Professional Documents
Culture Documents
Media
a(ribu,on
]
How
to
measure
cross-channel
media
a3ribu5on
and
access
impact
of
banner
view-throughs
[
Overview
]
Challenge:
establish
a
generic
model
to
accurately
a3ribute
conversion
credits
to
each
channel
and
analyse
the
true
impact
of
banner
view-throughs
in
par5cular.
Key
ques5ons:
Is
my
company
a3ribu5ng
the
right
amount
of
conversions
to
each
channel
or
some
channels
such
as
display
ads
geAng
to
much
credit?
What
would
be
the
op5mal
channel
mix
and
spend
to
op5mize
conversions?
Desired
outcome:
Iden5fy
the
op5mal
channel
mix
and
budget
for
each
and
if
possible
reduce
media
spend
in
some
channels
while
maintaining
eec5veness.
9/18/09
Datalicious
Pty
Ltd
2
[
Methodology
part
1
]
Development
of
a
single
source
of
truth
that
accurately
tracks
all
campaign
touch
points
leading
up
to
a
conversion
event
in
one
central
repor5ng
plaRorm
to
enable
de-duplica5on
across
mul5ple
channels
and
calcula5on
of
true
ROAS
(Return
On
Adver5sing
Spend).
Recorded
campaign
touch
points
should
include
responses
from
paid
and
organic
sources
as
well
as
online
and
oine
channels
plus
include
clicks
and
view- throughs
to
provide
maximum
accuracy.
Omniture
would
be
the
plaRorm
of
choice
due
to
its
exibility
and
accuracy
but
sa5sfying
results
could
also
be
achieved
using
other
plaRorms
such
as
Google
Analy5cs.
9/18/09
Datalicious
Pty
Ltd
3
ROAS
9/18/09
Datalicious
Pty
Ltd
4
Banner Ads
Ad Server
Email Blast
Email PlaGorm
Organic Search
Google Analy,cs
Instead of recording conversions for each channel in separate plaRorms, ul5mately leading to sales being recorded mul5ple 5mes due to the fact that more than one channel usually contributes to a nal conversion, all campaign touch points should be tracked and analyzed in one central analy5cs plaRorm to enable de-duplica5on and comparison of performance across channels.
9/18/09
Organic Search
Instead of recording conversions for each channel in separate plaRorms, ul5mately leading to sales being recorded mul5ple 5mes due to the fact that more than one channel usually contributes to a nal conversion, all campaign touch points should be tracked and analyzed in one central analy5cs plaRorm to enable de-duplica5on and comparison of performance across channels.
9/18/09
[
Methodology
part
2
]
Website
entry
survey
to
determine
the
origin
of
conversions
currently
a3ributed
to
direct
site
visits
as
well
as
branded
organic
and
paid
search
responses
(i.e.
search
terms
that
contain
the
brand).
Direct
website
visits
and
branded
searches
should
not
receive
credit
for
conversions
as
these
channels
most
likely
were
not
the
source
of
the
generated
awareness.
In
addi5on
to
recording
view-throughs
in
the
central
campaign
stacking,
an
ad
server
exposure
test
is
recommended
to
establish
the
incremental
increase
in
search
trac
and
conversions
due
to
display
campaigns
(i.e.
banners
ads
lead
to
an
increases
in
searches
if
they
are
not
clicked
on).
9/18/09
Datalicious
Pty
Ltd
7
Channel Word of Mouth Blogging & Social Media Newspaper Adver5sing Display Adver5sing Email Marke5ng Paid Search - Generic
Channel Straight to Site SEO - Branded SEM - Branded SEO - Generic SEM - Generic Display Adver5sing Aliate Marke5ng Referrals Email Marke5ng
Conversions
a3ributed
to
search
terms
that
contain
brand
keywords
and
direct
website
visits
are
most
likely
not
the
origina5ng
channel
that
generated
the
awareness
and
as
such
conversion
credits
should
be
re-allocated.
9/18/09
10% of users in control group, 90% in exposed group User tagged with segment
Audience Segmenta,on
Control
Exposed
N impressions
Control
Exposed
9/18/09
[
Methodology
part
3
]
Reality
check
and
deni5on
of
company
wide
cookie
expira5on
policies
for
the
various
channels
to
guarantee
accuracy
of
a3ribu5on
model.
Development
of
customized
success
a3ribu5on
model
following
the
principles
and
methods
suggested
by
Forrester
and
including
data
from
campaign
response
tracking,
website
entry
survey
and
ad
server
exposure
test.
One-o
summary
report
of
ndings
plus
automa5on
of
media
a3ribu5on
report
delivery
via
a
customized
online
dashboard
to
increase
ma3er
awareness
within
the
company.
9/18/09
Datalicious
Pty
Ltd
10
$100
Banner Ad $100
Paid Search
Email Blast
Success
$100
Banner Ad $100
Success
$100
Print Ad $33
Success
$100
9/18/09
9/18/09
13
[
About
Datalicious
]
9/18/09
Datalicious
Pty
Ltd
14
[
Con,nuous
op,miza,on
]
Data
Ac,on
Insights
9/18/09
15
Research
SEM
SEO
Tes5ng
Targe5ng
Analy,cs framework
9/18/09
16
Insights
Keyword
Research
Campaign
Repor,ng
Segmenta,on/Data
Mining
Quan,ta,ve
Research
Market/Consumer
Trends
Compe,tor
Analysis
Ac,on
Search
Lead
Media
Campaign
Op,misa,on
Internal
Search
Op,misa,on
Targe,ng/Merchandizing
A/B,
Mul,variate
Tes,ng
Sta
Training/Workshops
9/18/09
17
[ Challenging clients ]
9/18/09
18
cbartens@datalicious.com
Ques,ons Updates
twi3er.com/datalicious blog.datalicious.com
9/18/09
19