Contact us for more details and registration for this workshop.

The course

The workshop is specifically designed to address key issues in the minds of those with management and executive responsibilities and those who work alongside them. The objective of this course it to provide business professionals with a set of data and forecasting procedures and to demonstrate with illustrative examples how such procedures are used in data preparation and model building and forecasting. The course focuses on conceptual framework of models, and thus providing a clear vision of what each model represents, it’s underlying assumptions and what its model statistics imply. Such understanding helps forecasters and business professionals not only in evaluating models but also in selecting the right one for preparing forecasts. The limitations of each of the models are pointed out throughout the course as well. The main goal is to show the application of models in forecasting and the importance of quality of the data used to generate the forecasts. The most important thing for practicing forecasters and their management is to understand the concept, how a model works, how to evaluate it, and how to go about to improve upon it. The requirements of mathematics are kept to a minimum, as most of the forecasting software systems do this on their own.

There are many software packages in the market which have a built in expert system that automatically selects the ‘best’ model (aka ‘best fit’) and then provides the resulting forecasts. However, with the expert system comes the danger of ‘black box’ forecasting. The model selected by software may or may not be the best one. To avoid black box model building, it is important for forecasters to understand what goes behind each model, as well as how and why a given model is chosen. On this account, we will discuss in detail what goes behind each model. We will use real data examples throughout the course to give the live experience from variety of industry sectors. Our goal is to discuss those topics, which have a real application in business and are easy to apply.

This course provides a combination of best practices in data management, segmentation and cleansing approaches required for sound statistical forecasting and forecast presentation … blending years of experience as first-line practitioners and managers with years of management consulting practice as well.

Special segment of the workshop is dealing with application of different forecasting methods to make better business forecasts during times of high uncertainty including trade wars and increased tariffs.

IMPORTANT:

To translate and expedite classroom learning to real-life applications, participants are HIGHLY ENCOURAGED bringing their own laptops.
MS Excel DATA ANALYSIS and SOLVER Add-Ins MUST BE ACTIVATED.

BRING YOUR OWN DATA: we have allocated time to work with individuals on their specific need.

How this program can help companies in tough times like this?

In the face of fast-paced globalization of economies and increasingly intensified competition, rarely a day goes by without an announcement of a new product launch, an existing one discontinued or modified, a set of new features added to make products more effective, convenient and innovative. At the same time, when customers are demanding faster delivery and quicker service, organizations often lack the resources and the expertise in order to effectively manage a leaner supply chain required to meet these challenges. As the economic uncertainties unfold many executives with supply chain and logistics responsibilities are either working in “survival mode” or moving forward and beginning to make their supply chains more responsive, flexible, and efficient.

Making improvements in Demand Management is one of the most critical factors to be able to attain a lean Supply Chain, since a clear correlation exists between High Forecast Error and Missing Market Share, Lost Sales, Dissatisfied Customers, Obsolescence or Wasted Expenditures.

Industry leaders and decision makers need to re-examine the way businesses think about the future. Leadership in Demand Chain Management results in increased customer satisfaction, increased revenue, and increased profit levels on the bottom line. This sets your company apart from the competition.

This workshop also makes specific recommendations for those striving to move up the demand forecasting maturity curve by providing you with new tools that speed up the validation of your data used for forecasting, increasing effectiveness of use of the software you have invested in and significantly increasing productivity of your forecasting staff.

Who should attend?

EDs, Directors, Managers, Analysts, of:

  • Forecasting / Planning
  • New Product Forecasting
  • Supply Chain Management
  • Allocation and Planning
  • Load Forecasting
  • Strategic Planning
  • Demand Management Process
  • Brand Management
  • Promotions Planning
  • Finance
  • Production Planning
  • Merchandising
  • Product Life-Cycle
  • Trade Promotions
  • Retail Collaboration
  • Sales
  • Marketing
  • Market Research
  • Sales Analysis
  • Statistical Modeling


Note: The detailed content of the workshop may be adjusted without prior notice to better reflect singed-up participants needs.

Program Content

Module 1:
Approaching Forecasting
  • Forecasting Where?
  • Goal definition
  • Forecast Horizon and Updating
  • Forecast Use
  • Level of Automation
Data
  • Collection
  • Time Series Components
  • Visualizing Time Series
  • Interactive Visualization
  • Data Quality
  • Pre-processing & Cleansing
Case study and Exercise: Outlier recognition and treatment – Hands-On – Laptop required

A group activity based on a collection of FMCG, Service, B2B and Demographics data sets. The delegates will learn how to deploy basic normal distribution and statistical process control techniques to validate quality of the data before they attempt to generate forecasts using their systems. These techniques can be used with any forecasting software/system.
Performance evaluation
  • Data cleansing
Module 2:
Forecasting Methods Overview
  • How much data to use for different statistical models?
  • Choosing best model for forecasting based on different data patterns and properties
    • What are the model selection criteria?
    • Consideration of decisions for which model is developed
  • Exploring the essential elements of Time Series Models:
    • Inherent assumptions
    • When Time Series work and when it doesn’t work
    • Level, Trend, Seasonality, Cyclicality, and Randomness
Group Exercise: Exponential Smoothing and Averages – Hands-On – Laptop required

Delegates will explore simple forecasting models by building them in MS Excel. These models are common to all forecasting systems in the market. Many forecasters create higher forecast error by simply not understanding the impact of changing the models or their statistical parameters. By building these models in excel, delegates will solidify their knowledge of these critical base concepts.
Group Exercise and Discussion: Seasonal Decomposition – Hands-On – Laptop required

Delegates will analyze sales data to find patterns in past sales that they can use to predict future sales. As they find a pattern, delegates strip it away from the data, changing the shape of the data that remains. Then delegates will look for new patterns and strip them away. Delegates continue to strip away patterns until only unexplained fluctuations remain. Understanding this technique is fundamental to successful statistical forecasting using ANY commercially available software and system.
Group Exercise and Discussion: Triple Exponential Smoothing – Hands-On – Laptop required

Delegates will explore hands-on the most widely used statistical forecasting model. Understanding this model is fundamental to successful statistical forecasting using ANY commercially available software and system.
Cause and Effect Models
  • Exposing to key assumptions of Regression Models
  • When to utilize Regression Models for best forecasting result?
  • Steps in developing regression model for high accuracy forecasting
  • Taking in consideration of effects of holidays and promotional season on causal model forecasting
Case Study and Group Exercise – Hands-On – Laptop required

Advertising Expenditure vs. Market Share: Delegates will explore application of regression analysis as a ‘What-If?’ tool that can be used during their demand review meetings.

Inkjet Photo Paper Inc.: Delegates will develop regression model incorporating external data (sales of inkjet printers) into their forecasting model.

Case Study and Group Exercise: Forecasting in times of trade wars and new import tariffs: Delegates will develop several forecasting models to predict future based on changing business rules and decide on which method is best to use.
Module 3:
Forecasting of promotions
  • Understanding factors in developing promotions forecasts and sources of dad information for promotions forecasting
  • Developing promotional forecasting model that combines statistical time series forecasting with historical sales data and promotional information
  • Managing the process for unplanned and abnormal demand in market
  • Differentiating between seasonal profiles and promotional impacts
  • Separating promotional effects from forecasting baseline
  • Disclosing promotions forecast error
Group Exercise: Incorporating promotions into Statistical Forecast – Hands-On – Laptop required

Delegates will use dummy variables – a multiple regression forecasting technique – to incorporate impact of irregular promotional activity and other one-of-a-kind events that simple forecasting methods cannot capture.
Group Exercise: Regression for Promotional Planning – Hands-On – Laptop required

Delegates will use sales and price data along with dummy variables to develop “what-if’ forecasting model that will aid management in making decisions about the best pricing strategy for upcoming promotions.
Module 4:
New product forecasting
  • What are issues to consider when developing new product forecasts?
  • Understanding quantitative and qualitative methods for new product forecasting:
    • Customer survey
    • Consumer Panel
    • Test Marketing
    • Scanner Panel Data
  • Maximizing new product forecasting success rate and minimizing error rates through specifying new product success and failure factors
  • Early monitoring market response to new product introductions to minimise forecasting errors and potential inventory excess
Case study and Group Exercise: New Product Forecast Template – Hands-On – Laptop required

Delegates will discuss and understand the concepts of modeling forecast for new products that fall into line-extension category and incorporate marketing, sales and supply chain in the entire process.
Module 5:
Forecast error metrics
  • Detecting biased forecast as early as possible through determining forecasting error metrics and sources of error
  • Spotting forecasting problems and processing error remediation to take initial correction
  • Managing demand control processes to formally manage short term fluctuations
  • Measuring accuracy of the forecast, other vital demand information and controlling inbuilt levels of agreed inaccuracy
  • Discovering Forecast Value Add Analysis (FVA) to identify waste and factors that diminish accuracy
  • Monitoring accuracy of forecasting through exception reports in action
  • Leveraging on Error Measures to calculate safety stock level
Case study and Exercise – Hands-On – Laptop required

  • MAPE vs. WMAPE
  • Tracking Signal and Weighted Bias application
Module 6:
Communication and Maintenance
  • Future Forecast
  • Sales and Operations Planning process
  • Consensus Forecast process
Module 7:
Taking it Home
  • Group discussion and closing of workshop.

Registration

Please contact us to register. The number of seats is limited to ensure the quality of the workshop.