Costing as a Service - ULisboa · PDF fileOur research proposal is a cloud-based costing system that offers costing as a service using Time- ... TDABC, Cost Templates, Cloud Computing,

Costing as a Service

André Miguel Nunes Machado

Thesis to obtain the Master of Science Degree in

Information Systems and Computer Engineering

Supervisor: Prof. Miguel Leitão Bignolas Mira da Silva

Examination Committee

Chairperson: Prof. Miguel Nuno Dias Alves Pupo CorreiaSupervisor: Prof. Miguel Leitão Bignolas Mira da Silva

Member of the Committee: Prof. Maria do Rosário Gomes Osório Bernardo Poncesde Carvalho

July 2014

ii

Acknowledgments

Firstly, I want to express my gratitude towards Professor Miguel Mira da Silva for the opportunity of

researching in such an interesting field. The constant motivation, feedback and expertise contributed

significantly to the success of this research.

I would also like to mention the support, encouragement, patience and guidance of PhD. Carlos Mendes.

His knowledge was truly inspiring and essential to bring this work to this final stage. Undoubtedly, one

of the smartest persons I have known in my academic journey.

I would also like to mention and express my gratitude towards my family whose support and guidance

made me achieve most of my goals, including my academic achievements.

Last but not least, a special thanks to Raquel whose love, patience and companionship was truly out-

standing throughout the years.

iii

iv

Resumo

A consciencia e eficiencia de custos sempre foram algumas das principais preocupacoes para as

organizacoes de todos os sectores, mas nos ultimos anos a sua importancia cresceu devido a crise

economica e financeira global. Considerando a sua pequena dimensao e exposicao ao mercado, as

Pequenas e Medias Empresas (PME) precisam de consciencia e eficiencia de custos mais do que

nunca. No entanto, as metodologias de custeio eficientes e precisas estao fora do alcance da maioria

das PME. Nesta investigacao, propomos que o custeio deve ser oferecido como um servico para reduzir

o custo da analise de custos.

A nossa proposta e um sistema de custeio, cloud-based, que oferece o custeio como um servico usando

a metodologia Time-Driven Activity Based Costing (TDABC) e o conceito de Business Process Costing

Templates que, juntos, reduzem o custo da analise de custos, em particular para as PME. Utilizamos o

Design Science Research Methodology (DSRM) para conduzir a nossa investigacao. Esta proposta foi

demonstrada em tres organizacoes portuguesas e avaliada com informacoes obtidas a partir de entre-

vistas e resultados da instanciacao do sistema em todas as organizacoes.

Palavras-chave: Custeio, TDABC, Templates de Custos, Cloud Computing, Cloud Services, SaaS.

v

vi

Abstract

Cost awareness and cost efficiency has always been major concerns to organizations from all industries

but in the last few years its importance grew due to the global economic and financial crisis. Considering

their small size and market exposure, Small and Medium Enterprises (SMEs) need cost awareness and

efficiency more than ever. However, efficient and accurate costing methodologies are out of reach for

most SMEs. In this research we propose that costing should be offered as a service to reduce the cost

of cost analysis.

Our research proposal is a cloud-based costing system that offers costing as a service using Time-

Driven Activity Based Costing (TDABC) methodology and the concept of Business Process Costing

Templates that together reduce the cost of cost analysis, especially for SMEs. We used the Design

Science Research Methodology (DSRM) to conduct our research. This proposal was demonstrated in

three Portuguese organizations and evaluated with feedback gathered from interviews and results from

the system instantiation in all organizations.

Keywords: Costing, TDABC, Cost Templates, Cloud Computing, Cloud Services, SaaS.

vii

viii

Contents

Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii

Resumo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii

Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x

List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

1 Introduction 1

1.1 Research Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

1.2 Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2 Problem 7

2.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2.2 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2.3 Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

