uct (as opposed to opinions about the product) before committing to its
In both examples, however, even the development of a “proper prototype”
(a crude but functional version of the final product) would have required a
lot of upfront time and a significant investment in research and develop-
Their solution to the “proper prototype” problem was to pretend that they
had such a prototype. In the speech-to-text example, actual hardware and
software was replaced with a bit of trickery, and in the Pilot example it
was replaced by Hawkins’ imagination – fake it before you make it.
I found these two stories striking because they are so different from the
typical approach people and companies take when they have an innova-
tive idea they want to pursue. Most people fall in love with their idea
(their it) and assume it will be successful (the right it) so they just start
building it. They jump the gun – so to speak – and begin by focusing
and investing in the wrong things at the wrong time. More precisely,
they invest too much too soon to develop a first version of the product
with too many features, too much functionality and too much “polish.”
They presume to know what people will want. They assume that if they
build it right, people will want it. In most cases, these presumptions and
assumptions turn out to be both wrong and costly.
Pretotyping: A Word Is Born
The more I thought about the speech-to-text and Pilot experiments, the
more I became convinced that what those teams did was not only clever,
but a crucial step in the process of developing new and innovative prod-
ucts. A step that most people skip and often end up paying dearly for it.
Over a span of several months, I shared these two stories with dozens of
colleagues, friends, entrepreneurs, venture capitalists, engineers and
product managers. Surprisingly, none of them had heard of these exam-
ples before. All of them, however, were similarly impressed by the
clever solution of “faking it before making it,” and a good number of
them slapped their heads and said things like: “I wish I had done some-
thing similar before sinking years and millions in my last idea.”
I realized that I had stumbled into something valuable and important
that, while not new or original, was neither well known nor widely prac-
ticed. Quite the contrary. But this something did not have a term to de-
scribe it, and I thought it deserved and needed a name in order to be-
come better known, discussed and more widely adopted. So I started
thinking of a possible term for this concept. (Note: At the time I started
thinking about pretotyping I was not yet aware of Eric Ries’ great Lean
Startup Movement or the term MVP – Minimum Viable Product. More
about the relationship between pretotyping and MVP later.)
Since a core element in both examples is the act of pretending (the IBM
folks pretended to have built a speech-to-text system and Jeff Hawkins
pretended to have a Pilot in his shirt-pocket,) the first word that came to
mind was pretendotyping – yikes! My second attempt at coining a word
was even worse. Since the core idea is to quickly test an idea even be-
fore you invest in building a proper prototype, I came up with the word
preprototyping – double yikes! Fortunately, these two horrific words
contained the seed for a much better term. By dropping a few letters
here and there, I came up with pretotyping. Much better. The artifacts
that are produced by the process of pretotyping (e.g., Hawkins’ wood
block) would be called pretotypes.
I liked the words pretotyping and pretotype, but was I the first to use
them? Perhaps someone else had already been using them and had some
kind of “rights” to their use and meaning. I ran to Google and typed
“pretotyping” in the search box. To my delight, Google came back with
“Did you mean prototyping?” The search engine assumed that I had mis-
spelled the word so it gave me a bunch of results for prototyping instead
– a good sign. When I insisted that I did not mean prototyping and to
please give me results for pretotyping instead, Google returned a rela-
tively small amount of pages where people had misspelled the word pro-
totyping. Searching for pretotype yielded similar results. The coast was
clear. I had stumbled into a new word that nobody else was using.
Even better, the associated domain names pretotyping.[com, org] and
pretotype.[com,org] were also available. My first instinct was to pull
out my credit card and buy all of them, but I realized that by doing so I
would be violating the core message of pretotyping: I would be investing
in something before making sure that it was worth investing in it. Even
though it would only cost a few dollars to reserve the domain names, the
principle was at stake. I thought that pretotyping and pretotype were
great words to describe a great concept, but would other people feel the
same? I had to pretotype pretotyping.
Fortunately, as part of my job at Google, I have the opportunity to talk to
a lot of people about innovation and I give a lot of presentations on the
subject both to customers and to other Googlers. So, along with the
speech-to-text and Palm Pilot examples, I started using the words preto-
typing and pretotype in all of my presentations, meetings and discus-
sions. In every single case, people responded very positively to both the
concept and the words. They were sending me examples of their preto-
types, asking me for suggestions on how to pretotype their ideas and
they were even lobbying with their colleagues and managers to preto-
type ideas before starting to build them. It looked as if I was on the right
One day, I was forwarded an email from the head of one of the biggest
advertising agencies in the world who had attended my presentation on
innovation. He thanked the organizers for the presentation, said that he
and his team loved the concept of pretotyping and that “… the word pre-
totyping has now entered our company’s lexicon.”
That day, I knew I had enough evidence that pretotyping and pretotype
were the right words for the right concept, and I felt confident taking the
next step and investing a few dollars to buy the associated domain
Although I believe it’s more effective to use examples to explain preto-
typing, it’s worthwhile trying to define it.
Here’s a somewhat formal definition – the dry and boring kind you’d
find in a dictionary:
Pretotyping [pree-tuh-tahy-ping], verb: Testing the initial appeal and
actual usage of a potential new product by simulating its core experi-
ence with the smallest possible investment of time and money.
Here’s a less formal definition:
Pretotyping is a way to test an idea quickly and inexpensively by cre-
ating extremely simplified, mocked or virtual versions of that product
to help validate the premise that "If we build it, they will use it."
Here’s a very informal definition:
Pretotyping: Fake it and test it before you make it!
My favorite definition of pretotyping, however, is based on the subtitle
of this book:
Make sure – as quickly and as cheaply as you can – that you are
building the right it before you build it right
Pretotyping and Prototyping
Some people might argue that pretotyping is too close to prototyping
both in spirit and practice and, therefore, there is no need to differentiate
between the two nor to invent a new word. I thought about this issue a
lot. The problem I see is that the term prototyping covers a huge range
of the spectrum between the abstract idea for a product and the final
A prototype for a speech-to-text computer, for example, could include an
actual combination of hardware and software to digitize speech, break it
down into phonemes, convert the phonemes to possible words and sen-
tences, apply error corrections to those words and sentences, etc. Such a
prototype would take months or years of development and cost millions
of dollars. It would be a one-off and still be far from being a final prod-
uct, so it would definitely be considered a prototype. It would be exactly
what most people think of when they think of a prototype.
Mention the word prototype to someone, and they will imagine some-
thing primitive and with rough edges but they would expect it to be
somewhat functional and close to the final product. If Jeff Hawkins told
people he had a prototype for the Palm Pilot, those people would expect
to see something with batteries and an LCD screen, not a block of wood.
If IBM told its potential customer that it had a prototype speech-to-text
machine, they would not expect a human typist taking dictation in the
room next door.
Besides functionality, a key difference between pretotypes and proto-
types is that the cost and time-frame for pretotyping is at the lowest end
of the spectrum that is usually covered by prototyping. It’s acceptable
for a prototype to take months or years of development and cost millions
of dollars. It’s definitely not acceptable for a pretotype to take that long
or cost that much.
