Pharmaceutical forecasting: throwing darts or adding value?

Several clients have brought this review from McKinsey, the global management consultancy, to our attention over the last few weeks.  It’s prompted us to reflect on the forecasting we conduct with our clients and put down our opinions on the value of the forecasting process.  It also gives us the opportunity to listen to your opinions.

Throwing darts?

The article, in Nature Reviews Drug Discovery, reports on an in-depth piece of analysis into the accuracy of analyst’s drug forecasts.  These consensus forecasts were chosen for analysis as they represent the most complete data source available for looking at forecasts longitudinally and across a broad set of drugs.  The authors also argue that these forecasts are often relied upon by pharmaceutical companies and investors as the starting point for their own forecasts.

The principal finding is that most consensus forecasts are wrong – often substantially – and that accuracy remains an issue even several years post-launch.  Although it is not surprising that forecast accuracy is poor, the extent of the inaccuracy is particularly arresting.  One can only wonder what impact this has had on companies’ decision making.

Less about the number; more about thorough understanding and a simple, co-operative and iterative process.

At groupH, we have been helping our clients with their drug forecasting needs, and how this feeds into their decision making processes, for many years.  We recognise the limitations of the process and avoid at all costs the generation of falsely-precise forecasts.  We believe that the way to extract true value is to aim for a thorough understanding of the market and of the product – across the whole organisation – and create the ability to react and adapt to changes quickly and easily.  This generates an evolving, hopefully increasingly accurate, forecast range with which to feed into decision making.

Our view is to;

  1. Engage in in-depth scenario planning and competitor evaluations. E.g. What if;
    • Clinical data emerges that means our product will no longer be best in disease or best in class?
    • We don’t reach certain payer-orientated secondary end points? What happens to the value of our product?
    • A pipeline competitor’s clinical profile improves?
  2. Involve a cross-functional team, which meets regularly, to create a non-biased, consensus view across the organisation. Including;
    • Global Marketing/Business Insights.
    • Finance.
    • Regulatory.
    • Clinical.
    • Pricing/Market Access.
    • Marketing Companies/Country Affiliates.
  3. Model the forecast as simply as possible to;
    • Enable input from multiple users with minimum training requirements.
    • Create the ability to document and track assumptions transparently.
    • Refine the forecast rapidly when new information arises.

We know we’ll never be 100% accurate in our forecasting, but we aim to aid our clients in getting the most out of the process.

What are your thoughts and experiences?  We’d love to hear from you.

Want to Dig Deeper?

Pharmaceutical forecasting: throwing darts? M Cha, B Rifai and P Sarraf. Nature Reviews Drug Discovery. Vol 12.p737-8. October 2013.

Available to download at:

Further insightful commentary:

McKinsey Confirms Pharma Forecasts Fragile; Can Industry Learn To Survive Without Them (Forecasts, That Is)? D. Shaywitz. Forbes Pharma & Healthcare. October 2013.

Available to download at:

In defense of pharma forecasting. F. David.

Available to download at: