upload_3dp5dsk7w4jpmya4irtj4gz4et.pdf

NEYART Journal

ISSN: 2992-7161

Page

132

ASSESSMENT OF CUSTOMER SATISFACTION IN A SERVICE

COMPANY USING DISCRETE-EVENT SIMULATION

EVALUACIÓN DE LA SATISFACCIÓN DEL CLIENTE EN UNA

EMPRESA DE SERVICIOS APLICANDO LA SIMULACIÓN DE

EVENTOS DISCRETOS

Del Socorro Corral María

Tecnológico Nacional de México/I. T. De Chihuahua II

https://orcid.org/0009-0008-1754-0480

maria.c@chihuahua2.tecnm.mx

Silva Máynez Lucia Xiomara

Tecnológico Nacional de México/I. T. De Chihuahua II

https://orcid.org/0000-0002-0130-6592

lucia.sm@chihuahua2.tecnm.mx

Reyes Ledezma Daniela María

Tecnológico Nacional de México/I. T. De Chihuahua II

https://orcid.org/0009-0001-1885-6312

dani3la.reyes87@gmail.com

Torres Lozano Dayami Arisai

Tecnológico Nacional de México/I. T. De Chihuahua II

https://orcid.org/0009-0004-4394-2524

dayami.tl@chihuahua2.tecnm.mx

Medina Molina Yearim

Tecnológico Nacional de México/I. T. De Milpa Alta

https://orcid.org/0009-0000-4128-2524

yearim.mm@milpaalta.tecnm.mx

DOI:

https://doi.org/10.61273/neyart.v4i4.223

|

Received

: 29/05/2026

|

Accepted

: 03/08/2026

|

Published

: 01/09/2026

This work is licensed under

an international

Creative Commons Attribution 4.0 license.

NEYART Journal

ISSN: 2992-7161

Issue 4 | Vol. 4

–

No. 4 | July

–

December 2026 |

Page

133

Original Research Article

Abstract

—

This article presents an analysis of wait times in the university cafeteria using ProModel. The

simulation shows that even when the cafeteria’s capacity demand is met (in this case, with 75% efficiency

and no customer abandonment), the distribution of time within a single resource provider can operate

inefficiently. Customers wait an average of nearly 12 minutes to receive a service that takes 9 minutes

and 306 seconds longer than it should, which lowers their assessment of service quality.

The researchers used the SERVQUAL framework to determine that the issue was not a lack of resources,

but rather a misalignment between peak demand and service supply. Instead of hiring more employees,

their recommendation is to implement flexible scheduling with the current workforce during peak hours.

Finally, they recommended that future research integrate these simulations with direct assessments of

customer satisfaction.

Keywords

—

Mathematical

model,

Quantitative

method,

Service

enterprise,

Simulation,

Statistical

analysis.

Resumen

—

Este artículo presenta un Análisis del tiempo de espera en el comedor universitario utilizando

ProModel. Puede observarse a partir de la simulación que incluso cuando se satisface la demanda de

capacidad del comedor (en este caso, con un 75% de eficiencia y sin abandono de clientes), la distribución

del tiempo dentro de un único proveedor de recursos puede operar de manera ineficiente. Los clientes

permanecen en espera durante un promedio de casi 12 minutos para recibir un servicio que dura 9 minutos

306 segundos más de lo que debería tardar lo cual reduce su evaluación de la calidad.

Los investigadores utilizaron el marco SERVQUAL para determinar que no se trataba de una falta de

recursos, sino de una desalineación entre la demanda máxima y la oferta de servicio. En lugar de más

empleados, su recomendación es la programación flexible con la fuerza laboral actual durante las horas

pico.

Por

último,

recomendaron

que

las

investigaciones

futuras

integren

estas

simulaciones

con

evaluaciones directas de la satisfacción del consumidor.

Palabras Clave

—

Análisis estadístico, Empresa de servicios, Método cuantitativo, Modelo matemático,

Simulación.

INTRODUCTION

For enterprises that rely on providing services to generate revenue, customer satisfaction is the most core

performance indicator, which directly links three key dimensions that determine the survival or demise of

NEYART Journal

ISSN: 2992-7161

Issue 4 | Vol. 4

–

No. 4 | July

–

December 2026 |

Page

134

Original Research Article

an enterprise: the first is whether customers will continue to choose the enterprise's services, that is,

customer retention; the second is the public's evaluation of the enterprise's brand, that is, brand reputation;

the last is whether the enterprise can avoid the risk of being eliminated by the market and sustain its

operations long-term. To maintain its competitiveness in a fiercely competitive market, an enterprise must

never arbitrarily modify its existing service processes after all, even adjusting a queuing numbering rule

or

adding

a

new

service

window

could

trigger

chain

reactions

across

the

entire

system

and

cause

disruptions. Before actually implementing any changes to real-world processes, the enterprise must first

obtain a reliable tool to simulate in advance all scenarios that may arise in the entire service process after

the changes, identify potential problems, and only then proceed with implementation.

Discrete-event simulation, referred to as DES, is exactly such a quantitative analysis tool. It does not

require the enterprise to conduct any experiments in its physical brick-and-mortar stores or service outlets.

