The Summer 2010 issue of Foresight is now available. Here is Editor Len Tashman’s preview:
For so many years, we forecasters have developed and refined models for demand forecasts – forecasts for product and item sales, orders, shipments – without paying adequate attention to the details of how these forecasts are to be used in the supply chain. Even more basically, we have been casual in assessing how reliably the data histories represent the true demand for the organization’s output. Our special feature, Forecasting for the Supply Chain, addresses both of these key issues, and then some.
The feature section leads off with Mike Gilliland’s article, “Defining ‘Demand’ for Demand Forecasting.” Mike dispels the conventional wisdom that Orders and Shipments set upper and lower bounds for true demand. He emphasizes the need for forecasts that reflect the organization’s best guess at what is really going to happen after taking supply limitations into consideration.
John Boylan, Foresight’s Supply Chain Forecasting Editor, presents the challenge of “Choosing Levels of Aggregation for Supply-chain Forecasts.” John compares bottom-up, top-down, and hybrid approaches to forecasting a product hierarchy.
John teams up with Mohammad Ali to analyze “The Value of Forecast Information Sharing (FIS) in Supply Chains.” They show how the sharing of forecasts can upgrade the collaborative process, resulting in mitigation of “bullwhip” effects and achieving savings in inventory costs.
In our section on Forecast Model Building, Roy Batchelor examines the calculation of “Worst-Case Scenarios in Forecasting: How Bad Can Things Get?” Roy reveals that conventional models are not set up to tell us about the impact of extreme events—hence, their worst that actually plays out in the future. He illustrates what needs to be done to more accurately determine how bad things can get.
While 2010 is seeing many tight elections for national and local offices, our World of Forecasting section looks ahead to the U.S. presidential election in 2012. Allan Lichtman tells us that his “Keys to the White House” model is already signalling a victor and that only dramatic intervening events could change this prediction.
Roy Pearson devotes his Forecasting Intelligence Column to a commentary and extension of Adam Gordon’s Spring 2010 feature article, “A DEFT Approach to Trend Forecasting.” In “Looking Under the Hood of That Trend,” Roy finds that the forecaster’s tool kit contains related approaches for critically evaluating trends, including decomposition by causal forces, rulebased forecasting, and trend-impact analysis.
Our book review is a look ahead at mid-21st-century demographics in the USA. Reviewer Ira Sohn examines Joel Kotkin’s The Next Hundred Million: America in 2050 and finds it a worthy discussion of the likely changes in American living patterns across the next four decades.
We wrap up this issue with an excerpt from a letter to the editor that raises the question, “Should You Report Forecast Error or Forecast Accuracy?” This is an extension of our previous two surveys on forecast accuracy reporting (in the Summer 2008 and Winter 2009 issues) and will be the subject of a new survey this autumn.
Len and I are soliciting comments on this issue for a future article in Foresight. Please post any comments publicly here on The BFD, or privately to me (mike.gilliland@sas.com) or Len (lentashman@cs.com) directly. Next week I'll be posting my (highly favorable) comments on a compelling argument to use Max (Forecast,Actual) in the denominator of performance calculations, given by David Hawitt in a letter to the editor.
Forecasting 101 Webinar -- Now available for on-demand review
We drew nearly 400 live attendees to last week's Forecasting 101 Webinar , and it is now available for free on-demand review. The focus is on forecast value added (FVA) analysis, so this is most suitable as an introduction for those new to the FVA approach.
