Not sure how to explain (using the proper economics terminology) the analysis of the degree to which the employment of a particular tactic (A) translates to a particular outcome (B; say, revenue).
I had been thinking that the proper way of stating this would be
marginal product to A, but I think that MarginalProduct speaks to the amount of output derived through the use of one unit of A.
MarginalReturns might be the better or more correct term for this, as it pairs Returns with Output derived from production (as involved with MarginalProduct).
I had started to consider this with the analysis of GaryV's statement and how it applies to market dynamics, which are subject to change. My interest there is that whereas the employment of tactic A may provide a hefty return at time T1, it may provide far less at T2 — thus tactic A's usefulness may be proved according to whether or not this analysis shows X (your desired outcome).