- Demand forecasting is broken into several different types of forecasting ranging from statistical forecasting to sales forecasting.
- There are also many methods of performing demand forecasting that is covered in this article.
The descriptions of demand planning or forecasting seem to try to get into a lot of detail about the topic regarding specifying how demand planning is accomplished, whereas the term, is highly general.
Wikipedia does not have a definition of demand planning, at the time of this writing. The way they define demand forecasting
“Demand forecasting is the activity of estimating the quantity of a product or service that consumers will purchase.”
..is synonymous with demand planning.
Let us see if we can do a bit better.
I recently found this interesting story on the origin of the work forecasting. David Orrell covers the first forecasting office and how it was setup in Britain.
“The first meteorological office was setup in Britain in 1854. The Met. Office as it became to be known as headed by Admiral Robert Fitzroy–he who captained the Beagle and taken Darwin around the world. The ex-navy man saw that weather forecasting had the potential to save lives by warning mariners of storms, like the one that destroyed nearly thirty French and British vessels in the Crimean a month before his appointment. A network of forty weather stations was setup around the UK, and weather reports were published in London newspapers. In France, the chemist and accountant Antoine Lavoisier funded a chain of observation stations before being sent to the guillotine for his unpopular taxation activities. FitzRoy’s efforts were also not well received by the establishment. At the time, weather prediction was something practiced by astrologers, and it was not seen as a fit subject for scientific pursuits. The popular press enjoyed comparing the Met. Office’s inaccurate predictions to those from astrological sources, such as Zadkiel’s Almanac. The mainstream scientists saw all of this as a threat to their reputation. Fitzroy tried to blunt the comparison to astrology by avoiding loaded words like “prediction”; instead, he invented a new word of his own: forecast.” – David Orrell
Demand Forecasting Definition and What is Forecasting Versus Demand Planning?
- Demand Forecasting: This is the development of a projection as to future values. The method is not declared in the definition of forecasting. It is merely the determination of future values.
- What is Forecasting Encompassing?: Forecasting is a general term. There is weather forecasting, futures forecasting, etc.. However, demand planning applies specifically to supply chain forecasting. It encompasses any method that is employed to predict demand or sales.
- What is Forecasting Lead Time?: Forecasting must be completed before the event occurs, or before the one can actively do something about the forecasted event. In supply chain forecasting (or demand planning), the forecast must be finalized before the total lead time to bring the item into the supply network, or to provide it to the customer is complete.
Methods of Demand Forecasting
- Statistical Forecasting: Statistical forecasting is simply one type of demand planning. Demand planning is used interchangeably with the term forecasting, but demand planning is also a child term to forecasting (although in practical usage, it makes sense to use them interchangeably as the domain of the forecasting is assumed). However, the word forecasting, much like the word prediction is the broadest word in this space. Weather forecasters perform forecasting but do not perform demand planning. The terms demand planning, and forecasting becomes technically synonymous when the phrase “supply chain” is added to “forecasting.”
- Sales Forecasting: Sales forecasting is the process by which the sales group or sales department generates their forecast.
Supply Chain Forecasting Definition
Supply chain forecasting is also known as demand planning. While supply chain forecasting is its discipline, forecasting across areas (except for weather forecasting, which has its methodology based upon causal modeling and extensive weather computer modeling), often use the same forecasting methods.) Each area or domain of supply chain planning ends with “planning.” so “supply planning,” “production planning,” etc..
Demand Planning is literally that, planning for the demand for products that the company will receive in the future.
The Major Forecasting Approaches
Forecasting can be performed statistically, which means using sampling from the population (that is the demand history data points), or based upon judgment methods.
Judgment methods can adjust a statistical forecast, or they can simply be purely judgment based. The former approach is called manually changing the statistical forecast, and the second, when combined with other individuals, is called consensus-based forecasting. All of these methods are important to any forecasting definition.
A forecasting method is the category of forecasting which is being applied.
- Time series forecasting (such as exponential smoothing, and moving average)
- Consensus-based forecasting (the Delphi method, and prediction markets) are also methods.
