Pharmaceutical marketing researchers today need to shrink budgets while simultaneously increasing forecast precision. The logic behind this paradox is simple: Increased precision drives better marketing decisions that grow revenues. Shrinking budgets decrease costs. Growing revenues and decreasing costs both improve corporate profitability.
The simplest way to improve precision is to increase sample size but this increases cost. To improve precision without increasing sample, we need to observe more events without increasing the number of interviews we conduct. One method that creates more observable events per respondent is treatment simulation. In traditional marketing research, researchers base each estimate on one allocation per respondent. In simulation, researchers can ask a prescriber to treat multiple patients and then create a sample for analysis based on the total number of treatment events.
In the full article, Roger Green describes a validation exercise comparing the precision of allocation versus patient simulation methods.