3 Related Work 13

3.1 Costing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

3.1.1 Cost and Management Accounting . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

3.1.2 Cost Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

3.1.3 Costing Methodologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

3.1.4 Costing Templates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

3.2 Cloud Computing Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

3.2.1 Service-Oriented Computing and Cloud Computing . . . . . . . . . . . . . . . . . 21

4 Objectives 23

5 Proposal 25

5.1 Costing Service Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

5.2 Costing Service - Analysis Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

5.3 Costing Service Tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

5.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

ix

6 Demonstration 37

6.1 Social Security IT Institute . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

6.2 Defence Data Center . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

6.3 Card4B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

7 Evaluation 47

7.1 Interviews and Questionnaires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

7.1.1 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

7.2 Pries-Heje et al. Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

7.2.1 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

7.3 Osterle et al. Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

7.3.1 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

7.4 Demonstration Critical Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

8 Lessons Learned 55

8.1 Objectives Accomplishment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

9 Conclusion 59

9.1 Main Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

9.2 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

9.3 Communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

9.4 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

Bibliography 65

x

List of Tables

1.1 Mapping beetween DSRM and the Project Structure . . . . . . . . . . . . . . . . . . . . . 5

xi

xii

List of Figures

1.1 The DSRM Process Model (Peffers et al., 2007) . . . . . . . . . . . . . . . . . . . . . . . . 3

1.2 The DSRM process instantiated to our proposal (adapted from (Peffers et al., 2007)) . . . 5

2.1 Whale Curve of Profitability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

3.1 Cost elements diagram (adapted from (Blocher, 2005)) . . . . . . . . . . . . . . . . . . . 15

5.1 Costing Service - Analysis Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

5.2 Data Model of the Costing Tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

5.3 CaaSH - Dashboard - Costing Tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

5.4 CaaSH - Transactions Sample - Costing Tool . . . . . . . . . . . . . . . . . . . . . . . . . 32

5.5 CaaSH - Business Processes Sample - Costing Tool . . . . . . . . . . . . . . . . . . . . . 33

5.6 CaaSH - Business Processes Instances Samples - Costing Tool . . . . . . . . . . . . . . 33

5.7 CaaSH - Resources and Resource Pools Associations - Costing Tool . . . . . . . . . . . . 34

5.8 CaaSH - Chart Example - Costing Tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

6.1 CaaSH - SS IT Institute - Transactions Sample . . . . . . . . . . . . . . . . . . . . . . . . 39

6.2 CaaSH - SS IT Institute - Business Processes Sample . . . . . . . . . . . . . . . . . . . . 41

6.3 CaaSH - Defence Data Center - Transactions Sample . . . . . . . . . . . . . . . . . . . . 42

6.4 CaaSH - Defence Data Center - Business Processes Sample . . . . . . . . . . . . . . . . 44

7.1 Questionnaire Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

7.2 Strategies for DSR Evaluation (Pries-Heje et al., 2008) . . . . . . . . . . . . . . . . . . . . 51

xiii

xiv

Abbreviations

ABC Activity-Based Costing

AIS Accounting Information System

BPMN Business Process Model and Notation

DSRM Design Science Research Methodology

IT Information Technology

SME Small and Medium Enterprise

TDABC Time-Driven Activity Based Cost

xv

xvi

Chapter 1

Introduction

Enterprises are becoming increasingly complex and managing that complexity is a growing challenge.

Competition is fierce among these entities that always tried to differentiate between themselves through

a variety of factors, one of which is efficiency. Efficiency can be measured in a variety of aspects being

one cost efficiency. Cost efficiency has always been a major concern to organizations but in the last few

years its importance grew due to global economic and financial crisis. Due to their small size and market

exposure, Small and Medium Enterprises (SMEs) need cost efficiency more than ever (OCDE, 2009).

However, as organizational complexity grows, so does the complexity of cost analysis (Wileman, 2010)

since organizations have become too organized for statistics and too complex for analysis. Information

about how and where the money was spent is a concern of organizations across all industries. Knowl-

edge about costs distribution and true understanding of overhead costs allocation is essential for an

enterprise to focus on the most profitable products and services (Delloite, 2011). Only this information

can provide the needed tools to decision making.

In order to obtain detailed information about costs and overheads distribution several cost methodologies

were developed. These methodologies evolved and differentiated themselves from traditional cost ac-

counting systems to better distribute overhead costs that have been rising inside organizations in recent

years (Miller & Vollmann, 1985). The increasingly importance of overhead costs comes from the fact

that the industry has evolved from manufacturing to services (Skoda, 2009). This development implied

a substantial growth of overhead costs being currently accountable for the major share of total costs

(Miller & Vollmann, 1985, Skoda, 2009).

Organizations using these accurate costing methodologies know exactly where resources are being

spent and what is the profitability of their products and services. However, although these costing

methodologies deliver clear advantages to organizations, the adoption is far behind of what would be

expected. One of the reasons behind this lack of adoption is the high costs of these methodologies for

SMEs since they require time, expertise (internal know-how or consultancy services) and expensive and

1

very complex software solutions that are out of reach for the most of these organizations (Hall et al.,

2011). SMEs account for more than 98 percent of the European Economy (Ecorsys, 2012), therefore

these costing methodologies are out of reach for the majority of the European economic tissue.

However, while organizations have a clear need for cost reduction, the tools to provide them the correct

guidance on how to do it are out of their reach, forcing them to rely on traditional accounting systems.

This often leads to poorly supported decisions that usually have undesirable results. For instance, a

manager may decide to cease the production of a particular product since it appears to be unprofitable

when, in reality, a series of correlated aspects made it the sustainability factor of the organization. Thus,

the lack of awareness may lead an organization to its closure.

The research presented in this thesis follows the principles of Design Science Research Methodology

(DSRM) (Hevner et al., 2004, Peffers et al., 2007) and is structured according to the guidelines estab-

lished by this methodology. We will present a brief description of DSRM followed by the identification of

the problem, why is it relevant and what is the research question that guides this research. Next, we will

describe extensively the state of art of costing, costing methodologies and costing templates followed by

a summary of cloud services. Afterwards, we will define the objectives of a solution and then proceed

to the proposal section where we extensively describe our proposal to solve the identified problem and

that fulfills the goals defined.

After that, we will present the developed tool that supports the costing service and the demonstrations

made in three Portuguese organizations using that service. Finally, we will discuss the evaluation criteria

for this research followed by our conclusions.

1.1 Research Methodology

This section describes the research methodology used and the reasons behind such choice. We chose

to use Design Science Research Methodology (DSRM) (Hevner et al., 2004, Peffers et al., 2007) to

conduct our research.

The scientific view of design arises from the concepts found in The Sciences of the Artificial (Simon,

1996). One of those concepts is Design Science which was defined as the research and development

of artefacts and their relation to our physical, psychological, economic, and social environment. This

concept later motivated the development of design research that provided the needed guidelines so that

design-oriented professionals (e.g. engineers) could act on the basis of a systematic body of evidence

(van Aken & Romme, 2009). Hevner et al. later provided a reference on Design Science Research

(DSR) in Information Systems (Hevner et al., 2004).

In order to completely understand DSRM it is important to comprehend the concepts of both method-

ology and information system. According to Hevner et al. a methodology is a system of principles,

2

practices and procedures applied to a given branch of knowledge (Hevner et al., 2004). On the other

hand, Laudon & Laudon defined information system as a set of interrelated components that collect

(or retrieve), process, store, and distribute information to support decision making and control in an

organization (Laudon et al., 2012).

Our research is inserted in the intersection of the abovementioned concepts: Design Science and In-

formation Systems. We intend to research and develop an artefact to solve a problem related to the

intersection of IT and organizations. This is why we chose DSRM to help us conduct our research.

DSRM consists on an iterative process composed by six steps (Peffers et al., 2007), as seen in Fig-

ure 1.1, which help information systems researchers conduct, evaluate and present design science

research. Next, we will explain each of the DSRM steps (Peffers et al., 2007).

Figure 1.1: The DSRM Process Model (Peffers et al., 2007)

1. Problem identification and motivation: Consist on defining the specific research problem and

justifying the value of a solution. The problem should be atomized conceptually so that the solution

can capture its complexity. The problem definition will be used to develop an artefact to effectively

provide a solution to the proposed problem. Current state and importance of the problem should

be included in this section. This step corresponds to Chapters 1 to 3.

