Martec International Education Retail course material, since 1987.

Buying and Merchandising EL75

Advanced Demand Forecasting

The forecast already exists here, and the work is making it less wrong: price elasticity, halo and cannibalization, smoothing, seasonality, ARIMA.

Class (module) EL75: one unit you can assign inside your course.

Get free evaluator access

Class specification

2 h 00 minstudy time
102screens across 7 sections
19scored questions in the mastery test
70%to pass, with unlimited retakes
Module summaryIncludes a 23-page module summary
Class codeEL75
Course groupsBuying and Merchandising, Supply Chain Management, Consumer Goods Supply Chain Management, Direct to Consumer
Learning pathsOn no published learning path.
CertificatePassing this class's mastery test earns a certificate for this class, which the learner can print.
LevelAdvanced
PrerequisitesThose studying the Advanced class should be familiar with the content covered in the fundamentals class or should study it as a pre-requisite.
Narration82.9 minutes, 98 of 98 files captioned
Assessment19 scored questions (10 multiple choice, 3 multiple response and 6 of other types), 70 percent to pass, unlimited retakes
Practice questions15, not scored
Answer keyYes
Downloads included4 files
Professional credit2.2 CPE credits, NASBA QAS Self Study
Included with the classEach class includes 12 months' access to the Retail and Consumer Goods Industry Knowledge Bank.
DeliveryRuns in a browser from a link. No install, no LMS integration needed.

Refining a forecast that already runs is where this module starts. Something upstream has produced a number and the fundamentals class has explained how; the question here is how to make it less wrong. Price elasticity, halo and cannibalization, then exponential smoothing and the factors that drive it; then trend modifiers, seasonality factors and the ARIMA family; then profiles, causal methods and the strategy question of which model belongs where, before the output is handed on to the people who order against it. Assign it knowing that this is a calculating class rather than a descriptive one.

I

What the material actually looks like

The heaviest screen here is a chain whose stores raise their own replenishment orders, and the smoothing math sits well down the list.

The fullest screens in EL75, named as they appear in the class.
ScreenWhat it shows
Category Management Where Stores Submit Own Replenishment Ordersforecasting in a category management chain where the stores, not the center, raise their own replenishment orders
Operational Challenges in Causal Forecastingwhat gets in the way when causal methods meet a real business
HALO and Cannibalization Effectsone product's demand moving another's, set out as effects to allow for
Demand Not Salesthe distinction returning inside the strategy section rather than as a first lesson
Forecasting Profilesprofiles as an instrument, with the demand patterns they suit
Adjusting Smoothing Factors Based on Length of Sales Experiencethe smoothing factor tied to how much sales history a product has
Limitation of Exponential Smoothing in Trending Datawhere the method breaks down, shown against trending demand

The screen that decides whether your class can take this module is Adjusting Smoothing Factors Based on Length of Sales Experience: The math in this module concentrates in exponential smoothing and the trend and seasonality modifiers, and this is the screen where the parameter stops being a symbol and becomes a value somebody has to choose. Read it with the screens around it and you will know within a minute whether your class can take the level.

The class at a glance

EL75: 102 screens in 8 sections, the mastery test last 1 6 2 16 3 20 4 18 5 11 6 3 T 21 + 6

EL75 holds 102 screens in the order the module presents them: section 1, Introduction, 6 screens; the menu screen; section 2, Advanced Forecasting Techniques, 16 screens; section 3, Forecast Refinement: Trend and Seasonality, 20 screens; section 4, Advanced Forecast Profiles, 18 screens; section 5, Forecasting Strategy and Application, 11 screens; section 6, Summary, 3 screens; the mastery test of 21 screens, last; 6 screens the menu does not list.

  1. 1 Introduction, 6 screens
  2. M Menu, 1 screen (navigation)
  3. 2 Advanced Forecasting Techniques, 16 screens
  4. 3 Forecast Refinement: Trend and Seasonality, 20 screens
  5. 4 Advanced Forecast Profiles, 18 screens
  6. 5 Forecasting Strategy and Application, 11 screens
  7. 6 Summary, 3 screens
  8. T Mastery Test, 21 screens
  9. + Screens outside the menu, 6 screens the menu does not list
Module anatomy, EL75: each band is one section of the module as its menu lists them, in order, sized by its screen count; each tick is one screen; the band marked T is the mastery test, which comes last.

