Word - of - mouth
in your marketing mix:
what is the return?
Koen Pauwels
Associate ProfessorWhat does Word of Mouth do for you?
• WOM hot topic in research and business:
"By far the best user of marketing dollars is to see if you
can create some kind of person - to - person viral effect"
George Zachary, partner at Charles River Ventures
• But hard to measure its impact empirically:
1) What is the elasticity compared to markcom?
2) When is the main effect: short - or long - term?
3) How much $ does it get you to spread around?
Important as $ is flowing back to ads (GoDaddy!)
2*
Types of Word - of - Mouth Marketing
• Viral Marketing: Creating entertaining or informative
messages that are designed to be passed along in an
exponential fashion, often electronically or by email.
• Referral Programs: Creating tools that enable satisfied
customers to refer their friends.
• Community Marketing: Forming or supporting niche
communities that are likely to share interests about the
brand (such as user groups, fan clubs, and discussion
forums); providing tools, content, and information to
support those communities.
.womma.org) * From the Word of Associatiweb site wMouth Marketing on (ww
3Literature
• WOM research has been around for decades
• Primarily survey based, with very few empirical
studies on referral marketing
• Brief overview of empirical WOM studies:
4Godes and Mayzlin (2004) - studied how to use
Internet news group postings to measure WOM
4Chevalier and Mayzlin (2006) - fect of examine the ef
consumer reviews on relative sales of books at
Amazon.com and BarnesandNoble.com
4oo and Hanssens (2006) va, YVillanue - V compared CL
for customers acquired through WOM vs. traditional
channels
4Research Objective
• dynamic process Online referrals: a complex
4Actions of existing customers cause reaction of
yet - to - be customers. When the latter convert
they, in turn, may act by inviting new users, etc.
• s can WOM can be influenced as well a
ess of traditional influence the effectiven
n effortsmarketing communicatio
• need a tool to Marketing practitioners
te the effectiveness understand and evalua
of this process.
5Empirical context
• sites Social networking - the fastest growing
in the World Wide Webarena
4User - generated content attracts and retains visitors
4Very little use of traditional advertising; they instead
organize events, media appearances and WOM
6Site Example - friendster.com
ad • send a message
• post a comment
• upload pictures
• wa blogrite
• dowmusicnload
• invite a friend
7Online Referral Process
8Data
• 258 daily observations (time series) from a major
social networking site:
4Number of new signups
4Number of invitations sent by existing network members
4Number of promotional events
4Number of press and Web media news appearances
• Descriptive statistics
4Total of 23,852 invitations sent
4Total of 23,162 subscriptions received
4Daily average: 92 invitations
4Daily average: 89 new subscriptions
9Dynamic Interactions
LTRADITIONA
MARKETING
WOM REFERRALS SIGNUPSNEW
10Modeling Approach
• Four - AR system: Vvariable
11Results: Impulse Response Functions
12Results: Elasticities
• sticities:Calculating arc ela
where
change in new signups caused by one standard
deviation shock in marketing variable X,
and are mean values.
Short term Long term
WOM Referrals 0.068 0.532
Media 0.008 0.017
Events 0.008 0.026
13Economic Implications: Simulation
Cost per thousand impressions (CPM) ?
• Price quotes from several SN sites: CPM ~ $ 0.40
4Average number of pages viewed per visitor/month: 130
4Average page carries about 2 to 3 ads
4Therefore, average user contributes approximately 13
cents per month or approximately $1.50 a year
• Using our estimates of WOM elasticity of 0.53
4Each invitation sent is worth about 75 cents per year
4By sending out 10 invitations, a user brings about $7.50
14Summary
•Proposed an approach to evaluate the
effectiveness of electronic word - of - mouth
•Quantified the elasticity of referral marketing in
application to online SN site
•WOM referrals - a very strong impact on new
customer acquisition
42.5 times higher than the average advertising elasticity
4r 20 times highethan the elasticity for marketing events
(0.53 vs. 0.026)
4r 30 times highethan the elasticity for media
appearances (0.53 vs. 0.017)
•Economic implications - starting point to plan a
referral incentive program
15Future research
• What if you start giving incentives for WOM?
1) Diminishing returns from 'weaker' links?
But potential exists: Trusov (2006) finds that
subscribers only send out invitation to 1-2
friends, while they have 3-4 influential friends
2) Diminishing returns from selling your soul?
From organic WOM, to fertilized WOM, to BS
16THANK YOU !
QUESTIONS?
17