Designing and Implementing a Dashboard with Key Performance

Indicators for a Higher Education Institution

, Ana Azevedo

, José Azevedo

and Michelle Eiko Hayakawa

CEOS.PP, ISCAP, P.PORTO, Rua Jaime Lopes de Amorim, Matosinhos, Portugal

Departamento de Gestão e Apoio Técnico, IFTM, Cuiabá, Brazil

Keywords: Business Intelligence, BI, Performance Indicators KPIs, BI Software Tools, Business Analytics, Data

Warehouse, DW.

Abstract: In a context where public and private institutions continuously seek excellence in management and

improvement in service delivery, governance is one of the contemporary practices that has been adopted not

only as a model for evaluation and monitoring, but also for transparency. To this end, the institutionalization

of a governance model represents not only a challenge for public managers, but also the opportunity to make

use of information technology, especially Business Intelligence, as a way to ensure agility in accessing and

processing information. Through this work, we propose the construction of a prototype of a dashboard for the

BI system to be adopted as a practice of governance by results in a Brazilian public institution, focusing on

education, not only to support the decision making of managers, but also to allow efficiency and public

transparency, and to minimize the difficulties in access, in the management of various BI systems and the

informational asymmetries existing in the Public Administration.

1 INTRODUCTION

Access to information and the ability to put

knowledge to productive use have always been the

hallmark of successful people, companies and even

nations (Fraga, Erpen, & Varvakis, 2017). The

dynamics and competitiveness of the modern world

require that organizations, public or private, make

decisions quickly and based on knowledge obtained

from Business Intelligence (BI) systems. BI

represents an important tool in conducting business

effectiveness and innovation, because it allows

transforming the large volume of operational data into

complex and competitive information, capable of

assisting in decision making with the use of analytical

and interactive tools. In this context, BI is also an

important tool for the practice of Corporate

Governance (GC), because it allows the strategic

alignment of the organization, the measurement of

results and transparency of information, due to the

continuous use of indicators, built based on a set of

information. It is this set of information that allows

the manager to make a safe decision and minimize the

https://orcid.org/0000-0003-0882-3426

https://orcid.org/0000-0001-6951-4278

risks of failure in a project or in a strategic action.

From this point of view, it was found that the Federal

Institute of Education, Science and Technology of

Mato Grosso (IFMT) has great difficulty in accessing

data and building management indicators, and that

there is still a need to implement tools that help

managers in decision-making, promoting governance

and transparency of information.

The objective of this work is to propose and carry

out a proof of concept of a BI system for governance

by results, through performance indicators within the

IFMT, aiming to support the decision making

process.

The motivation for this research is due to the fact

that during the professional exercise at the IFTM, one

of the authors experienced moments of difficulty in

the rapid access to information for decision making,

due to the absence of a model focused on the

governance of results and performance indicators. In

this sense, besides the personal and professional

interest in the subject, it is observed that the relevance

of the research is related to the use of performance

indicators and governance. These are contemporary

Azevedo, A., Azevedo, J. and Hayakawa, M.

Designing and Implementing a Dashboard with Key Performance Indicators for a Higher Education Institution.

DOI: 10.5220/0010539501650172

In Proceedings of the 13th International Conference on Computer Supported Education (CSEDU 2021) - Volume 1, pages 165-172

ISBN: 978-989-758-502-9

165

themes that are beginning to be demanded by Public

Institutions, in addition to society itself.

The rest of the paper is organized in the following

way: Related topics will be approached, namely

Business Intelligence (BI), the components of a BI

system, and interfaces and ways of representing KPIs.

Following are the design, development, description,

and evaluation of the dashboard. The article ends with

a conclusion.