2. Definition of the objectives of a solution: In this step we infer the objectives of a solution from

the problem definition and knowledge of what is possible and feasible considering the related work

analysis and discussion. Objectives can be quantitative, e.g., terms in which a desirable solu-

tion would be better that current ones, or qualitative, e.g., a description of how a new artefact is

expected to support solutions to problems not previously addressed. Given that design is an incre-

mental process, not all problems become objectives since a solution may only intend to partially

solve them. This step corresponds to Chapters 1, 2 and 4.

3

3. Design and development: In this iteration we create the artefact according to the objectives

defined. An artefact may be a construct (vocabulary and symbols), model (abstractions and rep-

resentations), method (algorithms and practices) or an instantiation (implemented and prototype

systems). The artefact of this research is an instantiation. In order to create the artefact we need

to determine the artefact’s desired functionality and construct its architecture based on the related

work and theoretical foundations researched earlier. This step corresponds to Chapter 5.

4. Demonstration: Usefulness of the artefact is demonstrated by solving one or more instances of

the identified problem. This demonstration can be done through simulation, experimentation or

case studies. It should be explicit how to use the artefact to solve the problem. In our research we

intend to instantiate our artefact in three small organizations. This step corresponds to Chapter 6.

5. Evaluation: In this step, researchers should assess how well the artefact solves the identified

problem. Comparing the objectives defined with the actual results obtained in the demonstration

step should answer this question. This comparison includes measures, surveys, feedback, simu-

lations and the appraisal of the scientific community. If the results are not according to the expec-

tations, researchers might decide to iterate back to the design and development step (Pries-Heje

et al., 2008). This step corresponds to Chapter 7.

6. Communication: The last step of the process consists of communicating the problem and its

importance and how the artefact contributes to a solution to relevant audiences. This step is

crucial as only with support from experts in the field it is possible to ensure that the research is

correct. This can be accomplished by submitting scholarly publications, workshops or other means

that reach those interested on the research.

It is essential to guarantee that each of these steps is done sequentially to achieve the expected results.

Still, there are several starting points and more than a single iteration may occur before achieving the

end result. The DSRM process has four entry points:

1. Problem-centered: if the research starts from the definition of a problem;

2. Objective-centered: if a set of objectives is already defined;

3. Design and Development-centered: if already exists an artefact that lacks formal analysis in the

problem domain;

4. Client/Context initiated: if consists of evaluating the performance of a solution that was already

applied.

Our research started the DSRM process with a problem centred initiation since we started with the

identification of a problem. Afterwards, we followed orderly the next five steps to complete the full

research cycle. Figure 1.2 shows the DSRM process instantiated for our research in particular.

4

Figure 1.2: The DSRM process instantiated to our proposal (adapted from (Peffers et al., 2007))

The DSRM process provides the guidelines needed for the researcher to conduct the research. It

is a common and accepted framework for the researcher, reader and reviewer, providing a common

language and context. It is also worth mentioning that Design Science Research in Information Systems

was promoted by 111 full professors that signed a memorandum proposed by 10 authors (Osterle et al.,

2010).

1.2 Structure

So that we can provide a clear presentation of the contents of this research we aligned and structured

this document with the steps defined in the process model of DSRM. Therefore, the chapters of this

report will have a direct relationship to the steps and guidelines defined above. The relationship between

chapters and DSRM steps is expressed in Table 1.1.

Chapter DSRM Step

Introduction Problem Identification and Motivation

Problem Problem Identification and MotivationDefinition of Objectives of a Solution

Related Work Problem Identification and MotivationObjectives Definition of Objectives of a SolutionProposal Design and DevelopmentDemonstration DemonstrationEvaluation EvaluationConclusion Communication

Table 1.1: Mapping beetween DSRM and the Project Structure

5

6

Chapter 2

Problem

This section corresponds to the problem and motivation step of DSRM. This research started with prob-

lem identification which we will be describing in this section plus the motivation behind its resolution.

Finally we raise a research question based on the identified problem which, besides summarizing our

problem, represented a major guideline for the rest of the research.

2.1 Overview

Costing has been a major concern to all organizations since their genesis. As a competitive advantage,

cost efficiency has been something that all organizations tried to achieve in order to increase their profit

margins or reduce the price of their products or services. Cost efficiency is recognized as one of the

most important aspects in respect to the competitive advantages of an organization.

However, although cost efficiency is valuable to organizations in order to provide them the mechanisms

needed to increase their margins and profits, organizations cannot opt out and ignore cost efficiency, as

without it organizations cannot succeed.

Organizations are currently inserted in an economy that is becoming more and more competitive over

time (OCDE, 2009). Furthermore, global economy is facing a long period of uncertainty (OCDE, 2009).

This global economic crisis has increased many of the costs that organizations have while reducing

very significantly the demand for products and services (OCDE, 2009). On the side of increasing costs,

SMEs have an additional problem: They have now more difficulties on obtaining capital (Commission et

al., 2005) So, to sum up, what we have now are organizations with fewer margins to increase prices and

increased pressure to reduce costs. There is a critical need for cost awareness and cost efficiency.

Organizations are left with two options: increase their revenues by raising the price of their products and

services, while facing reduced demand, or reduce costs. However, reducing costs has a main hurdle:

7

Where should organizations cut and reduce costs.

Even though we recognise that efforts can be made on the side of revenue, our research will focus

only on costs, cost efficiency and cost awareness. We chose this path not only because we find that

a lot of improvements may occur on the costs side but also because there is an observable trend in

organizations in general: They are trying to be more efficient and cost-aware than ever before, instead

of simply raising the price of products and services, given that this is less sustainable on the long run.

Normally, organizations resort to cost accounting in order to analyse costs and achieve the desired cost

reductions. This approach as a major issue: traditional cost accounting systems give low detailed in-

formation and lack of the needed granularity to properly do cost analysis and, therefore, cost reduction.

