It doesn't make sense to divide the number of subscribers by the number of months - a subscriber is likely to be a long term customer (I appreciate the number of trial subs will be significant, but...), so x/4 seriously understates the monthly average.