Instead, it only requires building an identical simulation model of the real service system, and then this

virtual model can be used to calculate all operational metrics that the enterprise cares about: for example,

how many people will queue up, how long everyone will wait to receive service, the maximum business

volume that the entire system can handle, and whether resources such as labor and site space are being

utilized efficiently. The entire process of calculation and analysis will never disrupt the normal operation

of real business activities such as offline brick-and-mortar stores and online customer service, and will

not cause any inconvenience to actual customers. The following sections will sequentially explain the

background, rationale, objectives, and limitations of this study.

International Background

Currently,

a

method

called

discrete

event

simulation

is

commonly

used

internationally

to

optimize

customer service experience. A concrete application case of this method was implemented in Colombia

in 2020

—

researchers Heredia-Acevedo, Ceballos, and Sánchez-Torres used Simul8® software to build

a complete service process simulation model for a local SME operating in the fast food industry. Through

this model, they identified optimization directions that could shorten customers' waiting time, but at the

same

time,

they

also

pointed

out

the

limitations

of

this

method:

it

cannot

accurately

reproduce

the

complete processes of those pure service-oriented scenarios.

Similarly, Mendoza Casseres, González Conde, Corcho Martínez, and Berdugo Alonso used discrete-

event simulation to analyze the queuing

problem in the emergency room of a healthcare provider in

NEYART Journal

ISSN: 2992-7161

Issue 4 | Vol. 4

–

No. 4 | July

–

December 2026 |

Page

135

Original Research Article

Barranquilla, finding that disorganized scheduling directly affected patient stay times and referrals, which

negatively impacted patients’ perception of the service received.

In Brazil, da Rocha Nascimento, dos Santos, da Silva, Bueno, and Machado (2021) applied discrete-event

simulation using Arena software to improve queue management at a bank branch ( ), demonstrating that

reassigning staff significantly reduced wait times and increased satisfaction among users of the financial

service.

In Indonesia, Firmansyah and Saputra (2021) studied the effect of implementing a queueing system on

customer satisfaction, concluding that reducing perceived wait times has a direct and positive impact on

customers’ assessment of the service they received.

In Malaysia, a study published in the International Journal of Industrial Management applied discrete-

event

simulation

to

optimize

a

hypermarket’s

queuing

system,

showing

that

congestion

at

the

cash

registers, caused by insufficient staffing, significantly reduced customer satisfaction, and that simulating

different scenarios allowed for the proposal of a more efficient resource configuration (Application of

Discrete Event Simulation for Enhancing Queuing System at Lotus Alor Setar, 2025).

In Turkey, Aslan and Özderiir (2021) developed a simulation application for a university student dining

hall, demonstrating that modeling arrival and service times makes it possible to anticipate the formation

of lines and design strategies to improve the user experience of the food service.

In the United States, Curin, Vosko, Chan, and Tsimhoni (2005) used discrete-event simulation to reduce

service time at a high-demand fast-food restaurant on a college campus, achieving significant reductions

in wait times by reconfiguring the workstations.

Finally, Kumar (2005) analyzed, from an empirical perspective, the competitive impact of improving

service processes on customers’ waiting experiences in retail markets, demonstrating that the perception

of waiting time is a critical determinant of satisfaction and repurchase intent.

Latin American Context

In the national context (Peru), various studies have examined the application of discrete-event simulation

in service companies with the aim of improving customer service and satisfaction. Sotelo Seguil (2017),

in his thesis for the degree of Industrial Engineer at the Pontifical Catholic University of Peru, designed

and implemented a discrete-event simulation model to improve public safety services in the district of

NEYART Journal

ISSN: 2992-7161

Issue 4 | Vol. 4

–

No. 4 | July

–

December 2026 |

Page

136

Original Research Article

San Martín de Porres, concluding that the model made it possible to identify feasible improvements in

the allocation of operational resources.

Similarly, a study conducted at the Pontifical Catholic University of Peru used discrete-event simulation

to evaluate the level of service in the in-person customer service operations of a telecommunications

company, finding that simulating different staffing scenarios made it possible to reduce wait times and

increase the customer satisfaction metric reported by the company.

Similarly, at César Vallejo University, a discrete-event simulation model was developed to reduce wait

times in the outpatient department of a public hospital in Chimbote; the results showed reductions of over

90% in the average wait time in line, which led to a substantial improvement in the satisfaction of the

patients treated (Calla Delgado, 2020).

Complementarily,

Gutiérrez,

Ramos,

Uribe,

Ortega-Loayza,

Torres,

Montesinos,

and

León

(2009)

analyzed the relationship between wait time and user satisfaction at the central pharmacy of a general

hospital in Lima, concluding that there is a significant association between the two variables, which

supports the relevance of addressing service times as a strategy for service improvement.

In the consumer goods sector, a thesis from the Pontifical Catholic University of Peru applied discrete-

event simulation to the cookie production and packaging process at a consumer goods company, with the

goal of increasing the process’s productivity and profitability, demonstrating the tool’s usefulness beyond

purely production-related processes.

Meanwhile,

Tamashiro

and

Yacarini

Vadillo

(2023),

in

a

study

published

in

the

journal

*Industrial Data* of the National University of San Marcos, applied a discrete simulation model using

ProModel

software

to

improve

the

productivity

of

a

Peruvian

manufacturing

company’s

production

process,

after

first

evaluating

the

impact

of

these

improvements

on

the

satisfaction

of

the

process’s

internal customers.