Evaluator access opens the class itself, including every screen this figure counts.

II

What it covers

Introduction 6 screens

Six screens of setup. The objectives screen is the one worth reading, because Martec puts AI and machine learning into the module's stated aims, and the causal forecasting screens later make good on that promise.

Advanced Forecasting Techniques 16 screens

Price elasticity and the halo and cannibalization effects open it, and then the arithmetic starts. Exponential smoothing is defined, the smoothing factor is shown driving how responsive a forecast is, and the method is run twice, against demand that drifts slowly and against demand that shifts hard. Two further screens set the factor by length of sales history and by the size of the deviation, and the section ends on where smoothing stops working, which is data with a trend in it.

Forecast Refinement: Trend and Seasonality 20 screens

Twenty screens, and the arithmetic thickens as they go. A trend modifier is defined, calculated and then qualified; seasonality factors are applied and then adjusted; and the last stretch climbs to ARIMA and SARIMA, then SARIMAX and the business of selecting and managing its X variables, closing on backtesting.

Advanced Forecast Profiles 18 screens

Profiles first: what a profile forecasts and which demand patterns suit one. Then causal methods, where AI and machine learning enter them, what goes wrong operationally, and what the benefits and limitations actually are. The section widens at the end to where in the business the forecast is made: warehouse and dispatch level, e-commerce and omni-channel operations, gross demand against net demand, manufacturer direct to consumer, wholesale, and the category management case where stores raise their own replenishment orders.

Forecasting Strategy and Application 11 screens

Method selection treated as a decision, not a calculation. Forecast accuracy, the demand and sales distinction restated at this level, choosing the model that fits, and the considerations that sit around the choice. It ends on strategy inside category management, on the handover into allocation and replenishment, and on forecasting across the supply chain.

Summary 3 screens

Three screens: the module summary and a recap of the core principles.

Mastery Test 21 screens

The mastery test itself. Its composition is set out in full below, and no question, option or answer from it appears anywhere on this site.

A further 7 screens carry the terms of use, the help and contact panels, the certificate and the pop-out detail screens the menu does not list separately. They are counted in the 102 screen total in the specification above.

Learning objectives

Martec states the objectives as: By the end of the program, students will be able to do the following:

  • Explain more advanced forecasting models, sometimes assisted by AI or machine learning, and identify when they are most useful
  • Apply forecasting methods for direct to consumer operations, wholesale operations and to forecasting warehouse dispatches
  • Describe ways to improve forecasting strategy.

III

What a student finishes with

What a student has done by the end is mostly calculation, and the calculations are the specific ones a forecasting team argues about. A smoothing factor chosen, then chosen again for a shorter sales history and a larger deviation. A trend modifier calculated and then qualified. Seasonality factors applied and then adjusted. Backtesting run against the result. Above the arithmetic sit the questions the module treats as strategy: which model belongs on which demand pattern, where AI and machine learning add something to a causal method and where they add operational fragility instead, and how gross demand parts company with net demand once the forecast is being made for a warehouse, a wholesale account, a direct to consumer channel, or a store raising its own replenishment orders.

The certificate is issued by Martec. It records that the student passed this class's mastery test, and it claims nothing beyond that. Your institution decides what, if anything, it counts toward.

Passing this class's mastery test earns a certificate for this class, which the learner can print. How the Mastery Credential works.

IV

Where it fits a syllabus

You assign this where a class already has a forecast in front of it and the question has moved on to sharpening it: Martec sets the module at advanced level in Buying and Merchandising, states two hours, and asks for the fundamentals class first. A final year or graduate elective in supply chain, merchandising or retail analytics is where it lands in a university. The people it addresses already own a forecast: senior merchandisers, senior buyers, their managers, supply chain managers, and the consultants and system vendors who build the tools those people use.

Scope

It is an assignable unit inside a course you already teach: one class, its own mastery test, and a completion record. Your course, your reading list and your assessment stay yours.

Syllabus line, ready to adapt

EL75 Advanced Demand Forecasting, Martec International. Online class (module), about 2 hours of study, with a mastery test of 19 questions; a certificate is issued on passing at 70 percent. Assigned as a supplementary unit inside this course.

