Attribution IQ
Objective
Understand how Attribution IQ supports cross channel analysis.
In most cases people think of measuring the influence of marketing touches on driving conversions when talking about attribution. Attribution IQ supports this use case very well, but it can also be used for many other use cases. When used in CJA, we can measure the influence that any data point has on driving an event of interest, for example, let's measure the influence of web pages on driving calls.
Exercise
- Create a new panel with a Freeform Table and title it Attribution IQ. Ensure your date range is the last 6 full months.
- Drag the Page Name dimension & drop it into the panel, then add in the Sessions and Event: Call Center dial-ins metrics:
As expected, the “Calls” metric reports all 0's because there are no Call events that occur on the same event as a Page View.
Let's use Attribution IQ to configure how attribution works for this metric.
- Click on the gear to the right of the Calls metric in the table and click the checkbox to use non-default attribution model.
- In attribution model settings, select the Time Decay model and configure a 1-Day Half-life. Set the Lookback window to Person (Reporting Window)
- Click Apply.
This is a perfect model for what we're trying to do. Typically, customers will try to do something online for a while before they give up and call into the call centre. This means that they may hit a couple of pages before they call. We don't just want to give credit to the last page for driving the call. But we probably do want to give the most credit to the last page for driving the call.
We configured the time decay model to look back from the call event up to 1 day and give the most credit to the last page that a person saw prior to a call, and then incrementally less credit to each page behind that, up to 1 day minutes prior to the call.
The resulting table:
Based on this table, we can see the top pages that are driving calls into the call centre. Let's go deeper.
- In the Components menu, search for the Call Reason dimension and click on the arrow to the right to see the items within that dimension.
- In the "Call reason Items" view, click "Show items from last X months" until values show
- Select the top 4 call reasons (by holding down the shift key).
- Now drag them under the "Calls" metric in your table until the +Filter By prompt in blue appears, then drop them.
The result is a table that uses a 1-Day Time Decay model to give credit to pages for driving calls, broken out by the top 4 Call Reason types.
Marketers typically use this data to uncover the top pages driving calls and testing different versions of those pages (Adobe Target is a great option here) until they find a version that works best at keeping people in the cheaper, online channel.
Congratulations, this completes the exercise on Attribution IQ!