Prototypes are a necessary and incredibly useful tool that can and should
be used to answer many questions about a potential product, such as:
• Can we build it?
• Will it work at all?
• Will it work as intended?
• How small/big can we make it?
• How much would it cost to produce?
• How long will the batteries last?
• How will people use it?
• What will people use it for?
Pretotyping, on the other hand, focuses on answering one – very basic
and very important – question: Is this the right thing to build? Once
that question is answered positively, then it makes sense to move from
pretotyping to prototyping.
The conclusion I reached is that the term and practice of pretotyping de-
serves to stand on its own. Just as a startup is a specific type of early-
stage company, pretotyping can be viewed either as a specific subset of
prototyping or a prelude to it.
It Will Fail
You now have a rough idea of what pretotyping is about and we’ll get
into more details and examples a bit later, but before doing that I want to
spend some time explaining why pretotyping all your ideas is so impor-
Do you remember this sobering collection of statistics from earlier on?
• 90% of all mobile apps don’t make any money
• Four startups out of five lose money for the investors
• 80% of new restaurants close within one year
The actual numbers may vary, but the message is clear. Simply put:
Most new its are destined to fail – yours included. Most its are destined
to fail because they are the wrong it: ideas that may sound good in the-
ory but once developed turn out to be nowhere near as interesting, com-
pelling, or as useful as originally anticipated.
Pretotyping does not have the power to turn a wrong it into a right it
– nothing can do that. But pretotyping will help you identify wrong its
quickly and cheaply so you can keep trying new its (or variations on ex-
isting its) until you find the elusive right it.
Since failure is our enemy, and it’s important to “know thine enemy”,
let’s look at failure a bit closer.
The Law of Failure
The evidence for the really bad odds against new its is so compelling
and reliable that we can decree it to be a law:
The Law of Failure
Most new its will fail – even if they are flawlessly executed.
Where the word “most” stands for a disheartening high percentage (typi-
cally 70-80-90%) and its represents pretty much any category you can
think of: startups, restaurants, movies, books, soft drinks, TV shows, etc.
And, yes, your it is included in one of those categories and, yes, it has
the same lousy odds as everybody else’s its.
I can hear some of you complain: “But how does this law help us? It just
tells us that we’ll probably fail even – if we do a great job with our it. It
gives us crappy odds and leaves us hanging. All that does is lower our
morale and kill our enthusiasm.”
True, on the surface the Law of Failure does not appear to be very help-
ful. Strictly speaking, it’s not even a proper law. Can you imagine if
Newton had stated his observations on gravity as: “Most things will fall
if dropped?” Isaac, however, had it relatively easy. He was dealing with
an immutable and universal law of nature. The eventual market success
of any new product, on the other hand, has to contend with highly fickle,
mutable and (as often than not) irrational human behavior. In this con-
text, the probabilistic formulation of the first law is as good as it gets.
While far from perfect, I believe The Law of Failure is extremely impor-
tant. If you accept this law as true, or even “mostly true” or “possibly
true,” and that neither you nor your it are not exempt from the law, your
mindset should change from: “Let’s go for it! Let’s just build it, and go
for broke!” to a more cautious “Let’s try it. Let’s pretotype it!”
I know that “Go for it!” and “Go for broke!” have great romantic and he-
roic appeal. “Jumping in with both feet”, “betting the farm” and “damn-
ing the torpedoes” is how many legends are made – but it’s also how
catastrophic failures arise.
Having said that, I feel the need to mention that there may be cases when
you may want to decide that you don’t care what the odds are and you
just want to do or build your it regardless of what happens. I am not dis-
couraging that. I believe that, at least a few times in our lives, we should
all take some crazy risks and just go for it! There should be times when
you care more about creating your specific it than having the right it. If
that’s the case, laugh at the Law of Failure, throw caution to the wind,
throw this book in the wastepaper basket and throw your heart and soul
into it. Godspeed! I am rooting for you and I wish you success.
If, on the other hand, you are in a situation where you are not 100%
committed to a very specific it, and maximizing the odds for success is
critical, then give The Law of Failure the respect it deserves because ...
... Failure Is Not An Option
That’s right. For any given it, failure is not an option, but the most
We can’t get away from the Law of Failure, we can’t change the odds
for new its.
What we should, and will, do instead, is use The Law of Failure to our
advantage, the same way accountants use tax laws and Lady Gaga uses
How do we do that?
We invite failure, we seek it, we hunt it down and get it to show its ugly
face as soon as possible so we can determine if we are on the wrong
track and make the necessary adjustments.
We concoct some really cheap bait (bait) in the form of a pretotype.
Something that looks and smells like our it; something we can use to try
to trick the beast of failure into rearing its ugly little head. We trek to the
entrance of the dark and musty hole in the ground where the beast
dwells. We then dangle our pretotype bait at the entrance of the hole to
see if the beast of failure emerges from the shadows and inches close to
us, close enough that we can smell its fetid breath and catch a glimpse of
its cruel mouth and beady eyes. Close enough to make sure that it is the
real beast, then we toss our cheap bait to the beast as a sacrificial offer-
ing and run like hell in the opposite direction – before the beast can sink
its teeth into our flesh and drag us down to its miserable hole to feast on
The best thing you can do is feed the beast cheap little morsels of vari-
ous its. The beast likes to eat wrong its but – given half-a-chance – it
would LOVE to eat you! You must be ready to toss morsels of your its
to the beast and run away. If you don’t, if you get too attached to your
its, and invest too much time developing it before dealing with the beast,
you will probably end up having all your time and effort devoured by the
If we do this well, we’ll lose our bait (our pretotype), but we get to live
another day and try a different it – and keep trying until we come up
with a bait that fails to attract the beast of failure – a bait that might just
turn out to be the right it.
Pursuing your idea to the end, even if it turns out to be the wrong idea
may sound exciting and heroic. But pretotyping is no less exciting. In
pretotyping, you are still on a heroic and difficult quest – the quest for
the right it. Between you and the right it there’s the formidable beast of
failure. You cannot avoid dealing with the beast; you still have to fight it
– but with pretotyping your odds of success are much greater.
That’s the essence of our strategy – the essence of pretotyping. But
playing this game with failure makes sense only if the bait we use is
something cheap and inexpensive a pretotype that we put together in a
few hours or days and at a minimal cost – something we don’t mind
Three Ways To Fail
Failure is the most likely outcome for any given it, but not all failures
are created equal. There are three main ways to proceed with your it;
three ways to deal with the beast of failure:
• Do nothing with it
• Go for it (productype it)
• Give it a try (pretotype it)
The first way is to fail the is way of sloths and chicken: people or com-
panies too lazy, insecure or cowardly to put forward any effort or risk
anything. Dealing with failure by not trying is the surest way to always
fail. If you’ve read this far, I’m sure you are not in that category. You
are ready to build something.
