Every time a new drug launches, we write about sales forecasts. So many hundreds of millions by 2015; so many billions at peak (for the better prospects). Some drugs hit their early marks, and some few exceed them. Some fall spectacularly short, as our drug launch disasters report clearly shows.
Thing is, those forecasts are often inaccurate. Very often, according to a study published in Nature Reviews Drug Discovery, and highlighted in Forbes by columnist David Shaywitz. Researchers from McKinsey & Co. looked at 1,700 analyst forecasts on 260 drugs launched from 2002 to 2011. They found that most forecasts were wrong, and more than 60% of them were way, way off--more than 40% over or under. A sizable number of the forecasts overstated the case by more than 160%.
We'll get into the nitty-gritty on this study below. First, we'll suggest that, when you read the article, download the supplemental material at the bottom of the page. That's where you'll find a chart showing which drugmakers appear to best analysts at the forecasting business, at least when it comes to allocating R&D resources toward products that will pay off: Sanofi ($SNY), Novartis ($NVS) and Roche ($RHHBY). Worst? Eli Lilly ($LLY), Merck ($MRK) and Pfizer ($PFE).
One of the more confounding things about the overall analysis is that new information didn't improve accuracy by much. As drugs racked up years on the market--building up sales histories along the way--consensus forecasts got better, but only a few percentage points at a time. Sales projections compared with actual peak sales were off 71% a year before launch and 58% at launch, which are pretty dismal results. But even after 6 years on the market, the forecasts were 45% different from the actual results that panned out later. And forget learning from a predecessor drug's experiences; follow-on drug forecasts were no better than first-in-class projections.
There are some interesting details; for instance, cancer drugs were more likely to be underestimated, perhaps because of additional indications won over time, the authors said. Cardiovascular and central nervous system drugs were more likely to be pegged too high. Forecasts for Big Pharma's drugs were less likely to be skewed than those for smaller companies' drugs, but when they were off, the variance was higher--64% versus 54%. And Big Pharma drugs were more likely to be underestimated before launch, while products from small companies tended to be overestimated.
What does all this mean? As Forbes notes, drug companies make a lot of decisions based on sales forecasts, internal and otherwise. The numbers might determine how large a sales force is dedicated to a particular product, or how it's priced. Early forecasts can determine whether a drug is advanced through the R&D pipeline at all. So, improving the forecasts could save money--sometimes, significant money.
The study authors have a few ideas for that. We'll let you check out the article for those. In the meantime, check out our look at the drugs whose forecasts fell into the way, way off category.
Special Reports: Top 10 Drug Launch Disasters | Top 15 Drug Launch Superstars