If you have consistently used Google Analytics for eCommerce tracking you may have noticed disparities between data throughout your account. Whenever I come across this issue, notoriously common when tracking a campaign’s success, I find myself conflicted as to what results I should report on. After digging around, a little research and talking to Analytics experts (including Google) we felt it was time to defuse this situation.
First and foremost, its important to note that Google Analytics uses the Last Non-Direct Click model as its default attribution model. This attributes all revenue to whatever medium / source / campaign / etc.that last sent a visitor unless the last click was Direct (in which case its attributed to the previous interaction).
This is how revenue is attributed in any tab/report but Multi-Channel Funnel reports. On the other hand, if you’re digging deeper and considering attribution you can end up with varying data. In Multi-Channel Funnel reports (commonly used to analyze Assisted Conversion metrics) you will almost always see larger revenue reported – though this is dependent upon the Attribution model your instance uses. This is because in these reports attribution is generally not oriented as Last Non-Direct Click, and so you are realizing revenue from other marketing efforts (campaigns, sources, etc.).
Here’s a quick working example:
One of our client’s ran an Instagram campaign that had $25,000 reported (in revenue) when analyzing in the Acquisition tab. While this was considered a decent response and our benchmark for minimum performance, they were much happier (understandably!) when we considered attribution. This expanded, more telling approach highlighted ~$40,000 in revenue from Multi-Channel sources.
This indicates that the $25,000 encompasses all users who either:
1) Arrived from Instagram and purchased right then, or
2) Arrived from Instagram, came back via Direct, and purchased
In contrast, with a different attribution mode, Google Analytic’s reporting would not attribute the revenue solely to those two interactions/paths. Example being, that someone could have visited our client’s website 8 times, then clicked on the Instagram post and revenue would be more easily split. Where as the $25,000 represents all Last Non-Direct Click revenue from the Instagram posts, the ~$40,000 represents users who interacted with the post but converted via a different source later (aside from Direct). Fortunately this indicates that these posts had a positive influence on a much larger number of purchases and revenue.
With $25,000 in revenue generated from users right after they clicked the Instagram post, and $40,000 of revenue from users who converted at some later date / interaction – the campaign was considered a much bigger success and highlighted the value of multi-channel reporting.