In

the

financial

sector,

Alarcón

Bozzo

and

Díaz

Aroco

(2018),

in

their

undergraduate

thesis

at

the

Universidad Peruana del Norte, designed a simulation system to reduce wait times in the operations area

of

a

bank

branch

in

Cajamarca,

demonstrating

that

simulation

allows

for

anticipating

the

effects

of

different service configurations on the quality of service as perceived by the customer.

NEYART Journal

ISSN: 2992-7161

Issue 4 | Vol. 4

–

No. 4 | July

–

December 2026 |

Page

137

Original Research Article

Study Objective

The objective of this study is

to

evaluate customer satisfaction at

a service company by

applying a

discrete-event simulation model that represents the service process, identifies the critical factors affecting

wait and service times, and proposes improvement scenarios that contribute to increasing the level of

satisfaction perceived by customers.

Rationale for the Study

From a practical standpoint, this research is justified because it provides the service company under study

with an analytical tool that allows it to anticipate the behavior of its service processes without incurring

the

costs

or

risks

of

experimenting

directly

on

the

actual

system,

which

is

particularly

valuable

in

organizations where operational errors have an immediate impact on the customer experience. As noted

by da Rocha Nascimento, dos Santos, da Silva, Bueno, and Machado (2021), discrete-event simulation

allows organizations to test alternative service configurations and anticipate their effects on customer

satisfaction before implementing them, thereby providing management with objective information for

decision-making.

This study conducts a theoretical integration, combining queuing theory, simulation methods, and classic

evaluation

models

to

connect

the

quality

of

service

processes

with

customer

satisfaction,

filling

the

theoretical gap between the two.

The core foundational framework supporting the entire study is the SERVQUAL model, which was

proposed by Parasuraman, Zeithaml, and Berry in (1988). Under the definitions of this model, customer

satisfaction is not a vague perception score, but a clear gap value: on one side is the customer's expectation

of

the

service

before

receiving

it,

and

on

the

other

side

is

the

actual

perception

generated

after

the

customer completes the service experience. The gap between these two is the core source of satisfaction.

To

measure

this

gap,

assessments

must

be

carried

out

from

five

dimensions,

namely

reliability,

responsiveness, assurance, empathy, and tangibility. The simulation model used in the study can conduct

assessments one by one against the specific conditions of these five dimensions, and can also make

predictions for future changes in these dimensions.

The research method adopted in this study uses a set of modeling and simulation processes specifically

designed for the field of industrial engineering. This set of processes is not pieced together arbitrarily; it

has

undergone

validity

verification.

Anyone

who

operates

step

by

step

in

accordance

with

the

NEYART Journal

ISSN: 2992-7161

Issue 4 | Vol. 4

–

No. 4 | July

–

December 2026 |

Page

138

Original Research Article

requirements of the process can repeatedly produce the same results. The entire process contains six

interconnected steps in sequence: data collection, distribution fitting, conceptual model construction,

verification, validation, and scenario analysis.

As Law (2015) pointed out, to ensure that research results can truly and reliably reflect the actual system,

the two steps of verification and validation must be completed correctly and well

—

if mistakes are made

in these two steps, all conclusions drawn afterward cannot be aligned with the real system in reality, and

the results of the entire study will therefore be untenable.

Study Limitations

This

study

has

several

limitations,

these

deficiencies

directly

affect

the

scope

of

application

of

the

research conclusions, and the conclusions cannot be applied in all scenarios.

First, we only analyzed the situation of a single company, and the research results obtained cannot be

generalized to the scenarios of other organizations, let alone directly apply this set of conclusions to more

different environments.

Second,

all

the

research

data

we

used

only

came

from

a

fixed

period

of

time,

and

the

probability

distribution calculated from these data may shift with changes in customer demand and seasonal patterns,

which will not be consistent with the actual situation.

Third, when we measured customer satisfaction, we only relied on the two indicators of waiting time and

service level, and the subjective perceptions that affect users' core judgments on service quality were all

omitted and not included in the satisfaction assessment.

Finally, the simulation software we selected has its own limitations. It cannot support us to fully reproduce

the complexity of the real system that we need to restore in the research, and the level of detail that can

be ultimately achieved is restricted by the software itself from beginning to end.

DEVELOPMENT

Research Paradigm

This research was conducted under the positivist paradigm, which assumes the existence of an objective

reality that can be measured, quantified, and explained through the application of the scientific method.

This paradigm is relevant to the study insofar as the assessment of customer satisfaction is approached

through observable variables and a simulation model that allows for the numerical representation of the

service

system’s

behavior,

thereby

minimizing

the

researcher’s

subjective

influence

on

the

results

NEYART Journal

ISSN: 2992-7161

Issue 4 | Vol. 4

–

No. 4 | July

–

December 2026 |

Page

139

Original Research Article

obtained. In this regard, Hernández-Sampieri and Mendoza (2018) note that the quantitative approach,

characteristic

of

the

positivist

paradigm,

begins

with

an

idea

that

is

progressively

refined;

once

the

objectives and research questions are defined, hypotheses and variables are established, a plan is devised

to test them, and the variables are measured within a specific context, with the aim of identifying patterns

of behavior and testing theories through statistical analysis of the collected data.

Level of Research

The study was conducted at a descriptive level, since its purpose was not to deliberately manipulate the

variables of the real system, but rather to characterize the behavior of the customer service process at the

university cafeteria based on the observed data, and subsequently represent it using a simulation model

that would allow for a description of its performance in different operational scenarios. In this regard,

Arias

(2012)

defines

descriptive

research

as

research

whose

purpose

is

to

characterize

a

fact,

phenomenon, individual, or group in

order to

establish

its structure or behavior, without necessarily

explaining the causes that give rise to it.

Participants

This study was conducted on the premises of the campus cafeteria of a public university in Mexico. All

observation subjects were regular patrons of this cafetería long-time repeat customers who frequented the

establishment. The study focused on the details of the service these regular customers received at the

cafeteria.

Researchers

recorded

a

total

of

68

complete

sets

of

observation

data,

which

were

used

to

estimate three customer-related values: first, the time customers arrived at the store; second, the total

length of time customers spent from the start to the end of receiving service; third, the level of customers'

satisfaction with that service instance. The estimation results calculated from these observation data could

reach a confidence level of 90% this value is an indicator used to measure the reliability of the estimation

results, and the original text only provided this value without adding any additional explanations.

Procedure

Data

collection

took

place

during

the

January

–

June

2025

semester,

a

period

during

which

direct

observations were made at various times of high and low customer traffic, with the aim of capturing the

natural variability of the service process. During this stage, customer arrival times, service times at each

service point, the number of people in line, and relevant incidents in the process were recorded; this

information

was subsequently used as input for building

and calibrating the simulation

model. Data

NEYART Journal

ISSN: 2992-7161

Issue 4 | Vol. 4

–

No. 4 | July

–

December 2026 |

Page

140

Original Research Article

collection was conducted in a non-intrusive manner, taking care at all times not to disrupt the normal flow

of service.

Ethical Considerations

The study was conducted in accordance with the ethical principles governing research involving human

subjects, ensuring respect for the dignity, privacy, and well-being of the participants at all times. Customer

observation was conducted anonymously, without recording any personal data that would allow for their

identification , and the information collected was used exclusively for academic purposes and to improve

the service, safeguarding the confidentiality of the data obtained. Furthermore, prior authorization was

obtained from the authorities responsible for the university cafeteria to conduct the observation of the

process;

every

effort

was

made

not

to

interfere

with

the

establishment’s

normal

operations;

and

the

research was conducted at all times in accordance with the principles of scientific integrity, avoiding the

manipulation or alteration of the collected data and respecting the authorship of the sources consulted.

Procedure for Data Analysis and Statistical Analysis

Researchers first collected three types of data: the time customers arrived at the store, the time required

to prepare and serve each customer's order, and the store's actual operational performance. These data

could not be used directly; they first had to be fitted into an existing probability distribution model to

eventually

build a simulation model

capable of

replicating real

scenarios, laying the groundwork for

subsequent research.

After completing the model, they used the ProModel software to carry out specific simulation work. They

first reproduced the entire customer flow process of the coffee shop in the software

—

they could accurately

simulate the full movement path of customers from entering the store, placing orders, picking up meals,

to leaving. Through this simulation, they first evaluated the store's current actual operational status, then

projected the level of operational improvement that different types of adjustments to the store would bring

about respectively.

However, before using this model for any formal analysis, they did not dare to apply it directly; instead,

they conducted two rounds of verification and checks. They compared the results output by the model

with the store's real daily operational conditions, and then cross-checked them against various indicators

used to measure customer satisfaction. Only after confirming that the model's results were accurate and

fully capable of supporting subsequent analysis did they officially launch the follow-up research work.

NEYART Journal

ISSN: 2992-7161

Issue 4 | Vol. 4

–

No. 4 | July

–

December 2026 |

Page

141

Original Research Article

Data Analysis

This section presents and interprets the results obtained from running the discrete-event simulation model

developed in ProModel software, corresponding to the service process at the university cafeteria under

study. The analysis is organized around three complementary areas: the behavior of the customer flow

entering the system, the overall productivity of the service process, and the composition of the cycle time

experienced by the customer, with the aim of identifying the factors that influence perceived satisfaction

and of providing quantitative evidence to support opportunities for system improvement.

Customer Flow Behavior

Table 1 presents the results obtained:

Table 1

.

Customer Flow Behavior

.

Customer Flow Indicator

Value

Successful Arrivals (successfully served)

36

Arrivals in the system (in process at the end of the run)

12

Failed arrivals (rejected or abandoned)

0

Source

: Author’s

own analysis

.

During the simulated period, the system recorded a total of 36 successful arrivals

—

that is, customers who

entered the café and completed the service cycle without incident. Additionally, 12 arrivals remained in

the system at the end of the simulation run; these correspond to customers who were at some stage of the

process

—

waiting, being served, or leaving

—

without having yet completed their service cycle when the

observation period ended, and whose inclusion is important to avoid underestimating the actual load the

system faced. After this simulation, all customers who were originally supposed to access the service

successfully entered the system. Not a single customer failed to complete their visit; not one person was

left out.

Although everyone waited for a considerable amount of time, and many customers queued for a long

stretch before it was their turn to receive service, the carrying capacity of this system fully accommodated

the needs of all customers, not losing a single guest. Even though everyone waited a long time, the

NEYART Journal

ISSN: 2992-7161

Issue 4 | Vol. 4

–

No. 4 | July

–

December 2026 |

Page

142

Original Research Article

maximum workload the system could handle remained sufficient, and there was never a situation where

it could not accommodate more people.

