Month over month variance analysis


New Member
Hello all,
Search on web has been fruitless. I'm conducting a month over month analysis for sales volumes by variance buckets of 1) seasonality 2) demand and 3) number of days. For example, if sales in June were 100 and sales in July were 150, how much of this 50 variance is due to 1-3 above? My current approach is to get a centered moving average for each month based on 5 yrs worth of data to get the seasonality index. Then for both June and July I "deseasonize" the sales by the the associated monthly index to get adjusted difference, which I figure would be true demand excluding seasonality. But how do I account for the variance due to the extra day in July?