In this study the interaction of forecasting method (econometric versus exponential smoothing) and two situational factors are evaluated for their effects upon accuracy. Data from two independent sets of exanie quarterly forecasts for 19 classes of mail were used to test hypotheses.
Counter to expectations, the findings revealed that forecasting method did not interact with the forecast time horizon (short versus long term). However, as hypothesized, forecasting method interacted significantly with product/market definition (First Class versus other mail), an indicator of buyer sensitivity to marketing/environmental changes. Results are discussed in the context of future research on forecast accuracy.
KEY WORDS Forecast accuracy Forecasting methods Situational factors
Guidelines about the relative accuracy of alternative sales forecasting methods would be of value to managers in deciding which to use. Sales forecasts are frequently relied upon for marketing, financial, manufacturing, inventory, and other decisions. In multi-product situations with numerous SKUs, naive and extrapolation methods are often used (Dalrymple, 1987). This appears to be supported in the literature on forecasting. For example, in a review article on forecasting methods for marketing, Armstrong et al. (1987) offer the following: 'A substantial amount of research on extrapolation methods, beginning with Winters (1960), leads to the conclusion that, beyond a modest level, additional sophistication is detrimental-it does not improve accuracy, but it does increase costs and reduce understanding.' The evidence for this conclusion derives from studies comparing extrapolation with econometric ('additional sophistication') methods. While not refuting this conclusion, a few studies point to the possibility that factors inherent in marketing and business situations may yield different conclusions about the accuracy of various methods. For example, Schnaars (1984) considered such factors as the time horizon of the forecast, data availability, level of product aggregation, type of product (durable, nondurable), and the historical stability of the time series in comparing extrapolation with a simple random-walk model. He found differences by product-type and the stability of the historical series. These findings, and those of Bretschneider et al. (1989) on political and organizational influences on forecast accuracy, Brodie and deKluyver (1987) on market share, and Warshaw (1980) on the