For

the

upcoming

system

upgrade,

there

is

no

need

to

worry

about

the

system

being

unable

to

accommodate more users, nor to patch any capacity loopholes. We only need to focus on improving the

user experience for customers.

System Productivity

Table 2

presents the process’s productivity.

Table 2

.

System Productivity

.

Performance Indicator

Value

System Productivity

75%

Source

: Author’s

own work

.

The model estimated a productivity rate of 75% for the simulated service point, defined as the proportion

of the total simulation time during which the service resource was effectively occupied serving customers.

This value indicates a reasonably high utilization of the installed capacity , but at the same time shows

that a quarter of the server’s a

vailable time was not devoted to direct customer service, either due to

periods of inactivity or interruptions associated with the dynamics of the process. This level of production

efficiency is very common in service industries with fluctuating demand, and university canteens are a

typical example of such scenarios. After all, the number of diners in a university canteen inherently

fluctuates with class start and end times: during peak hours, long queues form at service counters, while

during off-peak hours, staff may not receive a single order for hours on end. This alternating pattern of

slack

and

busyness

in

production

efficiency

is

by

no

means

an

exceptional

case

across

all

service

industries where demand shifts constantly.

Using this single indicator alone cannot lead to the conclusion that the service suffers from low efficiency

or wasted resources. One cannot fixate on this single figure to hastily judge that the establishment’s

service is poorly run, that idle staff constitute a waste of resources

—

this single indicator is fundamentally

incapable of supporting such a conclusion.

To determine whether employees’ idle time actually impacts customer experience, this indicator must be

evaluated alongside the total duration of a customer’s visit, from enteri

ng to exiting the establishment.

Only by measuring the two indicators together can one clarify whether the idle time when staff are not

NEYART Journal

ISSN: 2992-7161

Issue 4 | Vol. 4

–

No. 4 | July

–

December 2026 |

Page

143

Original Research Article

working truly slows down diners’ entire process and diminishes their experience. Relying solely on the

aforementioned production efficiency indicator cannot yield a reliable conclusion.

Composition of the Customer Cycle Time

Table 3 presents the behavior of the process times.

Table 3

.

Breakdown of the customer cycle time

.

Cycle Time Component

Average Duration

Average Service Time (Total Cycle)

21.60 min

Waiting time in line

11.84 min

Actual service time (effective service)

8.85 min

Hold time

0.62 min

Server downtime

12.75 min

Source

: Author’s

own calculations

.

The average service time recorded by the model was 21.60 minutes, a figure that represents the complete

cycle experienced by the customer from the moment they enter the system until the end of their service.

When breaking down this value into its constituent parts, we see that the wait time in the queue amounts

to 11.84 minutes

—

a figure that accounts for the largest proportion of the total cycle and, in relative terms,

is the most decisive factor in customer satisfaction, since it corresponds to a period during which the user

remains inactive, waiting to be served, without receiving any direct value from the service.

In contrast, the actual service time

—

that is, the period during which the customer actually receives the

service

—

was

8.85

minutes,

a

figure

considerably

lower

than

the

wait

time.

This

ratio

highlights

a

significant imbalance within the process: for every minute the customer spends being served, they spend

nearly an additional minute and a half waiting for that service

—

a situation that, if left unaddressed, can

progressively erode the perception of service quality, even when, as noted in the previous section, no

customer actually abandoned the system. The process stalled for 0.62 minutes, which is a mere over 30

seconds of lag. This data indicates that the downstream link further along the sequence encountered a

temporary restriction, making it unable to process the incoming business at the originally planned speed.

In addition, there was a cumulative 12.75 minutes of server downtime. Adding up this period of shutdown,

the overall idle rate reached 25%

—

to put it plainly, throughout the entire service cycle, normal work

NEYART Journal

ISSN: 2992-7161

Issue 4 | Vol. 4

–

No. 4 | July

–

December 2026 |

Page

144

Original Research Article

could not be carried out for a full quarter of the time. This idle situation occurred because there were not

enough customers requiring service.

The core finding of this study is that two seemingly completely contradictory scenarios

—

long wait times

and extensive service idleness

—

actually coexist. On one hand, customers have to queue for a very long

time before receiving service; on the other hand, there is still a huge amount of time where servers or

service staff are idle and unable to start work. This coexistence of abnormal circumstances shows that all

problems stem from uneven demand distribution, not insufficient total service capacity. It is not that the

total volume of service that can be provided cannot support the visiting customers, but that the time periods

during which people in need of service arrive are extremely uneven: they either crowd in all at once,

clogging the entire process, or no one comes for ages, leaving staff waiting idly. The optimization work

to be carried out next must focus on two tasks: adjusting demand to make it more even, and dynamically

allocating resources, rather than blindly increasing the total service capacity. Do not follow the old path

of

expanding

capacity

by

adding

more

servers

and

more

staff.

First

smoothing

out

the

time

gaps

in

demand, then flexibly deploying backup resources in line with changes in passenger flow, is the correct

direction to solve the problem.

Discussion

The results obtained through the discrete-event simulation model allow us to compare the behavior of the

university cafeteria’s service system with the findings reported in previous research that has addressed

similar issues in different contexts and service sectors, thereby enabling us to assess both the consistency

of the results obtained and the specific contribution this study makes to existing knowledge regarding the

relationship between processing times and customer satisfaction.