Not least, most accounting systems are focused on mandatory state-demanded reports (Hicks & Cost-

ing, 2002) showing only large blocks of information totally misaligned with the organization’s business

processes. Therefore, when it comes to calculate the cost of a product or service, traditional method-

ologies give inaccurate values, mostly because they lack the needed granularity and differentiation of

information. Very often, such information is inaccurate because of wrong distribution of overhead costs

that normally lead to bad decisions (Lambert III & Chen, 1996). Correct distribution of overhead costs

is truly essential since they have grown from being a minor share of the total costs to the major one

(Skoda, 2009).

Presently, there are several costing methodologies to address the abovementioned problem. These

methodologies resort to the activities that occur inside the organization to design the flow of costs from

the inputs (e.g. material) to the outputs of an organization (products and services). Knowing the flow

of costs allow organizations to know exactly where money is being spent and what is the cost of a

singular product or service. This cost awareness allows organizations to take measures to improve their

efficiency.

The problem with these accurate costing methodologies is that they require a lot of expertise and are

normally supported by very expensive and complex software solutions (Hall et al., 2011). Whereas large

organizations can support the costs associated with the required expertise and software solutions, SMEs

cannot (Hall et al., 2011). It is crucial that SMEs have access to these accurate cost methodologies

since they operate in a market that is more competitive (Nandan, 2010) and they are more exposed to

the effects of an economic crisis (OCDE, 2009).

Thus, organizations, especially SMEs, need cost awareness to improve their efficiency. Accessing the

tools that provide this cost awareness is too expensive for them. Therefore, we summarize the main

problem of our research as:

Problem: The costs associated to cost analysis make it unaffordable for SMEs.

To solve this problem we propose that costing and cost analysis should be offered as a service instead of

as an investment in a one-time project. This approach should enable organizations to access accurate

8

costing methodologies because costs are diluted over time and the tools needed to perform this cost

analysis are also offered as a service. Our proposal will also give organizations the ability to do on-

demand cost analysis so that they can constantly evaluate the flow of costs as well as take measures to

improve their cost efficiency.

2.2 Motivation

In this section we try to accomplish what is described in DSRM (Hevner et al., 2004, Peffers et al., 2007)

as justifying the value of a solution in order to motivate the researcher and the audience of the research

to pursue the solution, accept its results and contribute to understand the research reasoning.

As we stated before in the previous section there is a strong motivation for organizations to do cost anal-

ysis. As the competitiveness grows, cost awareness becomes a matter of survival instead of an option.

Although organizations cannot opt out cost analysis, the tools needed to perform such analysis are too

expensive or, when widely available, provide wrong information leading to wrong decisions (Lambert III

& Chen, 1996). Our problem summarized this issue stating that accurate costing methodologies are

out of reach for the majority of the organizations. We shall now present the advantages of having the

right information, provided by the right and accurate costing methodologies, and what assessments may

organizations do with the resulting information.

As we seen before, cost awareness is useful to know the flow of costs within an organization from its

source until the final product or service is finished. Usually, these cost objects (products, services,

clients) have their profits distributed according to a variation of the 80-20 rule (also known as the Pareto

Rule) (Cooper & Kaplan, 1991) This distribution of profits (Figure 2.1) is often called the whale curve of

profitability (Nandan, 2010, Narayanan & Kaplan, 2001). This distribution evidences that the 20 percent

most profitable customers generate from 150 up to 225 percent of the profit while the next 60 percent

are breaking even and the last 20 percent destroy the surplus of profit (Guerreiro et al., 2008, Nandan,

2010, Sievanen et al., 2004). An example of this profit distribution can be seen graphically on Figure 2.1.

It becomes evident that one organization having this information would gain a competitive advantage

over those who don’t possess this knowledge (Sievanen et al., 2004). Having this information enables

an organization to identify which cost objects are being harmful to the business and then apply corrective

measures on the flow of costs of those objects, thus increasing the total profits (van Raaij, 2005). It is

also clear that the correctness of this information is crucial otherwise a company would be applying

corrective measures to profitable outputs (van Raaij et al., 2003).

Even if a product, service or client is profitable, managers are also interested in knowing the structure

and the source of the costs that compose the total cost of a product. Total cost of a product is composed

by different components that deducted from the price result on the margin (or profit). Identifying correctly

9

Figure 2.1: Whale Curve of Profitability

and accurately these components is crucial to the success of cost analysis since decisions relying on

this information could have significant impacts in an organization’s condition.

Another important aspect of cost analysis is efficiency of resources. That is, to see which resources are

under-performing or over-performing and why. Gathering this information entails knowing exactly the

consumption of each resource versus its capacity.

To sum up, our motivation for this research is that organizations, especially SMEs, need to be able to

make decisions based on accurate information provided by accurate costing methodologies. We want

to enable organizations to gather this precious information that could have a positive contribute to their

efficiency.

We will not focus on the profitability and capital aspects of the cost analysis within this document. How-

ever, we consider that it is essential to motivate the cost analysis with both the costs and profitability

aspects since they are linked when an organization thinks of cost analysis. Costing is the baseline for

profitability analysis so we motivate our research with both costs and profitability issues, even though

we will only focus on costing issues.

2.3 Research Questions

We propose in this thesis that cost analysis should be offered as a service with the intent of making cost

analysis more affordable regarding both capital and expertise. The proposed service should also be

easy to use so that SMEs can access accurate costing methodologies with minimum knowledge of cost

analysis. In order to provide such needs, our service will support the instantiation of business process

cost templates from different industries.

10

With this in mind, our research question which will serve as entry point for our research is:

RQ: How can cost analysis be offered as an accessible service?

Since we intend to make cost analysis less demanding both in terms of capital and financial expertise,

we chose the word “accessible” to define the set of capabilities that we want to achieve.

We believe that this question can be answered with the conception of an artefact that offers a costing ser-

vice using an accurate costing methodology and business process cost templates in a cloud computing

environment and business model.

