Download
business forecasting planning n.
Skip this Video
Loading SlideShow in 5 Seconds..
BUSINESS FORECASTING & PLANNING PowerPoint Presentation
Download Presentation
BUSINESS FORECASTING & PLANNING

BUSINESS FORECASTING & PLANNING

288 Views Download Presentation
Download Presentation

BUSINESS FORECASTING & PLANNING

- - - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript

  1. BUSINESSFORECASTING & PLANNING IBF Online Seminar – Fundamentals of Demand Planning & Forecasting Mark Lawless IBF Senior Consultant mlawless@ibf.org

  2. Biography Mark Lawless is a Senior Consultant for the Institute of Business Forecasting and the founder and Managing Principal in Marlaw Business Advisory Services. He has extensive experience in forecasting, planning, business process development, and business management. Mark has been associated with the Institute of Business Forecasting since its inception years in the 1980’s. He has held a number of C-level positions, including Chief Planning Officer, Chief Financial Officer, and Chief Operating Officer. During his company affiliations in wide range industries, he has been responsible for: • Development of planning and forecasting processes • Development of forecasting models • Selection and implementation of supporting automated systems • Presentation of forecasts and plans to all levels of management and to major investors and analyst groups • Remediation and continuous improvement of forecasting and planning processes and related forecasting models During his association with the Institute of Business Forecasting, he has published articles in the Journal of Business Forecasting and served as an editorial advisor to the publication. He has made a variety of presentations at IBF Conferences on topics of forecasting. He has served as IBF conference chairperson, conference keynote speaker, and moderator for IBF topical groups at IBF conferences. He participated in the development of the IBF Forecaster Certification Program, and has developed and run tutorials to prepare those taking the certification examination. He has been a key participant in the IBF In-House Training Program since its inception. Mark holds an undergraduate degree in Economics, and graduate degrees in Economics, Finance, and Accounting. He is an alumnus of Southern Illinois University (Edwardsville), Washington University (St. Louis), Boston College, and Bentley University. He is a member of Financial Executives International (FEI).

  3. Topic 1 Topic 1 Demand Forecasting & Planning Basics

  4. Topic 1 Demand Forecasting and Demand Planning is a Journey!!

  5. Topic 1 The Journey Determine the Destination Evaluate the Alternative Routes Plan the Trip – Length, Time, Needs, Equipment, etc. Organize and Prepare for Risks and Contingencies Be Prepared – Adapt to Changing Conditions Reach and Explore the Destination

  6. Topic 1

  7. Topic 1 People DATA Process Methods & Models Goals & Objectives Systems Communication & Reports Business Models

  8. Topic 1 How accurate are your demand plans and demand forecasts? How accurate should they be?

  9. Topic 1 Goal: To answer questions like the following regarding demand forecast accuracy…… • What level of accuracy is expected by company management? Is it reasonable? • How do you measure accuracy? • How accurate do your forecasts need to be? • What are the limits of forecast accuracy? • What affects the accuracy (and the error) of DP forecasts? • What are the effects of accuracy or of error? • What steps can be taken to improve forecast accuracy? • What steps can be taken to improve accuracy expectations of users and management?

  10. Topic 1 Difference Between Forecasting and Planning?

  11. Plans are built upon forecasts……… Demand Plan n. a desired outcome at a future time based upon targets and goals Topic 1 Demand Forecastn. an unbiased prediction or estimate of an actual Demand value at a future time

  12. Topic 1 Why be so concerned about about demand forecast accuracy?? • Downstream effects on business planning and business mandagement processes • Impact on important business decisions • Potential impact on business resources and business performance – operational and financial

  13. Topic 1 What is the environment in which the forecast is being created?

  14. Topic 1 Domain Experts DATA Forecast Process Methods & Models Company Goals & Objectives Forecasting Systems Key Business Assumptions Business Structure

  15. Topic 1 Customers & Consumers Industry Economy COMPANY Technology Government & Regulation Competitors

  16. Topic 1 The demand forecast is a foundation element of the demand plan! And other downstream plans…… The demand plan is only as good as the forecast upon which it is based

  17. Topic 1 Demand Forecast Demand Plan Financial Planning & Budgeting Related Planning Processes Sales & Operations Planning (S&OP) Inventory & Customer Service Planning

  18. Demand Forecast Development Structure Topic 1 Internal Information Sources External Information Sources Assumptions Create & Issue Forecasts Obtain Relevant Info. Formulate Problem Data Analysis & Cleaning Methods Evaluation and Testing Method Selection Correct Sources of Error Isolate and Evaluate Error

  19. What is the underlying work flow and business model being assumed? Channels of distribution? Topic 1 CUSTOMER ORDERS MANUFACTURER SHIPMENTS CUSTOMER HQ/BUYER CUSTOMER WAREHOUSE SHOPPING HOUSEHOLD RETAIL STORE CONSUMER TAKEAWAY aka CONSUMPTION SELL-THROUGH

  20. How Good Are the Assumptions of the Demand Forecast? Marketing Programs Sales Programs Pricing Product Relationships Competitor Actions Economic and Industry Environment Topic 1

  21. Who is participating in the forecasting? Bias?? Topic 1 Operations & Supply Chain Marketing & Sales Demand Forecasts & Plans Finance Key Management Seek Reliable, Unbiased, Domain Experts!!