- Causal forecasting is a method of forecasting where independent variables are selected in order to predict dependent variables.
Forecasting methods can be confused with a forecasting model, which is most often used to describe the specific mathematical formula used to develop a forecasting in either time series or causal forecasting. The model is differentiated from the forecasting parameters which then control the model in its development of the forecast.
How is a Forecast Method Different from a Forecast Model?
Forecasting models are the specific or exact procedure that was used to create the forecast. A three-period moving average is a model while a two-period moving average is another model, both of which are within the time series forecasting method.
Forecasting models are subordinate to, or children of parent forecasting methods.
Forecasting models are strongly associated with statistical forecasting, and not consensus-based forecasting or collaborative forecasting.
How is Forecasting Performed?
Forecasting is performed in software. Sometimes the software generates the forecast, and in other situations, the software serves to accumulate forecast input from individuals. The concept in statistical forecasting is that the application produces a forecast automatically, which is then adjusted using domain expertise (insight provided the individual) and that this results in a forecast improvement over a real statistically generated output.
Manual Intervention into the Forecast
There is a debate in the research literature and among practitioners as to how much manual intervention improves a statistical forecast, and the vast majority of companies themselves do not know if their manual adjustments do improve the forecast, as this article explains. On the other side of the continuum is consensus-based forecasting methods. Here the forecast is created based on the educated guesses of those who provide their forecasting input.
Consensus-based forecasting relies upon a variety of individuals to provide a forecast information. While companies frequently discuss using more consensus-based forecasting, in fact, its use is often limited to sales forecasts, new product introductions, and S&OP forecasting.
The category of software which explicitly addresses consensus-based forecasting is smaller than that address statistical forecasting.
Major Areas of Interest for Forecasting
The topics listed below are a sampling of some of the key areas of discussion and interest for supply chain forecasting. This includes the following
- The Data Loading and Management Aspect
- Forecasting Methods and Models
- Best Fit
- Management of the Removal of Outliers
- Historical Removal
- Error Management
Aggregation is used both for visibility, for grouping items for adjustment (up by 2%, or down by 3%), and for top-down forecasting, which allocates a forecast derived from a higher level to the individual product locations combinations. This area is currently one of the most important in supply chain forecasting, and even the largest companies often have a problem performing aggregation correctly.
The Data Loading and Management Aspect
Statistical forecasting systems use a lot of data. Of all the supply chain planning applications, demand planning applications have the most data because they often have to keep sales history for at least three years. Three years is necessary for seasonal pickup variations. If a seasonal variation is discovered, then usually a seasonal forecasting model can be applied. This also leads directly to the next topic:
Forecasting Methods and Models
The method is broadly the forecasting approach taken. A forecasting method would be whether a statistical approach or consensus-based forecasting approach is used. A model, on the other hand, is a particular technique that is used to produce the forecast. For instance, a trend model, or a level model, or Crostons, are all accurate forecasting models.
Best Fit Model Selection
In statistical prediction, there are several ways to find a model to fit. One is to choose a method manually, and another is to allow the system to choose the model. The second method is called best-fit functionality and exists in most supply chain statistical forecasting systems (although there are significant differences in how well the procedure works, and how easy it is to run. While at one point, there was an idea that best fit could be used exclusively to select the best model, practical historical experience shows this not to be true. Best fit procedures are now understood to work with a human model selection.
Management of the Removal of Outliers
In statistical forecasting, sometimes demand history is considered unrepresentative. Sometimes this is because the data is incorrect, or because the historical event is considered to be one-time phenomena. In this case, the concept presented is that by removing the outliers the forecast will improve. This is a very controversial area of forecasting and is covered in the post.
Historical Removal is less used as a method than outlier removal but is more scientific. It works backward from the earliest demand history points, and removes them and creates a forecast. It does this until the subsequent removal of periods no long improves the forecast. This is covered in the SCM Focus book on forecasting.