First, the complete absence of failed arrivals recorded in the system studied is consistent with the findings

reported by Heredia-Acevedo, Ceballos, and Sánchez-Torres (2020), who, by simulating the customer-

process at a fast-food SME in Colombia, concluded that discrete-event simulation allows for the early

identification of whether installed capacity is sufficient to absorb demand, thereby preventing customer

loss before intervening in the actual system. In the case of the café analyzed, this favorable result suggests

that, unlike other service contexts where saturation is the main problem, the establishment’s capacity does

not, on its own, represent a structural constraint on service.

NEYART Journal

ISSN: 2992-7161

Issue 4 | Vol. 4

–

No. 4 | July

–

December 2026 |

Page

145

Original Research Article

However, the finding regarding the marked difference between the waiting time in line (11.84 minutes)

and the actual service time (8.85 minutes) is consistent with the observations of Kumar (2005), who

argues that the perception of waiting

time is

a

critical

determinant

of customer satisfaction in

retail

settings, regardless of how efficiently the service itself is delivered. This approach takes on particular

relevance in the present study, insofar as the system evaluated does not exhibit obvious deficiencies in

the duration of the actual service, but rather in the proportion of time the customer remains inactive before

being served, which reaffirms that satisfaction does not depend solely on the server’s speed, but on the

comprehensive management of customer flow.

Similarly, the 75% productivity level obtained in the simulation, along with a server downtime of 12.75

minutes, aligns with the findings reported by da Rocha Nascimento, dos Santos, da Silva, Bueno, and

Machado (2021) in their study on queue management at a Brazilian bank branch, where they identified

that the coexistence of periods of staff inactivity with prolonged customer wait times is not due to a lack

of resources, but rather to an uneven distribution of demand throughout the business day. This consistency

reinforces

the

interpretation

that

the

central

problem

of

the

system

under

study

relates

to

the

synchronization between staff availability and peak customer traffic times, rather than to the quantity of

available resources.

Along the same lines, the findings are also consistent with the study by Firmansyah and Saputra (2021),

who concluded that reducing perceived wait times directly affects customers’ assessment of the service

received, even when the actual service time remains unchanged. Similarly, the behavior observed in the

composition of the customer cycle time is compar

able to that described for the “ ” queuing system at a

hypermarket

in

Malaysia,

where

congestion

caused

by

occasional

staff

shortages

at

certain

times

significantly reduced customer satisfaction

—

a situation

that was addressed by simulating

alternative

resource allocation scenarios (Application of Discrete Event Simulation for Enhancing Queuing System

at Lotus Alor Setar, 2025).

At the national level, the results of this study align with those reported by Calla Delgado (2020) in the

outpatient

department

of

a

public

hospital

in

Chimbote,

where

the

application

of

a

discrete-event

simulation model revealed that wait times were the most critical component of the care cycle, and with

the findings of Gutiérrez et al. (2009), who found a significant association between wait time and user

satisfaction in a hospital pharmacy service in Lima. This consistently supports the interpretation that wait

NEYART Journal

ISSN: 2992-7161

Issue 4 | Vol. 4

–

No. 4 | July

–

December 2026 |

Page

146

Original Research Article

time

—

and not necessarily service time

—

is the variable with the greatest impact on the perception of

quality in various types of service organizations within the Peruvian context.

Likewise, the results obtained at the university cafeteria are consistent with those reported in the research

conducted at the Pontifical Catholic University of Peru on in-person service at a telecommunications

company, in which simulating different staffing scenarios made it possible to reduce wait times and

improve the satisfaction indicator reported by the organization. This suggests that, as in the present study,

interventions aimed at balancing human resource availability with demand variability constitute a more

effective strategy than simply expanding installed capacity.

From a theoretical perspective, the results obtained can be explained in light of the SERVQUAL model

proposed

by

Parasuraman,

Zeithaml,

and

Berry

(1988),

in

which

the

responsiveness

dimension

—

understood as the willingness to attend to and provide service promptly

—

constitutes one of the most

decisive gaps between customer expectations and perceptions. The findings of this study suggest that,

although actual service time does not

represent

a weakness

in

the system,

the perception of service

responsiveness could be compromised by long wait times, which would reduce the overall evaluation of

this dimension by customers of the university coffee shop.

Finally, the methodological consistency of the results obtained is also supported by Law (2015), who

emphasizes that the validity of a simulation model depends largely on whether its outputs reliably reflect

the

relationships

that

actually

occur

in

the

real

system.

In

this

regard,

the

observed

correspondence

between the results of this study and the findings reported in research conducted in different sectors and

countries

—

healthcare,

banking,

telecommunications,

retail,

and

food

service

—

constitutes

further

evidence of the robustness of the model developed, insofar as the same behavioral pattern

—

characterized

by the predominance of wait time over service time

—

is consistently repeated across diverse service

systems.

Conclusions

The application of the discrete-event simulation model to the service process at the university cafeteria

made it possible to achieve the objective set forth in this study by quantitatively evaluating customer

satisfaction based on the observed behavior in terms of wait times, service times, and system occupancy,

without the need to directly intervene in the establishment’s actual operations during the study period.

NEYART Journal

ISSN: 2992-7161

Issue 4 | Vol. 4

–

No. 4 | July

–

December 2026 |

Page

147

Original Research Article

First, it is concluded that the system’s installed capacity is sufficient to meet the recorded demand, as no

failed arrivals occurred during the simulated period. This result rules out insufficient resources as the

primary

cause

of

potential

customer

dissatisfaction

and

shifts

the

focus

to

how

those

resources

are

distributed throughout the service day.

