Managers dealing with new products need to forecast sales growth, especially at the time when sales reach their peak, known as the peak sales time. They use a few initial years’ data to develop a forecast of the peak sales time and test for its statistical significance using traditional measures such as the t-statistic. Unfortunately, these traditional measures only indicate whether the forecast is significantly different from zero or not, not whether the forecast is accurate, that is, closer to the actual value. To determine the accuracy of the forecast, we develop a new metric called voice over noise (VON) that is derived from a diffusion model framework. VON is built on the premise that both the consumer voice (i.e., word-of-mouth that drives sales growth) and the market noise (i.e., distortions of the sales numbers) are important to consider in assessing the accuracy of the forecast; neither alone is sufficient. We empirically prove that a stronger VON indicates a more accurate forecast of peak sales time, which the traditional statistical measures cannot do. For our empirical test, we use 15 new products from the past 3 decades.
|Publication status||Published - Aug 2020|
SourceChina Europe International Business School (CEIBS)
- prepeak sales data
- peak sales time
- new product diffusion