BCP Business & Management ECRM 2022

Volume 40 (2023)

Application of Linear Programming on the Maximum-Profit of

Milk Tea Store
Yuewei Wang*
Department of Math, University of North Carolina at Chapel Hill, NC, US.
*Corresponding author: [email protected]
Abstract. The aim of this article is the application of the industrial profit optimization on milk tea store.
The linear programming model is used to determine the optimal production structure of amount of
raw material purchased. The first part optimizes only the single profit from the sales which are put
into the objective function. The second section of the article extends the model by the considering
factors of employee. This model will have more constraints. The model aims to meet the maximum
cost of raw materials, and meet the situation that the staff will have different abilities on making tea
and each tea for each person takes different time to complete. The model will be built based on real
data, and then will compare the real data results with the calculated results. Conclusions of the paper
demonstrate the applicability of the model and possibilities of its extensions.
Keywords: Optimization; Linear programming; Industrial cost management; Production cost.

1. Introduction
Milk tea stores are now a hot business all over the world. More and more people are choosing to
join the milk tea industry to start their own business. A variety of milk tea spots have come into
existence and the control of milk tea cost and sales and inventory is a very important part. In this
paper, a milk tea store will be analyzed that once opened but closed down in the United States in the
direction of ingredient purchasing and profit rationing, combined with the linear programming model
learned in the course [1].
Linear programming is an important model for optimization class problems. It has a mature theory
and is widely used by the class in practical applications. It has a unique advantage in studying the
optimal allocation of limited resources. It is convenient for users to make decision analysis in order
to make the best use of resources to obtain the best economic benefits [2]. There are also genetic
algorithms, neural networks, ant colony algorithms and other methods to find optimal solutions in
optimization class problems [3]. Linear programming is a common contemporary mathematical
model, which is very suitable for dealing with multi-species problems. There are a large number of
papers that use linear programming for industrial warehouse cost management as well as human
resource management [4,5]. Therefore, this paper will do multi-objective linear programming model
on the basis of milk tea cost control. All the data came from a milk tea store that has gone bankrupt
and has been made public with the owner's permission. This article will also examine why this milk
tea store went bankrupt by comparing the calculated results with the actual operating costs [6].

2. Model building and solving

2.1 Cost organization
We obtained through the Internet means that a brand of milk tea store has three hot commodities
which are whitening and skin care milk tea, skin clearing milk tea, fitness milk tea. The brand focuses
on nutrition, health, safe milk tea, which is popular with girls. The series of milk tea mainly consists
of five kinds of ingredients. They are milk, mung bean powder, pearl powder, honey, and lotus seed
powder. Each milk tea needs to consume different proportions of raw materials, and the profit margin
varies between each raw material. The relevant data is shown in the following Table 1.

358
BCP Business & Management ECRM 2022
Volume 40 (2023)

Table 1. Milk Tea Recipe and Profit Table

Name Milk Mung bean powder Pearl powder honey Lotus seed powder profit
Whightening tea 250 8 0.5 10 0 10
Skin-refreshing tea 250 10 0.5 10 10 8
Fitness tea 500 12 0 20 0 12

According to the feedback from the industry, the shelf life of each raw material is limited, and for
food safety. If too much raw material is supplied, it will cause the problem of raw material wastage,
which will cause additional burden to the enterprise. Therefore, this paper will focus on solving this
internal problem. At the same time, in order to take care of various tastes, the enterprise needs to
ensure enough at least 5 copies of whitening and skin care milk tea, 20 copies of skin clearing milk
tea and 30 copies of fitness milk tea every day. The maximum supply of milk is 30,000, the supply
of mung bean powder maximum is 5,000, the supply of pearl powder maximum is 160, honey 3,000,
lotus seed powder 3,000.
By analyzing the survey data and content of this paper, this paper will use linear programming as
the solution model of this paper. Here we set 𝑋1 copies of whitening milk tea, X2 copies of skin-
clearing milk tea and X3 copies of healthy milk tea are sold every day. The objective function is

