An Attribution Model is a set of rules that defines how credit is given to campaigns based on the breadcrumbs in a conversion path. There are several different Attribution Models that are used in digital marketing, First and Last Touch being the most common. In this article we'll explain the models Feathr makes available to you.
Example conversion path:
A prospect finds your website through an Invite link from Campaign A. They return a week later by clicking on a Feathr ad from Campaign B, but don't register. The next day they see a few ads from Campaign C and later return directly and register for your event.
In a Linear attribution model equal credit is given to each Campaign a prospect interacts with. In the example above, each of Campaigns A, B and C would receive 33.333% credit for the conversion.
Note: This is the default model in the new Feathr User Interface
In a Last Touch attribution model the last channel a prospect interacts with receives 100% of the conversion credit. In the example above, Campaign C would receive all the credit since it was the last Feathr campaign the Person interacted with before converting.
This attribution model assigns 100% of credit for a conversion to the first campaign a prospect interacted with. In our example above Campaign A would receive all credit.
Positional attribution is like a combination of First Touch, Last Touch and Linear. The first and last campaigns the Person interacts with in their Path to Conversion each receive 40% of the credit for the conversion. The remaining 20% is distributed evenly between the campaigns with breadcrumbs between the first and last interactions. In the example above, Campaign A and Campaign C would each get 40% credit, and Campaign B would get 20%.
Credit is distributed according to how much time elapsed between the campaign breadcrumbs and the conversion breadcrumb in Time Decay Attribution. The campaign with the most recent interaction would get the most credit, and the campaign with the oldest interaction would get none. The exact distribution depends on the length of the Path to Conversion and where the crumbs from each campaign fall in that path. In the example above, Campaign C would receive something like 75% of the credit, Campaign B would get something like 25% and Campaign A would get 0%.
In this model, all campaigns are given full credit for every conversion they influence by showing up in the Path to Conversion. In the above example, Campaigns A, B and C are each given 100% credit for the conversion. Naturally, this leads to problems with side-by-side and aggregate ROI reporting, as the ROI for a campaign relates directly to the credit the campaign is assigned for conversions. As a result, although this model is simple and easy to understand, it is not recommended in any but the most basic use-cases.