2 RELATED TOPICS

In this section, the main concepts related to this work

are introduced

2.1 Business Intelligence

Business Intelligence (BI) can be presented as an

architecture, a tool, a technology, or a system that

holds and stores data, analyses it using analytical

tools, and provides information and/or knowledge,

facilitating the production of reports, queries, and

fundamentally, allowing organizations to improve

their decision-making. In short, Business Intelligence

can be defined as a process that transforms data into

information and then into knowledge (Azevedo &

Santos, 2009). Turban & Volonino (2013)

conceptualize BI as a combination of software

architecture, databases, analytical tools, graphical

displays and decision-making methodologies, whose

objective is "to enable interactive access (sometimes

in real time) to data, to enable manipulation of data,

and to give business managers and analysts the ability

to conduct appropriate analyses." (Sharda, Delen &

Turban, 2018, p. 16) or improve the organization's

performance (Piedade, 2012). For this author, at the

strategic level, BI systems provide information on

several performance indicators that allow to verify

whether the strategic objectives of the organization

have or have not been achieved, supporting the

planning or redefinition of new methods of operation

and business. At the tactical level, BI demonstrates

how business processes are evolving, if there are

problems and what are the new business trends.

Lastly, at the operational level, BI can provide

information related to the activity of the organization,

its business or its customers. In short, for Reginato &

Nascimento (2007, p. 73), the tools of BI can "provide

a systemic view of the business and help with uniform

distribution of the data between users. Its main

objective is to transform large quantities of data into

quality information for decision making".

2.2 Components of a BI System

The four main components of a BI system, according

to Sharda, Delen, & Turban (2018), are the Data

Warehouse (DW), Business Analytics (BA),

Business Performance Management (BPM) and a

user interface for viewing the information (Figure 1).

DW is the place where all the data extracted from

the information systems are stored, which are

organized and oriented from the main subjects of the

organization, vary over time in the availability of

historical and current data, are integrated and have the

characteristic of being non-volatile; that is, it is not

allowed to change or delete the data after inclusion in

the DW (Loshin, 2012; Reginato & Nascimento,

2007; Sharda, Delen, and Turban, 2018).

It is important to highlight that before the data are

deposited in the DW, it is necessary to go through the

ETL (Extract, Transform and Load) process, that is,

it is the process that collects the relevant data from

transactional databases, spreadsheets, and text files,

transforms them into a pattern (through cleaning,

treatment and classification processes) and loads

them into the DW (Turban & Volonino, 2013).

Business Analyics are tools that help transform

data into knowledge, which can help stakeholders

make business decisions, promote revenue growth,

risk reduction, cost management and others (Sharda,

Delen, and Turban, 2018; Loshin 2012).

BPM uses analysis, reporting and BI queries to

optimize the overall performance of the organization.

This tool can be based on the Balanced Scorecard

(BSC) methodology, allowing for comparison, and

sharing of performance goals and results, so that

managers can quickly understand how the company's

activities are going. These systems use various types

of indicators to measure organizational performance.

The most commonly used indicators are KPIs or key

performance indicators.

Finally, there is the user interface that

corresponds to the way information is visualized,

made available as reports or web interfaces

(dashboards, reports, scorecards, spreadsheets and

other information transmission and visualization

tools, such as corporate portals and virtual reality

presentations (Côrte-Real, 2010; Sharda, Delen, and

Turban, 2018).

2.3 Interfaces and Ways of

Representing KPIs

For Costa (2012, p. 167), as a fundamental

requirement, a BI system must offer interfaces that

make it easier for the manager to interact and

CSEDU 2021 - 13th International Conference on Computer Supported Education

166

Figure 1: Components of a BI system.

understand data in order, for example, to provide

adequate information for decision-making. This

human-machine interaction (user-friendly interface)

allows the individual to generate reports and analyze

autonomously, thereby improving user satisfaction

and experience. These interfaces can facilitate access

to KPIs, but for the decision-making and success of

governance by results, it is important that business

managers have knowledge about performance

indicators, have a critical vision, and know the

objectives and strategic goals of the organization.

KPIs are usually represented by visual elements,

capable of demonstrating the performance, state or

status of an indicator and can often be related to

productivity, quality, capacity, time, profitability and

others. These types of KPIs are more dynamic,

interactive and easy for the user to understand, as they

allow them to quickly understand whether the

situation is correct or not.

KPIs are released on dashboards or printed

reports. In dashboards, KPIs are presented in a more

interactive and visual manner. Few (2007)

conceptualizes the dashboard as a visual

representation of the most important information to

achieve one or more goals, consolidated and

organized in a single view screen so that the

information can be monitored quickly and with user

participation.

3 DESIGN, DEVELOPMENT,

DESCRIPTION, AND

EVALUATION OF THE

DASHBOARD

This section presents the design, development,

description, and evaluation of the dashboard.

3.1 Design

To carry out this stage it is important to carry out the

planning, selection and definition of the constructive

aspects. For this purpose, a documental report

(included in the appendix) was prepared, containing

the planning, roadmap, and requirements for the

development of the BI system, in which the internal

characteristics of the researched institution, the

expected benefits for the solution of the problem and

the context in which the BI system will operate were

taken into consideration (user type and technology).

Thus, it is possible to design a BI system that

consists of two main elements: a) dashboards

containing the indicators created for the areas of

teaching, research, extension and management; and

b) a web page for hosting the performance indicators

and user access. The details of the development and

evaluation process are presented in the following

sections.

3.2 Development

Among the various types of BI tools available in the

market for the development of the dashboard to

display performance indicators, the starting point was

the classification carried out by Gartner, called

"Magic Quadrant" (Howson, Sallam et al, 2018), to

guide the process of choosing the software used to

develop the dashboard. The experimental phase of

this research included the installation and use of a

demonstration copy of the selected software. All

software has great potential for use and can be used

as BI tools in the institution, but some of these are

easier to use. The analysis of each of the selected

software tools is presented in Table 1.

Business

Analytics

Data Warehouse

User Interface

Data Sources

BPM

ETL

Designing and Implementing a Dashboard with Key Performance Indicators for a Higher Education Institution

167

Table 1: Evaluation of software characteristics and requirements.

Characteristics and requirements Power Bi QlikView Tableau Tableau Public Pentaho

Licence

Free * * * X X

Commercial X X X X

Operating system

Windows X X X X X

Linux X X X X

Mac OS X X X X X

Platform Desktop X X X X X

Web X X X X

Usability **

Ease of use Good Excellent Excellent Excellent Difícil

Data manipulation Good Good Excellent Excellent Difícil

Attractive Regular Excellent Excellent Excellent Regular

Product Cost Moderate High Moderate Não possui Do not

have

Market Small,

medium and

large

companies

Small,

medium

and large

companies

Small,

medium

and large

companies

Small, medium

and large

companies

Medium

and large

companies

* These softwares are free for a limited time.

** Scale used: difficult, regular, good, and excellent.

Initially, the main requirement for the

construction of the BI system was the use of an open-

source tool; however, throughout the elaboration of

this research, it was verified that we could opt for one

that was not open source as long as its cost ratio

benefit would justify it. Thus, after evaluating all

these constraints, Tableau Public was chosen.

DW was locally stored and built based on the

selection of data extracted from the Institution's

information systems, documents published on the

institutional website and open data from the Federal

Government. However, for the construction of some

indicators, a database with fictitious content was also

used, because some available information was

incomplete and others depended on specific or

exclusive access authorization; thus, for security

reasons and to avoid the information being not

corrupted, stolen or misused, a database with

fictitious content was chosen. This does not damage

the development of the research, since the intention is

to propose a model that can be used by the institution.

Before loading the data into the DW through the ETL

process, some information from the databases was

processed, transformed, and standardized. To

perform this action, it is possible to use Tableau Prep

or the Tableau Creator that includes, in addition to the

Tableau Desktop the Tableau Prep. With this, it is

possible to perform the cleaning and preparation of

the data and join and combine data between different

tables or databases.

However, in some databases, it was not necessary

to apply the ETL techniques because the data

obtained were fully capable of being applied in the BI

tool for building dashboards. However, depending on

the system or from the database, some data needed to

be transformed or standardized, such as by example

the names of the IFTM's Campi, as for these data

there were identified various forms of nomenclature:

codes, acronyms, full or abbreviated names, etc. The

information obtained in some of the institution’s

databases required the cleaning, association, and

preparation of data, but as the volume was not so

large, this process was performed with the tools made

available in Excel (filters, formulas, and supplements

such as Power Query and Power Pivot), as shown in

the set of images in Figure .

