Generically speaking, forecasting is the process of estimation in unknown situations. In a supply chain context, it centers on (customer) demand planning, which is a business-planning process used by sales teams to develop demand forecasts as input to service-planning processes, production, inventory planning, revenue and cash-flow planning.
Supply chain demand forecasting is a challenging problem, but it is generally possible to create good forecasts through an intelligent combination of judgmental and statistical methodologies. Statistical models based on solid historical data that take into account current market intelligence and global economic trends provide a good baseline, which, when tweaked by a seasoned expert who has intelligent insight into expected consumer response to a "one-time" event or environmental change can often deliver extremely reliable results on a continuous basis.
For more information on supply chain forecasting, see the following posts.