We want to test our artefact using field studies and data from real world organizations in multiple in-

dustries. We will evaluate our artefact with real-world testing and interviews while the research will be

evaluated according to the principles proposed by Osterle et al.(Pries-Heje et al., 2008) and the Pries-

Heje et al.framework (Osterle et al., 2010).

11

12

Chapter 3

Related Work

This chapter covers the second step of DSRM, definition of the objectives of a solution, in which we infer

the solution’s goals from the problem definition and the related work. We will provide an overview of the

tools and methods available that could contribute to solve the identified problem.

3.1 Costing

In this section we will provide an overview of Cost Accounting, what is it and what is it used for. The most

relevant concepts necessary to understand the proposal will be emphasized throughout this section.

Finally, we will present the most relevant costing methodologies analysed.

3.1.1 Cost and Management Accounting

Cost Accounting (or costing) can be defined as the process of collecting, classifying, assigning and

analysing the costs associated with the activity of an organization (Blocher, 2005).

According to Vanderbeck (Vanderbeck, 2012), Cost Accounting provides the detailed cost information

that management needs to control current operations and plan for the future. The goal of cost account-

ing is to gather all possible information so that it can be structured and used by management to take

decisions and measure the organization’s performance.

Management Accounting, on the other hand, is the internal business building role driven by account-

ing and finance professionals inside the organization. Main goals of Management Accounting include

helping the design and evaluation of business processes, budgeting and forecasting, implementing and

monitoring internal controls, and aggregating information to help drive the organization’s economic value

(Vanderbeck, 2012).

13

Information Systems built to support Cost Accounting are called Cost Accounting Information Systems.

These systems are a subset of Accounting Information Systems (AIS). Main goals of these systems

include recording transaction data and calculating the cost of the outputs of an organization (Vanderbeck,

2012).

3.1.2 Cost Concepts

Cost – An organization incurs a cost when it uses a resource for some purpose. E.g., an organization

that produces manufactured products incurs in a cost each and every time a raw material is used in

production or manufacturing labour time is consumed. Costs can be classified according to several

aspects: Nature (labour, material or expenses); Traceability (direct if economically traceable or indirect

if not); Behaviour (fixed if the total cost does not change according to volume (e.g. rental) or variable if

it is affected by volume (e.g. material) (Blocher, 2005, Vanderbeck, 2012).

Cost Pools – Meaningful groups in which costs are often assigned (Blocher, 2005).

Cost Driver – Any factor that as the effect of changing the amount of total cost. E.g. a cost driver is any

factor that causes a change in the cost of an activity. For an organization that competes on the basis

of cost leadership, management of key cost drivers is critical. Cost drivers are also known as allocation

bases (Blocher, 2005).

Cost Object – Any product, service, customer, activity, or organizational unit to which costs are assigned

for some management purpose. Products, services and customers are generally cost objects while man-

ufacturing departments are considered either cost pools or cost objects, depending on whether manage-

ment’s main focus is on the cost for the products or for the manufacturing departments. Cost objects are

essential to decision making, performance measurement and strategy implementation (Blocher, 2005).

Cost Assignment – The process of assigning resource costs to cost pools and/or cost pools to cost

objects. There are two types of assignment – direct tracing and allocation. The first is used for

assigning direct costs while the second is used for indirect costs. Costs are either direct costs or indirect

costs, regarding the cost pool or cost object being considered (Blocher, 2005).

Direct Cost – A cost that can be conveniently and economically traced directly to a cost pool or a cost

object. E.g., cost of raw materials required to manufacture a particular product is a direct cost because

it can be traced directly to the product (Blocher, 2005).

Indirect Cost – A cost without a convenient or economical traceability to a cost pool or from the cost

pool to the cost object. Costs related to management staff or material handling are good examples of

costs that cannot be usually traced to individual products and therefore are considered indirect costs.

Since this type of costs cannot be traced to the cost pool or cost object, the assignment of these costs

is made using cost drivers. This process is known as cost allocation (Blocher, 2005).

14

Product Costs – Only the costs necessary to complete the product are included: direct materials, direct

labour and overheads related to the product (Blocher, 2005).

Period Costs – All non-product expenses for managing the firm and selling the product that are not

involved in the manufacturing process such as general costs, selling costs, and administrative costs

(Blocher, 2005).

Figure 3.1 summarizes the most important concepts described above and their relationships.

Figure 3.1: Cost elements diagram (adapted from (Blocher, 2005))

3.1.3 Costing Methodologies

Activity Based Costing

In the early 80’s a new costing methodology was created with the intention of bringing forward a new

approach to cost information (Cooper & Kaplan, 1987) that was not limited to tax and financial accounting

which was the mainstream at the time (Hicks & Costing, 2002). Activity-Based Costing aims to solve the

problem related to the ability to trace overheads costs (e.g. IT, Administrative, Marketing) to activities

done within the organization that must reflect their cost on the organization’s outputs (products and

services). That is, organizational outputs (such as products or services) give rise not only to functional

activities or tasks that directly contribute to their production (e.g. manufacturing) but also to support

activities that are needed within the organization to support them (such as IT and Marketing).

In order to implement the ABC methodology an organization should start by collecting information about

all the functional and support costs, such as those abovementioned. Next, the trace between those

costs and the activities that raised the need of those resources must be built. Afterwards, these activities

should be routed to the organizational outputs that made those same activities necessary.

ABC methodology defines an activity as an action executed inside an organization (e.g. packaging or

delivering a product) that have a particular cost rate based on the cost of the resources allocated to that

activity. Allocation of resources to activities and then to the products or services of an organization is

15

done based on interviews to those involved in the activities as well as in some estimates provided by

the management team. This process results on splitting the costs related to the resources used by the

activities, using variables like percentage, headcount or any other unit that is found relevant to the cost

analysis. Finally, output costs are calculated by summing the costs of all the activities that were needed

to create the final product or service (Blocher, 2005).

Traditional costing methodologies assign overhead costs by volume, that is overhead costs or support

activities are distributed by products using some variable (or driver) that reflects capacity usage (e.g.

number of hours) regardless of the specificities of the product. On the other hand, ABC uses activities

which mean that different products may use a set of different activities and therefore a set of different

cost rates to calculate the final cost of a product or service.

