UNDERSTANDING GLOBAL SUPPLY CHAINS
of data (so it can identify two complete seasonal patterns) to begin forecasting effectively. Open model time-series (OMTS) techniques typically need at least four years of data, whereas regression typically needs at least five periods of data for each variable in the regression equation (so if we had sales as one variable and advertis- ing, price, and trade promotions as the three independent variables, we would need at least 4 variables times 5, or 20 periods of data). Of course, many companies have such a short life cycle for their products that many of these techniques are simply never practical.