Finding constant within Data Set

I have a data set that tells me the sales data for every month for all products.
Another data set that tells me the total quantity sold for all products.
Then in a third data set is just total sales / quantity. Equal to sale per product.

However, within the sales data set there is a fairly constant sales stream unrelated to the sale of the products. I am unable to back out what that constant figure is.
This in turn skews the total sales / quantity data. Whenever the quantity sales increase it disproportionally decreases sales/quantity as the distribution of the constant figure number gets distributed to more quantity.
Is there a way to segregate out the constant figure to a reasonable amount?
The constant figure also can change month to month but there is not much variation.

Thank you