Although ABC has some advantages over traditional costing systems it also has some pitfalls. First,

costs are calculated using individual subjective estimates from the management team and data pro-

vided from interviews with workers. The accuracy of these estimates may be questionable since, in

most cases, there is no evidence of the correctness of the data. Wrong estimates may lead to errors

and distort measurements (Kaplan & Anderson, 2007). Second, ABC requires not only the creation

of an activity for every task performed inside the organization but also its cost specification. Thus, the

complexity of the model grows with the number of activities. Finally, since it is common to have activities

with variable costs (e.g. special packing vs standard packing) and ABC defines activities as single tasks

with fixed cost rates, models tend to have many similar activities just to simulate variable costs.

Time-Driven Activity Based Costing

Time-Driven Activity Based Costing (TDABC) (Kaplan & Anderson, 2004, 2007) is an alternative costing

methodology to ABC developed to calculate the profitability of products and services focusing on as-

signing overhead costs to these cost outputs. This methodology was created to address the ABC pitfalls

described earlier.

The TDABC model simulates the actual processes used to perform work throughout an enterprise,

therefore capturing far more variation and complexity than a conventional ABC model. Such variation

and complexity is captured without significant demand for data estimates, storage, or processing capa-

bilities. This model aims simplicity and flexibility while focusing on assigning overhead costs (Kaplan &

Anderson, 2004, 2007).

Targeting simplicity, TDABC assigns resource costs directly to the cost objects requiring only two sets of

estimates, neither of which difficult to obtain: the cost of supplying resource capacity for the department

and the capacity usage by each transaction processed in the department (Kaplan & Anderson, 2007).

The cost of supplying resource capacity (or capacity cost rate) is the total expenses related to a particular

resource (e.g. IT Department) divided by its capacity (normally expressed in time). Total expenses of a

department may include costs such as personnel, supervision, occupancy, equipment and technology.

16

On the other hand, capacity is the time available from the employees to actually perform the tasks.

Capacity used by each transaction is the estimated consumption of capacity (typically unit times) needed

to execute a particular activity. After estimating these two values it is possible to calculate the cost of an

activity simply by multiplying the estimated consumption by the unit cost of a resource.

Regarding flexibility and concerning the limitations of ABC of each activity reflecting only one fac-

tor/condition (Dejnega, 2011), TDABC introduces the concept of time-equations. As we described ear-

lier, ABC required the creation of a new independent activity for each small variation needed to drive

the cost. However, in TDABC, linear equations are used to model the different resources consumed by

an activity. An activity may have different consumption of resources (such as time) depending on the

conditions that occurred in a particular instance of that activity. If we take as an example the packaging

of an order that takes longer when gift wrapping is requested, in ABC there would be two activities: one

for standard packaging and another for gift wrapping. However, in TDABC it is possible to express this

variation with the following equation:

Packaging = (3 + 5 ∗ ϕ) ∗ Logistics Dtp. CCR

ϕ = 1 if gift wrapping ∨ ϕ = 0 otherwise(3.1)

Finally, TDABC provides mechanisms to gather information about its own accuracy and to identify possi-

ble wastes or inefficiencies (Kaplan & Anderson, 2007). Given the capacity of a resource and the actual

capacity used within a time period it is possible to calculate waste or inefficiency. If the sum of times

used by a particular resource in the activities is below its total capacity it usually indicates that there is

some sort of waste or inefficiency. On the other hand, if it is above, it normally indicates overuse. Indeed,

discrepancies between capacity and its usage may also indicate errors in the model construction.

3.1.4 Costing Templates

Business Process Cost Templates

Business Process Cost Templates is a method to reduce the costs of adopting efficient costing method-

ologies, such as TDABC, through re-utilization and standardization of business processes for organiza-

tions inside the same field or industry (e.g. Pharmacies, Hospitals). The main goal of these templates is

to dilute the costs associated with the analysis required to implement a costing methodology, in particular

TDABC, making the adoption of such methodologies more affordable (Lourenco, 2013).

The method that creates a template for a particular field is composed of two distinct phases: a Modelling

Phase and an Application Phase. The first is done only once and is where the field or industry is

analysed and a generic cost model is developed using an organization within the same field or industry.

17

The second, results of the application of the template produced in the previous step. The template is

instantiated and the specificities of the organization are set. These specificities may include addition

or removal of activities, changing the coefficients in time-equations, or adding some unrepresented

condition. This adjustment is crucial since not all organizations are identical, even though they belong

to the same industry or field (Lourenco, 2013).

Regarding the Modelling Phase, it’s important to describe briefly the six steps needed to create the

template:

1. Identification of Resources: Consists in identifying all the resources used in the organization

and creating the allocation tree that represents how they organize themselves into resource pools.

In some way, these resource pools represent the departments that will execute the activities. This

structure of resources allows the identification of direct costs while assigning overhead costs to

functional departments.

2. Definition of Cost Structure: This step results in a tree representing multiple levels of aggregation

cost objects that are possible to define starting from the transaction level going as high as the

organization and business level.

3. Definition of Activities: Consists in identifying the activities performed by an industry. For each

identified activity, the usage of resources is modelled. Using BPMN and TDABC’s time-equations,

resources and their costs are associated to activities. It is important to note that each activity is

linked through the time-equation with one and only one resource pool.

4. Construction of Business Processes: Since the main goal of this method is to maximize re-

utilization, the same activity can be used in several different processes. Considering this approach,

the creation of business processes simply consists on selecting the activities that compose a

particular business process plus the specification of the occurrence of conditions and their order.

Business Process costs are obtained by adding the values (costs) that resulted from the time-

equation defined that link each activity to that Business Process.

5. Allocate Processes to Products: The cost of a product will result from the sum of the costs of

the processes that contribute to it. So, for each product, the processes that contribute in some way

for its creation must be allocated to it.

6. Definition of Segmentation Groups: Some levels of variation can be expressed in this step.

That is, inside some industry, there is sometimes some segmentation that can be based on a set

of variables such as equipment, geographical location, size, etc.

