Professional Documents
Culture Documents
Entry/exit decisions Resource allocations Location decisions Set sales objectives & evaluate performance Set forecast (% of potential)
Information sources
Secondary data Past sales data Primary data
Population of New York City: 8 million 4% between 18-22 = 320,000 60% high school graduates = 192,000 40% have income > $50,000 = 96,000
Compare proposed changes to current results Help set budgets Provide basis for monitoring results Aid in production planning
Considerations in forecasting
Customer behavior (past & future) Competitors behavior (past & future) Environmental trends Product strategies
Methods of forecasting
Judgment based Sales extrapolation Customer based Model based
Delphi method
Nave extrapolation / opinion (2nd most common) Sales force composite (3rd most common)
Appropriate for mature, static industry Average of three period sales over time Average of change in three period sales over time Alternative method to smooth data Forecast sales = a intercept + b slope (time)
Exponential smoothing
Sales = a intercept + b (advertising) + c (price) Develop model on half of past data Test model on other half of data
Ask what they would pay This price may be arbitrary Ask what they would pay Follow-up prices are coherent Ask what they would pay Ask what they would pay Ask what they would pay
Add feature: e.g., a videogame expansion card Add another feature: e.g., a Friendstar device Add another feature: e.g., a hard drive Add another feature: e.g., a Microsoft office
Conclusions
Forecasting is necessary, but difficult All methods have plusses and minuses
All are based on prior experience Will generally miss the turning points Have expected, best and worst forecasts for each
Be prepared!
Advanced Marketing BiMBA 2006