The second way to fail, is the exact opposite of the first. Instead of lazi-
ness, insecurity and cowardliness, we have excessive eagerness, hubris
and confidence. Dealing with failure by underestimating it will, more
often than not, lead to painful, costly and slow failure.
These first two types of failures are caused by too much thinking, too
much talking and too little reality too late. All its are born as ideas, but
if we don’t quickly shift from thinking and talking to putting something
concrete in front of our potential users and customers, our its run a real
risk of spending too much time in a very dangerous place I call Thought-
Thoughtland is a fictional place inhabited by two strange species that
float around and interact: ideas and opinions. More precisely: unrealized
ideas and opinions about those unrealized ideas.
Thoughtland is where every it begins life as a simple, pure, abstract idea.
As they float about in Thoughtland, its attract opinions that stick to
them, like barnacles to a ship.
Thoughtland is a very safe place for ideas because, until they are actu-
ally turned into something more tangible such as a rough prototype of an
app or the first draft of a book or script, they cannot fail. The only thing
an abstract idea can “produce” is an opinion: something even more ab-
stract and of even more dubious value.
But Thoughtland is a very dangerous place for creators, innovators, en-
trepreneurs and authors. The opinions that fester in Thoughtland and at-
tach to our idea can lead us to fail in two painful ways:
False Negative opinions about our it scare us into abandoning our idea
so we do nothing with it.
False Positive opinions about our it make us blind to The Law of Failure
and cause us to prematurely overcommit and go for it.
Let’s look at how these two scenarios happen.
The “Do Nothing” Scenario
Most its never leave Thoughtland. They remain forever in this limbo as
unrealized ideas. This is the saddest form of failure. Sure, the odds
were that that idea was a wrong it, but there was a small chance that this
idea was the next Palm, or Google or Twitter, and someone gave up on it
without even trying it. Sad. Sad. Sad.
A good percentage of its fail to see the light of day simply because their
ideators never get off their butts to do something about them. They be-
lieve the idea is a winner, other people tell them that the idea is a winner,
but they are too lazy/tired/busy/broke/inexperienced/afraid/(insert favor-
ite excuse) to do something about it. As we’ll see later, pretotyping can
help us deal with this particular situation.
The remaining percentage of its fail to see the light of day not because
we are lazy/tired/busy/…, but because, while in Thoughtland, our its at-
tracted enough negative opinions (possibly from ourselves but mostly
from others) that our belief in it first wavered then collapsed altogether.
This happens a lot and, unfortunately, it kills many right its. How does
this happen? Let me give you an example:
Let’s assume that Alice has an idea for a mobile app, something that uses
text messaging and allows people to create short messages – say 100-
200 characters max – that will be automatically broadcasted to our
friends, family or anyone who wants to follow us. Let’s call this app
Alice takes her idea for MultiTextBot, her it, to Thoughtland. Let’s see
Alice describes MultiTextBot to a couple of dozen friends and solicits
Almost all of her friends tell her that this is a lame idea and that
they’d never use it:
“Who cares about what you are doing?”
“Why would I follow you?”
“I don’t want to be followed.”
“What’s with that stupid 100-200 character limit?”
The few friends that are too nice to be totally negative make helpful
suggestions: “Perhaps if you got rid of that stupid character limit and
add the ability of sending photos and GPS coordinates before you
even think about launching it.”
Silly friends. What do they know anyway? Alice decides to take her it
to real professionals - to VCs. They’ll see how great it is.
The VCs don’t get it either. Some just pass: “Sorry kiddo, this is not
big enough for us, but good luck!” Some ask for user numbers, but
Alice has none: “Sorry, it’s just an idea at this time, but look at my
PowerPoint projections …” The VCs tell Alice to “… come back if
and when you have a million users then we’ll talk.”
Wow. How could Alice even think that this was a good idea. Good thing
she asked for opinions before quitting her job and started development
on this stupid app. She decides to forget about it. Phew, that was a close
This happens – a lot! Of course, since most its are not the right it, nega-
tive opinions kill a lot of bad ideas. But they also kill many innocent
and very promising right its.
Most of you have probably realized that Alice’s MultiTextBot idea is a
very thinly disguised description of Twitter – arguably one of the most
successful and world-changing products in history.
And yet, before Twitter’s usefulness and impact became obvious and ir-
refutable, the initial opinion and reaction of most people who heard the
idea – including a lot of VCs and smart investors – was negative: they
didn’t get it. There are still people who don’t get it, but that does not
matter because there are tens of millions who do get it and use Twitter
everyday. Twitter was the right it – but you wouldn’t have known it by
its reception in Thoughtland.
The “Go For It” Scenario
We’ve seen how negative opinions kill many right its in Thoughtland.
But that’s only half the story. Let’s look at how positive opinions get us
to overcommit to wrong its.
We need another example for our it.
How about this: Tom, a first-class software developer, has an idea for a
mobile app to help romantically-challenged guys like himself. This app
will automatically send thoughtful text messages to their significant
other at random times during the day. Thanks to this app – let’s call it the
HoneyTextBot – the significant other will receive text messages like: “Hi
honey. I’m thinking of you. Love. Your little hamster.” or “Hey babe, I
just text to say I LOVE YOU. XOXOXO”
Tom’s HoneyTextBot will make the significant other think that the
romantically-challenged guy is thinking about them at that moment
– even though he may be out drinking beer with his pals and watching
mud wrestling. How romantic!
This is Tom’s it – his new idea on the table.
Here’s what can happen to Tom’s idea in Thoughtland:
Tom mentions his idea for the HoneyTextBot app to a couple of doz-
ens of friends and colleagues (all males) to solicit their opinion. He
calls this his “market research.”
Most of Tom’s friends, let’s say 70%, think it’s a great idea and tell
Tom they will definitely buy the app for $1.99 and would use it regu-
Tom extrapolates from his “market research” and reaches the conclu-
sion that he could easily make millions with this app: “70% of guys
with phones and a significant others times * $1.99 = ... dunno exactly
... but it’s got to be a lot of money!”
Bolstered by such favorable expert opinions and careful financial
projections, Tom quits his job and spends 3 months and all his savings
writing a full-featured and highly-polished version of HoneyTextBot.
Tom is a great developer and has an excellent sense for design, so the
app looks beautiful and operates flawlessly. The first version can
send lovely little text messages in over 20 languages! To cover all
bases, and pre-empt any competition, Tom decides to develop and
launch the app on all the major mobile platforms (Android, iPhone,
Tom launches HoneyTextBot simultaneously on the Android, iPhone
and Blackberry market and …
… not much happens. Nobody seems interested in Tom’s beautifully
crafted app. Not even his friends. Of those two dozen friends – 70%
of whom of told him they’d buy and use HoneyTextBot – only three of
them actually bought it, and only after Tom reminded them several
times. After a week, two of them uninstalled it from their phone and
the third forgot it was even there.