Second, it is concluded that the factor that most significantly affects the total duration of the service cycle

is the wait time in the queue, which far exceeds the actual time spent being served. This disproportion

constitutes

the

study’s

most

significant

finding,

as

it

demonstrates

that

the

perception

of

customer

satisfaction in the evaluated system depends to a greater extent on the time customers spend waiting to be

served than on the duration or quality of the service received once service begins.

Third, it is concluded that the productivity level achieved

—

75%

—

along with the downtime identified on

the

server,

reveals

the

existence

of

periods

of

idle

capacity

that

coexist

with

moments

of

greater

congestion. This indicates that the system’s problem stems from a

y mismatch between the availability of

service resources and the variability of demand throughout the semester analyzed, rather than a structural

limitation in the number of staff or available service points.

Fourth, it is concluded that discrete-event simulation is an appropriate and reliable tool for evaluating

customer satisfaction in service companies, as it allows for the integrated representation of variables that,

in isolation, would not provide a sufficient explanation of the phenomenon under study, and enables the

projection of improvement scenarios on a quantitative basis before modifying the actual system.

Finally, it is concluded that the results obtained lay a solid foundation for guiding future interventions in

the university cafeteria’s servi

ce system, particularly regarding the dynamic management of staff during

peak periods, with the aim of reducing customer wait times without compromising the efficiency achieved

in the use of available resources.

As

a

first

line

of

future

research,

it

is

recommended

to

develop

and

evaluate,

using

discrete-event

simulation, different scenarios for dynamic staffing based on the peak time slots identified during the

semester, with the aim of determining the resource configuration that will significantly reduce customer

wait times without increasing

the service’s operating costs or affecting

the system’s current

level

of

productivity.

As a second line of future research, it is suggested to complement the quantitative approach of this study

by incorporating instruments for directly measuring customer satisfaction

—

such as surveys based on the

NEYART Journal

ISSN: 2992-7161

Issue 4 | Vol. 4

–

No. 4 | July

–

December 2026 |

Page

148

Original Research Article

SERVQUAL model

—

which would allow for comparing the results obtained through simulation with

users’ actual perceptions regarding the dimensions of reliabilit

y, responsiveness, security, empathy, and

tangible service elements, thereby enriching the comprehensive understanding of the phenomenon of

satisfaction in service companies.

Conflict of Interest

The authors declare that they have no conflicts of interest.

Data Availability

All datasets relevant to the results of this study are available in their entirety in the article.

Source of Funding

This study was not funded by any organization.

Statement on Generative AI

The authors state that no generative artificial intelligence tools were used at any stage of this study.

References

Alarcón Bozzo, G. C. A., y Díaz Aroco, T. (2018).

Diseño de un sistema de simulación para reducir el

tiempo de espera en el área de operaciones de la empresa INTERBANK Agencia Cajamarca

[Tesis

de

pregrado,

Universidad

Peruana

del

Norte].

Repositorio

institucional

UPN.

https://repositorio.upn.edu.pe/

Application of Discrete Event Simulation for Enhancing Queuing System at Lotus Alor Setar. (2025).

International

Journal

of

Industrial

Management,

19

(4),

197-209.

https://doi.org/10.15282/ijim.19.4.2025.11959

Arias, F. G. (2012).

El proyecto de investigación: Introducción a la metodología científica

(6.a ed.).

Editorial

Episteme.

https://www.formaciondocente.com.mx/06_RinconInvestigacion/01_Documentos/El%20Proyect

o%20de%20Investigacion.pdf

Aslan,

E.,

y

Özderiir,

N.

(2021).

Bir

Üniversite

Öğrenci

Yemekhanesinde

Simülasyon

Uygulaması.

Kahramanmaraş

Sütçü

İmam

Üniversitesi

Sosyal

Bilimler

Dergisi,

18

(2),

965-978.

https://doi.org/10.33437/ksusbd.629888

NEYART Journal

ISSN: 2992-7161

Issue 4 | Vol. 4

–

No. 4 | July

–

December 2026 |

Page

149

Original Research Article

Calla Delgado, V. F. (2020). Simulación de eventos discretos para reducir el tiempo de espera en el área

de

consulta

externa

de

un

hospital

público.

INGnosis,

6

(1),

16-26.

https://www.researchgate.net/publication/350455179

Curin, S. A., Vosko, J. S., Chan, E. W., y Tsimhoni, O. (2005).

Reducing service time at a busy fast food

restaurant

on

campus

.

Proceedings

of

the

2005

Winter

Simulation

Conference,

2628-2635.

https://doi.org/10.1109/WSC.2005.1574562

da Rocha Nascimento, M. A., dos Santos, L. M., da Silva, A. M., Bueno, R. C., y Machado, S. T. (2021).

Discrete event simulation applied to single queue management: A case study at a bank agency.

Independent

Journal

of

Management

&

Production,

12

(9),

S831-S842.

https://doi.org/10.14807/ijmp.v12i9.1632

Firmansyah, F., y Saputra, A. C. (2021). Effect of the implementation

of queue system on customer

satisfaction.

