3. Neil Patel
CrazyEgg, KissMetrics,
Hello Bar
Andrew Chen
Uber, Mohr Davidow
Brian Balfour
HubSpot
Sean Ellis
Qualaroo, Xobni, Dropbox
Dave Macclure
500 Startups
Alex Schultz
Facebook
Thanks Guys!
6. Discussion: Growth Hacking
1. Why is this topic in such high demand?
2. Do we all have the same definition?
3. What do you want to learn?
7. WIIFM? Want you to leave this with…
UNDERSTAND (1 hour)
• Understanding of macro trends
leading to it
• Understanding of the disciplines
involved
• Clear litmus test for what a Growth
Hack is
• Understand two main growth loops,
subtypes
• Understand channel vs platform
hijacks
• Have examples of loops and hijacks
• Understand k-factor, growth
multipliers, cycle time
DO (1 hour)
1. Market Fit
• Qualitative measurement
• Quantitative measurement
2. Growth Fit
• Compute Growth Fit Score
3. Find Core Loops
• Complete brainstorming exercises
towards discovering your main loops
4. Model Loops
• Model out your top loop
5. Growth Roadmap
• Start to fill out a growth roadmap
10. And there was still “Other”:
• Referrals
• Unknown origin
• Word of mouth
• Organic
?
11. Growth Harnesses The “Other”
Brand & Performance Marketing Growth Marketing
• Requires large budgets
• Costs are linear to sales
• Must “breakthrough” noise
Consumers share message
with other consumers
Company shouts message
to consumer
• Smaller, one-time investments
• Potentially exponential returns
• Referrals & content more persuasive
13. Growth Hacking is a Blend of Disciplines
• Experimentation
• Analytics
• Agility
• Coding
• API hacks
• Grey Hat
• Product Management
• Feature development
• Roadmap prioritization
• Gamification • Social Media
• Performance Marketing
• Mobile app chops
14. Is there a platform hijack?Is there a loop?
Identifying a Growth Hack: The Test
If you answer yes to either question, it’s a growth hack.
or
15. New user signs up
What’s a (Viral) Loop?
Some users
import address
book
Some users
send invites
Some friends
click through
S: Chen & Balfour, SVBR
26. LinkedIn with Contacts Uploader
LinkedIn (Few users in early days) Your Contacts List
Referral Loop
Invite new users
Connection request
existing users
Re-engagement Loop
Find Connections
Upload contacts
(Gmail, Outlook, etc.)
Accept connection
Suggest more
friends
Create account
27. New user signs up
What’s a (Content) Loop?
Some create
new content
Some of the content gets
indexed by Search Engines
Some searchers
see content and
click through
28. New user signs up
for LinkedIn
Content Loop for LinkedIn on Google
Creates Profile,
Indexed by
Google
Someone searches for
“John P. Appleseed”
Visit to John’s
profile
29. Content Loops
Examples:
• Social Profiles & Directories
– LinkedIn, Facebook, FindLaw
• Reviews, Curation
– Amazon Review
• Marketplaces
– Craigslist, Ebay, AirBNB
– CafePress, Thingaverse
– SoundCloud
• Information
– Slideshare, YouTube
– Quora, JustAnswers, Wikis
– Medium, Udemy
• Ad Sponsored Products
– Yahoo Websites, Wix, VistaPrint
Do your users generate
content?
Is that content indexed and
public?
and
If both of these are true, then
you have a content loop.
31. Channels vs Platform Hijacks
Traditional Channels
• Linear
• Obvious to marketer
• Stable
• No special tools
required
• Sets off prospect’s
skepticism radar
“Growth” Platform Hijacks
• Exploding in size (hopefully)
• Non-obvious at first
• Small window of
exploitation (grey hat)
• May require engineering
toolset
• Flies under the ”I’m being
sold to” radar
32. A user of a different platform can
interact with something you’ve put
on there.
What’s a Platform Hijack?
To take full
advantage they
sign up for
your offering.
They take actions within
your offering
Those actions
lead to more
chances for
others to
interact on the
other platform
S: Chen & Balfour, SVBR
33. Birth & Death of a Growth Channel
Chat bots?
Direct Mail
Overcrowded,
Noisy, Spammy,
Expensive
Email / Newsletters
Facebook Post
App Store
Vchat?
