Starting in 1997 and continuing over more than a decade, Ventana built a number of different models for one of the world’s best known pharmaceuticals companies. These models were designed to make contingent predictions of drug sales using inputs such as the attributes of products and the mix of promotion alternatives available. Several of the compounds whose future market response we studied did not proceed past clinical trials, so model predictions could not be tested. But one drug that successfully passed clinical trials achieved one of the most successful launches in the history of the industry. While the role of the model in enabling the success is hard to establish, many people at the company felt the model played a very important role, including the senior marketing vice president of the company. The initial use of the model was not to forecast sales but rather to structure the “war games” used to plan the marketing campaign that launched the product. It served this purpose well, and the exercise explicated many aspects of the market in a self consistent way that had not been conceived before the games. The model was controversial because it predicted that with the right promotion campaign the new drug could be a “blockbuster.” This was not inconsistent with the hopes of many people within the company, but there were many others who thought the drug would be only modestly successful. (The company’s offering was not the first product in a new drug class for serving the market.)
Ventana turned the model over to its client to use during and after the drug launch. We received reports from two individuals that the pre-launch model predicted the post-launch sales quite accurately. Ventana was unable to verify these reports directly, but we were able to determine that the model accurately predicted the market share of the company’s drug in its drug class for about 18 months after the launch.
One aspect of the model was almost universally dismissed by the company’s sales department (as distinct from its marketing department). This was regarding the importance of pharmaceutical sales representatives. Company managers believed that the sales force drove sales in all phases of a product’s lifecycle. Ventana’s modeling confirmed the importance of the sales force during and immediately after the launch of a new product, but they also revealed that the dynamics of subsequent sales efficiency were more counterintuitive than most people realized. It took managers a few years to internalize the findings, but by 2002 there was generally good agreement on the impact of promotion to physicians.