  22. Assess the potential sources of bias…… Risks that people may perceive Natural tendencies and behaviors Expected use of the information Relationship with You and others Incentives and other drivers Topic 1

  23. How do perspectives vary that may create bias? Supply Chain Marketing Sales Finance & Accounting General Management Topic 1

  24. Supply Chain View Demand by item/SKU Inventory Requirements and Costs Material Labor Production efficiency and capacity Topic 1

  25. Marketing View Category/market size Types of consumers/end users Target market characteristics Price trends New product development Competitive factors Seasonal factors Topic 1

  26. Sales View Customer service & satisfaction Customer trends Geographic differences Price trends Competitive factors Topic 1

  27. Finance & Accounting View Accounting Finance Treasury Budgeting Capital Investment Monetization of Business Actions and Plans Shareholder and Lender Relations Business Capitalization Topic 1

  28. General Management View Business expansion Business investments Merger/acquisition transactions Strategic actions Competitive positioning Economic conditions Financial market conditions Business capitalization Topic 1

  29. Natural BIAS of each function depends upon their responsibilities and the expected use of the info…. Topic 1 Function Because DEMAND SUPPLY • Marketing May call high Want idea to go forward • Sales May call high Want to ensure product available for their customer May call low Then can exceed quota if based on forecast • Operations May call high Do not want to be out of stock May call low Do not want to have too much inventory, to “compensate” for Marketing/Sales optimism Assess risk of “missing” demand vs. “missing” supply p. 29

  30. Some products are harder to forecast and plan than others…… Topic 1 And… Some products are not reliably forecastable!! • Products with highly volatile demand • New products • Highly promoted products • Products with many substitute products available • Internal • External • Products with a short life cycle • Products with intermittent demand

  31. An Approach: Identify degree of “forecastability” for products using Coefficient of Variation Topic 1 Coefficient of Variation (COV) COV = Standard Deviation/Average Value • Ensure that outliers and missing data are adjusted for • Ensure that trend, cyclicality, and seasonality are isolated from the data • Choose a threshold value for COV, usually a value between .7 and 1.0 • Identify those products with an adjusted COV > Chosen Threshold • If COV > Chosen Threshold, separate for other forecasting approaches or for hedging strategies

  32. Data Analysis & Data Cleaning Data Plot Central Tendency - Mean Variation – Volatility of Demand Systematic Variation Trend Seasonality Cyclicality Data Issues for Data Cleaning Topic 1

  33. Data Cleaning Issues Missing Values Outliers Data Shifts Structural changes Non-Linear Series Promotional, Marketing, and Sales Program Synchronization Topic 1

  34. Topic 1 Types of Models Model Families Quantitative71% Qualitative/Judgmental17% Other 12% • Decision trees • Sales force estimates • Executive opinion • Surveys & market research Time Series53% Cause & Effect 17% • Naïve • Trend • Moving Average • Filter • Smoothing • Decomposition • ARMA/ARIMA • Regression • Econometrics • Neural network Source: IBF Survey 2010

  35. Univariate Time Series Model Elements Topic 1 Y = time series L = level component T = Trend component S = seasonal component N = random noise component

  36. Topic 1 Steps in the Model Selection Process Forecasting with the Model Verification – Ex Post Forecasting - Ex Post Error Evaluation • Ex-ante forecasting • Error measurement, analysis and model improvement Estimation Specification of the Model • Consider models that support the underlying business • Match the method with the data pattern

  37. Topic 1 Error Measurements Error Measures Squared Error Measures • Error • Absolute Error • Mean Error (ME) • Mean Absolute Deviation (MAD) • Mean Percent Error (MPE) • Mean Absolute Percent Error (MAPE) • Weighted Mean Absolute Percent Error (WMAPE) • Squared Error (SE) • Mean Squared Error (MSE) • Root Mean Squared Error (RMSE) Used to evaluate forecasts and models Used primarily to evaluate models

  38. Topic 1 Forecast Error Formulas

  39. Error increases with product detail and time horizon Topic 1 All IndustriesMean Absolute Percent Error(MAPE) Source: IBF 2010 Survey

  40. What Are the Common Sources of Error? Topic 1 • Process Problems • Biased Estimates • Data Problems • Lack of or Poor Data Cleaning • Quality and Reliability of Assumptions Made • Poor Method Selection • Poorly Specified Models • Inherent Demand Volatility • Excessive Promotional Activity • Unstable Business, Economic, Competitor, and Political Environment

  41. Topic 1 Things To Know About Forecasting Errors

  42. Topic 1 Cost of Error – Reduced Net Cash Flow • Inventory carrying costs • Product costs • Storage costs • Handling & shipping costs • Interest expense on borrowed funds • Excess & obsolete inventory exposure • Greater mark-downs & discounts • Operating efficiency reduction • Revenue loss/gross profit loss • Reduction in customer satisfaction/repeat purchases

  43. Topic 2 Topic 2 Data Management & Data Cleaning

  44. Data Issues Missing data Outliers Shifts, structural changes Changing factors Trend Seasonal Cyclical Topic 2

  45. Example: Missing Value Topic 2

  46. Examples: Outliers Topic 2

  47. Example: Structural Shift Topic 2

  48. Topic 2 Example: Change in Seasonality Company X Percent of Sales by Quarter 2002-2005

  49. Data Analysis Checklist How much data do you have? How reliable is the data? Are there data definition changes? Are the data aggregated? Disaggregated? Do the time periods synchronize and line-up? Are we missing any values? Are there outliers? Do we know why? Are there structural shifts in the data? Topic 2 • What patterns does the data exhibit? • Randomness & volatility • Trend: linear, non-linear • Seasonality (weekly, monthly, quarterly) • Cyclicality • Are there events and company programs affecting the data? • What phase of the product life cycle is reflected in the data? • Are the data normally distributed? Are they distributed otherwise?

  50. Topic 2 Normally Distributed Data Series