Sales Forecasting Methods
This course is led by Régis Bourbonnais, Professor at Paris-Dauphine and recognised expert in sales forecasting
Eurodecision is a French state-approved occupational training organisation (accreditation no. 11780765678).
Course objectives
During this seminar you will discover the methods available for making sales forecasts, and will understand the limits and uses of a forecasting system in order for you to set up, within your company, the forecasting technique best suited to your business sector requirements.
Following this course, you will be able to:
- understand the calculation algorithms appropriate for the sector and their expected accuracy levels
- Classify items according to the required forecasting method
- know how to correct abnormal values in a sales history
- configure forecasting software on two important points: the level of application for the calculation algorithm and the rules for breaking down forecasts
- use forecasting methodologies suitable to your products
- define alert tools and dashboards
- know the objective to be achieved concerning the forecast quality
- define the organisation (validation of the forecast in-house by the Supply Chain through a collaborative process)
Intended audience
All companies analysing the feasibility of sales forecasts based on statistical tools, or wishing to improve their forecasting processes. The course is specifically intended for those involved in the optimisation of the supply chain, production, purchasing, and marketing.
Teaching methods used
Each participant uses a computer to solve the case studies in an Excel worksheet. Trainees may bring sales histories for class analysis during the course. Each participant will receive a copy of the book entitled "Sales Forecasting".
Detailed outline
Day One
The first day presents the various forecasting methods and illustrates them extensively through case studies in Excel. The objective of the first day is to:
- show the simplicity of implementing the main methods
- thoroughly understand their operation
- grasp their limits of use
Morning 9:00- 12:45
I. Diversity of Forecasting Problems
II. How to Analyse a Sales History
- Data presentation
- Moving averages
- How to detect and correct abnormal values
- Breaking down chronological series
- Trends
III. How to Analyse Seasonal Variation and Why
- Calculating seasonal coefficients
- Fixed and additive seasonal coefficients
- Multiplicative and sliding seasonal coefficients
- Test for detecting a seasonal variation
- Calculating the seasonal variation by item category
- Seasonal variation and promotions
- Forecasting by extrapolation of trends and seasonal variation
Afternoon 14:00 - 17:30
IV. The Most Widely-Used Forecasting Method: Exponential Smoothing
- General principles
- Calculation methods
- First-order exponential smoothing: stationary series
- Second-order exponential smoothing: linear trends
- Triple smoothing: quadratic trends
- Holt's Model: different adaptability between means and trends
- Holt-Winters Model: including seasonal variation
- Optimisation of the smoothness ratio
- Self-regulation of a smoothing system
- How to factor in promotions
- Forecasting for items with erratic demand
V. A More Original Method: The Explanatory Approach
- The concept of correlation
- Building an explanatory model
- Including explanatory factors
- The statistical interpretation of a model
- The logic of the upstream sector
- The advanced indicator model
- How to select explanatory factors
- Example: Forecasting visits to the Eiffel Tower
- Calculating elasticity (sales/prices, sales/advertising, etc.)
VI. Test
Day Two
Morning 9:00- 12:45
VII. Measuring the Quality of a Forecast
- Indicators for a given item
- How to judge the overall quality of the forecast
- Sample dashboards and alert tools
- How to define a forecast quality objective
VIII. Organising the Information System and Choosing a Forecasting Technique
- Aggregate level to use to calculate a forecast
- How to ensure consistency between monthly and weekly forecasts
- Integrating the order book when preparing a forecast
- How to calculate daily forecasts
Afternoon 14:00 - 17:000
IV. Choosing an IT Solution
- Overview of the main forecasting software packages
- How to choose a package
X. Sample Applications by Sector
Based on participants' requests, the trainer can analyse a few of the sales histories provided by attendees or can provide examples of forecasting systems used in different sectors: agri-foodstuffs, textile-clothing, electrical equipment, mass distribution, distribution of automotive supplies, etc.
XI. Reasons for Making Sales Forecasts
- Flexibility and forecasts
- Forecast errors and safety stocks
XII. In-Depth Round Table Discussion of Sales Forecasting Problems
XIII. Conclusion – Course Assessment
Trainer
This course is led by Régis Bourbonnais, Lecturer at Paris-Dauphine University where he teaches Econometrics and forecasting techniques. Régis Bourbonnais is Director of the Masters Supply Chain Programme: Management and the Economy of Networks. He has both academic and business experience in sales forecasting.
He has co-ordinated the design and implementation of several forecasting systems in various industries: fine chemistry, construction, agri-foodstuffs, mass distribution, automobile, etc.
Régis is the author of a book entitled "Sales Forecasting", as well as "Econometrics", (Editions Dunod, 7th edit. 2009), "How to optimise procurement" in collaboration with Philippe Vallin (Editions Economica, 2nd edit. 2006) and "Analysis of Time Series" (Editions Dunod, 2nd edit. 2008).
This customised course is held on your company premises.
Eurodecision also offers a slightly modified version of this course to participants from different companies, with presentation of examples adapted to your business, modified practice assignments, etc.