Arthur G. Cook, ZS Associates

Forecasting for the Pharmaceutical Industry: Models for New Product and In-Market Forecasting and How to Use Them is premised on the four challenges of the subject: accuracy, bias, over-generalization and over-detail.

		Predicting the future is quite difficult and forecasting accuracy is generally challenged by uncertainties around key assumptions. But uncertainty can be turned into a positive factor—informing decision-makers about uncertainty to increase their awareness in their decision-making.

		Forecasts always contain bias. Sometimes this is the result of data collection methodologies, sometimes it is contributed by the depth of conviction of key opinion leaders, and sometimes it is simply the result of a loud, vocalized stance. Bias is inevitable, but hidden bias is the bane of the forecaster … and contributes to sub-optimal decision making.

		The balance between “too general” and “too detailed” is difficult to define a priori. An example of a “too general” forecast is one that is a simple spreadsheet where patients are first multiplied by share and then by price per therapy to generate revenue. This is a valid forecast algorithm, but once more subtle questions are asked—such as, are there different patient segments, how is share constructed and what is the compliance rate&mdash;the forecast loses its utility. At the other end of the spectrum are forecasts that are “too detailed.” These tend to take the form of complex algorithms and multi-paged spreadsheets with many formulae. This is also a valid forecast algorithm, but the transparency and transferability of the forecast itself becomes quite low.

		In Forecasting for the Pharmaceutical Industry, we address these four main challenges. First, we transform the uncertainty in forecasting into a perspective that enables more informed decisions. We tackle bias head on, identifying the most prevalent sources of bias in pharmaceutical forecasting and suggesting methods for control of bias factors. The final challenge—balance between “too general” and “too detailed”&mdash;is a function of corporate culture and what is acceptable within an organization. We can, however, examine the characteristics of each extreme and suggest approaches to balanced forecasting.

		Forecasting feeds into and influences many other functional areas within an organization: sales, marketing, manufacturing, finance, portfolio management and others. These linkages may be unidirectional (where forecasts feed into decisions made by the other functional areas) or bidirectional (where the forecast is used to quantify the effects of market changes envisioned by other functional areas).

		One way to consider the role of forecasting in an organization is to map out the decisions influenced by the forecast. The inputs of the forecast to the various planning processes range from guidance on clinical trials (value of the target product profile claims) to guiding business development activities to both long- and short-term brand planning activities, to detailed manufacturing planning.

		Thus, the many functional and planning areas that interact with the forecast create tremendous pressure on the model construct, the analytics, the inputs and the outputs of the forecast. The danger is that every function develops its own model for its own planning purposes, based on a different set of assumptions with little or no consistency across the board. The key challenge to forecasting is to create a process where the needs of function can be made without compromising the integrity of the forecast approach. 

		Forecasting for the Pharmaceutical Industry: Models for New Product and In-Market Forecasting and How to Use Them is published by Gower Publishing Co., Burlington, VT.

		Arthur G. Cook is a Principal with ZS Associates, a global sales and marketing firm (

). Previously, he was Director of Global Forecasting for Syntex Pharmaceuticals. Art has been involved in pharmaceutical forecasting for over 30 years, has worked with many major pharmaceutical companies on their forecasting processes, and has created forecasts for over 150 therapy areas.

Articles

The second edition of an essential business planner's library