1CM270

Tutorial 6
Module 3: Pricing and Revenue Management
Stochastic and deterministic price-differentiation, Customized pricing

1. Suppose you are selling 2 hats over a selling season of 2 periods. In each period, exactly 1
customer arrives. Both customers have the same exponential willingess to pay for the hat. In
particular, the probability that they will buy the hat at price p is given by

b(p) = e−0.5p .

(a) Compute the optimal price in the first period before the first customer arrives, i.e. p∗1 (2),
as well as the cumulative expected revenue over the two periods given that you follow the
optimal pricing policy, i.e. v1∗ (2).
(b) Compute the optimal static price and the associated expected revenue. Compare this with
the optimal dynamic pricing policy and associated cumulative revenue. What conclusions
can you draw?

Solution:

(a) We work backwards. We start in period t = 2, and find

p∗2 (2) = arg max p · e−0.5p = 2, with v2∗ (2) = 2/e.

p

Note that since there is only 1 customer arriving, we also have p∗2 (1) = p∗2 (2) and v2∗ (2) =
v2∗ (1). We continue with t = 1, and find

p∗1 (2) = arg max p · e−0.5p + e−0.5p · 2/e + (1 − e−0.5p ) · 2/e = 2, with v2∗ (2) = 4/e.
p

and

p∗1 (1) = arg max p · e−0.5p + (1 − e−0.5p ) · 2/e = 2 + 2/e, with v2∗ (1) = 2(1 + e−1/e )/e.
p

Note that technically we do not need to compute p∗1 (1) since we have 2 hats in period 1.
(b) In this case you charge a single price. Since we are selling 2 products over two periods, we
can compute the expected revenue as the expectation of a binomial random variable with
two trials and success probability b(p). Hence, the expected revenue under a static price p
is given by
E[revenue|p] = p · e−0.5p · 2.
If we maximize this by taking derivaties and setting equal to 0, we find p∗ = 2 with an
expected revenue of 4/e. Note that this is equal to v2∗ (2). The reason for this is that the
number of periods to sell is equal to the number of products, and customers always arrive.
So at most we can sell K = T products, and we sell each product with probability b(p).

2. Implement the backward induction algorithm for the dynamic pricing problem of the lecture.
Verify that for the sample instance on slide 21 the cumulative expected revenue over the entire
selling season under the optimal policy equals v1∗ (10) ≈ 28.04.

1
3. [Adapted from Philips (2021)] A barber charges $12 per haircut and works Saturday through
Thursday. He can perform up to 20 haircuts a day. He currently performs an average of 12
haircuts per day during the weekdays (Monday through Thursday). On Saturdays and Sundays,
he does 20 haircuts per day and turns 10 potential customers away each day. These customers
all go to the competition.
The barber is considering raising his prices on weekends. He estimates that for every $1 he raises
his price, he will lose an additional 10% of his customer base (including his turnaways) compared
to the current setting. He estimates that 20% of his remaining weekend customers would move
to a weekday to save $1, 40% would move to a weekday to save $2, and 60% would move to a
weekday to save $3.
Assuming he needs to price in increments of $1, should he charge a differential weekend price? If
so, what should the weekend price be? (Assume he continues to charge $12 on weekdays.) How
much revenue (if any) would he gain from his policy?
Solution:
We first compute the current revenue:
Weekdays (Monday to Thursday):
4 × 12 × 12 = $576
Weekends (Saturday and Sunday):
2 × 20 × 12 = $480
Total weekly revenue:
576 + 480 = $1056

Now we compute the revenues under the different price increases:

(a) Increase by $1: He loses 10% of his 30 customers (20 current + 10 turnaways) = 3 customers
per weekend day. 20% of the remaining 27 customers (5.4 customers per weekend day) move
to weekdays. Total customers moving to weekdays over the weekend =
2 × 5.4 = 10.8
Weekend revenue:
2 × min((27 − 5.4), 20) × 13 = $520
Weekday revenue increase:
10.8
4 × (12 + ) × 12 = $705.6
4
Total weekly revenue:
520 + 705.6 = $1225.6
(b) Increase by $2: He loses 20% of his 30 customers = 6 customers per weekend day. 40%
of the remaining 24 customers (9.6 customers per weekend day) move to weekdays. Total
customers moving to weekdays over the weekend =
2 × 9.6 = 19.2
Weekend revenue:
2 × (24 − 9.6) × 14 = $403.2
Weekday revenue increase:
19.2
4 × (12 + ) × 12 = $806.4
4
Total weekly revenue:
403.2 + 806.4 = $1209.6

