CTL.SC1x -Supply Chain & Logistics Fundamentals
Introduction to Demand Planning & Forecasting
MIT Center for Transportation & Logistics Demand Process – Three Key Questions
Demand Planning n Product & Packaging What should we do to shape and n Promotions create demand for our product? n Pricing n Place
Demand Forecasting What should we expect demand to n Strategic, Tactical, Operational be given the demand plan in place? n Considers internal & external factors n Baseline, unbiased, & unconstrained
Demand Management n Balances demand & supply How do we prepare for and act n Sales & Operations Planning (S&OP) on demand when it materializes? n Bridges both sides of a firm
Material adapted from Lapide, L. (2006) Course Notes, ESD.260 Logistics Systems.
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 2 Forecasting Levels
Level Horizon Purposes
• Business Planning Strategic Year/Years • Capacity Planning • Investment Strategies
• Brand Plans Quarterly • Budgeting • Sales Planning Tactical • Manpower Planning • Short-term Capacity Planning Months/Weeks • Master Planning • Inventory Planning
Operational • Transportation Planning Days/Hours • Production Planning • Inventory Deployment
Material adapted from Lapide, L. (2006) Course Notes, ESD.260 Logistics Systems.
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 3 Agenda
• Forecasting Truisms • Subjective vs. Objective Approaches • Forecast Quality • Forecasting Metrics
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics Forecasting Truisms 1: Forecasts are always wrong
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 5 1. Forecasts are always wrong
Why?
n Demand is essentially a continuous variable
n Every estimate has an “error band”
n Forecasts are highly disaggregated w Typically SKU-Location-Time forecasts
n Things happen . . . OK, so what can we do?
n Don’t fixate on the point value
n Use range forecasts
n Capture error of forecasts
n Use buffer capacity or stock
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 6 Forecasting Truisms 2: Aggregated forecasts are more accurate
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 7 2. Aggregated forecasts are more accurate
• Aggregation by SKU, Time, Location, etc. • Coefficient of Variation (CV)
n Definition: Standard Deviation / Mean = σ/µ
n Provides a relative measure of volatility or uncertainty
n CV is non-negative and higher CV indicates higher volatility
Red: µ=100, σ=45, CV=0.45 Blue: µ=100, σ=1, CV=0.01 200 180 160 140 120 100 80 Daily Demand Demand Daily 60 40 20 - 2/26/11 3/28/11 4/27/11 5/27/11 6/26/11 7/26/11 8/25/11
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 8 Aggregating by SKU • Coffee Cups and Lids @ the Sandwich Shop n Large ~N(80, 30) CV = 0.38
n Medium ~N(450, 210) CV = 0.47
n Small ~N(250, 110) CV = 0.44
800
700
600
500 Small 400 Medium
300 Large
200
100
- 8/14/13 9/13/13 10/13/13 11/12/13 12/12/13 1/11/14 2/10/14 3/12/14 4/11/14 5/11/14 6/10/14
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 9 Aggregating by SKU • What if I design cups with a common lid? Large ~N(80, 30) CV=0.38 • Common Lid ~N(780, 239) CV = 0.31 Med. ~N(450, 210) CV=0.47 Small ~N(250, 110) CV=0.44 n µ = (80 + 450 + 250) = 780 units/day 2 2 2 Lids ~N(780, 239) CV=0.31 n σ = sqrt(30 + 210 + 110 ) = 239 units/day
1,600 1,400 1,200 1,000 800 600 400 200 - 8/14/13 9/13/13 10/13/13 11/12/13 12/12/13 1/11/14 2/10/14 3/12/14 4/11/14 5/11/14 6/10/14
Example of Modularity or Parts Commonality • Reduces the relative variability • Increases forecasting accuracy • Lowers safety stock requirements
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 10 • Forecasts with longer time buckets have better Aggregating by Time forecast accuracy. 1,600 Daily Demand for Lids ~N(780, 239) CV=0.31 • The time bucket used 1,200 should match the situation.
800
400
- 8/14/13 9/13/13 10/13/13 11/12/13 12/12/13 1/11/14 2/10/14 3/12/14 4/11/14 5/11/14 6/10/14
Weekly Demand for Lids ~N(5458, 632) CV=0.12 8,000
6,000
4,000
2,000
- 1 5 9 13 17 21 25 29 33 37 41 45 49
30,000 Monthly Demand for Lids ~N(21840, 1264) CV=0.06 25,000 20,000 15,000 10,000 5,000 - 1 2 3 4 5 6 7 8 9 10 11 12 CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 11 CV reduces as we Aggregating by Locations aggregate over SKUs, time, or locations. • Suppose we have three sandwich shops
n Weekly lid demand at each ~N(5458, 632) CV=0.12
~N(5458, 632) ~N(5458, 632) ~N(5458, 632)
~N(16374, 1095)
• What if demand is pooled at a common Distribution Center?
n Weekly lid demand at DC ~N(16374, 1095) CV=0.07
σ σ n σ CVind CVind = CVagg = = = µ µn µ n n
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 12 Forecasting Truisms 3: Shorter horizon forecasts are more accurate
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 13 3. Shorter horizon forecasts are more accurate
?