A major topic of interest in prediction is how the error is measured and managed. An error measurement is important in being able to determine how much the forecast accuracy can be improved. Surprisingly, many companies do not know their forecast error at different levels or in different contexts. Forecast error quotations that are used inside of companies are notoriously unreliable. Forecast error management is its specific topic, which is described in more detail in this article.
What is forecasting is a very good question to ask. Something that is called forecasting, such as demand sensing is not forecasting.
Search Our Other Forecasting Basics Content
Statistical forecasting is covered in the following book.
Forecasting Software Book
Providing A Better Understanding of Forecasting Software
This book explains the critical aspects of supply chain forecasting. The book is designed to allow the reader to get more out of their current forecasting system, as well as explain some of the best functionality in forecasting, which may not be resident in the reader’s current system, but how they can be accessed at low-cost.
The book breaks down what is often taught as a complex subject into simple terms and provides information that can be immediately put to use by practitioners. One of the only books to have a variety of supply chain forecasting vendors showcased.
Getting the Leading Edge
The book also provides the reader with a look into the forefront of forecasting. Several concepts that are covered, while currently available in forecasting software, have yet to be widely implemented or even written about. The book moves smoothly between ideas to screen shots and descriptions of how the filters are configured and used. This provides the reader with some of the most intriguing areas of functionality within a variety of applications.
- Chapter 1: Introduction
- Chapter 2: Where Forecasting Fits Within the Supply Chain Planning Footprint
- Chapter 3: Statistical Forecasting Explained
- Chapter 4: Why Attributes-based Forecasting is the Future of Statistical Forecasting
- Chapter 5: The Statistical Forecasting Data Layer
- Chapter 6: Removing Demand History and Outliers
- Chapter 7: Consensus-based Forecasting Explained
- Chapter 8: Collaborative Forecasting Explained
- Chapter 9: Bias Removal
- Chapter 10: Effective Forecast Error Management
- Chapter 11: Lifecycle Planning
- Chapter 12: Forecastable Versus Unforecastable Products
- Chapter 13: Why Companies Select the Wrong Forecasting Software
- Chapter 14: Conclusion
- Appendix A:
- Appendix B: Forecast Locking
- Appendix C: The Lewandowski Algorithm.
Sales forecasting is covered in the following book.
Sales Forecasting Book
The Problems with Combining Forecasts
In most companies, the statistical and sales forecast are poorly integrated, and in fact, most companies do not know how to combine them. Strange questions are often asked such as “does the final forecast match the sales forecast?” without appropriate consideration to the accuracy of each input.
Effectively combining statistical and sales forecasting requires determining which input to the forecast have the most “right” to be represented – which comes down to those that best improve forecast accuracy.
Is Everyone Focused on Forecast Accuracy?
Statistical forecasts and sales forecasts come from different parts of the company, parts that have very different incentives. Forecast accuracy is not always on the top of the agenda for all parties involved in forecasting.
By reading this book you will:
- See the common misunderstandings that undermine being able to combine these different forecast types.
- Learn how to effectively measure the accuracy of the various inputs to the forecast.
- Learn how the concept of Forecast Value Add plays into the method of combining the two forecast types.
- Learn how to effectively run competitions between the best-fit statistical forecast, homegrown statistical models, the sales forecast, the consensus forecast, and how to find the winning approach per forecasted item.
- Learn how CRM supports (or does not support) the sales forecasting process.
- Learn the importance of the quality of statistical forecast in improving the creation and use of the sales forecast.
- Gain an understanding of both the business and the software perspective on how to combine statistical and sales forecasting.
- Chapter 1: Introduction
- Chapter 2 Where Demand Planning Fits within the Supply Chain Planning Footprint
- Chapter 3: The Common Problems with Statistical Forecasting
- Chapter 4: Introduction to Best Fit Forecasting
- Chapter 5: Comparing Best Fit to Home Grown Statistical Forecasting Methods
- Chapter 6: Sales Forecasting
- Chapter 7: Sales Forecasting and CRM
- Chapter 8: Conclusion