How to Improve Forecasting with Limited Data
The historical performance of comparable products combined with an understanding of known variables can help managers develop more accurate revenue forecasts in nascent markets where data is scarce.
Forecasting in Uncertain Environments: An Application to the U.S. Motion Picture Industry
Anirban Mukherjee and Vrinda Kadiyali
Samuel Curtis Johnson Graduate School of Management at Cornell University, Research Paper No. 10-07
This study examined a market that is notoriously difficult to forecast — the movie industry. By applying a rigorous methodology and paying more attention to even scarce data for comparable products and seasonal variables, the authors have developed an approach that provides measurable results for forecasting with little historical data. The resulting methodology can be useful for managers in dynamic industries such as technology and biotech.
Though past performance is no indicator of future success, the historical performance of comparable products — even if the time period is short — combined with an understanding of known variables, such as seasonal demand for products, can help managers develop more accurate revenue forecasts in nascent markets where data is scarce.