6. Who is this guy?
- Forbes 30 under 30
entrepreneur
- Former operations
research PhD
- National Science
Foundation
fellow at UC Berkeley
- Raised $18M from
investors:
YC, A16Z, 500, Sierra, etc.
7. What we built
- Sales technology startup
- Social enterprise
- Google, Box, hundreds of
customers (mid-market &
enterprise)
- ACV $50K, millions in
revenue
- 50 local employees,
500 global, 40 countries
10. !!!!
• How can we help groups with minimal (technology)
literacy participate in the crowdsourcing economy?
Original Problem
11. Original Idea (2010)
Let’s build a company that takes the
easiest jobs from Amazon Mechanical
Turk and makes them available to the
poorest people in the
world through their cell phones!
13. User interviews Personas! Scenarios!
The Process!
Lo-fi prototyping User testing! High-fidelity
prototypes!
14. Every single part of our original
idea was wrong!
• There were not many image tagging
jobs on Mechanical Turk
• The jobs were too hard for workers
living in a slum in India
• The jobs were paid so low that we
couldn’t make any money from them
15. Pivot (2010)!
• Instead of scraping Mechanical Turk,
let’s charge customers directly to do
data entry for them
18. This idea was still wrong!
• Yes, you can get very poor people to do
data entry on a mobile phone
• Why would you, though? BPO
outsourcing is far better and less
complicated.
• We built a system, but only found one
paying customer for $600.
19. Pivot (Summer 2011)!
“Mechanical Turk is really
terrible – people hate to use it.”
Let’s build the world’s best
crowdsourcing platform – a
replacement for Mechanical
Turk!!
20. We built an awesome, sophisticated
replacement for Mechanical Turk,
starting with an API….!
22. Over a year, we added lots of
great features to the product,
and plenty of people were trying
them out.!
23. Over a year, we added lots of
great features to the product,
and web users seemed to be
trying them out.!
The problem: we had nearly zero
paying customers.!
24. It took us a year to realize:!
- Almost nobody uses Mechanical
Turk (<$5M market)
- Mechanical Turk customers are
cheap and don’t have much money,
not recurring customers
- The people who want to outsource
work aren’t software developers
using APIs – they’re businesspeople
25. The problem!
- We had spent over a year
building beautiful, highly
effective products and ignoring
the most important fact:
customers did not want to use
them!
38. This is an important idea.
“The #1 Company-
Killer is Lack of
a Market.”
-Ben Horowitz,
39. This is an important idea.
“The #1 Company-
Killer is Lack of
a Market.”
-Ben Horowitz, There’s just one
mistake that kills
startups – not making
something users
want.
-Paul Graham, YC
42. It’s simple:
customers are buying what
you’re selling.
How do you know you have
product-market fit?
You know where to
get more customers!
43. raise a little capital
experiments
customer development
spending money slowly to learn
as quickly as possible
raise a lot of capital
acquire customers fast
clear connection of spending
money to customer growth
spending money as fast as
possible to grow customers as
fast as possible
before
product market fit
after
product market fit
47. customer = someone who
pays you for the product
market = the big universe of
customers that you plan to serve
48. Example: LeadGenius (B2B)
Our customer is a head of sales running
a B2B sales team of 10 or more people.
Our market is Fortune 5000 companies
spending money on outbound sales.
49. Example: Shortcut
(ocean cargo marketplace)
Our customers are ship owners with
spare capacity and individual
manufacturers who want to transport
cargo.
Our market is the international short
sea shipping market – everyone who is
spending money on shipping freight.
50. Example: Carstomize
(machine learning for car
configuration)
Our customer is the head of
marketing at car manufacturers.
Our market is going to be the
ten largest automakers in the
world.
54. TAM: TOTAL ADDRESSABLE MARKET
How much money could you make
if absolutely every possible
customer in the world used this
product?
55. TAM: TOTAL ADDRESSABLE MARKET
How much money is being spent
on this problem globally today?
56. TAM: TOTAL ADDRESSABLE MARKET
How much money is being spent
on this problem globally today?
“global shipping is a $400B market”
57. TAM: TOTAL ADDRESSABLE MARKET
How much money is being spent
on this problem globally today?
“global shipping is a $400B market”
“new car sales is a $570B market”
59. EXAMPLE: DOLLAR SHAVE CLUB
Product:
Cheap mail-order
razors for men
TAM:
Shave & razor industry
60. EXAMPLE: DOLLAR SHAVE CLUB
Product:
Cheap mail-order
razors for men
TAM:
Shave & razor industry
$34B global market!
61. EXAMPLE: DOLLAR SHAVE CLUB
Product:
Cheap mail-order
razors for men
TAM:
Shave & razor industry
$34B global market!
includes: Fancy razors, store-bought razors,
mail-order razors
63. SAM: SERVICEABLE
ADDRESSABLE MARKET
How much of that market can we
serve with our specific product?
“online truck sales are a $50B market”
“short sea shipping is a $70B market”
64. EXAMPLE: DOLLAR SHAVE CLUB
TAM was the full
shaving industry.
SAM is the size of
the cheap/
disposable razor
category – $10B.
SAM is the smaller sector that you’ll go after.
65. SOM: SHARE OF MARKET
How much revenue will we
realistically be able to make in our
first few years?
66. SOM: SHARE OF MARKET
How much revenue will we
realistically be able to make in our
first few years?
Bottom up analysis:
value per customer X
number of customers we reach
67. EXAMPLE: DOLLAR SHAVE CLUB
There are 90M US men over 21 (research)
We’ll get 5M people to see our ads (strategy)
50% are men (average)
75% use disposable razors (research)
10% turn into customers after
seeing ads (experiments)
68. EXAMPLE: DOLLAR SHAVE CLUB
5M x 50% x 75% x 10%= 187,500 customers
187,500 customers x $60/year = $18M revenue
$18M revenue / SAM of $10B
= 0.2% SOM
69. Remember… these are just
estimates! You don’t need to be
perfect.
You’re trying to convince yourself
and prove to investors that you
have a viable business.
73. 6 million businesses in the US with >1 employee
100,000 midsize businesses in the US with >
$10M revenue
Fortune 500, Fortune 5000, Global 2000 with
multiple sub-organizations
Estimating Market Sizes
74. Facebook (B2C)
Ample consumer data
Market sizing information
Good for SOM
Also… you can use it
directly when launching!
Estimating Market Sizes
84. VCs like businesses that can
scale up to $100M per year
in revenue.
That means you should target a
total addressable market of
$1 billion or more.
How big should your market be?
93. Good initial markets:
- are underserved, feel the pain
- are narrowly-defined and tractable
by your team
- let you find hardcore customers
who love you!
94. Quotes
“Build something a small number of people
want a large amount.”
“When you have an idea for a startup, ask
yourself: who wants this right now? Who wants
this so much that they'll use it even when it's a
crappy version one made by a two-person
startup they've never heard of?”
- Paul Graham, YC
95. Things to remember:
- Your market will grow over time.
- Think big and impressive when
you’re setting your vision – aim for
the biggest possible overall market.
- Think small and doable when you're
setting your initial market.