H O W L I F E T I M E V A L U E D E T E R M I N E S Y O U R
S U C C E S S A N D H O W T O I M P A C T I T
Lifetime Value
Presentation Overview
Relevant background
Why Lifetime Value
Importance beyond social media
Virality
Retention
Monetization
The Cost Side
LTV varies among customers
Uncertainty of LTV
F R O M S T A R T - U P T O B I G C O
Relevant Background
• Grew to top-5 casual game company
• Initiated and negotiated sale to Playdom, which was then rolled into $570 million Disney
acquisition
Co-founded Merscom CCO, led all marketing/sales/distribution
• Responsible for Europe, Latin America, Russia and India
• Grew it from scratch to 25 percent of Playdom’s revenue
GM of Playdom’s International Publishing team
• Joint venture of EW Scripps and Capitol Broadcasting
• Launched Facebook and mobile games
CEO of FiveOneNine Games
• Lead UA, analytics, monetization and community
• Social Casino space
Currently Chief Growth Officer at Spooky Cool Labs and Chairman of
Globalization Committee at NC Central’s School of Business
Been there, done that
T H E T H R E E M O S T I M P O R T A N T L E T T E R S F O R
Y O U R B U S I N E S S : L T V
Why Lifetime Value
Based on three performance metrics
Monetization
Virality
Retention
Interdepence
Relationship between LTV and CPA
• Success
LTV > CPA
• Failure
LTV < CPA
Think of LTV starting day 1
Green-light
Design and
develop
focused on
LTV
Beta and
other testing
to optimize
LTV
Post launch
to focus on
improving
LTV
C R U C I A L T O A N Y B U S I N E S S
Importance beyond social media
Online offerings
 Netflix
 Ebates
 Farmville
B2B
Square
Rackspace
Bronto
SaaS
 Salesforce
 HootSuite
 Basecamp
Retail
Restaurants
Department
stores
Car dealers
W O R D O F M O U T H ’ S E X P O N E N T I A L E F F E C T
Virality
Definition of virality
 K-score
 K=i*conv% (conversion percentage), where “i” is the
number of invites sent out by each new customer and
“conv%” is the percentage of invites that convert into
costumers
An advanced look
Importance of virality
Lowers cost of
customer
acquisition
Exponential
growth
Improving Virality
K=i*conv%
Increasing I
• Generate virality quickly
• Cater to Connectors
Increasing conv%
• Quality of communication
• Provide value for virality
C R U C I A L A N D H A R D T O F I X
Retention
Definition of retention
Customer
lifetime
N-day retention
Churn rate
Importance of retention
If they do not come
back, monetization
improvements are
virtually useless
Improving retention
 Product quality
 Get in their heads
 Make it social
 Make it global
 Use email and
advertising for re-
engagement
S H O W M E T H E M O N E Y
Monetization
Definition of monetization
ARPU (Average
revenue per user)
ARPDAU
(Average
revenuer per
daily active user)
Percentage of
customers who
monetize
Average
transaction
Average number
of monetization
events per
customer
Importance of monetization
Improving monetization
Product
quality
Value
More
selection
Balancing
Shopping
experience
Promotions
and sales
D O N O T F O R G E T V A R I A B L E C O S T S
The Cost Side
Cost Drivers
• Hosting
• Support
Running Costs
• Platform fees
• Payment processors
Payment Processing
• Engine, such as Epic’s UDK
• Analytics
Software Royalties
• Properties
• Talent
IP Licensing
T H E R E I S N O S I N G L E L T V
LTV varies among customers
Cohorts
Time of year
Stage of product lifecycle
Holidays
Segments
Age
Sex
Income
Interests
Sources
Incentived
ads
Targeted
search ads
Television
Virality Press
Cross
promotion
I T I S A P R E D I C T I O N
Uncertainty of LTV
Uncertainty Principle
Quantum Mechanics
• The universe is random
• Perfect predictions are impossible if the universe is random
Not a function that creates a value
You are predicting a future event
Create a range, not a number
• Albert Pujols is likely to hit 30-40 home runs is more accurate than Pujols is
likey to hit 36 home runs
Models are simplifications of the world
Risk vs Uncertainty
Risk
Something you can put a
price on
Uncertainty
Risk that is hard to
measure
Do not
confuse
uncertainty
for risk
Correlation of past data
does not create certainty
The major difference between a thing that might go wrong and a thing that cannot possibly go wrong is
that when a thing that cannot possibly go wrong goes wrong it usually turns out to be impossible to get at
or repair,” wrote Douglas Adams in The Hitchhiker’s Guide to the Galaxy series.
Wrong assumptions can have profound effects
• Independence of variables
• Mortgage industry
Chaos Theory
• Not a synonym for the game industry
• A small change in initial conditions can produce a large
and unexpected divergence in outcomes
• Major risk when modeling against past performance
Do not discount qualitative information
More data is better than less
This includes non-quantitive measures
Billy Beane has dramatically increased scouting
The Smell Test
Avoid Overfitting
Mistaking noise for signal
Fitting a statistical model to
match past observations
Test how much of the
variability of the data is
accounted for by your model
Solutions
• Think probabilistically
• Distribution shows honest uncertainty
Create LTV range
• Can validate assumptions
A/B Test
• Regularly (weekly or monthly) compare data with predictions
• “When the facts change, I change my mind”, John Maynard Keynes
Surveillance
Avoid Overfitting
Include qualitative data
L L O Y D @ V E R U S E N T E R T A I N M E N T G R O U P . C O M
W W W . L L O Y D M E L N I C K . C O M
@ L L O Y D M E L N I C K
Thank you

Yetizen presentation on LTV

  • 1.