How could it be that an it that solicited such positive opinions turned out
to be such a flop? How did Tom’s 70%-will-buy projection turn into
0.0002%-actually-bought? Well, that’s what happens when you make
your decisions based on what you “learn” in Thoughtland.
In this case, Tom’s Thoughtland-based analysis gave us a false positive.
While dwelling in Thoughtland Tom was misled into thinking that his it
was the right it. Thinking he had the right it, Tom quit his job, and spent
three months developing the full-blown app – three versions of it. Tom
didn’t just skip the pretotyping phase, he even skipped the prototyping
stage. He went straight from idea to what I call a productype.
Productyping is the evil twin of pretotyping. If pretotyping can be
summarized as, “make sure you are building the right it before you build
it right,” productyping can be summarize as, “build it right even if you
are not sure you are building the right it.”
What was Tom thinking? He is a smart guy. Why did he invest even
one penny in creating more than one version of HoneyTextBot? Why
did he even bother to internationalize it and provide support for multiple
What happened is that, buoyed by such positive opinions, Tom ignored
the Law of Failure. He assumed success and decided to go all out, all at
This happens a lot. When our own infatuation with our it is combined
with false positives from Thoughtland it’s hard to resist going for it.
Besides doesn’t “going for it!” sound good? Doesn’t it feel good to say it
and do it? Isn’t that The American Way? Yes. Yes. Yes. It feels good –
make that GREAT – at first.
Overoptimistic individuals, by the way, are not the only ones who fall
for this trap. Seasoned professionals at major companies fall for it just
as often; they go from Thoughtland to productype in one fell swoop.
Productyping is the way most new products are developed.
Productyping is the reason why most failures are slow, painful and
Get It The Hell Out Of Thoughtland ASAP
All its – wrong its and right its – are born in Thoughtland. But, as we’ve
seen, spending too much time in Thoughtland will easily mislead us to
prematurely abandon potentially good ideas or overcommit and over-
invest in potentially bad ideas. In other words:
• Do nothing with it
• Go for it (productyping)
As we know, chances are that our it is not the right it, but the place to
make that determination is not in Thoughtland but in the real world
where, instead of useless opinions, we can collect actual usage and mar-
We must not let our it fester in Thoughtland, we have to get it out of
there as soon and as cheaply as possible. And that’s where pretotyping –
the third and best way to deal with the beast of failure – comes in:
• Give it a try (pretotyping)
Enough preambles, justification, explanation and definitions. Time to
get to the meat of this book – the actual creation and testing of preto-
First, I’ll introduce you to some basic types of pretotyping, then we’ll
look at ways to test them and finally I’ll combine everything we’ve
learned into a few complete examples.
A Hodgepodge of Pretotyping Techniques
Some day, if this book turns out to be the right it, I will invest time and
effort to create an extensive, well-structured, formal and official-looking
taxonomy of pretotyping techniques. At that time, I will give each tech-
nique a fancy name, describe ideal usage scenarios and provide plenty of
examples. But since this version of the book is still a pretotype, what
you’ll be getting is more of a hodgepodge of just a few different ways
you can pretotype your ideas with a rough explanation of when and how
to use them.
Here’s a quick summary of the techniques we’ll talk about:
The Mechanical Turk – Replace complex and expensive comput-
ers or machines with human beings.
The Pinocchio – Build a non-functional, “lifeless”, version of the
The Minimum Viable Product (or Stripped Tease) – Create a
functional version of it, but stripped down to its most basic func-
The Provincial – Before launching world-wide, run a test on a
very small sample.
The Fake Door – Create a fake “entry” for a product that doesn’t
yet exist in any form.
The Pretend-to-Own – Before investing in buying whatever you
need for your it, rent or borrow it first.
The Re-label – Put a different label on an existing product that
looks like the product you want to create.
Feel free to use, abuse, misuse or confuse any of these techniques.
Combine, refine, re-define and add to them to your heart’s content. If
you come up with an interesting pretotyping technique or suggestion let
me know about it (firstname.lastname@example.org); describe it and suggest a name
for it, and I might just include it in future versions of the book or feature
it on my blog (pretotyping.blogspot.com).
Now a few more words on each technique.
The Mechanical Turk
This pretotyping technique borrows its name from the famous Mechani-
cal Turk chess-playing “machine” that was touring the world in late 18th
century. People were led to believe that the “Turk” was a mechanical
contraption (an automaton) programmed to play chess. In reality, inside
the box there was a small, but talented, chess player making the moves
by manipulating the mannequin.
A Mechanical Turk pretotype is ideal for situations where you can re-
place costly, complex or yet-to-be-developed technology with a hidden
human being performing the functions of that technology.
The IBM speech-to-text experiment is a perfect example of this tech-
nique: Developing even a high-quality speech-to-text engine would have
taken years and a huge investment, but a human typist, hidden in another
room the same way the chess player was hidden inside the Mechanical
Turk contraption, easily simulated that complex functionality.
This pretotyping technique was Inspired by Jeff Hawkins’ wood block
Palm Pilot pretotype and has been named after the wood puppet who, af-
ter being visited by the Blue Fairy, becomes a real boy.
A Pinocchio pretotype is best suited for situation where things like size,
shape, weight, portability, etc., are important and where one’s imagina-
tion can be used to fill in the blanks – much the same way Hawkins’ pre-
tended that his wood block had the functionality required to schedule
appointments, store phone numbers and keep notes.
The Minimum Viable Product (or The Stripped Tease)
The term Minimum Viable Product (MVP) was introduced and popular-
ized by Eric Ries, the creator of The Lean Startup movement and one of
my personal heroes.
As the name suggests, this technique involves creating a working preto-
type – an actual product – but with features and functionality stripped
down to the bare minimum in order to: “… collect the maximum amount
of validated learning about customers with the least effort."
Since they involve some actual, if basic, functionality, MVPs typically
require more work than Mechanical Turk or Pinocchio pretotypes. But
an MVP can be developed much more quickly because it dispenses with
all non-critical features. An MVP for an online family diary application,
for example, should only support text entries (and perhaps uploading of
pictures), but it should not bother to provide support for different text
fonts, uploading of videos or different types sharing. Such features may
be nice, and even required, for the success of the final product but
should only be added once initial testing indicates that the online family
diary is the right it.
Note: As I mentioned before, I learned about Lean Startups and MVPs a few
months after I’d been talking about pretotypes and building them. During a
workshop, I built a “Stripped Tease” (my name at the time) pretotype for a mo-
bile app and someone said to me: “Hey, how is this different from Eric Ries’
MVP concept?” I did not have a good answer at the time. But after learning
more about MVPs and Eric Ries work, it’s clear that MVPs and pretotypes (as
well as the overall Lean Startup methodology) are all focused on helping crea-
tors, innovators and entrepreneurs avoid the same basic mistake: investing a lot
of time and money to develop products for which there is no market – or not
enough of a market to justify the investment.
