The importance of data-driven attribution 

An attribution model is a set of rules that determine which touchpoints in a conversion path receive credit for sales and conversions. But with more than one model to choose from, how can brands be sure they are fairly assigning the full and correct value to every channel on a customer’s path to purchase?

Attribution models have traditionally and - all too often - arbitrarily assigned value to touchpoints in a user journey based on their position within the path to a pre-defined outcome; most likely a sale. In the case of the much maligned but ubiquitous last-click model, the interaction parameters and resulting decision-making are often a series of binary questions (was there a click: yes/no, did they convert: yes/no, was there a succeeding click: yes/no etc.), with the value in full being assigned to whoever got in front of the user last, without any real consideration to the preceding touchpoints. Much like a striker receiving all the credit for scoring the winning goal whilst neglecting to acknowledge the assist from the midfielder who passed two onrushing defenders to execute an inch-perfect through ball.

How can we overcome this?

Data-driven attribution models attempt to redistribute value to other touchpoints by looking holistically at the user journey and assigning a value to every touchpoint based on a variety of different attributes. The more data used in the input of the model creates more sophisticated ways to analyse the value of each channel, all of which is routed in concrete mathematics.


If, for example, the majority of conversions began with an email as the first touchpoint and had a random channel as the touchpoint immediately preceding the conversion, then common sense dictates email is the channel providing value. A last click attribution model would assign value only to the last touchpoint, therefore ascribing email zero value.

However, more complex models will also factor in what a user did onsite after clicking on or seeing a specific channel’s messaging. If, for example, a user clicked on that initial email and spent a large portion of that browsing session reading about a product, would we perceive that the email was of a greater value in the context of the resultant sale? To take this further, what value do we attribute to the content of the site?

Two of the great benefits of digital advertising are the precision of the audience targeting and the ability to measure performance. However, too often the performance of each channel is viewed in silo, without any real acknowledgement of the bearing each channel has on the other when a user is exposed to a combination of messaging across multiple mediums.

With clients increasingly demanding that their marketing and advertising budgets deliver real business outcomes, the need to understand users that arrive onsite as a result of engaging with their acquisition efforts and tailoring the onsite experience accordingly has never been greater.

Reassigning value across every channel

Attribution modelling in digital marketing has historically only applied to paid customer acquisition channels. But in a world where the onsite experience needs to be connected to these channels, it is too limiting to only assign value to spend generating traffic, as the onsite touchpoints that a user engages with will tell a story too.

Adding in the salary of your content writer to your attribution model might sound like a strange idea, but if a 3rd party vendor is paid to dynamically manipulate page content and programmatically alter the user experience of a website based on data signals such as user behaviour and traffic source, then this is a channel that should be plugged in to a marketer’s attribution model, just as affiliate content sites would be.

In the next 12 months, we expect more marketers to be looking holistically at their marketing budget allocation and assigning genuine business value to each. Whether that’s scrutinising internal resource allocation, or holding partners to account, we expect marketers to dive in to the nuance of how every touchpoint on the path to conversion interacts and influences each other, and ultimately contributes to their business’ bottom line.

Digital Marketing


Ebook: Beyond programmatic advertising

Discover how customer experience is shaping the programmatic advertising landscape in 2019

Download eBook
Beyond programmatic advertising

Posted by Aj Hill

Director of Strategy