Chapter 3 Demand Forecasting Slides prepared by Laurel Donaldson Douglas College Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. Learning Objectives LO 1 Identify uses of demand forecasts, distinguish between forecasting time frames, describe common features of forecasts, list the elements of a good forecast and steps of forecasting process, and contrast different forecasting approaches. LO 2 Describe at least three judgmental forecasting methods. LO 3 Describe the components of a time series model, and explain averaging techniques and solve typical problems. LO 4 Describe trend forecasting and solve typical problems. LO 5 Describe seasonality forecasting and solve typical problems. LO 6 Describe associative models and solve typical problems. LO 7 Describe three measures of forecast accuracy, and two ways of controlling forecasts, and solve typical problems. LO 8 Identify the major factors to consider when choosing a forecasting technique. Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 2 Chapter Outline What is forecasting? Features common to all forecasts Elements of a good forecast Steps in the forecasting process Approaches to forecasting Judgmental methods Time series models Associative models Accuracy and control of forecasts Choosing a forecasting technique Excel Templates Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 3 What is Forecasting? A demand forecast is an estimate of demand expected over a future time period I see that you will get a 100 in OM this semester. Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 4 How big a facility do I need to manufacture a new videophone? How much money do I need to run operations of my Need to FORECAST demand! accounting office? How many pairs of white shoes should I order for the summer season in my store? How many operators should I schedule next month for my call centre? How much lettuce should I buy for next week in my restaurant? Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 5 3 Uses for Forecasts: Design the System • long term (annual) • (types of products & services to offer, capacities, equipment, location) Use of the System • medium term (monthly) • (inventory, workforce levels, planning production) Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. Schedule the System • short term (daily, weekly) • (production, purchasing, staff scheduling) 6 Features of Forecasts Assumes causal system past ==> future Forecasts rarely perfect because of randomness Forecasts more accurate for groups vs. individuals Forecast accuracy decreases as time horizon increases Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 7 Elements of a Good Forecast Compatible Useful time horizon Meaningful Reliable Accurate Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. Easy to understand & use 8 Steps in the Forecasting Process 1 Determine purpose of forecast 2 Establish a time horizon 3 Select a forecasting technique 4 Obtain, clean and analyze data 5 Make the forecast 6 Monitor the forecast Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 9 Approaches to Forecasting Judgmental non-quantitative analysis of subjective inputs considers “soft” information such as human factors, experience, gut instinct Quantitative Time series models extends historical patterns of numerical data Associative models create equations with explanatory variables to predict the future Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 10 Judgmental Methods Executive opinions pool opinions of high-level executives long term strategic or new product development Expert opinions Delphi method: iterative questionnaires circulated until consensus is reached. technological forecasting Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 11 Judgmental Methods Sales force opinions based on direct customer contact Consumer surveys questionnaires or focus groups Historical analogies use demand for a similar product Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 12 What is a Time Series? Time series: a time ordered sequence of observations taken at regular intervals of time The following 6 patterns could be identified in a time series: Level: (average) horizontal pattern Trend: steady upward or downward movement Seasonality: regular variations related to time of year or day Cycles: wavelike variations lasting more than one year Irregular variations: caused by unusual circumstances, not reflective of typical behaviour Random variations: residual variations after all other behaviours are accounted for (called noise) Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 13 Patterns of a Time Series Demand for snowboards Seasonal peaks (winters) Trend component Actual demand line Random variation Year 1 Year 2 Year 3 Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. Year 4 14 Time series models Naive methods Averaging methods Moving average Weighted moving average Exponential smoothing Trend models Linear and non-linear trend Trend adjusted exponential smoothing Techniques for seasonality Techniques for cycles Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 15 Naive Methods Next period = last period Simple to use and understand Very low cost Low accuracy Stable time series data : Ft At 1 Seasonal variation s : Ft At n Data with tren d : Ft At 1 At 1 At 2 F = forecast A = actual Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 16 Naive Method - Example Uh, give me a minute.... We sold 250 wheels last week.... Now, next week we should sell.... Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 17 Naive Method with Trend: Example 2 years ago we sold 50 memberships. Last year we sold 75 memberships. This year we expect to sell … 100 Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 18 Averaging Methods Demand in previous n periods Moving : Ft Average n Weighted Moving Average Weight :F t period n Demand Weights period n Exponentia l : Ft Ft 1 At 1 Ft 1 Smoothing F = forecast A = actual = smoothing constant Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 19 Moving Average average of last few actual data values, updated each period easy to calculate and understand smoothes bumps, lags behind changes choose number of periods to include fewer data points = more sensitive to changes more data points = smoother, less responsive Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 20 Moving Average - Example Compute a three-period moving average forecast for period 6, given the demand below Period 1 2 3 4 5 Demand 42 40 F3 F4 F5 43 40 41 41.33 F6 43 3 3 40 41 Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 21 Weighted Moving Average - Example Compute a 4-period weighted moving average forecast for period 6 using a weight of 0.4 for the most recent period, 0.3 for the next, 0.2 for the next, and 0.1 for the next. Period Demand 1 42 2 3 40 43 4 40 5 41 Weight 0.1 0.1 F2 0.2 F3 0.3 F4 0.4 F5 41 F6 0.2 0.1 0.2 0.3 0.4 0.3 0.4 The choice of weights may involve the use of trial and error to find a suitable weighting scheme Weights must add up to 100% Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 22 Moving Average Example Period 1 2 3 4 5 6 7 Demand Forecast 9 12 14 16 19 23 26 (9 + 12 + 14)/3 = 11 2/3 (12 + 14 + 16)/3 = 14 (14 + 16 + 19)/3 = 16 1/3 (16 + 19 + 23)/3 = 19 1/3 Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 23 Quantity Graph of Moving Average 30 28 26 24 22 20 18 16 14 12 10 Moving Average Forecast – – – – – – – – – – – Actual Sales | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. | | | 10 11 12 24 Moving Average Example Apply weights of .5 for most recent period, then .3, then .2 Period 1 2 3 4 5 6 7 Demand 9 12 14 16 19 23 26 Forecast [(.5 x 14) + (.3 x 12) + (.2 x 9)] = 12.4 [(.5 x 16) + (.3 x 14) + (.2 x 12)] = 14.6 [(.5 x 19) + (.3 x 16) + (.2 x 14)] = 17.1 [(.5 x 23) + (.3 x 19) + (.2 x16)] = 20.4 Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 25 Moving Average And Weighted Moving Average Weighted moving average 30 – 25 – Quantity 20 – Actual sales 15 – Moving average 10 – 5 – | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. | | | 10 11 12 26 Exponential Smoothing sophisticated weighted moving average weights decline exponentially most recent data weighted most subjectively choose smoothing constant ranges from 0 to 1 (commonly .05 to .5) widely used easy to use easy to alter weighting Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 27 Exponential Smoothing Formula Forecast = previous forecast plus a percentage of the forecast error Actual - Forecast is the error term is the % feedback Ft = Ft-1 + (At-1 - Ft-1) F = forecast A = actual Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 28 Exponential Smoothing: Alternate Formula Forecast = previous forecast plus a percentage of the forecast error is the weight on actual demand 1 is the weight on previous forecast Ft = (1 - Ft-1 + (At-1) F = forecast A = actual Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 29 Exponential Smoothing: Example Forecasted demand = 142 video games Actual demand = 153 Smoothing constant = .20 New forecast = .2 (153) + (1 - .2)(142) = 30.6 + 113.6 = 144.2 ≈ 144 games Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 30 Exponential Smoothing: Example Forecasted demand = 142 video games Actual demand = 153 Smoothing constant = .20 New forecast = 142 + .2 (153 - 142) = 30.6 + 113.6 = 144.2 ≈ 144 games Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 31 Exponential Smoothing: Example Prepare a forecast using smoothing constant = 0.40. What is the starting point? average of several periods of actual data subjective estimate (for this example, use 60) first actual value (naïve approach) Period 1 Actual 65 2 3 55 58 4 64 Forecast 60 F2 F1 0.4A1 F1 F3 F2 0.4A 2 F2 Calculatio ns 60 0.465 - 60 62 62 0.455 - 62 59.2 F4 F3 0.4A 3 F3 59.2 0.458 - 59.2 58.72 Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 32 Exponential Smoothing: Your Turn! What are the exponential smoothing forecasts for periods 2-5 using =0.5? Use naïve approach for 1st week Week 1 2 3 4 5 Demand 820 775 680 655 Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 33 Exponential Smoothing: Your Turn! F2=(.5)(820)+(1 - .5)(820) =820 F3=(.5)(775)+(1 - 0.5)(820)=797.5 Week 1 2 3 4 5 Demand 820 775 680 655 0.5 820.00 820.00 797.50 738.75 696.88 Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 34 Selecting a Smoothing Constant Demand 225 – Actual demand 200 – a = .5 175 – 150 – | 1 | 2 | 3 | 4 | | 5 6 Period Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. a = .1 | | 7 8 | 9 35 Choosing When demand is fairly stable, use a lower value for smoothes out random fluctuations When demand increasing or decreasing, use a higher value for more responsive to real changes Try to find balance trial and error can change over time. Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 36 True or False? A moving average forecast tends to be more responsive to changes in the data series when more data points are included in the average. False As compared to a simple moving average, the weighted moving average is more reflective of the recent changes. True A smoothing constant of .1 will cause an exponential smoothing forecast to react more quickly to a sudden change than a value of .3 will. False Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 37 Excel: Exponential Smoothing Solved Problem 1: Excel Template Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 38 Techniques for Trend – Develop an equation that describes the trend – Look at historical data Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 39 Nonlinear Trends Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 40 Linear Trend Equation – – Fit a trend line to a series of historical data Use regression to find the equation of the line (called the Least Squares Line) Equation : yt a bt Slope : b n ty t y n t t 2 2 y b t y - intercept : a n Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 41 Linear Trend Demand Actual observation Deviation Deviation Deviation Deviation Deviation Deviation Deviation Points on the line yt a bt Time Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 42 Linear Trend: Example week t 1 Sales y 150 t2 1 ty 150 2 3 4 157 162 166 4 9 16 314 486 664 5 t 15 177 y 812 25 t 2 55 885 ty 2,499 t 2 225 5 2,499 15 812 12,495 12,180 6.3 5 55 225 275 225 812 6.3 15 a 143.5 5 yt a bt 143.5 6.3t b Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 43 Excel - Linear Trend Scatter with Trendline Insert Chart 160 Scatter 140 y = 10.171x - 20242 120 Highlight data range Sales 100 80 Right Click on a data point 60 40 20 0 1996 1997 1998 1999 2000 2001 2002 2003 2004 Year Add Trendline Type: Linear Options: Show equation on chart Or Insert Functions: =SLOPE(Range of y's,Range of x's) =INTERCEPT(Range of y's,Range of x's) Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 44 Excel - Linear Trend Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 45 Trend-Adjusted Exponential Smoothing select values (usually through trial and error) for = smoothing constant for average b = smoothing constant for trend estimate starting smoothed average and smoothed trend use most recent data Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 46 Trend-Adjusted Exponential Smoothing TAFt+1 = St + Tt (3–6) St = TAFt +α(At TAFt) (3–7) Tt = Tt-1 + b( St St-1 Tt-1) where St = smoothed average at the end of period t Tt = smoothed trend at the end of period t Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 47 Trend-Adjusted Forecast: Example Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 48 Techniques for Seasonality Additive or Multiplicative Model quantity added to average or trend or proportion x average or trend Additive Model Demand Demand = Trend + Seasonality Multiplicative Model Demand = Trend x Seasonality time Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 49 Using Seasonal Relatives Seasonal Relative (or index) = proportion of average or trend for a season in the multiplicative model seasonal relative of 1.2 = 20% above average Deseasonalize remove seasonal component to more clearly see other components divide by seasonal relative Reseasonalize adjust the forecast for seasonal component multiply by seasonal relative Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 50 Times Series Decomposition 1. Compute the seasonal relatives. 2. De-seasonalize the demand data. 3. Fit a model to de-seasonalized demand data, e.g., moving average or trend. 4. Forecast using this model and the de-seasonalized demand data. 5. Re-seasonalize the deseasonalized forecasts. Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 51 Techniques for Seasonality - Example Predict quarterly demand for a certain loveseat The series has both trend and seasonality. Quarterly relatives : Q1 = 1.20, Q2 = 1.10, Q3 = 0.75, Q4 = 0.95. Trend equation yt=124+7.5t (t = 1 in first quarter of 2003) Predict demand for quarter 3 of 2006 for quarter 3 of 2006 t 15 y15 124 7.5 15 236.5 F15 236.5 0.75 177.38 Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 52 Associative Forecasting If I want to predict ridership originating from a new train station, what data might I look at? 1. Find (predictor) variables that are associated with ridership at other stations. 