The output from the modelling phase is a generic template for some industry that can be used in the

application phase where it is instantiated for a specific organization. The Application Phase consists of

five steps:

18

1. Resource Cost Gathering: The cost hierarchy obtained in the modelling phase represents the

resources of the organization and how support costs are allocated. Total cost and practical capac-

ity for each resource should be gathered in order to apply the template to a specific organization.

Afterwards, these costs will be reflected on the activities, processes and products.

2. Segmentation Variables choice: In this step it is necessary to choose the correct segment of the

template by choosing the one which variations better reflect the reality of the organization.

3. Application of Template: After gathering the specific resource costs, applying the template con-

sists only on instantiate the processes, activities and time-equations.

4. Adjustment of Template: Even though organizations belong to the same industry and have com-

mon processes, they are not the same organization. Costs may be different as well as the set of

activities or the sequence of processes. Adjustments to adapt the template for a particular orga-

nization are done within this step. These adjustments include adding or removing processes or

activities and changing the drivers in the time-equations, in order to reflect the characteristics of

the organization being instantiated.

5. Calculation of Costs: Calculations for products and services are done using the rules defined by

the TDABC methodology. To run the model, the method receives as input data from transactions

that is needed for the time-equation variables defined in the modelling phase.

3.2 Cloud Computing Services

Cloud computing is one of the major technological challenges for organizations nowadays (Dillon et al.,

2010, Marston et al., 2011, Wei & Blake, 2010). Cloud Computing promise to deliver all the functionality

of existing IT services while dramatically reducing the upfront costs of computing that prevent many

organizations, especially SMEs, from deploying many cutting-edge IT services (Armbrust et al., 2010).

Traditionally, SMEs had to make substantial upfront capital investments to procure the needed IT infras-

tructure and expertise (e.g. developers and system administrators) leading to high ownership costs of

needed IT services (Garg et al., 2013). Therefore, Cloud Computing is often seen as a solution for this

problem as it is based on the premise that virtualization and shared computing resources lower the costs

with infrastructure and expertise needed by the organizations (Marston et al., 2011).

Moreover, another economic appeal of Cloud Computing is the CapEx to OpEx transition meaning that

capital expenses are converted to operating expenses (Rafique et al., 2011). Expressions such as pay-

as-you-go or pay-per-use are the ones that more effectively capture the economic benefit to the buyer

(Armbrust et al., 2010). The global Cloud Computing industry is expected to be a 100 billion dollar

business by 2014 (Columbus, 2013).

19

The National Institute of Standards and Technology (NIST) defines Cloud Computing as: a model for

enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing

resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned

and released with minimal management effort or service provider interaction.(Mell & Grance, 2011)

NIST also states that this Cloud model is composed of five essential characteristics, three service mod-

els, and four deployment models.

The five essential characteristics are (Mell & Grance, 2011):

1. On-demand self-service: A consumer can unilaterally provision computing capabilities (e.g. net-

work storage) as needed automatically without requiring human interaction with each service

provider.

2. Broad network access: Capabilities are available over the network and accessed through stan-

dard mechanisms that promote use by heterogeneous client platforms (e.g., mobile phones and

workstations).

3. Resource pooling: The provider’s computing resources (e.g. storage and server time) are pooled

to serve multiple consumers using a multi-tenant model, with different physical and virtual re-

sources dynamically assigned and reassigned according to consumer demand. Generally, the

consumer has no control or knowledge over the exact location of the provided resources but may

be able to specify location at a higher level of abstraction (e.g. country or datacenter).

4. Rapid elasticity: Capabilities can be elastically provisioned and released, automatically or with

human interaction, to rapidly scale in or out according to the demand. To the consumer, the

capabilities available often appear to be unlimited and can be acquired in any quantity at any time.

5. Measured service: Cloud systems automatically control and optimize resource use at some level

of abstraction appropriate to the type of service (e.g., storage or active user accounts). Resource

usage can be monitored, controlled, and reported, providing transparency for both the provider

and consumer of the service.

Along with the five essential characteristics the NIST also defines three Cloud Services models (Mell &

Grance, 2011):

1. Software as a Service (SaaS): The supplier’s applications run in a cloud infrastructure and are

provided to clients. A client may access those applications using various devices through a thin

client interface, such as a web-browser. The consumer does not manage or control the underlying

cloud.

2. Platform as a Service (PaaS): Clients may deploy self-created or acquired applications to the

cloud infrastructure using the services and the tools supported by the provider. The consumer

20

does not manage or control the underlying cloud.

3. Infrastructure as a Service (IaaS): Clients are provided with processing, storage, network, and

other fundamental computing resources where the consumer is able to deploy and run arbitrary

software, including operating systems and applications. The consumer does not manage or control

the underlying cloud but has control over the characteristics abovementioned.

These Cloud Service models are offered in five deployment models (Mell & Grance, 2011):

1. Private cloud: The cloud infrastructure is provisioned for exclusive use by a single organization.

It may be owned, managed, and operated by the organization, by a third party or by a combination

of both.

2. Community cloud: The cloud infrastructure is provisioned for exclusive use by a specific com-

munity of consumers from organizations that have shared concerns (e.g., security requirements).

It may be owned, managed, and operated by one of the organizations, by a third party or by a

combination of both.

3. Public cloud: The cloud infrastructure is provisioned for open use by the general public. The

organization that owns, manages and operates the infrastructure may sell Cloud services.

4. Hybrid cloud: The cloud infrastructure is a composition of two or more distinct deployment models

(private, community, or public) that remain unique entities. They are bound together by standard-

ized or proprietary technology that enables data and application portability.

3.2.1 Service-Oriented Computing and Cloud Computing

As competitiveness grows, organizations are required to constantly modify their IT systems by adding

new features or deleting old ones (Wei & Blake, 2010). Traditional software lifecycle models failed to

address this requirement for continuous integration (Blake, 2007). Service-oriented computing aims to

construct rapid and low-cost applications without sacrificing security and reliability, being this one of the

major challenges of a service-oriented approach (Papazoglou et al., 2008). A service is different from a

traditional software artefact since it is autonomous, self-described, reusable, and highly portable (Wei &

Blake, 2010).