To ensure the security of this process and to

facilitate association and data sizing, the Star Schema

(multidimensional modeling) was adopted for the

design of the DW. Kimbal and Ross (2013) state that

the multidimensional model consists of fact tables

linked to dimension tables by means of primary and

foreign keys. According to Fortulan & Gonçalves

Filho (2005, p. 58), the Fact Table is the main one,

where the occurrences are stored, and the Dimension

Tables provide a description of the data and have only

one primary key. The Star Schema for the execution

of the budget expenditure is presented in Figure . It is

important to highlight that this type of relationship is

automatically performed by Tableau, when the user

performs the selection of the dimensions, as

demonstrated for example in Figure.

After loading the data into Tableau Public, other

dimensional modeling was also performed through

CSEDU 2021 - 13th International Conference on Computer Supported Education

168

groupings, intersections, or associations, as shown in

Figure and Figure. However, it is important to note

that for the dimensional modeling process, Tableau

Public itself carries out the steps in a satisfactory

manner, without the need for additional and parallel

tools. In any case, the actions performed were

important for the enrichment and improvement of

knowledge, as it helped in the work developed by

Tableau.

Figure 2: Star scheme for the execution of the budget expenditure.

Figure 3: Use of MSExcel

tools for data processing.

Designing and Implementing a Dashboard with Key Performance Indicators for a Higher Education Institution

169

Figure 4: Screen with dimensions and measurements.

Figure 5: Relationship of databases.

Figure 6: Connections and interconnections with the various DM.

After the construction of all dashboards, they

were made available on a web page, which was

conceived with the intention of being aggregated on

the institutional page of the IFMT. However, since

this proposal is academic and not institutional in

nature, this platform for dissemination and

visualization of the performance indicators of the

institution was built and hosted in a different location.

There was no damage to the development of the

proposed project and the test application has potential

CSEDU 2021 - 13th International Conference on Computer Supported Education

170

for future use by the Institution, if so desired, because

the tools and technologies chosen are compatible.

3.3 Short Description

The dashboard of the web platform was divided into

four areas: a) teaching, b) research, c) extension, and

d) management. Figure presents an example of a

dashboard containing the performance indicators for

these four areas. Each graph that makes up the

dashboard is individually built into a Tableau

spreadsheet. These spreadsheets allow the insertion of

metrics, filters, dimensions, details, masks,

measurements, and calculated fields. Each sheet is

then sized and visually organized on the panel, thus

composing the thematic part of each area of the BI

system.

Several types of visual elements were used, such

as geographic maps, word clouds, buses, pizza, lines,

and areas, with the intention of making the navigation

pleasant and varied for the user, which is more

attractive and dynamic. It is important to note that not

all pages of the dashboard presented in this paper for

reasons of confidentiality.

3.4 Evaluation

Once the artifact was built, several tests and

experiments were conducted to verify its

performance. The prototype was made available for

experimentation and evaluation, and to make the

proof of concept. One of the authors, accompanied

the users selected for the experiments, and the

problems pointed out by those users were approached

to update the dashboard. The feedback obtained from

initial users was very positive.

4 CONCLUSION

The use of the BI tool Tableau Public to implement

the proposed indicators proved to be an efficient

solution, as it allows the use of Data Marts composed

of several and varied types of data, as well as flexible,

intuitive, and fast analyses.

The results obtained in the proof of concept

demonstrate that the proposed model has the potential

to assist managers in the decision-making process,

Figure 7: Dashboard with the governance indicators.

Designing and Implementing a Dashboard with Key Performance Indicators for a Higher Education Institution

171

monitoring performance and results, promoting

transparency of actions and stimulating the search for

improvement in service delivery. Similarly, it was

demonstrated that BI systems and tools are important

instruments in the construction of this model because

they ensure reliability, minimize errors, and provide

dynamics and agility in the processing and

presentation of information, in addition to optimizing

time and manpower.

During the research, some difficulties and

limitations were encountered. There were difficulties

in accessing the data in the databases because of the

wide variety and types of data available, as well as the

security procedures and access restrictions. Likewise,

the process of choosing the best software to build the

performance indicators requires a considerable

amount of time and study of each system.

This research has potential for future work,

namely the evolution of the artifact, since it requires

improvements and new implementations, not only to

overcome the limitations currently existing, but also

in the construction of new performance indicators.

ACKNOWLEDGEMENTS

This work is financed by Portuguese national funds

through FCT - Fundação para a Ciência e Tecnologia,

under the project UIDB/05422/2020.

REFERENCES

Azevedo, A. & Santos, M.F. (2009). Business Intelligence:

State of the Art, Trends, and Open Issues. In

Proceedings of the 1st KMIS 2009, 296-300.

https://doi.org/10.5220/0002303602960300

Bueno, R. L. P., Brelàz, G. de, & Salinas, N. S. C. (2016).

Administração pública brasileira no século 21: seis

grandes desafios. Revista do Serviço Público, 67(0), 7–

28. https://doi.org/10.21874/rsp.v67i0.1152

Côrte-Real, N. (2010). Avaliação da maturidade da

business intelligence nas organizações (Dissertação).

Universidade Nova de Lisboa, Lisboa. Recuperado 18

abril, 2018 de https://run.unl.pt/bitstream/10362/7477/

1/TEGI0269.pdf.

Costa, S. A. R. da. (2012). Sistema de business intelligence

como suporte à gestão estratégica (Dissertação).

Universidade do Minho, Minho. Recuperado 18 abril,

2018 de http://repositorium.sdum.uminho.pt/handle/

1822/25810.

Fortulan, M. R., & Gonçalves Filho, E. V. (2005). Uma

proposta de aplicação de business intelligence no chão-

de-fábrica. Gestão & Produção, 12(1), 55–66.

https://doi.org/10.1590/S0104-530X2005000100006.

Fraga, B. D., Erpen, J. G., & Varvakis, G. (2017). Business

Intelligence: métodos e técnicas de gestão do

conhecimento e as tendências para avanços do capital

intelectual. Business Intelligence, 7(1), 15.

Howson, C., Sallam, R.L., Richardson, J.L., Tapadinhas, J.,

Idoine, C.J., & Woodward, A. (2018). Magic Quadrant

for Analytics and Business Intelligence Platforms.

Gartner Report ID: G00326555 https://www.gartner.

com/en/documents/3861464/magic-quadrant-for-

analytics-and-business-intelligence-p

Kimball, R., & Ross, M. (2013). The Data Warehouse

Toolkit: The Definitive Guide to Dimensional

Modeling. John Wiley & Sons.

Loshin, D. (2012). Business Intelligence: The Savvy

Manager’s Guide. Newnes.

Piedade, M. B. (2012). Business Intelligence no suporte ao

conceito e à prática de student relationship

management em instituições de ensino superior

(Doutoramento). Universidade do Minho, Minho.

Recuperado em 13 abril, 2018 de http://repositorium.s

dum.uminho.pt/handle/1822/20461.

Reginato, L., & Nascimento, A. M. (2007). Um estudo de

caso envolvendo Business Intelligence como

instrumento de apoio à controladoria. Revista

Contabilidade & Finanças, 18, 69–83.

https://doi.org/10.1590/S1519-70772007000300007.

Sharda, R., Delen, D. & Turban. E. (2018). Business

Intelligence, Analytics, and Data Science: a Managerial

Perspective – Fourth Edition. Upper Saddle River, NJ:

Pearson Education, Inc.

Turban, E., & Volonino, L. (2013). Tecnologia da

Informação para Gestão: em busca de um melhor

desempenho estratégico e operacional (8

ed). São

Paulo: Bookman Editora.

CSEDU 2021 - 13th International Conference on Computer Supported Education

172