The long memory of newspapers' subscriptions : between the short-run and persistence response

The mainstream of marketing time series analysis has shifted from classical short-range dependence (ARMA, transfer functions and VAR models). However, in cases where purchase decisions entail some commitment (e.g., a subscription selling periodic use of a product or service), sales response entails a long-term effect is not permanent. Long-memory assumes that shocks to a time series have neither a persistent nor a short-run transitory effect, but that they last for a long time and decay slowly with time. Many marketing policies face a short-memory response at the individual customer level but display a considerable degree of persistence at the aggregate level. The aggregation of short-run individual decisions made by heterogeneous customers can show a long-memory pattern. In today's highly competitive newspaper industry, loyal, ongoing customers are a key to obtain stable and long-term profits. Often newspapers obtain a loyal customer base through subscriptions. This paper proposes a long-memory model to study the long-term sales response dynamics in subscription markets. The model accounts for the heterogeneity of the individual responses and distinguishes between both trend and long-memory components pattern of subscriptions. This model permits more accurate predictions of subscription sales than those obtained using persistence models