If you are interested and intrigued by pretotyping and this book, then you must
buy, read and follow Eric Ries’ book The Lean Startup. It’s a proper book and
something everyone should read, whether they work in a startup or a Fortune
In many cases, the major costs associated with a product are not in de-
veloping the basic functionality, but in scaling the product to support and
make it useful for a large number of users. A Provincial pretotype pro-
vides the core features of the intended final product, but limits its scope
(and scale) to support a small subset of the ultimate target market. As
always, this is best explained with an example.
Let’s assume that Sandra has an idea for a mobile application that helps
people find restaurants that serve only organic food. Let’s call Sandra’s
it the Organic Eater Helper.
One of the most expensive and time consuming aspects of this app
would be the creation and maintenance of a national database of restau-
rants that meet the requirements of serving only organic foods. There
may be thousands of such restaurants across the country, and to include
them all, and write the code to automatically keep the list up to date,
Sandra would have to do a lot of work – unnecessary and wasted work if
it turns out that the Organic Eater Helper app is not the right it.
A Provincial pretotype would be developed as follows: Sandra should
start by focusing on a particular city or county – ideally this is where she
already lives. Since there will probably be only a few organic restau-
rants in the area she selected, the development of the application is
greatly simplified. Sandra can hardwire the names and location of the
restaurant directly in the app instead of having to write code to poll a
central database with thousands of restaurants and only return the ones
closest to the user’s location.
In addition to simplifying and accelerating the development of the preto-
type app, the provincial approach will also simplify and accelerate San-
dra’s marketing and testing effort. Instead of advertising the app nation-
ally, she can focus on a smaller region and save a lot of money and still
learn whether or not her app might be the right it.
The Fake Door Pretotype
The name of this technique comes from a presentation by Jess Lee, co-
founder and VP of Products for Polyvore. Great name. Thank you Jess!
With a Fake Door pretotype, the only requirement is to create an “entry”
point for a potential product (or new feature). The product (or feature)
does not have to exist at all. In Jess’s words: “In a web product, what
this means is that you pretend that a feature exists and you see if any-
body clicks on it.”
Fake Door pretotypes are useful for determining the level of interest for
On the Internet, a Fake Door can be implemented as a link, a button on a
web page, or a web ad for your it.
Let’s assume that Sandy is thinking about writing a book on squirrel
watching (a worrisome variation on the already worrisome and mysteri-
ously popular hobby of bird-watching.) Before she invests months of
precious time away from her actual squirrel watching pursuit to write
The Complete Squirrel Watcher, Sandy can use a Fake Door pretotype to
determine the level of interest in such a tome by creating a web ad –
something like this:
The Complete Squirrel Watcher.
The only book for serious squirrelers.
Only $9.98. Click here for more information.
She can then pay for Google AdWords to serve her ad on squirrel-related
websites or whenever people search online for “squirrel watching.”
We’ll elaborate more on this particular example in the Putting It All To-
gether chapter – I am sure that you, and the dozens of squirrel watchers
out there can’t wait.
Some its may require a major upfront investments, in such cases, it’s
critical that you pretotype the idea by borrowing or renting those expen-
A new business that requires a physical store, for example, should not
commit to a 5-year lease until they are sure that the idea is viable. In-
stead, they could try to get a 3-month deal on some un-leased space or –
even better – arrange to squeeze their display inside another store that
may attract the same type of buyers.
The idea for a new green car rental company that only rents electric cars
should be tested by either renting or borrowing a few electric cars for a
few weeks – not buying a fleet of them upfront.
You get the idea. Be a cheapskate until you know you have the right it.
Unless you are a borderline psychopath, some of these techniques may
bother you from an ethical standpoint. Is it really right to create a “fake
door” for example, just to see if people click on it?
I thought about this quite a bit and came to the following conclusion:
Wrong its are responsible for huge amount of waste. They waste the
time of the smart people who develop them as well as the money and
natural resources that should have been used to build something better
and more useful. Time, money and resources invested in wrong its are
time, money and resources stolen from right its.
Think of all the products you’ve bought and used only once or twice be-
fore throwing them away and regretting the purchase. Think of all the
unsold products that end up in landfills.
Pretotyping can save you, and your potential customers, from wasting a
lot of time and money on wrong its.
Use your judgment and sense of ethics when developing and testing pre-
totypes and you should sleep well at night.
Pretotypes are created for one reason, and one reason only – to help us
determine the level of interest and people’s reaction to our it. The data
we collect with the pretotypes will help us determine if our idea is the
The only effective way to know if an it is the right it is to test it. Not in
Thoughtland, parading an abstract idea and collecting subjective opin-
ions, but in the real-world with a concrete pretotype used to collect data
from actual users.
Data Beats Opinions
At Google we have a couple of important core beliefs: “data beats opin-
ions” and “say it with numbers.”
But what kind of data should we collect with our pretotypes and what
numbers should we “say it” with?
It’s impossible to come up with a fixed set of metrics that will apply
equally well to all its. The success of a book, for example, is typically
measured by how many copies it has sold, and a movie by its box office
receipts. The success of a web-based service such Google GMail, on the
other hand, is best measured not by how many people signup a GMail
account, but by how many people use their account regularly (e.g., 7-
day active users.)
While there isn’t a universally applicable set of success metrics, there
are some common guidelines that can, with some modifications, be ap-
plied to most its.
Since this is a pretotype version of the book (see section on MVP), I will
only introduce two basic but important and useful metrics: Initial Level
of Interest and Ongoing Level of Interest.
Initial Level of Interest (ILI)
The first metric you should try to collect on any it is what I call the Ini-
tial Level of Interest or ILI.
The ILI metric is a simple ratio:
ILI = number of actions taken / number of opportunities for action
number of opportunities for action offered represents the number
of people who have been offered an opportunity to take some posi-
tive action associated with your pretotype,
number of action taken represents the number of people who have
actually taken you up on that opportunity.
As always, an example should help to make things clear.
Adam is a nudist and occasional skydiver and he is so passionate about
his two “hobbies” that he’s thinking of quitting his job as an accountant
(especially since they won’t let him work in the buff), buying a plane
and starting the world’s first nude sky-diving business: Birthsuit Skydiv-
Before Adam resigns from his job and buys that Cessna, it would be a
terrifically good idea (to put it mildly) for him to see what the level of
interest in his idea is. Is nude skydiving the right it? We know that there
are many nudists and many skydivers, but how many nudists would like
to skydive and how many skydivers would like to jump from a plane
with nothing but a parachute? Here’s what Adam should do to deter-
mine the level of interest.
There are online forums for both nudists and skydivers, and let’s assume
that Adam is already a member of at least a couple of them.
Adam could write the following post in the local nudist forum:
Fellow nudist, I am renting a charter flight for a nude skydive. The cost is
$100 per jump. No skydiving experience necessary, and I promise you won’t
land in a field of cactii. The first jump will be a month from now, Saturday,
May 31st in Santa Barbara. To sign up please send me an email with the
names and number of nudists in your party, and I’ll respond with the details.
Space is limited, so first-come, first-served will apply.
