Data-Driven Attribution

Data-Driven Attribution
What is it….and why should I use it?
INTRODUCTION
The customer journey is becoming increasingly
complex
‘6 out of 10 online users start their shopping activity on one device but
continue or complete on another’
‘75% of visitor conversions happen over multiple devices’’.
Without visibility of this multi-touch, multi-visit journey, marketers
cannot untangle the cost and value associated with each channel
and device, and therefore cannot understand how to best market.
The solution is to give marketers visibility of their entire customer
journey, and the associated costs.
THE PROBLEM?
Without a detailed, visitor-level understanding of the
multi-visit multi-channel and multi-device customer
journey, marketers:
Don’t understand the real cost of
acquisition or lifetime value of their
customers
Don’t know the optimal routes to
conversion of their highest value
customers
Can’t optimise this cost relative to their
customers’ lifetime value
Lack clarity on the marketing channels in
which to invest and where to rationalise
Are unsure at which stage of the buying
cycle they’re losing customers
What is a Attribution?
The process of accurately assigning value to each
digital marketing touchpoint across the complete
under journey, providing a greater understanding of
what combination of events drive conversions.
What Companies Are Currently Doing
‘Out of the box’ attribution modelling assigns value based on predetermined criteria.
All credit is attribution to the last
channel interaction
Con: You don’t know where your
customer comes from
All credit attributed to the first
channel interaction
Con: The majority of purchases
don’t happen on the first visit
Assigns 100% value to last Google
AdWord…
Con: Google Analytics telling you how
well Google AdWords are performing
Last Click
Linear
First Click
Time Decay
Assigns equal value to every interaction,
regardless of the role they actually played in
the conversion
Con: Rewarding all channels equally is like
everyone winning first prize - unrealistic
Value credited relative to proximity
to conversion
Con: Still optimising for last clicks
Last AdWords Click
Position Based
Default value attributed 40% to first touch,
40% to the last and 20% spread across all
others
Con: Still not a reflection of the actual
customer journey
‘Data Integration Is The Key To
Fixing Attribution’
- Himanshu Sharma
For attribution to be a success, you need to make sure
you are capturing all the relevant data…
These data sources can be, but are not limited to:
Website
visitor
data
CRM
data
Paid
search
data
Display
advertising
data
Point of
sale data
Affiliate
cost data
Data from
customer
service
Phone
calls data
Data from
internal
order
system
Google
merchant
center
Financial
data
Google
Webmaster
tools
Data-driven attribution uses analytics to understand the
true customer lifetime value
Incoming Brand Cost
Incoming Channel Cost
It identifies a set of unique user
events that contribute in some manner
to a conversion.
It assigns the actual value and cost to
each of these events, rather than using
an arbitrary metric like the simple ‘out
of the box’ models
Visit 1
Visit 2
Visit 3
Visit 4
Visit 5
Channel-specific models
Data-driven models
Look at real visitor data to
determine the role an individual
channel has played in a conversion.
Look across all visitors’ entire
journeys and assign real values
to each of the touchpoints.
Does not consider the interaction
between various channels.
Does consider the interaction
between various channels.
Leads to inaccurate attribution of
marketing spend.
Leads to accurate attribution of
marketing spend.
Our Preferred Data-Driven Model:
The Markov Model
All the channels are ‘states’, and the probability that a customer moves from
one channel to another is a ‘transition probability’
Looking through all the customer journeys, and their various paths to
conversions, gives the channels ‘transition probability’
Remove a channel and seeing how many conversions are expected.
The difference between the actual number of conversions and expected
conversions without that channel is the channels’ value!
Are you ready to make the move to
data-driven attribution?
Fospha uniquely combines a detailed understanding of the economics
of the customer journey, identifies what to do to improve it and gives
you the tools to make those improvements quickly.
[email protected]
www.fospha.com