Max = X1 + X2 + X3 , (1)

and it is subject to

250X1 + 250X2 + 500X3 ≤ 30000 (2)

8X1 + 10X2 + 12X3 ≤ 5000 (3)

0.5X1 + 0.5X2 + 0X3 ≤ 160 (4)

10X1 + 10X2 + 20X3 ≤ 3000 (5)

0X1 + 10X2 + 0X3 ≤ 3000. (6)

Here we use Python's PuPl library to solve the problem (see the attachment for the code). The final
result is as follows

Max value 𝑧 = 920.0, 𝑋1 : 40.0, 𝑋2 : 20.0, 𝑋3 : 30.0. (7)

Ultimately, we get that selling 40 servings of whitening and skin-clearing milk tea, 20 servings of
skin-clearing milk tea, and 30 servings of fitness milk tea will achieve the greatest profit given a
specific supply of ingredients.
2.2 Considering manual cost and other cost
At first the owner of the milk tea store was not very willing to disclose, but only gave a part of the
data. After the calculation was completed, the data was compared with the data of the owner of the
milk tea store and found to be similar to the actual situation [7]. However, the owner only disclosed
the subsequent data. So there is a more complicated model. Table 2 is available based on the data.

359
BCP Business & Management ECRM 2022
Volume 40 (2023)

Table 2. Components and Profits

Milk Mung Bean Pearl Lotus Sed
Name MilkSugar tea Honey Profit
essence powder Powder Powder
Whitening Tea 250 18 30 120 8 0.5 10 0 10
Skin-Refreshing
250 18 30 120 10 0.5 10 10 8
Tea
Fitness Tea 500 18 30 120 12 0 20 0 12
Signature Tea 10 18 30 120 0 0 0 0 10
Bubble Tea 10 18 30 120 0 0 0 0 10
Lychee Jelly Tea 10 18 30 120 1 1 0 0 12
Pudding Tea 10 18 30 120 1 1 0 0 12
Red Bean Tea 10 18 30 120 1 1 0 0 12
Brown Sugar Tea 10 18 30 120 1 1 0 0 12
Sencha tea 10 18 30 120 1 1 0 0 11
Jasmine Tea 10 18 30 120 2 1 0 0 12
Caramel
10 18 30 120 3 1 0 0 11
Macchiato
Chocolate Tea 10 18 30 120 1 0 0 0 14
Yuan Yang Tea 10 18 30 120 2 1 1 0 12
Cake Tea 10 18 30 120 3 2 1 0 9

The store has three employees, each of whom processes different types of milk tea at different
times, which has an impact on sales. Customers who came to the store to check the queue were too
many and would give up buying the relevant products [8]. Therefore, a reasonable arrangement of
staff scheduling and production of related raw materials is crucial to the output of the store. The staff's
production time for each type of milk tea is as follows in Table 3.

Table 3. Time for Making Each Milk Tea

Name A B C
Whitening Tea 1 2 1
Skin-Refreshing Tea 1.5 1.5 1
Fitness Tea 2 1.5 1
Signature Tea 1 2 1
Bubble Tea 1.5 1.5 1
Coconut Jelly Tea 2 1.5 1
Pudding Tea 1 2 1
Red Bean Tea 1.5 1.5 2
Brown sugar Tea 2 1.5 2
Sencha tea 1 2 2
Jasmine Tea 1.5 1.5 2
Caramel Macchiato 2 1.5 2
Chocolate Tea 1 2 2
Yuan Yang Tea 1.5 1.5 2
Cake Tea 2 1.5 2
Salary per hour 65 70 80

To simplify the model, the hourly rate is set to p here, and the corresponding labor cost per unit of
no-minute milk tea is represented by the following equation: to make the model more general, all
ingredients are represented here by 𝑥𝑖 ∈ [1, 𝑁] in Table 4. Employees are denoted by 𝑒𝑖 ∈ [1, 𝑁].

