The objective of the O2 Brand Graph pilot was to mine social media data in a way that would allow us to connect it to audience studies.
This presentation is an initial exploration of how we can use social media to augment a segmentation model with real-time data. Instead of tracking contents by keywords (“horizontal” tracking – any content mentioning specific keywords and keyphrases), we looked into mining social media contents and behaviours by audiences (“vertical” tracking – any content generated from a set of sources, regardless of the features of the content).
Kenya Coconut Production Presentation by Dr. Lalith Perera
Mapping the Brand Graph: a study of the O2 audience on Twitter [UPDATED]
1. Digital
Shoreditch
2012
#ds12
Brand
Day,
London,
30th
May,
2012
THE
BRAND
GRAPH
Dynamic
audience
mapping
through
social
media
data
Francesco
D’Orazio,
Chief
Innova9on
Officer
&
Head
of
FACE
Labs,
FACE,
@abc3d
Jess
Owens,
Senior
Analyst,
FACE,
@hautepop
5. “We
talk
of
the
rela9onships
consumers
have
with
our
brands
as
if
they
were
primary
but
the
data
points
to
things
being
otherwise.
Consumers’
most
valuable
rela9onships
are
not
with
brands
but
with
other
consumers”
Mark
Earls
6.
7. As
a
result,
we
don’t
really
know
who
we
are
talking
to…
9. Social
data
allows
us
to
see
how
interac9ons
play
out
at
individual
level
(microscope)
10. Social
data
allow
us
to
see
how
interac9ons
play
out
at
network
level
(the
macroscope)
11. The
telescope
helped
us
understand
the
infinitely
great.
The
microscope
helped
understand
the
infinitely
small.
Today
we
are
confronted
with
another
infinite:
the
infinitely
complex.
12. Social
science.
Computa9onal
Social
data.
Social
Science
Computa9on.
17. Social
Media
vs
Social
Data
Social media is the set
of applications and
platforms allowing Social data is the
people to participate in collective information
online social activities
produced by millions of
people as they actively
participate in online
social activities.
17
18. Cra]ing
quality
insights
requires
custom
data,
algorithms
and
analysis
Off
the
s
h
e
l
f
approaches
a
r
e
not
good
enough
23. What
we
found
Demographics
Informa9on
Flows
Interests
Influence
Dynamics
Behaviours
A[tudes
towards
topics,
brands
etc.
24. Introducing
the
Brand
Graph
+
Social
Graph
Interest
Graph
“…the
correla9on
of
the
network
of
people
who
are
connected
to
the
brand
(Social
Graph)
and
the
network
of
interests,
topics
and
ac9vi9es
of
that
group
of
people
(Interest
Graph).”
25. @O2
followers
(magenta
nodes)
Followers
of
the
@O2
followers
(magenta
nodes)
Mutual
connec9ons
within
the
graph
(blue
links)
26.
27. A
sample
of
the
top
1000
@O2
followers
with
less
than
2000
followers
generated
1
million
nodes
(primary
and
secondary
followers)
and
600K
horizontal
connec9ons
within
the
graph
35. When
did
they
join
Twiger?
#signups
6,000
5,000
4,000
3,000
2,000
1,000
0
Jul
06
Jan
09
Nov
11
36. How
o]en
do
they
tweet?
#tweets
10,000
9,000
8,000
7,000
6,000
5,000
4,000
3,000
2,000
1,000
0
0
10
20
30
40
50
60
70
80
90
100
percen9les
37. Where
are
they
twee9ng
from?
No[ngham
21,415
Edinburgh
22,833
Belfast
24,169
Leeds
26,702
Glasgow
35,150
Series1
Liverpool
40,161
Birmingham
41,265
Greater
London
79,643
Manchester
84,055
London
342,630
38. When
do
they
tweet?
