Session 17
April 12th, 2018
Measuring Forecast Accuracy
You are a marketing analyst for McDonalds and have the following sales forecasts ($M)
using two methods.
Tracking signal
• Ratio of cumulative error and MAD
• TS = ∑(Actual – Forecast) / MAD
1 100 60
2 100 130
3 200 200
4 200 270
5 400 340
3
2 Actual exceeds forecast
Tracking Signal
1
0
-1 Actual is less than forecast
Method 1
-2
-3
0 1 2 3 4 5
Month
Revenues 1000
COGS 800
Net Margin 50
COGS 800
A Namin, BT Ratchford, GP Soysal, An empirical analysis of demand variations and markdown policies for
fashion retailers, Journal of Retailing and Consumer Services 38, 126-136
12-Apr-2018 ISOM Operations Management 21
Current Industry Practice
Airlines Hotel
Video: Marriott
Media &
Car rental
broadcasting
Retailing
Gas pipelines
n Leisure
n 2 months in advance
n Discount rate
Business Strategy:
• Offer two fare classes f1 > f2
• Passengers that buy before a specific threshold pay f2,
otherwise they pay f1.
y1
y1: Protection level for
high-value or business y2
travelers y3
DB>20
Increase protection Sold at full
level y1 from 20 to 21?
Price later $7950
Yes – protect 21
seats
DB ≤ 20
Not sold by
May 29
Uncertain demand: Demand for high fare Uncertain demand: Demand for
tickets newspapers
Understocking cost = full fare – discounted Understocking cost = retail selling price –
fare purchase cost
• The previous solution can be extended to multiple fare classes, where now the
controls are given by a nested structure of booking limits, where each class has
access to the capacity of lower classes
• Cost
o Direct cost of the compensation
o Travel arrangement cost
o The ill-will cost
$105 - $405 = $-
#No show ≤ 10 300
!" = −!%&"
!'.)*+ =
− !'.,-% = −'. .*
Uncertain demand: Demand for high fare Uncertain demand: Demand for
tickets newspapers
Understocking cost = full fare – discounted Understocking cost = retail selling price –