360
BCP Business & Management ECRM 2022
Volume 40 (2023)

The employee list each employee makes 𝑁(𝑐𝑖) ∈ [0, 𝑁] kinds of milk tea for sale every day. The
types of milk tea are denoted by 𝐶(𝑖) ∈ [1, 𝑁]. The following objective function is obtained.

Table 4. Symbols and meanings

Symbol Explanation
𝑒𝑖 ∈ [1，𝑛] Number of employees
𝑥𝑖 ∈ [1，𝑛] Number of raw materials needed for each milk tea
𝐶𝑗 ∈ [1，𝑛] Number of milk tea
𝑁𝑗 ∈ [0，𝑛] Number of sold out milk tea
𝑃𝑗 ∈ [1，𝑛] Profit for each milk tea
𝑅𝑖 ∈ [1，𝑛] Single day storage for raw materials
𝑇𝑒𝑐 Time for employee making each tea
𝑡 Work time

Objective function is
n n

Max = ∑ ∑ Pj Ni (8)
i=0 j=0

and it is subject to several restrictions as shown below.

Number of all milk tea sold

N = 𝑁 1 + 𝑁2 ⋯ + 𝑁𝑗. (9)

Number of raw materials used

𝑥𝑖 × 𝑁𝑗 ≤ 𝑅𝑖 . (10)

Time constraint
n n

∑ ∑ Tj Ni ≤ e × 𝑡. (11)
i=0 j=0

The model is
max 10𝑥1 + 11𝑥10 + 12𝑥11 + 11𝑥12 + 14𝑥13 + 12𝑥14 + 9𝑥15
+8𝑥2 + 12𝑥3 + 10𝑥4 + 10𝑥5 + 12𝑥6 + 12𝑥7 + 12𝑥8 + 12𝑥9 + 0 (12)
subject to
𝐶1 : 10𝑥1 + 10𝑥1 0 + 10𝑥11 + 10𝑥12 + 10𝑥13 + 10𝑥14 + 10𝑥15 + 8𝑥2 + 12𝑥3 + 10𝑥4 + 10𝑥5
+ 10𝑥6 + 10𝑥7 + 10𝑥8 + 10𝑥9 ≤ 30000
𝐶2 : 18𝑥1 + 18𝑥10 + 18𝑥11 + 18𝑥12 + 18𝑥13 + 18𝑥14 + 18𝑥15 + 18𝑥2 + 18𝑥3 + 18𝑥4 + 18𝑥5
+ 18𝑥6 + 18𝑥7 + 18𝑥8 + 18𝑥9 ≤ 30000
𝐶3 : 30𝑥1 + 30𝑥10 + 30𝑥11 + 30𝑥12 + 30𝑥13 + 30𝑥14 + 30𝑥15 + 30𝑥2 + 30𝑥3 + 30𝑥4 + 30𝑥5
+ 30𝑥6 + 30𝑥7 + 30𝑥8 + 30𝑥9 ≤ 30000
𝐶4 : 10𝑥1 + 120𝑥10 + 120𝑥11 + 120𝑥12 + 120𝑥13 + 120𝑥14 + 120𝑥15 + 8𝑥2 + 12𝑥3 + 120𝑥4
+ 120𝑥5 + 120𝑥6 + 120𝑥7 + 120𝑥8 + 120𝑥9 ≤ 300000
𝐶5 : 𝑥10 + 𝑥11 + 𝑥12 + 𝑥13 + 𝑥14 + 𝑥15 + 𝑥4 + 𝑥5 + 𝑥6 + 𝑥7 + 𝑥8 + 𝑥9 ≤ 3000000.
𝐶6 : 𝑥1 ≥ 30, 𝑥2 ≥ 20, 𝑥3 ≥ 30, 𝑥4 ≥ 30, 𝑥5 ≥ 30, 𝑥6 ≥ 30, 𝑥7 ≥ 30, 𝑥8 ≥ 30, 𝑥9 ≥ 30，𝑥10
≥ 30，𝑥11 ≥ 30，𝑥12 ≥ 30，𝑥13 ≥ 30，𝑥14 ≥ 30，𝑥15 ≥ 30.