A cool thing about online forums is that most of them show you how
many people have read each post. This gives Adam the first number he
needs (i.e., the number of people who have read his post and had an op-
portunity to act by sending Adam an email telling him that they were in-
Let’s assume that a week after Adam posted his message, he sees that
1,490 people read his post (this is the opportunity for action offered
number) and that he has received 2 emails replies saying that they
wanted to sign up (this is the number of actions taken).
The ILI in this case and for this group would be: 2 / 1490 = 0.0013 or
Not very encouraging, but also not too surprising since most people (in-
cluding nudists) are naturally reluctant to jump from a perfectly good
plane. At this point Adam can send the two responders a message saying
that he’s sorry, but due to a lack of interest the nude parachuting event
has been cancelled.
Before abandoning his idea, however, Adam should post a similar offer
on the local skydiving forum. Something like this:
Fellow skydivers, aren’t you bored of the same old, same old, types of
jumps? To make things interesting, I am renting a charter flight for a nude
skydive. The cost is $100 per jump. I promise you won’t land in a field of
cactii, but on a nude beach – imagine the surprise! The first jump will be a
month from now, Saturday, May 31st in Santa Barbara. If you want to sign
up please send me an email with the number and names of people in your
party, and I’ll respond with the details. Space is limited so first-come, first-
served will apply.
Let’s assume that after a week, 898 skydivers had read his post and 112
replied that they wanted to signup.
The ILI in this case would be: 112/898 = 12.5% – much better. Now we
With this simple Fake Door pretotype and ILI metric, and less than an
hour worth of “work”, our nudist skydiving friend Adam has already
collected some very valuable data:
Skydivers are a much better target market (by a factor of about 100) for
his idea than nudists.
The ILI for skydivers is pretty high, over 10% with 10,000s of skydivers
in the US this number is good enough to pursue the idea further.
A percentage of the skydivers who replied said that they were very inter-
ested and were ready and eager to sign up. This is a very strong right it
ILI data is very powerful and easy to interpret and act upon when used
for comparisons with similar ILI. In Adam’s case, ILI data unambigu-
ously indicates that skydivers are a much better target market than nud-
ists. It’s much harder to know, however, if a given ILI is good enough to
proceed. For some its, an ILI of 12.5% might be considered great, for
others it might not. While it’s important and easy to collect data and
calculate the ILI, interpreting it will usually require some judgment and
Things are looking good for the Birthsuit Skydiving idea but, as we’ll see
ILI is just an early indicator of right it potential. Let’s investigate what
Adam should pretotype and measure next.
Note: I have a sneaky suspicion that there might be FAA regulations against
nude skydiving. Since this book is a pretotype, I did not investigate this mat-
ter at length. And, just to be sure, I am neither endorsing nor suggesting
that nude skydiving is a good idea – so don’t try this it at home. But if you
do, don’t blame me for giving you the idea or send me photos of you doing it.
Ongoing Level of Interest (OLI)
For some its, where success does not necessarily depend on repeat busi-
ness (e.g., a book or an arcade-style game app), a good result based on
an Initial Level of Interest (ILI) may be enough to proceed to the next
step. But there are many its where success does depend on repeat pur-
chases, return visits, or ongoing usage by the same group of people who
were initially interested in it. This is particularly important if the run-
ning of the business requires the upfront purchasing of some expensive
equipment or committing to some significant recurring costs.
Unlike the ILI, the Ongoing Level of Interest (OLI) is best represented
by a time-based graph (or table) rather than by a single number. Each
point/entry in the graph/table represents the level of interest at a particu-
lar date. What you should be looking for in the OLI graph/table is a
trend. Does interest fade to zero after a while? Does it drop a bit but
then steadies at an acceptable rate? Does it go up? In the first case you
probably have a wrong it, the second case could go either way and may
deserves a bit more study, and the third case is a promising indication
that you just might have a right it.
As always, this is much easier to explain with an example. Let’s pick up
where we left off with Adam and his nude skydiving business.
In the case of Birthsuit Skydiving, Adam would be foolish to quit his job
and buy that Cessna airplane just based solely on his ILI numbers. Even
if more than 10% of all skydivers were interested in a trying a nude
jump, if none of them come back for more this would be a short-lived
Before making any major decisions (like quitting his job) or investments
(like buying an airplane), Adam would be wise to check the Ongoing
Level of Interest (OLI) in his idea.
Fake Door pretotypes are great to test the ILI, but you need something
more concrete and substantial to test the OLI. Most people will not con-
tinue to open Fake Doors. The Pretend-to-Own pretotype would fit the
bill quite well in this case.
Instead of buying a plane, Adam should just rent one on an as-needed
basis. Renting a plane by the day may cost too much to make it a viable
long-term business option for Birthsuit Skydiving, he may even be losing
a few hundred dollars each time. But until Adam is convinced that his
nude skydiving idea is going to fly, it’s better for him to lose a few hun-
dred dollars testing it, rather than drop tens of thousand of dollars up-
front hoping that he has the right it. Remember The Law of Failure,
even with a positive ILI result, the odds are still against Adam’s it.
Let’s assume that Adam follows the pretotyping protocol, advertises the
flights on his local skydiving forum every week and, over a period of
two months, he runs 8 flights: one flight every Saturday.
Here’s is OLI data after the two months:
Flight # Signups Revenue Cost Proﬁt/(Loss)
1 21 $210 $250 -$40
2 20 $250 $250 $0
3 28 $280 $250 $30
4 17 $170 $250 -$80
5 7 $70 $250 -$180
6 3 $30 $250 -$220
7 0 $0 $0 $0
8 0 $0 $0 $0
Total 101 $1,100 $1,500 -$490
Sorry Adam! Things were looking good for a while – you even man-
aged to make a small profit on your third flight – but I’m afraid that this
nude skydiving thing may not be the right it.
A high ILI is great, but if the success of your it depends on ongoing us-
age, you should test the ongoing level of interest if there are significant
investments associated with your it. In Adam’s case, pretotyping is sug-
gesting that Birthsuit Skydiving may work well as a fun hobby or side
activity, but at this time it would be unwise for him to quit his job, buy a
plane and try to make a living with it. Pretotyping saved the day – and
saved us from the risk of having a nude parachutist landing in our back-
Put It All
Finally all the pieces are in place, so we can go through a couple of ex-
amples of creating and testing pretotypes, and making decisions based
on them. As you go through the examples, don’t be surprised if you can
come up with different ways to pretotype and test these ideas, there is no
single best approach. I’d be surprised if you can’t think of other ways of
approaching the same pretotyping challenges.
Example 1: The Complete Squirrel Watcher
Let’s build on our example from the Fake Door pretotype. As you might
recall, Sandy is thinking about writing a book on squirrel watching.
Since she would have to invest months of precious time away from her
actual squirrel watching pursuit to write The Complete Squirrel Watcher,
it would be a good idea for her to pretotype the book.
