I am building a marketing mix model (MMM) for an online casino, but they regularly run promotions and marketing campaigns that overlap. Trying to determine a baseline level of sales has been difficult. I'm not the most experienced with the wide range of programming-assisted statistical methods available for the task, so I've attempted to manage the project using Python-specifically the fbprophet package (facebook-prophet) to make this easier for me.
My questions are:
Can I estimate the baseline level of sales if I know specific start and end dates for some of the promotions, but not all of them?
Would I need to create dummy variables for all promotions, or can I just model them like a holiday in fbprophet?
Would I be better off decomposing to separate seasonality, from trend, and other noise to estimate a baseline instead?
I appreciate any guidance I can get.