Forecasting the upfront market has always been a challenging process for NBC. In the past the network used historical patterns, expert knowledge, and intuition to forecast demand: then later it used time – series forecasting models based on historical demand. However, these models were unsatisfactory because of the unique nature of NBC ‘s demand population of advertisers. NBC Ultimately developed a unique approach to forecasting its upfront market for demand, which includes a combination of the Delphi technique seeks to develop a consensus (or at least a compromise ) forecast from among a group of experts, while a grass roots approach to forecasting asks the individuals closest to the final customer’s purchasing plans. In their forecasting approach NBC used its (over 100) account executives, who interact closely with the network’s (over 400) advertisers to build a knowledge base in order to estimate individual advertiser’ s demands , aggregate this demand into a total demand forecast, and then continuously and iteratively update demand estimates based on the account executives expertise. Previous forecasting methods resulted in forecast errors of 5 to 12%, whereas this new forecasting