    H O WL I F E T I M E V A L U E D E T E R M I N E S Y O U R S U C C E S S A N D H O W T O I M P A C T I T Lifetime Value
  • 2.
    Presentation Overview Relevant background WhyLifetime Value Importance beyond social media Virality Retention Monetization The Cost Side LTV varies among customers Uncertainty of LTV
  • 3.
    F R OM S T A R T - U P T O B I G C O Relevant Background
  • 4.
    • Grew totop-5 casual game company • Initiated and negotiated sale to Playdom, which was then rolled into $570 million Disney acquisition Co-founded Merscom CCO, led all marketing/sales/distribution • Responsible for Europe, Latin America, Russia and India • Grew it from scratch to 25 percent of Playdom’s revenue GM of Playdom’s International Publishing team • Joint venture of EW Scripps and Capitol Broadcasting • Launched Facebook and mobile games CEO of FiveOneNine Games • Lead UA, analytics, monetization and community • Social Casino space Currently Chief Growth Officer at Spooky Cool Labs and Chairman of Globalization Committee at NC Central’s School of Business Been there, done that
  • 5.
    T H ET H R E E M O S T I M P O R T A N T L E T T E R S F O R Y O U R B U S I N E S S : L T V Why Lifetime Value
  • 6.
    Based on threeperformance metrics Monetization Virality Retention
  • 7.
  • 8.
    Relationship between LTVand CPA • Success LTV > CPA • Failure LTV < CPA
  • 9.
    Think of LTVstarting day 1 Green-light Design and develop focused on LTV Beta and other testing to optimize LTV Post launch to focus on improving LTV
  • 10.
    C R UC I A L T O A N Y B U S I N E S S Importance beyond social media
  • 11.
  • 12.
  • 13.
  • 14.
  • 15.
    W O RD O F M O U T H ’ S E X P O N E N T I A L E F F E C T Virality
  • 16.
    Definition of virality K-score  K=i*conv% (conversion percentage), where “i” is the number of invites sent out by each new customer and “conv%” is the percentage of invites that convert into costumers
  • 17.
  • 18.
    Importance of virality Lowerscost of customer acquisition Exponential growth
  • 19.
    Improving Virality K=i*conv% Increasing I •Generate virality quickly • Cater to Connectors Increasing conv% • Quality of communication • Provide value for virality
  • 20.
    C R UC I A L A N D H A R D T O F I X Retention
  • 21.
  • 22.
    Importance of retention Ifthey do not come back, monetization improvements are virtually useless
  • 23.
    Improving retention  Productquality  Get in their heads  Make it social  Make it global  Use email and advertising for re- engagement
  • 24.
    S H OW M E T H E M O N E Y Monetization
  • 25.
    Definition of monetization ARPU(Average revenue per user) ARPDAU (Average revenuer per daily active user) Percentage of customers who monetize Average transaction Average number of monetization events per customer
  • 26.
  • 27.
  • 28.
    D O NO T F O R G E T V A R I A B L E C O S T S The Cost Side
  • 29.
    Cost Drivers • Hosting •Support Running Costs • Platform fees • Payment processors Payment Processing • Engine, such as Epic’s UDK • Analytics Software Royalties • Properties • Talent IP Licensing
  • 30.
    T H ER E I S N O S I N G L E L T V LTV varies among customers
  • 31.
    Cohorts Time of year Stageof product lifecycle Holidays
  • 32.
  • 33.
  • 34.
    I T IS A P R E D I C T I O N Uncertainty of LTV
  • 35.
    Uncertainty Principle Quantum Mechanics •The universe is random • Perfect predictions are impossible if the universe is random Not a function that creates a value You are predicting a future event Create a range, not a number • Albert Pujols is likely to hit 30-40 home runs is more accurate than Pujols is likey to hit 36 home runs Models are simplifications of the world
  • 36.
    Risk vs Uncertainty Risk Somethingyou can put a price on Uncertainty Risk that is hard to measure Do not confuse uncertainty for risk Correlation of past data does not create certainty
  • 37.
    The major differencebetween a thing that might go wrong and a thing that cannot possibly go wrong is that when a thing that cannot possibly go wrong goes wrong it usually turns out to be impossible to get at or repair,” wrote Douglas Adams in The Hitchhiker’s Guide to the Galaxy series. Wrong assumptions can have profound effects • Independence of variables • Mortgage industry Chaos Theory • Not a synonym for the game industry • A small change in initial conditions can produce a large and unexpected divergence in outcomes • Major risk when modeling against past performance
  • 38.
    Do not discountqualitative information More data is better than less This includes non-quantitive measures Billy Beane has dramatically increased scouting The Smell Test
  • 39.
    Avoid Overfitting Mistaking noisefor signal Fitting a statistical model to match past observations Test how much of the variability of the data is accounted for by your model
  • 40.
    Solutions • Think probabilistically •Distribution shows honest uncertainty Create LTV range • Can validate assumptions A/B Test • Regularly (weekly or monthly) compare data with predictions • “When the facts change, I change my mind”, John Maynard Keynes Surveillance Avoid Overfitting Include qualitative data
  • 41.
    L L OY D @ V E R U S E N T E R T A I N M E N T G R O U P . C O M W W W . L L O Y D M E L N I C K . C O M @ L L O Y D M E L N I C K Thank you