Early, Clever,
Has Breakout,
Inexpensive
34. AirBNB Posting to Craigslist
Fill out
listing
Cross-
listed on
Craigslist
AirBNB (Fewer eyeballs) Craigslist (Lots of eyeballs)
Posted on
AirBNB
Visit
listing
Rent
space
Content LoopA small subset will
decide to post their
own listing on
AirBNB, continuing
the loop
Links back to AirBNB
36. AirBNB Poaching off Craigslist
Fill out
listing
Filled out
listing on
Craigslist
AirBNB (Fewer listings) Craigslist (Lots of listing)
Posted on
AirBNB
Emailed CL user to
suggest posting
on AirBNB too
42. Growth Loops
Initial Acquisition
Growth Loops
New users
Conversion
process
Visits
Impressions
Linear + Loops
Cycle Time
Invites
Optimize…
1. Conversion Rate 2. # of Invites 3. Cycle Time
45. Growth Multiplier
When k > 1When k < 1 When k = 1
• Each loop produces
fewer users than the
one before.
• Declining system.
• Still very good to
have a loop at all!
• To figure out the
total # of users or $s
from one round of
(linear) inputs, use
the Growth
Multiplier formula
• Each loop produces
more users than the
one before.
• Would go to infinity,
except you
eventually saturate
the market and run
out of potential new
users. Or
competition creeks
in. Or the platforms
shut you off.
• Each loop produces
the same number as
the one before.
• Linear growth of
users with each loop
• Goes on forever (in
theory)
47. Growth Multiplier
If k < 1, Growth Multiplier = 1/(1-k)
Tells you how many users you will get from your initial set of users.
Sometimes it only takes 4 loops to cap out…
Sometimes it takes 20 loops to cap out…
48. Growth Multiplier
Gives you leverage on conversion improvements
It’s just a 10% increase in conv rate but a 67% increase in users!
Initially your conversion rate is 20%
But then you optimize it through testing and get 22% conversion rate.
8.3 / 5 = 1.67
49. More complete example
You will likely have more detailed steps and need to track and optimize they all.
But they easily map back the simplified model.
50. Example – LZ FB shout-out
This loop has a low k-factor, but the growth hack is cheap to implement and raises revenue 6%.
51. Cycle Time
Cycle Time is the 3rd variable to optimize in growth loops. This one is key to managing cash flows.
52. WIIFM? Want you to leave this with…
UNDERSTAND (1 hour)
• Understanding of macro trends
leading to it
• Understanding of the disciplines
involved
• Clear litmus test for what a Growth
Hack is
• Understand two main growth loops,
subtypes
• Understand channel vs platform
hijacks
• Have examples of loops and hijacks
• Understand k-factor, growth
multipliers, cycle time
DO (1 hour)
1. Market Fit
• Qualitative measurement
• Quantitative measurement
2. Growth Fit
• Compute Growth Fit Score
3. Find Core Loops
• Complete brainstorming exercises
towards discovering your main loops
4. Model Loops
• Model out your top loop
5. Growth Roadmap
• Start to fill out a growth roadmap
55. Product Market-Fit Assessment
How would you feel if you could no longer
use [product]?
o Very disappointed
o Somewhat disappointed
o Not disappointed (it isn’t really that useful)
o N/A – I no longer use [product]
S: Sean Ellis http://www.startup-marketing.com/using-survey-io/
At least 40%
need to answer
in the green.
40%
Sean Ellis
56. Product Market-Fit Assessment
S: Sean Ellis http://www.startup-marketing.com/using-survey-io/
40%
Where are you?
Proceed to Growth Fit Score
Yes, it’s over 40%
Check engine. Narrow in on any users who answered
in the green. Find point of delight and build off it. Or
fix blockers to delight. If still under 40%, pivot.
No, it’s under 40%
We haven’t collected this yet.
Stop. Plan out how you will measure this.
57. Plan for Measuring Market-Fit - LZ
1. What defines the “core” of your offering?
2. How can you identify people who have used the
core of your offering at least twice?
3. And within in the last 2 weeks?
58. Setup Market-Fit Survey Now
Activity: Go here now and setup your survey: http://www.survey.io/
How will you automate and motivate the survey invites?
59. “Retention is the single most
important thing for growth.”
- Alex Schultz
Alex Schultz
Facebook
60. Quantitative Check on Product-Fit Viability
There is a segment of the population
that finds this offering compelling to
use regularly.
There is no segment of users that
finds it compelling for ongoing use.
Hypothetical Scenario where new user growth is in steady state
S: Alex Schultz https://www.youtube.com/watch?v=n_yHZ_vKjno
61. Quantitative Check on Product-Fit Viability
Adding an increasing number of
users at the top of the funnel
doesn’t help if the product doesn’t
have fit.
