You are on page 1of 28

# Revenue Management

Session 18
April 17th, 2018
Last Class

## 17-Apr-2018 ISOM Operations Management 2

Today’s Class
• Airline RM: Overbooking

• Price-based RM

• Willingness-to-pay

• Revenue maximization

## 17-Apr-2018 ISOM Operations Management 3

Common RM Tactics: Overbooking

## Suppose there are 100 seats on a flight from Hong

Kong to Singapore.

## The number of people who book tickets but do not

show up: Normally distributed with mean of 20 and
standard deviation of 10.

## 17-Apr-2018 ISOM Operations Management 4

Treatment of Overbooked Passengers
• Volunteers
o First seek customers willing to take a later flight in return for
compensation.
• Involuntary denied boarding
o Travel arrangement with a different flight or with another airline
o Compensation depending on the arrival time (may include meal
and lodging)

• Cost
o Direct cost of the compensation
o Travel arrangement cost
o The ill-will cost

## 17-Apr-2018 ISOM Operations Management 5

Optimal Overbooking Level

## Cu = Cost of underestimating no-shows

(when actual no-shows ≥ # of overbooked customers)
Co = Cost of overestimating no-shows
(when actual no-shows < # of overbooked customers)

## 17-Apr-2018 ISOM Operations Management 6

Optimal Overbooking Level

## • Compensation for each passenger denied boarding

• Arrangement for travel on another airline: \$200
• Free air ticket: \$105
• ill-will cost: \$100

## 17-Apr-2018 ISOM Operations Management 7

Marginal Analysis
n Cost of overbooking a passenger = \$405 (if he/she shows up)
Revenue
no
0

## Increase the no. of

overbooking from 10 to
11? #No show > 10
\$105
yes

\$105 - \$405 = \$-
#No show ≤ 10 300

## 17-Apr-2018 ISOM Operations Management 8

Rephrasing the Problem into Newsvendor Model
• How many seats should the airline overbook for this flight?
o Overage cost (too many seats overbooked)
• Co = \$405 - \$105 = \$300
o Underage cost (too little seats overbooked)
• Cu = \$105
• Optimal level of overbooking (X) satisfies
Cu 105
Pr Demand≤X = = =0.259
Cu +Co 105+300
• Since Demand is N(20, 102),
X∗ =20+10z∗ = 13.5
where z∗ =norminv 0.259, 10, 102 =−0.65 or use the standard normal
table
17-Apr-2018 ISOM Operations Management 9
Using z-table

!" = −!%&"

!'.)*+ =
− !'.,-% = −'. .*

## 17-Apr-2018 ISOM Operations Management 10

Summary
Revenue management with capacity Newsvendor
controls with single product
Decision: protection level for high fare Decision: order quantity

Uncertain demand: Demand for high fare Uncertain demand: Demand for
tickets newspapers

## Overstocking cost = discounted fare Overstocking cost = purchase cost –

salvage value

Understocking cost = full fare – discounted Understocking cost = retail selling price –

## 17-Apr-2018 ISOM Operations Management 11

Price Equilibrium

Demand Curve

D(p)

P*

## 17-Apr-2018 ISOM Operations Management 13

Willingness-to-Pay (WTP) – Discrete

• Given a price P*

## • w(p): frequency of population willing to pay p for a unit of the product

• w(p): probability a customer is willing to pay (exactly) p for a unit of the product
Price 0 1 2 3
Number of 10 60 20 10
customers
Frequency .1 .6 .2 .1

## 17-Apr-2018 ISOM Operations Management 14

WTP and Purchase Decision

∑%:()*( "(\$% )

## 17-Apr-2018 ISOM Operations Management 15

Revenue Maximization based on WTP
• Pricing a textbook

80 10 0.1
70 12 0.12
60 14 0.14
50 28 0.28
40 20 0.2
30 10 0.1
20 6 0.06

## 17-Apr-2018 ISOM Operations Management 16

Revenue Maximization based on WTP
• Pricing a textbook
Demand if
Price (P) Frequency WTP (w)
price = P
80 10 0.1 100*0.1 = 10
70 12 0.12 100*0.22 = 22
60 14 0.14 36
50 28 0.28 64
40 20 0.2 84
30 10 0.1 94
20 6 0.06 100

## 17-Apr-2018 ISOM Operations Management 17

Revenue Maximization based on WTP
• Pricing a textbook
Demand if
Price (P) Frequency WTP (w) Revenue
price = P
80 10 0.1 10 800
70 12 0.12 22 1540
60 14 0.14 36 2160
50 28 0.28 64 3200
40 20 0.2 84 3360
30 10 0.1 94 2820
20 6 0.06 100 2000
Revenue maximized at price = \$40
17-Apr-2018 ISOM Operations Management 18
Willingness-to-Pay – Continuous

W(p)

\$%\$&
Probability customer will purchase at !" = \$%\$

1
!−!

! !" ! D(p)
Price

! !

## 17-Apr-2018 ISOM Operations Management 19

Revenue Maximization—Linear Demand Approximation

## • Suppose the demand (# of textbooks) is a linear function of price p

D(p)=a−bp
• The revenue at price p is given by
R p =p×D p =p(a−bp)
• The revenue-maximizing price p∗ satisfies
dR(p∗ )
=0
dp
or equivalently,
∗ a
p =
2b
17-Apr-2018 ISOM Operations Management 20
Revenue Maximization Linear Demand Approximation

## • Estimation of the linear demand model using least squared error

method
• a! ,b" =(141.1, 1.65)
P Line Fit Plot
150

100
D

50 D
Predicted D
0
0 20 40 60 80 100
P

## • The optimal price based on the linear demand model is

p∗ = a! ⁄2b" =\$42.75

## 17-Apr-2018 ISOM Operations Management 21

Markdown Pricing vs Sales Promotions
• Markdown is a permanent reduction in price whereas promotions are temporary

## 17-Apr-2018 ISOM Operations Management 22

Reasons for Markdowns
• Fixed inventory (capacity) must be sold by a certain date
• E.g., Halloween costumes, tickets for musicals

• Obsolescence
• E.g., markdown before the arrival of the next generation of iPad

• Time of use
• E.g., markdown for winter coats in February

• Deterioration
• E.g., 1-day old bread/bakery sold at half price

## 17-Apr-2018 ISOM Operations Management 23

Markdown Pricing
• Regular season vs. markdown season

Markdown season
DM(p) = 20 - 4p
pM = ?

Regular season
DR(p) = 10 – p
pR = ?

## 17-Apr-2018 ISOM Operations Management 24

Bundle Pricing
• Two cable buyers, “sports lover” and “history lover”
• Willingness-to-pay for the two buyers

## ESPN History channel

Sports lover \$10 \$3
History lover \$3 \$10

## • Assume price = 90% of WTP

• Sports lovers would buy ESPN and history lovers would buy History
channel; hence, total revenue = \$9+\$9=\$18
• A bundle of the two channels costs \$11.7; hence, total revenue =
\$11.7+\$11.7=\$23.4
Both buyers and sellers benefit from bundling

## 17-Apr-2018 ISOM Operations Management 25

Psychological Pricing: Decoy Effect

## 17-Apr-2018 ISOM Operations Management 26

Psychological Pricing: The Magic Number 9
• Charm prices (\$49, \$79, \$1.49 and so on) are reported to boost sales
by an average of 24% relative to nearby prices