1
2
3
4
5
x
6
7
8
x
9
10
11
12
13
14
15
16
x
17
18
19
20
x
21
22
23
24
0
50000
100000
150000
200000
250000
46. @O2
followers
seem
largely
indifferent
to
the
UK
Top
10
box
office
films
Johnny
English
Reborn
69
The
Lion
King
743
A
mixture
of
movie
Real
Steel
27
chager
and
dvd
release,
The
Three
Musketeers
26
including
RT
compe99on
at
Play.com
Tinker
Tailor
Soldier
Spy
105
Footloose
45
Dolphin
Tale
4
Midnight
in
Paris
26
Don't
be
Afraid
of
the
Dark
20
What theyʼre actually talking about:"
Drive
78
Twilight:
Breaking
Dawn
3865
Dr
Who
2352
Tin9n
1374
47. How
does
@O2
fit
into
this
landscape?
Mentions of @O2"
@O2 follower activity"
Total: 3,120,371 contents in a month from @O2ʼs followers"
Of this, 7,523 included “@O2” – thatʼs 0.24%!
@O2ʼs followerstweet @O2 approx 350 times per day
Overall @O2 sees 1330 mentions/day"
49. THE
BRAND
GRAPH
Dynamic
audience
mapping
through
social
media
data
Francesco
D’Orazio,
CIO,
FACE,
@abc3d
Jess
Owens,
Senior
Analyst,
FACE,
@hautepop
Notes de l'éditeur
We wanted to identify subcommunities within the O2 audience on Twitter.BecauseTwitter is an interest graph, we assumed that following someone implied sharing the interest of the followed user.Therefore a subcommunity would be identified by a high concentration of horizontal connections within the graph.
To get this information we had to map:58k users following O2;Who was following each of the 58k users;Who else in the graph any of the users was following other than O2 or the primary O2 follower.For the sake of this exercise we looked at a sample of 1000 users.We selected the top users with less than 2000 followers. We then mapped their connection to O2.We then mapped who was following them.Finally we mapped how the primary and secondary followers were connected to each other user in the graph.We ended up plotting a graph of 1 million nodes, 1 million primary connections and 574278 horizontal connections within the graph.
The blue links represent how primary and secondary followers are connected to each other within the graph.
By looking at the density of the connections we could then identify hubs within the audience, points of high concentration of similar interests.
Once we knew where the hubs were we than isolated then and looked into the clusters.We identifies 10 users with the highest number of connections to other primary and secondary followers within the O2 social graph.
We then went on profiling the clusters.
It is possible to use social media to study audiences > not just opinions or topics.With the O2 Brand Graph - map and understand the audience of a brand online - mining social media in a different way / we could do a similar exercise for axe, dove
Low Klout scores (1-10) – it’s about my passions & interestsHigh Klout scores (51-60) – it’s professionalSame = light blueMusicLondon (though higher Klout is a bit more London-centric), also “UK”Fan, love (although there’s a lot more love from low-Klout users, who tell you more about themselves + their interests)Red = difference: the high Klout guys talk about their professions a lot more.“Marketing” is in bothBut “blogger”, “journalist”, “editor” and “writer” are high Klout onlyAs our “social media”, “digital”, “director”Green = difference: the low Klout followers tell you about themselvesMarriedFamilyChildrenStudentIt’s also low-Klout people who talk about “technology” and “mobile” as interests – high Klout talk “digital” (from job titles).
Love……my … family / friends / job / music / life / hubby / kids / dog…to … travel / have fun / laugh / talk…music … and fashion/ films / photography / music…the … arsenal / toon[love football lower down the list]Like “Love”, “fan” is one of the most common words in user profiles (active Twitter users).@O2 followers are “fans of”……the Wanted (boyband)…the Arsenal…the WWE…The Only Way Is Essex…the great outdoors…the F1…MUSIC (lots of music fans)…all things funny / digital / apple / tech / motorsport / Japanese….F1…Football…LFC / Liverpool FC…Cricket…Arsenal….Totenham Hotspur