How to increase forecasting effectiveness
Bob Draper, head of global forecasting at Eli Lilly, outlines four key ways to turn data into insight
By on Nov 10, 2009Bob Draper, head of global forecasting at Eli Lilly, outlines four key ways to turn data into insight
Pharma needs effective and accurate forecasts, but those often are two different and sometimes cyclical requirements, says Bob Draper, head of global forecasting at Eli Lilly.
Accuracy is about whether the forecast is right, Draper says, while effectiveness is the influence the forecast has on decisions the right decisions. Accuracy is driven by rigid processes and a reliance on models, but complexity often compromises response time. In contrast, the effectiveness that earns forecasting a seat at the table where key decisions are made comes from flexible processes, simple models and quick response times. As an industry, were awash in data from lots of sources, most of which dont agree, Draper says. And that leads to information hoarding and getting bogged down in trying to determine what the right source of data is, rather than the insight the data could bring.
In addition, the forecasting processes most pharmas have in place tend to be a function of the tools theyre using, rather than what meets the organizations needs. And the only way to move past the politics of decisions is to be truly transparent about the information used and where it comes from. Good forecasting gets the right information to people in the right format, Draper says. Out of every forecast, theres data and then theres insight that makes that data useful. You need both to make a forecast effective within the organization.
Lilly has looked at four ways of increasing effectiveness: increasing the credibility of forecasters themselves, improving insights, fostering fast, effective communication, and removing politics from decision making.
Good governance and processes that lead to data-driven opinions enable forecasters to be much more credible. You must exude trust, Draper stresses. That gives you leeway. Once youve established that credibility, people are willing to listen to you. But documentation is critical to back the specific assumptions and rationale you use, so that when changes happen, its easy to know where you started.
Good forecasters should strive to spend more time thinking about and communicating insights. A good story is always more effective than data, so you need good insight, Draper explains. Rather than saying you have a 9.7% share of market, for instance, talk about having the same share of market as the leading brand. You need to link themes across products and therapeutic areas, not just across geographies and product indications.
Then those insights must be communicated quickly and effectively. You need a single point of contact thats capable of rapid information retrieval, Draper advises. And you need standard formats of communication. Most companies have a PowerPoint culture, where the insights only become real once they are on a slide or one-page document. So, we need to provide information in those kinds of formats.
Finally, strong forecasts can help remove the politics from decision making by reducing the amount of opinion that creeps in, Draper says: The key is to minimize information hoarding and increase transparency.
That can be achieved by providing:
- A single point of information that includes widely agreed assumptions on key data points
- Up to date information that minimizes version control and creates confidence that the data is live and trustworthy
- Multiple data formats, including PowerPoint, Excel and Word
- 24/7 online access that establishes complete transparency by providing access to and inviting comments on assumptions behind the data
- Consistency that enables decision makers to draw parallels across products
According to Draper, Lilly has found that although positioning the organization to increase the effectiveness of forecasts is a long process, forecasting credibility is elevated when forecasters can increase the time they spend thinking and deriving real insight for use in decision making. By improving data access and automating many of the data-driven aspects of the process, weve reduced errors and increased the credibility and effectiveness of our forecasts, he says. Communication also has increased significantly, helping to flatten the workload for our forecasters and improve response times.
For more on forecasting best practice from Lilly and other participants in eyeforpharmas 3rd annual Pharma Forecasting Excellence USA Summit 2009, visit http://www.eyeforpharma.com/forecastingusa09/index.shtml.