I have a dataset with month sales for 6 months per customer. I want to compare the sales average of the first 3 months versus the last 3, but I know this averages may be influenced by seasonality of the sales.
I have the "historical" sales for these months for previous years (like 5 years). Is there a way I can use this previous data for detrending the sales of my dataset (present year)?
Right now, I'm approaching it this way:
For example, I can know that, historically, Jan-March sales are 5% higher than Apr-June. So when I calculate the variation for these periods of months, I substract 5% from my calculation. However, this doesnt detrend month by month... only the comparison between groups of months.