2. Associated = correlated = as one moves the other moves 3. Create a model that shows the relationship between the predictor variables and the predicted variable (e.g. ridership) 4. Technique is regression analysis • • Simple linear regression with one variable Multiple regression (can be non-linear) 5. Test the model to see which variables most useful in predicting ridership (look at r2) 6. Use the model to predict ridership, given values of the predictor variables. Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 53 Associative Models Predictor variables (x): used to predict values of the variable of interest (y) (also called independent variables) Linear regression: process of finding a straight line that best fits a set of points on a graph (use the Least Squares Equation) Multiple regression: models with more than one predictor variable (computations complex, created with computer) Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 54 Simple Linear Regression X 7 2 6 4 14 15 16 12 14 20 15 7 Y 15 10 13 15 25 27 24 20 27 44 34 17 Computed relationship 50 40 30 20 10 0 0 5 10 15 20 25 would a linear model be reasonable? Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 55 Excel: Simple Linear Regression Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 56 Correlation and Excel Correlation coefficient (r): measure of the strength of relationship between two variables ranges from -1 to +1 -1 = two variables move together in same direction +1 = two variables move together in opposite direction =CORREL(Range of y values, Range of x values) r2 measures proportion of variation in the values of y that is “explained” by the predictor variables in the regression model ranges from 0 to 1 higher values = more useful predictors =RSQ(Range of y values, Range of x values) Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 57 Linear Regression Assumptions Predictions are being made only within the range of observed values relationship may be non-linear outside that range y-intercept often not meaningful Variations around the line are random and normally distributed For best results: Always plot the data to verify linearity Small correlation may imply that other variables are important Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 58 Accuracy and Control of Forecasts Error = Actual value - Forecast value +ve = forecast too low, -ve = too high Three measures of forecasts are used: Mean absolute deviation (MAD) Mean squared error (MSE) Mean absolute percent error (MAPE) Control charts plot errors to see if within pre-set control limits Tracking signal Ratio of cumulative error and MAD Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 59 Error, MAD, MSE and MAPE Error e Actual Forecast At Ft Actual Forecast MAD n MSE 2 Actual Forecast n Actual Forecast Actual MAPE 100 n Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. % 60 MAD, MSE and MAPE MAD • Easy to compute • Weights errors linearly MSE MAPE • Squares error • More weight to large errors • Puts errors in perspective • above 70% satisfactory Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 61 Error, MAD, MSE and MAPE: Example Compute MAD, MSE, and MAPE for the following data. 100 Actual Forecast e e e 1 2 217 213 215 216 2 3 2 3 4 9 A 0.92% 1.41% 3 4 216 210 215 214 1 4 1 4 10 1 16 30 0.46% 1.90% 4.69% n 2 e n e Period e MAD 2.5 MSE 2 7.5 e 100 A 1.17% MAPE n Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 62 Forecast Errors bias = the sum of the forecast errors +ve bias = frequent underestimation -ve bias = frequent overestimation possible sources of error include: Model may be inadequate (things have changed) Incorrect use of forecasting technique Irregular variations Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 63 Controlling the Forecasting Process Control chart A visual tool for monitoring forecast errors Used to detect non-randomness in errors Set limits that are multiples of the √MSE Forecasting errors are “in control” when only random errors, no errors from identifiable causes “in control” if All errors are within control limits No patterns (e.g. trends or cycles) are present errors outside limit = need corrective action Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 64 Control Chart Error Upper limit Range of acceptable variation 0 Lower limit Time Need for corrective action Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 65 Controlling Forecasts: Control Limits Standard deviation = of error s Control Limits MSE 2 e n = 0 ± 2 (or 3) s 95% of all errors should be within 2s 97.7% of all errors should be within 3s Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 66 Control Chart Example A F A-F Month (Sales) (Forecast) Error 1 2 3 4 5 6 s 90 95 115 100 125 140 100 100 100 110 110 110 1575 6 -10 -5 +15 -10 +15 +30 MSE 100 25 225 100 225 900 1575 262.5 16.2 Errors should be within ± 2(16.2). Lower limit = -32.4 Upper limit = 32.4 Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 67 Control Chart Example 32 24 16 8 0 -8 0 1 2 3 4 5 6 7 -16 -24 -32 All the errors are within the control limits Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 68 Pharmacy Forecast Control: Your Turn! Below is a pharmacy’s actual sales and forecasted demand for a certain prescription drug for 5 months. How accurate is their forecast? Calculate MAD and MSE and create a control chart. Month 1 2 3 4 5 Sales Forecast 220 n/a 250 255 210 205 300 320 325 315 Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 69 Pharmacy Forecast Control: Your Turn! Sales 220 250 210 300 325 Month 1 2 3 4 5 A MAD = t n - Ft Forecast n/a 255 205 320 315 40 = = 10 MSE = 4 Sq. Error Abs Error A Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. t n 5 5 20 10 25 25 400 100 40 550 - Ft 2 550 = = 137.5 4 70 Pharmacy Forecast Control: Your Turn! s 550 137.5 11.7 4 Errors should be within ± 2(11.7). Lower limit = -23.4 Upper limit = 23.4 23.4 15.6 7.8 0 0 1 2 3 4 5 -7.8 -15.6 -23.4 All the errors are within the control limits Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 71 Tracking Signal Tracking signal ratio of cumulative error to MAD can be plotted on a control chart investigate if TS > 4 (Actual -forecast) Tracking signal = MAD Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 72 True or False? When error values fall outside the limits of a control chart, this signals a need for corrective action Ans: True When all errors plotted on a control chart are either all positive, or all negative, this shows that the forecasting technique is performing adequately. Ans: False A random pattern of errors within the limits of a control chart signals a need for corrective action. Ans: False Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 73 Choosing a Forecasting Technique No single technique works in every situation Two most important factors Cost Accuracy Other factors include availability of: Historical data Computers Time needed to gather and analyze the data Forecast horizon Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 74 Choosing a Forecast Technique Forecasting Amount of Historical Data Method Simple exponential smoothing Trendadjusted exponential smoothing Data Pattern Forecast Horizon Preparation time Complexity 5 to 10 observations Data should be stationary Short Short Little sophistication 10 to 15 observations Trend Short to medium Short Moderate sophistication Trend Short Regression 10 to 20 Short, Moderate sophistication medium, Trend long models Seasonal Enough to see seasonal Short to Short to Moderate patterns medium moderate sophistication 3 peaks and troughs Causal 10 Can handle Medium or Long Considerable complex patterns long sophistication regression observations development models per time, short time implementation independent variable Source: J. Holton Wilson and D. Allison-Koerber, “Combining Subjective and Objective Forecasts Improves Results,” Journal of Business Forecasting Methods & Systems, 11(3) Fall 1992, p. 4. Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 75 Choosing a Forecast Technique Factor 1. Frequency Short Term daily, weekly Medium Term monthly, quarterly Long Term annual 2. Level of aggregation Item Product family Total output 3. Type of model Smoothing Trend Trend Seasonal Regression 4. Degree of management involvement Low Moderate Managerial Judgment Trend Regression High 5. Cost per forecast Low Moderate High Source: C. L. Jain, “Benchmarking Forecasting Models,” Journal of Business Forecasting Methods & Systems, Fall 2002, pp. 18–20, 30. Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 76 Which technique? Sales for a product have been fairly consistent over several years, although showing a steady upward trend. The company wants to understand what drives sales. The best forecasting technique would be: A) trend models B) judgmental methods C) moving averages D) regression models E) exponential smoothing techniques Ans: D Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 77 Learning Checklist Describe at least three judgmental forecasting methods. Describe the components of a time series model, and explain averaging techniques and solve typical problems. Describe trend forecasting and solve typical problems. Describe seasonality forecasting and solve typical problems. Describe associative models and solve typical problems. Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 78 Learning Checklist Identify uses of demand forecasts Distinguish between forecasting time frames Describe common features of forecasts List the elements of a good forecast and steps of forecasting process, Contrast different forecasting approaches. Describe three measures of forecast accuracy, and two ways of controlling forecasts, and solve typical problems. Identify the major factors to consider when choosing a forecasting technique. Copyright © 2010 by The McGraw-Hill Companies, Inc. All rights reserved. 79