What are some best practices for calculating revenue / conversations from View Through attributed users, especially on FB?
We all know FB's results can be very inaccurate, especially when a product / app is already heavily advertised by non-paid sources on their network (Instagram pages with extremely large audiences etc).
How do you handle this? And also how does your workflow allow to keep this in check?
Our current method is to run lift tests within FB on a few key campaign types (Re-engagement, cold audience and brand aligned audiences). We will then get the lift % from these tests and then divide the result by this %. In some cases only 20-30% of the results can be actually attributed to the campaign.
I would love to know other people's thoughts on the matter. I'm all for looking awesome using view through ROI results, but we all know this is not scalable!