OUR TAKES ON:

An accurate measurement of marginal (not average) promotion response is required to determine the optimal size of a sales force. Unless the profitability of making an extra call is well understood, it is not possible to assess whether the sales organization has too many or too few sales reps.

The current (over)size of many sales forces in the industry is the result of "benchmarking" and biased estimates of physicians' promotion responsiveness.

Most retrospective data analyses fall victim to a "selection bias." This bias is a result of the sales reps directing their calls to the best targets based on their field knowledge that the data analyst in HQ lacks. Any physician-level retrospective regression analysis that compares script volume to product details would be ignoring the rep's role in selecting the more responsive physicians, and hence, overestimate promotion response.

Controlled field experiments are the most reliable way of obtaining unbiased estimates of marginal promotion response. And if designed appropriately, these experiments can be made non-disruptive to the sales force and easy to implement.

Territory design is more art than science. However, that doesn't mean that (most of) the objectives cannot be formulated in mathematical terms. The key challenge is being able to address multiple objectives at the same time such as balancing "potential" and workload. This requires pushing the limits on currently available software to obtain more than the cookie-cutter territory alignment.

The compensation plan is often a scapegoat. Even the best plan can't fix a flawed go-to-market system, or territories that are significantly out of balance. On the other hand, poor compensation plans can reduce the effectiveness of an otherwise well designed selling system.

The most important physician targeting dimension is promotion responsiveness. The ultimate goal of the targeting system is to rank physicians in order of their expected response (in terms of incremental prescriptions) to an additional sales call. All other variables (see Targeting Dimensions under Marketing Effectiveness) are proxies for promotion responsiveness that need to be used in the absence of a reliable measure for that particular metric.

A side note: As legislation or industry self-regulation starts blocking pharmaceutical sales reps' access to physician-level prescription data, accurate targeting from headquarters will become more crucial than ever

For new product launches, identifying the early adopters is key to success. Hence, practically all brand teams undertake some sort of an adoption analysis to prioritize their targets before going to market. However, when chosen without the proper statistical analyses, analog products used in modeling adoption often turn out to have little or no predictive power.

When even small improvements in early targeting efficiency can translate into large gains, it is critical to spend the additional effort in picking analogs that (i) have demonstrated predictive power against other products, and (ii) have a high likelihood of predicting the brand to be launched.

Towards the other end of the brand's life-cycle, expiration of marketing exclusivity doesn't mean the product should be cut-off all sales support at the point of generic entry. By identifying physicians who show a higher propensity to stay with branded products, significant value can be derived from the product for a sustained period even with substantially reduced sales effort.

"Lead brand" oriented call plans often direct sales reps in a manner that is not quite "customer centric." As a result, some physicians that are important for the portfolio as a whole are missed, and others are presented a primary detail on a less relevant brand for them.

With the appropriate optimization algorithm, tailored call plans can be developed to ensure that the brand emphasized in each call will be the one of most interest to each respective physician. This way, the efficiency and effectiveness of the sales organization can be improved without any significant damage the lead brand.