361
BCP Business & Management ECRM 2022
Volume 40 (2023)

In addition, the variables 𝑥𝑖 (1 ≤ 𝑖 ≤ 15) are all integers.

The maximize output is 𝑧 = 12830.0，while for each variable: {𝑥1 : 30.0, 𝑥10 : 30.0, 𝑥11 : 30.0,
𝑥12 : 30.0 , 𝑥13 : 620.0 , 𝑥14 : 0.0 , 𝑥15 : 30.0 , 𝑥2 : 20.0 , 𝑥3 : 30.0 , 𝑥4 : 30.0 , 𝑥5 : 30.0 , 𝑥6 : 30.0 ,
𝑥7 : 30.0, 𝑥8 : 30.0, 𝑥9 : 30.0}. From this, the number of individual raw materials that should be
purchased can be derived, which is the maximum profit when maximizing both employee efficiency
and output. This calculation was compared with the actual operating profit, and after deducting some
fixed costs, the operating profit was actually lower than the calculated result, so it is likely that this
led to the bankruptcy of the store.

3. Conclusion
The aim of the paper is to create proposal of an economic model, which would optimize production
of a milk tea company with respect to a maximization of the profit from selling its production.
Through the application of the model on the real data to get the final result, the maximized profit is
smaller than cost in real life, which leads to the bankruptcy of the milk tea store [9]. If the owner of
the store can buy the similar quantity of each type raw material as calculated, he can earn economic
profit. Because the store itself is owned by the owner, the effect of rent is not considered in this
calculation, and only the cost of goods and labor is included in the calculation [10].
The model could be further extended. It is possible to alter it for conditions of specific milk tea
store with different types of products or extend the model structure, so the scope would be wider. One
of proposal is that the model could be enriched by analysis of different elasticities such as price
elasticity, especially prices of different rental among different places and number of substitutes
among consumer choices.

References
[1] Gary D, Ferrier C, Knox-Lovell A. Measuring cost efficiency in banking: Econometric and linear
programming evidence, Journal of Econometrics, 1990, 46: 229-245.
[2] Frederick V. The Minimum-Cost Dairy Feed (An Application of "Linear Programming"), Journal of Farm
Economics, 1951, 33(3), 299-310.
[3] Amjady N. Short-term hourly load forecasting using time-series modeling with peak load estimation
capability. IEEE Transactions on Power Systems, 2001, 16(3): 498-505.
[4] Fred H, Sidney W. Hess A Linear Programming Approach to Production and Employment Scheduling.
Management Technology, 1960, 1: 46-51.
[5] Vishwa N. M., Ram B. M., Peter K. A. et al. Profit Optimization Using Linear Programming Model: A
Case Study of Ethiopian Chemical Company. American Jour. of Biological and Environmental Stats.,
2015, 1(2): 51-57.
[6] Charnes A. Cooper Management Models and Industrial Applications of Linear Programming.
Management Science 1957, 4(1):38-91.
[7] Aigner D. J., Chu S.-F. On Estimating the Industry Production Function. American Economic Review
1968, 58(4): 826–839.
[8] Lovell C.A.K., Grosskopf S., Ley E. et al. Linear programming approaches to the measurement and
analysis of productive efficiency. Top, 1994, 2: 175–248.
[9] Assessing competition in the banking industry: A multi-product approach. Journal of Banking and Finance,
2015, 50: 340-362.
[10] Baki S M, Cheng J K. A Linear Programming Model for Product Mix Profit Maximization in A Small
Medium Enterprise Company. International Journal of Industrial Management, 2021, 9(1): 64-73.

362