Before diving into attribution model analysis, it would be helpful to set up the dashboard to your desired view. To do so, you can utilize several settings:
1. Define how you want the data to be represented
There are two different ways to represent the data for attribution model analysis:
Revenue
Conversion
Based on this choice, all subsequent metrics, tables, and graphs in the dashboard will be adjusted accordingly.
For example, if you want to analyze the attribution model through the prism of:
Revenue: The metrics in the Performance summary block will show: Ad spend, Revenue, ROAS.
Conversions: The metrics in the Performance summary block will show: Ad spend, Conversions, CPA.
To select the representation mode, simply choose the desired value from the 'More' menu at the top of the page: 'Show revenue' or 'Show conversions'.
By default, when you access the attribution model dashboard, the data representation is set to Revenue. If you prefer to analyze based on Conversions, you can easily switch the view as demonstrated in the video above.
2. Specify the data you would like to view on the dashboard
There may be a large amount of data available for analysis. To simplify the results, you can use a global filter located at the top of the page to narrow down the data.
Applying the Global filter will impact the data throughout the entire dashboard, including the Performance summary, as well as all graphs and tables.
In addition to using the Global filter, you can also utilize a local filter that is available for specific parts of the dashboard, such as tables and graphs.
3. Choose the timeline for analysis
You can select any timeline for running the analysis. Simply click on the timeline dropdown at the top of the page. This will open a calendar where you can select the desired dates:
Please note that the attribution model runs on the available data. If you select a timeline where no data is available for specific dates, those dates will be unavailable for selection.
If the model is paused, the dashboard will display data from before the date when the model was put into the paused state.