?
?
?
21 22 23 24 1 2 3 4 5 6 7 8 9 10 11 12
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 14 3. Shorter horizon forecasts are more accurate
• Postponed final customization to closer time of consumption • Risk pooling of component (e.g., ham) increases forecast accuracy.
?
21 22 23 24 1 2 3 4 5 6 7 8 9 10 11 12
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 15 Forecasting Truisms
• Forecasts are always wrong è Use ranges & track forecast error • Aggregated forecasts are more accurate è Risk pooling reduces CV • Shorter time horizon forecasts are more accurate è Postpone customization until as late as possible
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 16 Subjective & Objective Approaches
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 17 Fundamental Forecasting Approaches Subjective Objective Judgmental Causal / Relational n Sales force surveys n Econometric Models n Jury of experts n Leading Indicators n Delphi techniques n Input-Output Models
Experimental Time Series n Customer surveys n “Black Box” Approach n Focus group sessions n Past predicts the future n Test marketing n Identify patterns
Often times, you will need to use a combination of approaches
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics Forecasting Quality
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 19 Cost of Forecasting vs Inaccuracy
ß Overly Naïve Models à ß Good Region à ß Excessive Causal Models à Total Cost Cost Cost of Errors Cost of Forecasting In Forecast
Forecast Accuracy
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics How do we determine if a forecast is good? • What metrics should we use? • Example - Which is a better forecast?
n Squares & triangles are different forecasts
n Circles are actual values
1100
1000
900
time
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics Accuracy versus Bias
n Accuracy - Closeness to actual observations
n Bias - Persistent tendency to over or under predict
Accurate
Not Accurate
Biased Not Biased CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics Forecasting Metrics
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 23 Forecasting Metrics et = At – Ft Mean Deviation n Mean Absolute n (MD) Deviation (MAD) e et ∑ t ∑ t 1 MD = t=1 MAD = = n n Mean Squared Root Mean Error (MSE) n Squared n e2 e2 ∑ t Error (RMSE) ∑ t MSE = t=1 RMSE = t=1 n n
Mean Percent n n e et Mean Absolute t Error (MPE) Percent Error (MAPE) ∑ A ∑ A MPE = t=1 t MAPE = t=1 t n n Notation: At = Actual value for obs. t et = Error for observation t Ft = Forecasted value for obs. t n = Number of observations CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics Example: Forecasting Bagels
• For the bagel forecast and actual values shown below, find the:
n Mean Absolute Deviation (MAD) n e n Root Mean Square of Error (RMSE) ∑ t t=1 n Mean Absolute Percent Error (MAPE) MAD = n
Forecast Actual n e2 Monday 50 43 ∑ t RMSE = t=1 Tuesday 50 42 n Wednesday 50 66 Thursday 50 38 n et Friday 75 86 ∑ A MAPE = t=1 t n
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 25 Example: Forecasting Bagels
• Solution: 90 1. Graph it. Forecast Actual 2. Extend data table: 80
w Error: et=At-Ft 70
w Abs[error] = |et| 60 w Sqr[error] = e2
w AbsPct[error] = |et/At| 50 Daily Bagel Demand Demand Bagel Daily 3. Sum and find means 40 2 Ft At et |et| e |et/At| 30 Monday 50 43 -7 7 49 16.3% Monday Tuesday Wednesday Thursday Friday Tuesday 50 42 -8 8 64 19.0% Wednesday 50 66 16 16 256 24.2% MAD = 54/5 = 10.8 Thursday 50 38 -12 12 144 31.6% RMSE = sqrt(126.8) = 11.3 Friday 75 86 11 11 121 12.8% MAPE = 104%/5 = 21% Sum 0 54 634 104% Mean 0 10.8 126.8 21% CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 26 Key Points from Lesson
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics 27 Key Points
• Forecasting is a means not an end • Forecasting Truisms
n Forecasts are always wrong
n Aggregated forecasts are more accurate
n Shorter horizon forecasts are more accurate • Subjective & Objective Approaches
n Judgmental & experimental
n Causal & time series • Forecasting metrics
n Capture both bias & accuracy
n MAD, RMSE, MAPE
CTL.SC1x - Supply Chain and Logistics Fundamentals Lesson: Demand Forecasting Basics CTL.SC1x -Supply Chain & Logistics Fundamentals Questions, Comments, Suggestions? Use the Discussion!
“Janie” Photo courtesy Yankee Golden Retriever Rescue (www.ygrr.org) MIT Center for Transportation & Logistics [email protected]