Investigating the Properties of the Eskin/Kalwani & Silk Model of Repeat Buying for New Products

A given overall aggregate sales history could be the realization of very different purchasing scenarios. For example, a low sales level for a new product could be the result of: (i) many consumers making a trial purchase but very few of them making a repeat purchase (because the product does not meet their expectations), or (ii) a low trial rate but a high level of repeat purchasing amongst the triers (because the product meets a real need among a relatively small set of buyers). Without a proper trial and repeat sales decomposition, it is impossible to determine which of these scenarios (or the infinite other possible combinations of trial and repeat patterns that could result in the same aggregate sales curve) best describes the early sales data. Consequently, the product manager might have great difficulty in developing an accurate forecast and choosing the appropriate course of action. In short, any forecast of future sales based purely on the aggregate sales data is going to be highly suspect.