Operations Management / Supply Chain Management

Module 03: Forecasting

Forecasting is an integral component of Demand Management.  Forecasts are used to develop and target actions to support expected business opportunities or potential changes in the marketplace.  Forecasts are dynamic and change with these different actions of the firm or actions beyond the firm’s influence (natural disasters, changes in demographics, etc.)

Forecasting is defined by APICS as “the process of making, checking, correcting and using forecasts.  It also includes determination of planning horizon.”  A forecast is defined as “ An estimate of future demand.  A forecast can be constructed using quantitative methods, qualitative methods or a combination of methods.  Forecasts can be based on intrinsic (internal) or extrinsic (external) factors.  Various forecasting techniques attempt to predict one or more of the four components of demand: cyclical, random, seasonal and trend.”  APICS Dictionary 13th Ed.


Forecasts are essential to any business.  Forecast data collection, methodology and use depends on the planning process involved, business characteristics and specific needs of the organization, it’s customers and suppliers.  We will look at different requirements and approaches to forecasting the following section.  However, we must first reflect on some basic characteristics of forecasts to put things into perspective.

  • Forecasts are rarely 100% accurate over-time:  In fact, if the forecast is 100% accurate, it is probably an accident.  There is variation in every process and this impacts predictability of results.  Very dynamic demand results in very erratic activity and high forecast errors.  More stable demand is easier to forecast, with better accuracy, but errors still result.
  • Forecasts should include an estimate of error:  Since forecasts are not 100% correct, it is extremely important to capture the degree of variation.  A forecast that is +/- 10% over time is very predictable where as one that is +/- 200% over time is not.  Business functions, in using forecast data, must understand this variation in order to create optimum supply and capacity utilization plans.
  • Forecasts are more accurate for product groups and families: It is much easier to create a fairly accurate forecast of total product sales for a business than it is to create one for an individual product family.  It is further extremely difficult to create a relatively accurate forecast for an individual item at the individual customer level.  So, what do we ask the sales team to do?  We ask for “accurate” forecasts for each one of their customers – the worst-case scenario.  What about make-to-order products?  A customer order is a type of forecast isn’t it?  Can a customer change the quantity of an order?  Can a customer cancel an order or add one or more extra line items?  Can a customer change the desired delivery date?  All of these add a degree of uncertainty to even real-live orders.  In a make-to-order environment forecasts are used for planning materials, people, and machines.
  • Forecasts are more accurate for nearer periods of time:  Think of the weather – it is much easier to give an accurate forecast of what will happen this afternoon than it is for the same day two years from now.  Will it snow today?  Hmm, let me check out the window.  Will it snow on January 2, 2018?  Who knows!!   This means that it is more accurate to forecast what will sell next week than it is to forecast sales six months out.  Many times this is used as an excuse for even generating a forecast.  People will say, “We do not have any idea what the market will be like in six months.”  However, for products with long lead times you must create a forecast in order to have expected supplier responsiveness and adequate long-term capacity.

Forecasting is essential for the following purposes: to plan for the future by reducing uncertainty; to anticipate and manage change; to increase communication / integration of planning teams; to anticipate inventory and capacity demands and manage lead times; to project costs of operations into budgeting processes; and to improve competitiveness and productivity through decreased costs and improved delivery and responsiveness to customer needs.