Hypothetical Scenario where new user growth increases 2% per day
62. “We spent a bajillion bucks on the Tables ad. And
now we have ½ Million installs. But…installs isn’t
really the metric that matters. We need Daily Active
Users – people who come back to the platform.”
63. 2. Growth Fit Score
Personal or embarrassing topic Natural topic of conversation
Long Lead Time Rapid Iteration
In person interaction Collects lots of data
Incremental features Game changer
Consumption only Outputs user generated content
Hard to locate prospects Unwalled garden of prospects exists
Vitamin Painkiller
Solo usage Facilitates interaction
FasterSlower
64. Personal or embarrassing topic Natural topic of conversation
Long Lead Time Rapid Iteration
In person interaction Collects lots of data
Incremental features Game changer
Consumption only Outputs user generated content
Hard to locate prospects Unwalled garden of prospects exists
Vitamin Painkiller
Solo usage Facilitates interaction
66. Sharing
Q A Applications for Growth
What about your app
would make it a topic
of conversation?
What moment in the
experience can you
nudge user to share?
What human emotion
can you tap to
motivate sharing?
What added non-
monetary incentives
can you offer to propel
more sharing?
What sort of monetary
incentives could you
offer? How would you
model out the ROI?
67. Interactions
Q A Growth Implications
What person-to-person
interactions does your
app already facilitate?
What interactions could
your app facilitate?
What sort of value can
you create for those
persons on the non-user
side of the interaction?
How much activation
energy would that
require from the non-
user and how can you
leverage that?
68. Content
Q A Applications for Growth
What kind of content
does your app create?
What kind of content
could it create?
What kind of user
generated content may
be related to this same
problem?
69. Platform Hijacks
Q A Growth Implications
What desktop apps do your
prospects use often?
What website do your
prospects use often?
What mobile apps do your
prospects use often?
Where do your prospects
live & work? Where do they
play?
Who are the biggest
influencers for your
prospects?
Who are the biggest
intermediaries for your
prospects?
From where do your
prospects draw inspiration?
70. Platform Hijacks
Q A Growth Implications
Which platforms that you
identified provide some
level of programmatic
access?
Which platforms that you
identified can you build
your own programmatic
access into?
How can you flow your
content into their platform?
How can you flow their
users into your platform?
79. Growth Roadmap
0. Market-Fit
1. Acquisition
2. Activation
• Paid vs Trial vs Freemium etc.
• Tripwire sale vs Core sale
3. Referrals & Hijacks
• Find main loops and hijacks
• Raise k-factor
– Tests to increase i
– Tests to increase c
– Tests to shorten cycle time
4. Revenue
• Pricing, packaging
• X-sells, upsells
5. Retention
• Hooked framework
• Octalysis framework
Potential
impact from
growth model
Evaluate ideas based on:
Change of it
succeeding
Time & Effort
to test it
x
x
Perhaps most
important. Go big
or go home.
Hard to predict
Important, but don’t
only test the easy things
or you could miss out on
biggest opportunities
Tips on
Weighting
80. Growth Roadmap
Experiments to: Potential Value
from k-factor model
Likelihood to
Succeed
Time & Money to
test it
Score
Increase # of invites
Increase conversion
Decrease cycle time
Editor's Notes
There are some of the trailblazers and teachers to which I owe my education.
B
Traditional funnel starts with non-users to hopefully generate some users.
Growth funnel starts with a user, and that users generates more users
Note how companies leverage more than one viral loop type.
Other parties will think, “Hey what’s this. Can I sign up too?”
Standard double sided or double loop referral
Content tends to be a slower loop than referrals.
Content tends to be a slower loop than referrals.
Circa 2010
Not really a hack when it’s a bus dev deal, but gives an interesting data-point on what to include in considering a channel or platform.
In my example, we started at the bottom – didn’t have a good collection system. So retooled NPS measures. Then got a result that was under 40%. Put together a 10 point plan to fix. Lastly, after implementing and testing all those measures, we got the NPS where it needed to be in order to proceed with activities like social referrals.
So Sean Ellis says get above 40% on the qualitative survey of usefulness
Alex Schultz takes it a step further, and says to measure MAUs vs Days Since Signup.
If you retention curve goes asymptic to a line parallel with the X axis, you have market fit
If your retention curve slopes down vs X axis, it will intercept it, and you’re not ready to do growth hacking
It’s ok if the green line is lower, so long as it basically flattens out to go parallel to x axis.
They got 500K installs but only 19K DAUs.
Some offerings have the right dynamics to go fast
Others can leverage growth strategies, but should not expect the same results
Some offerings have the right dynamics to go fast
Others can leverage growth strategies, but should not expect the same results