The Attribution Modeling Tool

The Attribution Modeling Tool (AMT) enables you to compare different models of assigning credit to channels.

There are several attribution models to choose from: 

• Default models of assigning credit for display clicks and impressions.
• Data-driven models, trained from your own data.
• Custom models, where you configure clicks, impressions, and lookback windows, based on your business needs. 

The Attribution Modeling Tool lets you compare up to three attribution models at a time, which can help you identify display advertisements that may be over- or undervalued in your current attribution model and adjust your investments accordingly.

This article shows you how to use the Attribution Modeling Tool. To get a deeper understanding of attribution models, see About attribution modeling. You can also learn more about custom attribution models or data-driven attribution models.

Use the Attribution Modeling Tool

To use the Attribution Modeling Tool:

1. In Reporting, click the Attribution tab.

2. Click Attribution Modeling Tool in the left-hand navigation.

3. Choose a Floodlight configuration.

4. (Optional) Choose one or more Floodlight activities. By default, All are selected.

5. (Optional) Choose up to three attribution models to compare. By default, the first attribution model is Floodlight. Click on the model to change it, or click on the blank attribution model after vs to choose a second model for comparison. You can also add a third model.

   You can choose any of the default attribution models. You can also choose a custom attribution model if one has already been created, or you can create a new custom attribution model. Learn more about custom attribution models

6. (Optional) Choose a primary and secondary dimension. Primary dimensions are listed across the top of the data table. Click on a dimension to choose it. Secondary dimensions are available in the Secondary dimension dropdown list.

Interpret the results and adjust your campaigns

When you're evaluating the effectiveness of your channels, use attribution models to reflect your advertising goals and business models. Whichever models you use, consider testing your assumptions by experimenting. Invest more or less in a given channel based on what the attribution models show, then check your results again to see how it's working.

If you're using more than one attribution model, you can compare the number of conversions or the amount of revenue attributed to your channels according to the assumptions in each model. The % change columns show you the differences between the first model and the second or third model. Click these columns to sort in descending or ascending order, quickly surfacing the channels, campaigns, sites, and placements that show the biggest shifts between different models.