Disclosure line for course materials

Advanced Demand Forecasting, class (module) EL75. Martec International Ltd, Bradford-on-Tone, Taunton, Somerset. Online course material; the copyright line inside the class states its date.

V

How it is assessed

Assessment, EL75
Scored questions19
Question types10 multiple choice, 3 multiple response and 6 of other types
ScoringScored out of 100. The threshold is applied to that percentage.
Pass threshold70 percent
Answer options shuffled15 of 19
Question orderFixed. The same questions are asked in the same order on every attempt.
Question bankThere is no question bank and no random draw. Every authored test item is served on every attempt.
Time limitNone.
AttemptsUnlimited. Retakes reset the whole test.
NavigationFree movement throughout. No linear lock, no per-screen gate, and no minimum view requirement before the mastery test opens.
Unscored practice questions15, placed through the module

The practice questions appear through the module as the student works, and none of them is scored. Nothing a student does before the mastery test counts toward the mark.

VI

Assessment integrity: what this test measures

Nineteen questions are scored. Six of them are of unrecorded type, so the mark cannot be read back into what sort of knowing it sampled. That matters more in a calculating module like this one than in a descriptive one, because nineteen recognitions spread across a hundred and two screens do not separate a student who can define a seasonality factor from one who can choose a smoothing factor for a product with two months of history, and those are different capabilities taught on different screens. The fifteen practice questions a student meets on the way leave no record. Give each student a short sales history and ask for a smoothing factor with the reasoning behind it, which is the evidence the score does not carry.

VII

Using this alongside AI assistants

A student with two months of sales history asks an assistant for a smoothing factor and can get a number back with a paragraph of justification under it that never mentions how much history there is. Two screens in this module exist for that exact question: 'Adjusting Smoothing Factors Based on Length of Sales Experience' ties the value to how much sales history the product has, and 'Smoothing Factors Based on Deviation Magnitude' ties the same value to the size of the deviation being reacted to. Unless the student supplies both inputs, an assistant is working without them. 'Limitation of Exponential Smoothing in Trending Data' compounds the problem, because the failure it describes is silent: the arithmetic completes, the output looks like a forecast, and it is wrong in the same direction every week. Reproducing a method correctly does not show whether it was the right one to reach for, and that choice is the judgment to grade. The vocabulary is where the same assistant is genuinely useful: it explains exponential smoothing, defines a seasonality factor and sets out what SARIMAX adds over SARIMA reliably, which is why that material is not worth grading.

VIII

What comes with it

Every file below sits in the module's resources tab, reachable from the first screen. None of them is gated. Four documents and no data set, which for a module that asks a student to pick a smoothing factor and calculate a trend modifier means the numbers to practice on are not in the class.

Downloadable files included with EL75
FileFormatWhat it is for
Advanced Demand Forecasting SummaryPDFThe module reduced to a handout.
Knowledge Bank Links Advanced Demand ForecastingDOCXCollected external reading.
CPE Credits for Advanced Demand ForecastingPDFThe claim details for US CPAs.
Guide on navigating Martec e-learning coursesPDFHow the player works, issued once rather than per module.

IX

What students need

A browser and a link. Nothing installs, no account is created on your systems, and there is no LMS integration to arrange: students open the link, register with a name and email, work through the class, and the mastery test marks itself.

The material is drawn on a fixed 1220 by 658 stage and scales as a whole to fit the screen, so it does not reflow. Text in the denser diagrams cannot be read on a phone held upright. Use a tablet, a laptop, or a phone held sideways with pinch zoom.

All 98 of its narration files carry a caption track.

X

Scope

This class has a defined edge: for each boundary below, either another class covers the neighboring material, or the line names who this class is written for.

  • The introductory method set is not retaught. Moving averages, ABC analysis and the basic replenishment parameters belong to EL74, one level below this on the same subject and the stated prerequisite. The demand and sales distinction does reappear here, but inside the strategy section rather than as a first lesson.
  • The forecast is where it stops. This module covers how the output feeds allocation and replenishment, but the replenishment models themselves, MIN and MAX, sell one get one, fixed and dynamic model stock, are taught in EL77, in the same group.

XII

How each reader gets access

XIII

See it for yourself

The way to judge whether this class suits your course is to request free evaluator access from Brian Hume, who issues a code.

Share by email