2
 (c) Increase by $3: He loses 30% of his 30 customers = 9 customers per weekend day. 60% of
the remaining 21 customers (12.6 customers per weekend day) move to weekdays. Total
customers moving to weekdays over the weekend =

2 × 12.6 = 25.2

Weekend revenue:
2 × (21 − 12.6) × 15 = $252
Weekday revenue increase:
25.2
4 × (12 + ) × 12 = $878.4
4
Total weekly revenue:
252 + 878.4 = $1130.4

By comparing the different price increases, we see that the barber would achieve the highest
weekly revenue by increasing his weekend price by $1, resulting in a weekly revenue of $1225.6.

4. Consider the theme park example on slide 6 of the lecture. Assume now that the theme park
owner invested in expanding his park so that he could accommodate up to 1,500 customers each
day. The price-response curves remain the same.

(a) What single price would maximize his total revenue, and what is his corresponding atten-
dance and revenue for each day?
(b) What variable prices should he charge for each day to maximize expected revenue, and
what are the corresponding attendance and revenue for each day?

Solution:
The table below gives the optimal static single price and the optimal static variable prices for
each day as well the corresponding attendance and revenues. You can compute these prices
using any optimization solver, including the solver add-in in excel.

Table 1: Revenue Comparison

Single Price Case Variable Price Case
Day Price Attendance Revenue Price Attendance Revenue
Sunday 24.79 1500 37, 188 25.81 1500 38, 709
Monday 24.79 260 6, 456 15.00 750 11, 250
Tuesday 24.79 408 10, 123 17.50 700 12, 250
Wednesday 24.79 469 11, 621 17.98 755 13, 572
Thursday 24.79 696 17, 254 19.01 1000 19, 011
Friday 24.79 1122 27, 806 22.48 1250 28, 103
Saturday 24.79 1500 37, 188 30.00 1500 45, 000
TOTAL 5955 147, 635 7455 167, 896, 572

5. [Adapted from Philips (2021)] A department store has 700 pairs of purple capri stretch
pants that it will mark down over the next four weeks (afterwards they are worthless). The
store manager knows that demand by week for the next four weeks will be linear, with the
following price-response functions:

• Week 1: d1 (p1 ) = 1000 − 100p1

• Week 2: d2 (p2 ) = 800 − 100p2
• Week 3: d3 (p3 ) = 700 − 100p3

3
 • Week 4: d4 (p4 ) = 600 − 100p4

Assume that the demands in the different weeks are independent—that is, that customers who
do not buy in a given week do not come back in subsequent weeks. You may assume fractional
sales.

(a) What is the optimum price the retailer should charge per pair if he can only set one price
for all four weeks? What is the corresponding revenue?
(b) Assume he can charge a different price each week. What are the optimum prices by week
he should charge? What is the corresponding revenue?

Solution:

(a) This can be formulated as an optimization problem. Let p be the single price to be offered,
and let x1 , x2 , x3 , and x4 be the amount sold in periods 1, 2, 3, and 4, respectively. Then,
the optimization problem is:

4
X
max p xi
p
i=1

Subject to:

x1 ≤ 1000 − 100p
x2 ≤ 800 − 100p
x3 ≤ 700 − 100p
x4 ≤ 600 − 100p
x1 + x2 + x3 + x4 ≤ 700
p, x1 , x2 , x3 , x4 ≥ 0

Using an optimization solver such as the excel solver, the optimal price can be found to be
6.00. The amounts sold in each period are x1 = 400, x2 = 200, x3 = 100, x4 = 0, and the
total revenue is 4200.
An alternative approach is to create the aggregate price response function x = 3, 100 − 400p
by summing the four functions. The optimal unconstrained price in this case is 3.875.
However, at this price, demand would be 1, 550 units, which is greater than the available
supply of 700. Thus, the revenue-optimizing price is the price at which exactly 700 units
are sold, that is the price such that 3, 100 − 400p = 700, which gives the correct answer of
p = $6.00.
(b) This is an optimization problem with a structure similar to the previous one; however,
we now have four different prices to calculate: p1 , p2 , p3 , and p4 . The corresponding
optimization problem is:

4
X
max (xi pi )
p1 ,p2 ,p3 ,p4
i=1

Subject to:

4
 x1 ≤ 1000 − 100p1
x2 ≤ 800 − 100p2
x3 ≤ 700 − 100p3
x4 ≤ 600 − 100p4
x1 + x2 + x3 + x4 ≤ 700
x1 , x2 , x3 , x4 ≥ 0
p1 ≥ p2 ≥ p3 ≥ p4 ≥ 0

Using an optimization solver, we find the following optimal prices and corresponding sales
and revenue:

Period Price ($) Sales Revenue ($)

1 7.125 287.5 2,048.44
2 6.125 187.5 1,148.44
3 5.625 137.5 773.44
4 5.125 87.5 448.44
Total 700 4,418.75

Table 2: Sales and Revenue Table

Note that variable markdown pricing increased revenue from $4,200 with a single price to
$4,418.75 for multiple prices.

6. [Adapted from Philips (2021)] HP is bidding against Lenovo for an order of 150 laptops.
HP’s unit cost is $1,000 per laptop, and, based on previous experience, HP’s belief is that
Lenovo’s bid will be uniformly distributed between $1,200 and $1,500 per unit.

(a) What is HP’s optimal price per unit to bid? What is HP’s corresponding probability of
winning the bid and expected contribution?
(b) Assume that Lenovo is a preferred supplier to this customer and, as a result, HP believes
that Lenovo enjoys a $200 per unit premium for this deal. This means that there should
be at least a difference of $200 between the bid of HP and the bid of Lenovo for HP to win
the bid. Assuming that HP has the same distribution on Lenovo’s bid as previously stated,
what is HP’s optimal price per unit to bid? What is HP’s corresponding probability of
winning the bid and its expected contribution?

Solution:

(a) HP’s probability of winning the bid as a function of the price is ρ(p) = 5 − p/300 for
bids in the relevant range of $1,200 to $1,500 (with 1 below the range and 0 above the
range). Therefore, HP’s expected contribution from a bid of p is (p − 1000) · (5 − p/300).
Differentiating this expression to determine the first order condition for the contribution-
maximizing p gives 25/3 − p∗ /150 = 0 or p∗ =$1,250. The corresponding probability of
winning the bid is 5 − 1250/300 or 83.3% and the expected contribution is .83 · 150 · 250 =
$31,125.
(b) If Lenovo has premium of $200 per unit, this means the bid response function shifts to
the left by $200. Indeed, HP’s bid must be at least $200 less than Lenovo’s bid to win.
Thus HP’s probability of winning the bid is now ρ(p) = 13/3 − p/300 for bids in the
relavant range $1,000 to $1,300. In this case, HP’s expected contribution from a bid of p is
(p − 1000) · (13/3 − p/300). Differentiating again to determine the first order condition for

5
 the contribution maximizing p gives 23/3 − p∗ /150 = 0 or p∗ = $1, 150. The corresponding
probability of winning the deal is 13/3–1150/300 = 50% and the expected contribution is
.5 · 150 · 150 = $11,250.

7. You work as an operations manager at OfficeComfort, a company specializing in ergonomic

office chairs. Currently, you’re in a competitive bidding situation against two other companies,
OfficeElegance and ChairMasters. The objective is to secure a contract to supply 150 ergonomic
office chairs to a corporate client. The production cost for each of your office chairs is $135 per
unit.

(a) Suppose that you anticipate that both OfficeElegance’s bid and ChairMasters’ bid will vary
uniformly between $150 and $200 per unit. What is the optimal price per unit to bid?
(b) For what price is the probability that you win the bid at least 90%?

Solution:

(a) Your probability of winning the bid as a function of the price is ρ(p) = (4 − p/50)2 for
bids in the relevant range of $150 to $200 (with 1 below the range and 0 above the range).
Therefore, your expected contribution from a bid of p is 150 · (p − 135) · (4 − p/50)2 .
Differentiating this expression to determine the first order condition for the contribution-
maximizing p gives p∗ =$156.67. The corresponding probability of winning the bid is 75.1%
and the expected contribution $2,441.1.
(b) The probability of winning the bid as a function of the price is ρ(p) = (4 − p/50)2 . Thus we
set 0.9 = (4 − p/50)2 and solve for p within the range 150 to 200 and find $152.57. Thus for
prices of $152.57 and lower, the probability of winning the bid is at least 90%. Note that
this price is quite low, and very close to the lower bounds of the other two competitors.