Should I use expected return on investment for the Net, direct mail flyers, cold calling, etc., and then divide the money
or just dump all of it into the best (calculated) option?

If this were a single choice to be made of many I could probably handle it using Decision Theory.