Our own experience tells us, that some people are more influential then others. A good run CRM system should therefore identify those “Opinion Leader”, “task makers” or “influencer” to treat them differently. But do they even exist? Lutz Finger (@LutzFinger) will look at anecdotal evidence as well as research to clarify how to measure influence. The findings:
# Influencers are often not wide-spreading… they are local.
# There not only a few dominating the rest… it is more network and peer pressure which influences.
# Influencers are often overestimated… 50% is homophily.
# Influence very much depends on the topic.
Rather than to look for influencers look for Broadcaster. However be aware that broadcaster only create awareness and not necessarily intention.
The video from the presentation: http://youtu.be/HgCNwUytjZI
3. Quiz
Please
raise
your
hand,
if….
• Would
you
choose
a
doctor
because
friends
recommended
it?
• Would
you
choose
a
computer
program
because
friends
recommended
it?
• Would
you
have
gone
to
a
movie
because
your
friends
recommended
the
movie?
4. Success
Stories
● 100
influencers
in
social
media
● Offered
Test
Drives
and
“broadcasted”
their
comments
● YouTube
videos
related
to
this
campaign
generated
over
6.5
million
views
…
and
was
it
successful?
● Ford
received
more
than
50,000
requests
for
informaQon
● 10,000
cars
were
sold
in
the
first
six
days
Awareness Trigger Intention
Source:
"Demys.fying
social
media,"
Influencer
McKinsey
Quarterly,
April
2012
Picture:
YouTube
channel
8. Since
1955
● opinion
leaders
(Katz
and
Lazarsfelt
1955)
● influenQals
(Merton
1968)
● Influencers
(Rand
2004)
● e-‐fluenQals
(Burson-‐Marsteller
2001)
● hubs
(Rosen
2000)
● Mavens
&
connectors
(Gladwell
2000)
Source:
Amazon.com
Website
Law of the Few
Social epidemics (are)… driven by the
efforts of a handful of exceptional
people
(Malcom
Gladwell,
2000,
The
Tipping
Point)
9. Influence
can
be
local
Why?
● By
Chance?
● Influence?
● Socio-‐Economical?
Source:
h3p://www.doh.wa.gov/
10. Yahoo!
Go
IntroducQon
2007
QuesQon:
• friends
influence
friends
to
adopt
• friends
with
similar
backgrounds
and
tastes
adopt
(homophily)
Sinan
Arala
et.al.:
• Analyzed
27.4
million
users
• 89
million
disQnct
relaQonships
• 14
bn
page
views
• 3.9
bn
messages
• Over
5
month
Source:
Sinan
Arala,
Lev
Muchnik,
and
Arun
Sundararajan
(2009):
PNAS
Vol
106,
no
51
11. Influence
comes
from
many
FracOon
of
Adopters
vs.
Non-‐adopters
Number
of
Adopters
in
Network
• Network
externality
are
the
main
driver
for
Influence
-‐
contagious
• Homophily
is
accounQng
for
50%
of
the
adopQon
effect
Source:
Sinan
Arala,
Lev
Muchnik,
and
Arun
Sundararajan
(2009):
PNAS
Vol
106,
no
51
12. Influence
depends
on
the
Topic
Not
all
Topics
are:
• 4
years
• 1001
Students
on
Facebook
• tradiQonal
Self-‐reported
Data
• How
did
taste
Spread
Source:
orgtheory.net
Source:
Kevin
Lewisa,
Marco
Gonzaleza
and
Jason
Kaufman
(2012):
PNAS
Vol
109,
no
1
13. The
Myth
of
Influencer
local
● Influence
can
be
many (like
peer
pressure)
● Influence
is
done
by
● Influence
might
be
in
reality
homophily
● Influence
depends
on
topic
13
14. BUT
why….
Aja
Dior
M.
omgg,
my
aunt
.ffany
who
work
for
whitney
houston
just
found
whitney
houston
dead
in
the
tub.
such
ashamed
&
45
min
sad
:(
Or
why….
Source:
TED
talk
by
Kevin
Allocca
6
month
16. How
to
MEASURE
broadcast
● Reach
● AcQvity
● Engagement
16
17. Youth
Olympic
Games
AcOvity
versus
Response
6000
Response
5000
● Reach
4000
PromoQng
YOG
Social
Media
AcQvity
● AcQvity
3000
● Engagement
2000
1000
0
Athletes
Role
Young
Young
YOG
Models
Ambassadors
Reporters
Ambassador
Source:
Fisheye
Analy.cs
17
18. CRM
Strategy
● Re-‐Check
your
Influencer
Strategy:
─ It
might
be
local
Intent
─ It
might
be
more
based
on
peer
pressure
─ It
might
not
fit
all
of
your
products
● Look
out
for
the
Broadcaster
─ Most
reach
measure
will
work
Awareness
─ Be
sure
that
you
measure
engagement
─ Stay
away
from
acQvity
Source:
Fisheye
Analy.cs
19. Thanks
If
you
liked
it.
Follow
me
on
Twi3er
@LutzFinger.
Book
by
O’Reilly
Media
To
be
published
in
Spring
2013