In Sandy’s case, since the success of a book is primarily determined by
how many people buy it (i.e., it does not really depend on repeat pur-
chases) all we need is a pretotype to find out the Initial Level of Interest
(ILI). Fake Door pretotypes are ideal for this. And here’s how Sandy
could go about it:
For $10, she can buy TheCompleteSquirrelWatcher.com domain and
create a landing page that says:
Fellow squirrel enthusiasts,
Thank you for your interest in “The Complete Squirrel Watcher”.
I am hard at work on the book, but it’s not quite ready for publication.
To reserve a copy at the special pre-order price of $9.98 send an email to:
and I’ll let you know as soon as the book is available.
The price will be $9.98
In the meantime, happy squirrel watching and don’t forget your rabies shots!
Sandy (Squirrelgirl) Watson
Sandy can then craft a web ad, for example:
Do you like stalking squirrels?
The official book for serious squirrel watchers
by Sandy Watson. Only $9.98
For a few dollars, she can place the ad on websites dedicated to squirrels
or have it show up when people use a search engine to search for any-
thing related to squirrels. When people click on her ad, they are redi-
rected to her website.
This Fake Door pretotype would cost less than $50 and take just a cou-
ple of hours of work requiring minimal technical skills.
Once this pretotype is in place, Sandy can let the ad run for a month or
so, after which she can analyze the data provided from the online ad
Assume this is the data generated by her pretotype:
People who have seen the ad: 23,402
People who have clicked on the ad: 634
People who sent an email saying to buy the book: 230
There are a couple of interesting ILI ratios here.
The first is an indication of how many people who go to squirrel pages
or search for squirrel are interested enough to click on an ad for a book
on squirrel watching. This first ILI can be calculated as follows:
ILI 1 = number of clicks on ad / number of ad impressions (i.e.,
how many people have seen the ad)
In this case, ILI1 = 634 / 23,402 = 2.7%
This is not great, but not too bad either.
The second ILI ratio gives her the percentage of people who, after click-
ing on the ad, are interested enough in the book to send an email to
ILI2 = number of emails / number page visits to the landing page
In this case, ILI2 = 36% (230 / 634)
This is very encouraging, a whopping 36% of the people who visit
Sandy’s TheCompleteSquirrelWatcher.com web page send her an email
to reserve a copy of the book. Of course not all of them will follow
through, but this is still a very good number.
Now comes the difficult decision. Should Sandy go ahead and write her
book based on this data?
That depends a lot on her expectations for the book. The data indicates
that the book is unlikely to land a spot on The New York Times’s best-
seller list – not enough people seem that interested in squirrels. But that
was never Sandy’s expectation. For her, becoming an authority on the
subject and selling a few hundred copies of her self-published book each
year – enough to pay for her squirrel watching gear expeditions – would
be good enough. In that case, the data from her pretotype suggests that
The Complete Squirrel Watcher will probably the right it for enough
people to make Sandy happy.
Example 2: Bob’s Rate This Plate App
For this example, let’s assume that Bob is a nutritionist who wants to
create a mobile app that analyzes a photo of a meal and returns a nutri-
tional analysis and some kind of score from “A: Healthy and nutritious”
to “F: Junk food”? Let’s call this it the Bob’s Rate This Plate app.
Bob talks to friends and many other people about this app, and most of
them tell him that it’s a great idea and that they would definitely use it.
Fortunately, Bob has heard about Thoughtland and knows how mislead-
ing opinions can be. He does not know for sure how many people would
use such an app or be willing to pay for it. Would users even remember
to stop and take a photo of the food before they start digging in? Would
they use it a few times – just for fun – and then never again?
Bob also realizes that developing an actual working software system to
automatically analyze a meal based on a picture of it would definitely
take a lot of work and money – and it may never get to the point where
it’s good or accurate enough to be useful (a problem similar to the one
one faced by the IBM team with their speech-to-text idea.)
There are a lot of open questions that have to be answered and expensive
technology to develop. This it definitely calls for some pretotyping.
First step: Fake Door and Pinocchio pretotypes
By now, you should not be surprised that, as a first step, I would recom-
mend for Bob to build some kind of Fake Door pretotype to measure ILI
(see previous example for how to do that.)
Let’s assume that the ILI data is encouraging. However, Bob’s vision
for, and definition of, success for this app requires not only initial inter-
est, but ongoing usage (i.e., a promising Ongoing Level of Interest). If
what the app requires is cumbersome or awkward to do people may not
stick with it. Heck, would Bob himself stick with it? Would he remem-
ber to take photos of his food before he starts eating it? Would he be
embarrassed to do it in front of people, especially in a restaurant?
Would he only take photos of his healthy meals and conveniently forget
to record that banana split?
If we don’t believe and don’t use our it ourselves, how can we sincerely
convince, or expect, other people to do that? To answer this question,
Bob should follow Jeff Hawkins’ Palm Pilot pretotype example and de-
velop a Pinocchio pretotype to test the idea on himself. Since Bob al-
ready has a smartphone with a camera, he does not have to go out and
build a wood-block like Hawkins did. He can simply pretend that his
phone’s camera app is the app he wants to build and fill the blanks with
If Bob discovers that, after a few of days of using his Pinocchio preto-
type, his initial enthusiasm for the idea starts to wane and he takes fewer
and fewer photos, then he might have a problem. Of course, he could try
to explain the failure away, “this app is not for me, it’s for my clients, I
already know what I should eat, I don’t need it.” He might be right in
this particular case, but he should still be concerned about it. The “I
won’t use it, but others will” argument is a giant red flag with “wrong it”
written all over it: not something to be dismissed lightly.
However, to continue our example, let’s assume that Bob quickly be-
comes so used to taking photos of his food before eating that it becomes
a habit for him and he does it consistently and automatically. Not only
that, but when he does it in front of other people, they ask him about it
and say that they would love an app like that for themselves. He also
starts to post his photos on an online album so he can keep track of eve-
rything he has eaten and mails them to a nutritionist colleague so she can
give her feedback on his diet. This is a good sign. Bob now knows that
he’d use the app himself on an ongoing basis and he found it useful
enough to “implement” a couple of new “features” (i.e., posting the pho-
tos on an online album and sending them to a colleague.)
His first two pretotypes tested well, the ILI was good and his personal
OLI was also very good, now it’s time to see if enough other people
would use the app on an ongoing basis.
Bob needs to get an idea of what the OLI is and the Fake Door pretotype
won’t work for that, nor would a simple Pinocchio pretotype (the Pinoc-
chio requires a lot of pretending and imagination about the intended fea-
tures and functionality, they are great for convincing the creator of the
idea who can fill those blanks, but not so good for collecting data from
other users.) What Bob needs is a simple, but functional, pretotype. Un-
fortunately, Bob is a nutritionist and not a programmer. Before investing
in hiring a programmer, is there a faster and cheaper pretotype he can
create that will still give him some OLI data? You betcha!
