The Science of Forecasting

Historical data is the starting point for forecasting.  Your company's sales information is crucial, along with industry data for the same periods in time.  What factors explain variations in your sales over time?  Most sales are sensitive to both internal factors (advertising and promotions, availability of product, and actions by your competitors) as well as external factors (consumer sentiment, interest rates, and seasonal factors). The first step in any forecasting activity is interpreting historical data.

Examples of "Science" Activities:

Extracting sales data from financial applications
Analyzing seasonal sales patterns
Estimating industry sales and calculating market share
Building models to explain how various external factors affect sales
Determining price elasticity under various conditions

The Art of Forecasting

Creating a forecast requires making assumptions.  Although no one knows exactly what the future holds, it is possible to make informed, intelligent assumptions.  What factors will affect your sales in the future?  If, in Step 1, you have been able to build a model that explains sales to date, you may be tempted to conclude that the future will simply be a continuation of the past.  The Art of Forecasting requires more thoughtful interpretation of future conditions.  Both internal factors (new production facilities, different channels of distribution) and external factors (changing consumer demographics, unemployment rates, global developments) must be considered.

Examples of "Art" Activities:

Forecasting economic factors such as housing activity, unemployment, interest rates, and consumer confidence
Predicting manufacturing capacity and cost of goods produced
Estimating emerging or declining markets for products
Forecasting consumer demographics such as housing formations, age/income trends, and product penetration
Identifying and quantifying the impact of new technologies

The Management of Forecasting

In order for a forecast to be useful, it must be integrated into all relevant aspects of your business.  Typically, this includes manufacturing, distribution, finance, marketing, and sales.  A good forecast is not developed as a self-fulfilling prophecy for setting higher sales targets, cutting prices, or building a new plant.  A forecast is neutral.  If the "science" has been conducted carefully and the "art" has been conducted thoughtfully, the forecast will gain acceptance as a starting point, rather than a means to a predetermined end.  As internal and external conditions change, the forecast can be adjusted accordingly.

Examples of "Management" Activities

Requesting input from each business division regarding future developments that will impact their operations
Seeking input from each business division regarding the accuracy and completeness of their data
Building consensus for the forecasting activity as it occurs, rather than presenting it as a "done deal".
Holding regular forecast update meetings with input from all divisions.
Emphasizing the neutrality of the forecast.
Educating staff regarding the tools and techniques used to develop the forecast.