In any given day, there can be hundreds to thousands of different marketing campaigns running. Marketing needs visibility into the effectiveness of the various campaigns to stay in front of the competition and find more ways to attract customers. One way to measure the effectiveness of all these campaigns is through attribution. Assuming this information exists, how can user leverage this information? Generally, attribution is simply measured directly against spend, but sometimes we start with a different question: What’s so special about these customers? This dashboard looks at attribution, but for a particular set of customers.
In the example shown here, I simulate a more typical direct response customer group. This is evidence by that there’s very little attribution prior to the customer cohort in question. Typically, if a business has a longer lead time (e.g. car buying), there will be more of a gradual build of prior to the cohort in question.
Components of the Dashboard
The main time series graph shows the overall distribution of attribution in context of the customer cohort. The multiple shades of red correspond to the category of attribution that’s assigned. You can think of an attribution category as the type of customer behavior it influenced at that point in time. For example, this can be general awareness, reminding, or purchasing. The number of categories is completely dependent on the type of attribution model being used – in this case there are 4 categories.
Attribution by Marketing Channels
From this time series graph, the spikes in attribution can be correlated with specific campaigns or promotions run by the marketing team. This can be validated by the (blue) table in the bottom right corner. In this section, a more detailed breakout of every marketing channel is presented, with the corresponding total attribution value. The shading of each cell is dependent on the impact that particular marketing channel had with respect to all other values in the table. This helps call out high and low impact channels, while still providing the actual numbers.
Finally, the bottom left graph presents a normalized view of the cohort. In this view, we display the overall distribution (blue bars) and cumulative distribution (orange line) of the cohort by the number of days it took to convert and the number of touchpoints (ads / emails / etc.) that this particular cohort experienced. This is useful to see in the event the user wants to narrow in on what campaign might have a more immediate effect versus a long-term effect (notice the filters at the top).
This is by no means a complete view of this dashboard and I simply cannot do it justice within a single image. There remains interactive filters and content that allows a user to narrow in on a specific marketing channel and highlight its importance with respect to the whole. Further, marketing channel cost information is excluded in this example, but can be overlaid for a direct attribution to cost ratio estimation. This is just an example of the family of reports that surround attribution.