Super-cheapo, low-tech, Mechanical Turk pretotype
Since Bob is a nutritionist and he has over 500 clients, he can ask a few
of his clients (say, 50 of them, about 10%) if they’d be interested in par-
ticipating in a one month experiment. All they have to do is to take a
photo of each of their meals before they start to eat and email that photo
to him. In return, at the end of each day, Bob will send them an email
with an nutritional grade along with some comments and suggestions on
how to improve their diet. Nothing too fancy or time consuming; some-
thing along these lines:
Thank you for helping me test Rate This Plate:
Here are your ratings for today:
Breakfast: F (eggs and bacon, c’mon you know better than that.)
Lunch: B (salad good, blue cheese dressing bad)
Dinner: A- (chicken and veggies looked healthy, but you get a minus for that
Please try to eat some fruits and veggies for the next few meals.
Let’s say that 30 (out of 50) of Bob’s clients agree to do the experiment
(ILI = 30/50, or 60%). At first, Bob is disappointed, even though this ILI
is high, he was hoping that all of his clients would agree to join – or at
least 80-90%. After talking with the clients who declined the opportu-
nity to join experiment, he learns a few things that he hadn’t thought of.
Many of his clients, for example, don’t have a mobile phone with a data
plan, so they can’t email photos to him. And a few clients felt very un-
comfortable sharing actual photos of their meals with him – or anyone
other person – but they’d be OK if their meals were analyzed by a com-
puter. Good things to know and keep in mind as he progresses.
When the experiment starts, Bob sends his 30 volunteer clients instruc-
tions on what to do (i.e. Take a photo of your everything you eat and
mail it to email@example.com) and, as the emails with
photo start coming in (~80/day) he rates the meals and sends his emails
with their grade and a nutritional analysis. A lot of work, but since he’s
not a programmer, it was faster and cheaper for him to do it this way.
After a month of running the experiment, Bob has a pretty good OLI ta-
(out of 30)
1 28 234
2 24 198
3 22 168
4 22 172
As it always happens, some people who said they would participate
didn’t send even one photo and, as time passed, some other volunteers
dropped off. By the end of then month, however, he still had over 2/3rds
of volunteers actively submitting photos. This is encouraging.
Even more encouraging, a lot of the users are sending him requests for
new features and functionality: “Hey Bob, can you send me my average
GPA?” “If I forget to take a photo can I just send you a description of
my meal?” “Can you send me a menu for each day that will guarantee
me an ‘A’?”
Some, on the other hand, complain: “Bob, I don’t have good phone re-
ception in our cafeteria, it sucks that I have to go outdoor to email you
the photo – while my food gets cold.”
When you don’t hear from users, chances are that they are either not us-
ing your product, or don’t care enough to send feedback on how to en-
hance it or improve it. Getting feedback, good or bad, is a great sign.
They care enough to suggest or complain.
Things are looking good for Bob: strong OLI and lots of user feedback.
Bob’s Rate This Plate app might just be a right it.
There is still a little issue of revenue and profitability. Bob wants to
make a business out of it. Would the people who have been using the
app for free be willing to pay for the service? How much would they be
willing to pay: $10/month – perhaps even $30/month? By now, I am
sure you know how he would answer that question. He still has 450 cli-
ents to experiment on. He can ask 100 if they would sign up for the
service at $10/month and another 100 to if they’d sign up at $30/month
and them measure the ILI and OLI for both.
Only a couple of clients signed up for the $30/month service but, sur-
prisingly, 42 of his clients signed up for the $10/month service – more
than he could handle manually. It was time to invest in automation. Un-
fortunately, he realized that the technology for automatically analyzing a
meal based on just a photo was at least a few years away. But he found
out that he could train college students who, for $15/hr, would do almost
as good a job as he did. He ran the numbers and realized that he could
make a nice profit of $4/patient each month.
After a few months running the service for his patients and making a
profit, Bob decides to go big – his it was the right it. He hires a devel-
oper to create a custom app (instead of the clumsy email-based preto-
type) and trains more students to handle the load.
Bob’s Rate This Plate app was the right it, and because of it there are
quite a few more healthy people out there.
Don’t you love a happy ending?
Now Go Make It
Even though we’ve gone through a lot of material very quickly, and I’ve
subjected you to some rather unusual examples, I hope that I was suc-
cessful in answering the following questions:
• What is pretotyping?
• Why is it important?
• What are some of the pretotyping techniques that you can use?
• What data to collect and what metrics to use with your preto-
Now it’s your turn!
I am sure you have quite a few its you want to try. Pretotyping will help
you in two ways:
• If your it has been held captive in Thoughtland for a while, preto-
typing should make it much easier for you to get started. Ignore
the naysayers and get off your butt. Pretotype it and see what
• If you are getting ready to take a big risk or make a big invest-
ment in your it, pretotyping will help you get started more
quickly. It will also provide you with valuable data that will ei-
ther give you more confidence that your it is the right it, or it will
help you realize that you should make some changes to your it, or
pursue a completely different it.
In all cases, feel free to keep in touch with me (firstname.lastname@example.org) to
let me know how it’s going for you and if I can help you in any way if
you decide to give pretotyping a try.
May you find your right it, and if you see the beast of failure, please tell
it Alberto said “Hello!”
Is this book the right it?
Over the past two years, I have given dozens of presentations and dem-
onstrations on pretotyping to thousands of people. I’ve practiced preto-
typing for my job and I have started to help other people and organiza-
tions successfully pretotype their own ideas.
The overwhelmingly positive reaction to pretotyping has surprised me.
People love the idea, they understand how and why it works, they want
to know more about it, and – according to what many have told me – it
has dramatically changed the way they think about pursuing and invest-
ing in new ideas and innovations. I have strong evidence that, when pre-
sented and explained live (usually with plenty of examples and hands-on
demonstrations), pretotyping is the right it.
Many people, intrigued by the concept of pretotyping and wanting to
learn more about it, have asked me to write a book on the subject. Writ-
ing a book, however, is no easy task (at least not for me) and one that re-
quires a significant investment of time, energy and concentration. On
top of that, most published books fail in the market – they are wrong its.
That’s why I decided to treat the idea this book as an it – an idea to test –
and to pretotype this it (i.e., pretotype the idea of a book on pretotyping
– a pretobook.)
Instead of spending months writing, editing, perfecting and polishing
hundreds of pages (and sacrificing trees for a hefty book that few might
read), I gave myself a few days to create a written version of my preto-
typing presentations and workshops. The result is the slim book (or
eBook) you are reading now.
Hopefully, the core idea, message and practice of pretotyping will be
able to shine through my somewhat clumsy writing, poor organization of
the material and painfully obvious lack of professional editing. If a book
on pretotyping – at least one written by me – is the right it, even this
version with a few rough edges, should achieve some level of success
and popularity. Of course, I’d love to see it succeed wildly, but I know
the odds are against it.
The Pretotyping Manifesto
Make sure you are building the right it
before you build it right
innovators beat ideas
pretotypes beat productypes
data beats opinions
now beats later
doing beats talking
simple beats complex
commitment beats committees
Alberto Savoia on Google+