David Binetti, Votizen

The Sacred, The
Profane, and the Pivot
A Brief Case Study of Applied
CDM, AARRR Metrics, and Lean Principles

(c) 2009 David Binetti
Monday, March 7, 2011

Deﬁnitions
Sacred
Steve Blank’s Customer Development Model (CDM).
I follow it religiously.
Profane
Dave McClure swears every third word. Oh, and
his Acquisition, Activation, Retention, Referral,
Revenue (AARRR) model is awesome.
Pivot
Eric Ries’s concept of strategic shifts predicated on
learnings. (Sorry, no pithy witticism for this.)


CDM (Brief Version)

A) Establish Hypotheses

Test -> (Iterate or Exit)

B) Sell

Test -> (Iterate or Exit)


CDM (Discovery - Phase A)
Category Hypothesis

Product State what you are selling.

Problem State the customer problem you are solving.

Channel State how you are going to sell your product.

State how you are going to generate demand for
Demand
your product.
State the market type (existing, re-segmented,
Market Type
new) you’re in.

Competitive State the competition you face.


CDM (Validation - Phase B)

I could explain in detail, but
basically boils down to:
prove you can sell by actually
selling before you scale.


AARRR Model
(AKA, Metrics for Pirates)

Metric Description
Acquisition Users visit

Activation Users sign up

Retention Users come back

Referral Users like and refer

Revenue User pay something for something


AARRR Model (cont.)
The Model measures:

Conversion rates for each step

Expected value at each step (not
included here.)

A ‘one-slide pitch deck’.

Have these numbers, and you need
nothing else.


Slight Differences

The CDM has you look at and
test the whole business model.
The AARRR assumes the model
rolls up in the metrics.


How They Complement

The CDM teaches you to test
assumptions, but isn’t as speciﬁc
on how to measure results.
The AARRR gives metrics to
know what’s good, but not how
to identify what’s broken.


Case: Votizen

Social Network for Registered Voters

Civic Participation Made Simple and
Meaningful

Voter-Focused (as opposed to politician-
focused) approach to $5B Political Market.

Independent, non-partisan, simple and fun!


CDM (Brief)
Category Hypothesis
Social Network for Registered Voters (“LinkedIn
Product
for Politics”)
Traditional campaign techniques #fail in a web
Problem
2.0 world.
The internet is the primary source of voter
Channel
information.

Demand Social networks are both the product and the
demand-generator.
Resegmented; displace current market for
Market Type
direct-mail & robocalls.

Competitive Hyper-fragmented.


AARRR (Brief)
Metric Description
Creates account on system. (Precursor-
Acquisition
requirement)
Certiﬁes voting record (Authenticates - gives
Activation
voter the power)

Referrals Forwards to friends (Acquisition model)

Uses system to affect change (Validates lifetime
Retention
value)
Private! (Sorry, not appropriate for a public
Revenue
document.)


Applied
Started with MVP - Minimum Viable Product.

Just enough to get feedback (cost: $1206)

Metric MVP
Acquisition 5%
Activation 17%
Referrals -
Retention -
Revenue -


Improve
Tried a bunch of A/B testing on home page.

Found an optimal message.

Learning: When given the option to explore,
people explored. When given a direction,
they took the direction.

If you want something done, make it part
of registration ﬂow.


Applied
Signiﬁcant, immediate jump in initial metrics.
(Spend ~$10K)

Metric MVP V.1
Acquisition 5% 17%
Activation 17% 90%
Referrals - 4%
Retention - 5%
Revenue - -


Improve Retention and Referrals

Same Process, Different Results

A/B, Feedback, Analytics, etc.

Incremental improvement on retention
and referrals.

Was improving, but marginal value of
efforts were asymptotic.

Time for a pivot!


The Pivot
Eric Ries concept: shift direction while
grounded in learning

Implementation of “back to start of CDM”

I took learning and created new MVP.

So, $10K sunk cost -- buh-bye! (ugh)

Needed to ﬁre current developers,
hire new skills. (double-ugh)


The Pivot (cont.)

But, not a waste!

Time, money and effort brought learning.

Look beyond the features to the problem.

“I always wanted to get more involved.
This makes it so much easier. Thanks!”


Applied
New product launched as MVP - @2gov on Twitter

Contact all public ofﬁcials at a single address.

Results? Much better R&R, and better conversion.

Metric MVP V.1 New
Product
Acquisition 5% 17% 42%
Activation 17% 90% 83%
Referrals - 4% 54%
Retention - 5% 21%
Revenue - - -


Applied
Next step: test revenue assumptions

Provide a channel for voters to support causes and
candidates in which they believe.

Another pivot! - No data yet

Metric MVP V.1 New Next
Product Iteration
Acquisition 5% 17% 42% ?
Activation 17% 90% 83% ?
Referrals - 4% 54% ?
Retention - 5% 21% ?
Revenue - - - ?


Takeaways
Set up your CDM hypotheses so that you know
what you are trying to accomplish and how to
tell if you are ﬁnished or need to start over.

The AARRR approach serves as a great ‘startup
dashboard.’ Numbers don’t lie (as long as you
don’t lie to yourself.)

Should you need to iterate, then pivot -- don’t
ﬂail. Measure success by lessons learned, not
by the results themselves (oh, and restate CDM.)


Learn from the Source

CDM: http://www.steveblank.com

AARRR: http:/ /www.slideshare.net/
Startonomics/startup-metrics-for-pirates-
presentation

Pivot: http://
www.startuplessonslearned.com/2009/06/
pivot-dont-jump-to-new-vision.html


Feedback Wanted!

Would love feedback!

Want to get better, so criticism is preferred.

“If you don’t have anything bad to say,
then don’t say anything.”

dbinetti@gmail.com


David Binetti, Votizen

More Related Content

What's hot (20)

Similar to David Binetti, Votizen (20)

More from Sheila Goodman (13)

Recently uploaded (20)

David Binetti, Votizen