 |
Timing Effects 37. How do you calculate the incremental effect of marketing expenditures over time? How do you know what's driving what over the long term? The way we estimate the long-term effect of marketing is that we look at the correlation between current sales and a weighted average of marketing expenditure over previous months, up to the current month. The weighting scheme can differ in how much it emphasizes previous versus current expenditure. If the best explanation of current sales comes from a set of weights that emphasizes previous expenditures, then the long-term effect of marketing is higher. If the weights emphasize the current expenditure, then there isn't a very long-term effect of marketing. The future effects of previous expenditures are then added up to account for the total effect of marketing. 38. Is there a lead/lag issue in the analysis? For example, does journal advertising in January impact Rxs in January? The lag effect of all marketing expenditures was taken into account to determine the long-term effect of marketing. The analysis is designed to determine whether previous marketing expenditures ("lagged marketing expenditure") influences current sales. (See Question 37 for more detail on how this is done.) Additionally, it should be noted that we do not consider potential lead effects that would say, for example, that sales in the current month affect marketing in subsequent months. This would be particularly relevant for brand-specific analyses, which we hope to investigate in the future. 39. Given DTC's long-term effect, if one stops spending on DTC after 2-3 months, does this long-term effect then overstate the effect of other activity's accuracy in those later months? No. The regression distinguishes between the long-term effects of DTC, and the current and long-term effects of the other marketing expenditures. The effects of DTC, DET, PME, and JAD are quantified separately. 40. How did you account for the "launch period" phenomenon, where typically high spends at launch go down over time, corresponding to an increase in Rxs, thereby creating a false-negative correlation? Each brand is assumed to have its own "trend," which can either be increasing or decreasing. Marketing expenditures cause deviations from this trend. For example, a brand could be generally trending upward because it's a new product, but if advertising decreases, the regression will not conclude that a decrease in advertising caused an increase in sales. 41. What is the average/median time for ROI to be 80% (80-20 rule for ROI lead time)? For detailing and journal advertising, it takes between 2 and 3 months for 80% of the total effect to accumulate (75% by end of month 2; 88% by end of month 3). For DTC and PME, it takes between 15 and 16 months for 80% of the total effect to accumulate (79% by end of month 15; 81% by end of month 16). So, in general, detailing and journal advertising effects accumulate more quickly than those of DTC and PME. |
ARPP Analysis | RAPP Study | AMP/ABM Study | Medical Journal Advertising
Additionanl Reading | Other Sites | Contact | Site Map | Home
