Build the attribution panel

In this video, we’ll show you how to configure and use the attribution panel in Customer Journey Analytics. The attribution panel lets you compare and visualize different attribution models to get a clear view of which interactions and touchpoints drive conversions. It also helps you perform a deeper analysis by examining the impact of one or more touchpoints within specific timeframes. This is achieved by using various lookback windows, which define the amount of time a conversion should look back to include touchpoints. Let’s go to the interface and see how this works. In the Analysis workspace, you can find the attribution panel under the Panels tab in the left navigation. Drag and drop it onto the canvas. Before building, check the selected data view in the top right corner. The data view controls which components, such as dimensions and metrics, are available. Let’s also set up the date range. The default is this month, but let’s switch it to the last 90 days for a more consistent view over time. Now let’s look at the input fields. They help you build the attribution analysis tailored to your particular use case. For this example, let’s say we work at an international retail company that runs multiple campaigns across various channels. These include digital, such as social media, in-app and email messaging, and so on, as well as traditional, such as print and podcasts. As a marketing analyst, we want to understand which channels actually drive online and in-app orders. First, we need to define the success metric. This is the end goal the attribution will be measured against. For our example, we select Orders. Channel is the touchpoint dimension that helps break down which interactions contributed to the conversion.

Select Marketing channel here. Next, we’ll select the attribution models we want to compare. As you can see, there are several, and you’ll choose them based on your business model, conversion timeline, and so on. There are obvious ones like first and last values getting all the credit, but there are also other models like the U-shaped, giving the majority of credit to both first and last, with some given to other dimensions in the middle. Review the documentation for definitions of all models. For our example, let’s select the last and first touch to see which channels are best at initiating customer journeys and which are most effective at driving conversions. Then, we look at linear to get a balanced view of the entire customer experience. Container can be set to person or session. The person attribution level looks at the entire customer journey, regardless of how many sessions a person had. The session attribution level is limited to the boundaries of a single session. The session option respects the modified session timeout configured in a data view. Since we want to analyze long-term influence across multiple visits, we’ll leave the set person. Lookback window is the last input we’ll define to build our attribution report. Again, a lookback window is the amount of time a conversion should look back to include touchpoints. If a dimension item is set outside of the lookback window, the value won’t be included in any attribution calculations. You can select a predefined number of days to look back from when a conversion happened, or specify custom time, where you can choose between minutes, hours, days, weeks, months, and quarters. For our example, let’s select the last 30 days. Now that we have our input set up, we’re ready for the fun part. Click build, and the panel renders the attribution visualizations and data for the dimension and metric selected. First, we’ll see the total number of conversions that occurred over the time window we’ve specified. In our case, it’s the total number of orders that happened within the last 30 days. The attribution comparison bar gives us a good view of how different marketing channels receive credit under the attribution model selected. For example, here, you can see how first touch orders give most credit to display. This is a bar chart by default, but much like visualizations in other panels in the analysis workspace, you can customize its appearance and settings by clicking the gear icon here. The freeform table here is like any other freeform table in the analysis workspace. You can drag and drop new components here to further break down your report, or remove them as needed. The rows and columns in the freeform table control the visualizations you see in the video. If we select the first three rows here, the bar chart will dynamically update to show only selected data. The overlap Venn diagram displays how often a conversion was influenced by multiple channels. The size of each circle represents the frequency of conversions involving that dimension. For example, here, we can see how often customers interacted with both display and podcast channels before placing orders. This is a great way to identify and analyze multi-touch interactions. By default, the diagram displays the top three dimensions, but selecting different rows in the freeform table updates it. Performance detail is a scatter visualization that compares up to three attribution models. Trended performance shows the trend of attributed conversions for the top dimension item, which in our case is display. Selecting a different row in the freeform table will update the visualization to reflect your selection. Finally, the channel flow lets you see which channels across a person’s journey were the most commonly interacted with, and in what order. It allows you to dive deeper and explore the conversion path, for example, to see how often a person moves from email to podcast. This is great for understanding how different channels work together and which sequences are the most effective. You can always update your attribution panel as needed. Simply click the configure panel icon here at the top, and the input fields appear. Let’s say we want to add the U-shaped attribution panel to the report, which will help us balance the extremes of the first and last touch models, and expand the lookback window to 90 days. Click build again, and the panel rebuilds to reflect your changes. You can further manipulate your data by filtering it with a segment. You can drag and drop a segment from your components list, or you can create a quick segment from here. For example, let’s say we want to narrow down our analysis to a specific day of the week. Simply click create quick segment and search for day of the week from the dimensions list. Let’s say for this example that we want to see the insights for Monday.

Once done, click apply, and you can see the quick segment has been added to the panel to narrow down our analysis. Now, you should know how to use the attribution panel to compare various attribution models. We hope this will help you build effective analysis and optimize your marketing efforts. Thanks for watching.