Multiple Regression Analysis of a time series

Hello everyone! I'm trying to use multiple regression analysis of a time series, but I've run into a problem interpreting the data.

First, I've chosen the top three variables that I felt would affect the Y variable.
Second, I converted each raw number into a percent change from the previous month (why? because that was the suggestion my stats book made).
Third, I divided by the seasonal index to get the Seasonal Adjusted Value of the percent change.

Now, did I overadjust by doing step 2&3? Should I have just skipped step 2?
And what does this leave me currently? How can I use this to estimate the next 2-3 months out?