Bongaya

Journal

for

Research

in

Management,

4

(1),

1-7.

https://doi.org/10.37888/bjrm.v4i1.245

Gutiérrez, E., Ramos, W., Uribe, M., Ortega-Loayza, A., Torres, C., Montesinos, D., y León, O. (2009).

Tiempo de espera y su relación con la satisfacción de los usuarios de la farmacia central de un

hospital general de Lima.

Revista Peruana de Medicina Experimental y Salud Pública, 26

(1), 61-

65.

http://www.scielo.org.pe/scielo.php?script=sci_arttext&pid=S1726-46342009000100011

Hernández-Sampieri, R., y Mendoza, C. (2018).

Metodología de la investigación: Las rutas cuantitativa,

cualitativa

y

mixta

.

McGraw-Hill

Interamericana.

http://www.biblioteca.cij.gob.mx/Archivos/Materiales_de_consulta/Drogas_de_Abuso/Articulos

/SampieriLasRutas.pdf

Heredia-Acevedo, D., Ceballos, Y. F., y Sánchez-Torres, G. (2020). Modelo de simulación de eventos

discretos para el análisis y mejora del proceso de atención al cliente.

Investigación e Innovación

en Ingenierías, 8

(2), 44-61.

https://doi.org/10.17081/invinno.8.2.3639

Kumar, P. (2005). The competitive impact of service process improvement: Examining customers' waiting

experiences

in

retail

markets.

Journal

of

Retailing,

81

(3),

171-180.

https://doi.org/10.1016/j.jretai.2005.07.004

Law, A. M. (2015).

Simulation modeling and analysis

(5.a ed.). McGraw-Hill Education.

NEYART Journal

ISSN: 2992-7161

Issue 4 | Vol. 4

–

No. 4 | July

–

December 2026 |

Page

150

Original Research Article

Mendoza Casseres, D. A., González Conde, M., Corcho Martínez, R. A., y Berdugo Alonso, A. (2018).

Aplicación

de la simulación

discreta en

el

área

de urgencias de una institución

prestadora de

servicios

de

salud.

INGENIARE

,

(21),

55-71.

https://dialnet.unirioja.es/servlet/articulo?codigo=6118796

Parasuraman,

A.,

Zeithaml,

V.

A.,

y

Berry,

L.

(1988).

SERVQUAL:

A

multiple-item

scale

for

measuring consumer perceptions of service quality.

Journal of Retailing, 64

(1), 12-40.

Pontificia Universidad Católica del Perú. (2015).

Diagnóstico y mejora de procesos utilizando simulación

de

eventos

discretos

en

una

empresa

de

consumo

masivo

[Tesis

de

pregrado].

Repositorio

institucional PUCP.

https://tesis.pucp.edu.pe/repositorio/handle/20.500.12404/6229

Pontificia Universidad Católica del Perú. (2016).

Mejora del nivel de servicio en la atención presencial

en

una

empresa

de

telecomunicaciones

empleando

simulación

de

eventos

discretos

[Tesis

de

pregrado]. Repositorio institucional PUCP.

https://repositorio.pucp.edu.pe/items/61592350-0670-

47d5-9a22-b0784cd77a96

Sotelo Seguil, M. G. (2017).

Diagnóstico y mejora para el servicio de la seguridad ciudadana en el

distrito de San Martín de Porres mediante simulación de eventos discretos

[Tesis de pregrado,

Pontificia

Universidad

Católica

del

Perú].

Repositorio

institucional

PUCP.

https://tesis.pucp.edu.pe/repositorio/handle/20.500.12404/9055

Tamashiro

Tamashiro,

E.,

y

Yacarini

Vadillo,

C.

J.

(2023).

Aplicación

de

un

modelo

de

simulación

discreta para mejorar la productividad del proceso de producción en una empresa manufacturera.

Industrial Data, 26(1), 303-332.

https://www.redalyc.org/journal/816/81676820017/html/

Universidad César Vallejo. (2019).

Diseño de un modelo de simulación de eventos discretos para reducir

el tiempo de espera del cliente en la empresa Super Taxi Elegant

[Tesis de pregrado]. Repositorio

institucional UCV.

https://hdl.handle.net/20.500.12692/29070

Collaborative Work Table

Role

Author(s)

Conceptualization

Del Socorro Corral María

Methodology

Silva Máynez Lucia Xiomara, García Aguirre Ilse Aydee

Software

Reyes Ledezma Daniela María, Medina Molina Yearim

Validation

Del Socorro Corral María

NEYART Journal

ISSN: 2992-7161

Issue 4 | Vol. 4

–

No. 4 | July

–

December 2026 |

Page

151

Original Research Article

Formal Analysis

Silva Máynez Lucia Xiomara, García Aguirre Ilse Aydee

Research

Reyes Ledezma Daniela María, Medina Molina Yearim

Resources

Del Socorro Corral María

Data Curation

Silva Máynez Lucia Xiomara, García Aguirre Ilse Aydee

Writing - Preparation of the original

draft

Reyes Ledezma Daniela María, Medina Molina Yearim

Writing - Review and editing

Del Socorro Corral María

Visualization

Silva Máynez Lucia Xiomara, García Aguirre Ilse Aydee

Supervision

Reyes Ledezma Daniela María, Medina Molina Yearim

Project Management

Del Socorro Corral María

Fundraising

Silva Máynez Lucia Xiomara, García Aguirre Ilse Aydee