Attribution modeling

Data-driven attribution methodology

There are two main parts to the Data-Driven Attribution methodology: (1) analyzing all of your available path data to develop custom conversion probability models, and (2) applying to that probabilistic data set a sophisticated algorithm that assigns partial conversion credit to your marketing touchpoints.

In this article:

Develop conversion probability models from all available path data

Data-Driven Attribution uses all available path data (from your floodlight configuration) —including data from both converting and non-converting users—to understand how the presence of particular marketing touchpoints impact your users’ probability of conversion. The resulting probability models show how likely a user is to convert at any particular point in the path, given a particular sequence of events.

Algorithmically assign conversion credit to marketing touchpoints

Data-Driven Attribution then applies to this probabilistic data set an algorithm based on a concept from cooperative game theory called the Shapley Value. The Shapley Value was developed by the economics Nobel Laureate Lloyd S. Shapley as an approach to fairly distributing the output of a team among the constituent team members.

In the case of Data-Driven Attribution, the “team” being analyzed has marketing touchpoints (e.g., Organic Search, Display, and Email) as “team members,” and the “output” of the team is conversions. The Data-Driven Attribution algorithm computes the counterfactual gains of each marketing touchpoint—that is, it compares the conversion probability of similar users who were exposed to these touchpoints, to the probability when one of the touchpoints does not occur in the path.

The actual calculation of conversion credit for each touchpoint depends on comparing all of the different permutations of touchpoints and normalizing across them. This means that the Data-Driven Attribution algorithm takes into account the order in which each touchpoint occurs and assigns different credit for different path positions. For example, Display preceding Paid Search is modeled separately than Paid Search preceding Display.


In the following high-level example, the combination of Organic Search, Display, and Email leads to a 3% probability of conversion. When Display is removed, the probability drops to 2%. The observed 50% increase when Display is present serves as the basis for attribution.

illustration of display increasing likelihood of purchase

Was this helpful?
How can we improve it?

Need more help?

Sign in for additional support options to quickly solve your issue