Future period sales can be
forecast by combining univariate forecasts for price, marketing, economic and
competitive factors and substituting them in the sales response function from
the marketing-mix model.
Once weights are generated
through marketing-mix models that measure relationships between marketing input
variables and sales, forecasts can be generated scoring these weights against
future values of the marketing input and competitive variables. Future values of
input variables can be generated using automatic univariate forecasting
algorithms (for e.g. ARIMA).
Longer term forecasts can be
generated by infusing demographic shifts in the consumer target markets.
Forecasting is useful as a strategic tool to plan pricing and marketing budget
allocation, and assess the impact of competitive activities.
MetriScient™ 