21

22

Chapter 4

Objectives

This section corresponds to the Objectives Definition step of Design Science Research Methodology

(DSRM).

As we stated before in the problem section the main issue with cost analysis is that it is too expensive for

the majority of the organizations in both human and financial capital. However, we identified a clear need

for cost analysis, that use accurate costing methodologies, which help organizations stay aware of the

flow of costs throughout the organization. This awareness helps the process of decision making since

managers need correct and accurate information about costs and profitability so that they can decide

which products or services should be improved or ceased, in order to maintain the sustainability of the

organization.

We also identified another problem regarding cost analysis. It is often offered as a one-time big project

instead of being offered as a service. Constant changes in business processes and in the markets where

organizations operate urge the need of a costing service, instead of a one-time analysis. Managers

should be able to test new strategies as well as refine processes, activities and costs of resources, that

also change frequently over time.

Therefore, the main objective of this proposal is to present a cost analysis service, targeted to SMEs,

which solve the abovementioned problems. This service should be more affordable and easier to use

than current solutions being also less dependent on external consultancy services. Our intent is that

these characteristics will allow smaller organizations to conduct a bigger share of the cost analysis

process.

Besides this main goal, we also want to achieve some other more specific objectives:

• The service should be easily available to organizations;

• The service should support any kind of organization or industry;

23

• The service should not require specialized expertise upon the moment of application on an orga-

nization;

• The service should be more affordable than traditional methods from the beginning of the cost

analysis until its end, i.e., the moment when results are obtained.

24

Chapter 5

Proposal

This section maps to the design and development step of Design Science Research Methodology

(DSRM) and its output is an artefact, is this case an instantiation, that aims to solve the identified

problem and the objectives defined earlier in this document.

As we identified earlier in the problem chapter, SMEs cannot access the tools needed to perform cost

analysis using accurate cost methodologies (such as TDABC) due to the high costs associated with

the required time, expertise (internal know-how or consultancy services) and especially, expensive and

very complex software solutions. Our intention with this proposal is to provide the means and the tools

needed to create a costing service that provides access to an accurate costing methodology (TDABC)

without the complexity and the higher costs associated with current solutions.

Considering this, we can briefly describe our proposal as a cloud-based costing service that uses

Time-Driven Activity Based Costing (TDABC) methodology and the concept of costing templates

to reduce the costs associated with cost analysis.

In order to accomplish all the previously settled goals we developed a cloud-based costing service, i.e.,

a web-based system that supports an accurate costing methodology, TDABC, and the Business Process

Cost Templates described in the Related Work chapter (Chapter 3). We will explain the features that our

system includes in order to fulfil the objectives defined earlier (Chapter 4). Afterwards, we will discuss

how these features accomplished those objectives.

5.1 Costing Service Features

1. Time-Driven Activity Based Costing methodology: According to what was discussed in the

Related Work section, our proposal uses Time-Driven Activity Based Costing to perform the cost-

ing analysis. Therefore, our service supports all the needed guidelines proposed by Kaplan and

25

Anderson (Kaplan & Anderson, 2004, 2007).

2. Business Process Costing Templates: Another core feature of the developed tool is the support

of Business Process Costing Templates. This costing templates result from the application of

the method explained in the Related Work chapter (Chapter 3). Our system supports both the

Modelling Phase and the Application Phase meaning that the template can be created or loaded

inside the application and edited as needed. Being a cloud-based costing service, not dependent

on the industry of application, the tool is generic. In order to provide this generalization the tool

was built with respect to both the guidelines proposed by TDABC (Kaplan & Anderson, 2007) and

Business Process Costing Templates (Lourenco, 2013).

3. Creation/Edition of Business Processes and Time-Equations: Although the need of this fea-

ture is also patent in the above ones, we shall explain and detail further this feature. One of the

main issues with current Account Information Systems that support TDABC models is the cumber-

some method of defining and editing business processes and time-equations. The complexity of

time equations tend to raise as the number of conditions and business processes defined grows.

This issue is usually solved resorting to external consulting services that offer the full analysis and

software configuration in a one-time project. Resorting to consultancy services raise the costs of

cost analysis as we stated before. Therefore, we implemented form-based editor of time-equations

and business processes within our tool. This form-based editor helps and guides the users by al-

lowing them to only select values that are meaningful to the analysis and that comply with the

guidelines defined in both TDABC (Kaplan & Anderson, 2007) and Costing Templates (Lourenco,

2013).

4. What-if Analysis: With the intent of helping managers test their decisions, our tool has the ability

to do simple what-if analysis as proposed by Kaplan and Anderson (Kaplan & Anderson, 2004,

2007). What-if analysis will help managers and decision-makers evaluate the effect of reengi-

neering a business process, reducing raw material costs or changing a time-equation conditions.

Our support of what-if analysis, although very simple, allows the cost analysis to be an ongoing

process, validating the need of costing as a service.

5. Data Integration: In order to avoid manual insertion of data to perform the cost analysis, our tool

supports automatic feed of data. This consists on providing the user with a template spreadsheet

formatted according to the guidelines of both TDABC and Costing Templates which is the model

expected by the system to perform the cost analysis. The implemented tool also supports data to

be imported from web services if they meet the requirements of the costing template. We explored

the possibility of automatic and direct ERP or Service Desk software integration but since the most

of the software investigated are able to export the needed data in CSV or spreadsheet format, we

opted to consider only this type of automatic data feed.

So that the information generated by this tool may be used by other tools and considering that

26

integration isn’t only about data in, we developed the ability to export, via JSON endpoints, all the

information produced by our tool. This means that every HTML containing information related to

the cost analysis can be exported or integrated with other tools via JSON endpoints. This feature

strengthens the usefulness of the service since it can be used between two other services or tools

to perform accurate cost analysis.

6. Data Visualization: To help managers conduct the costing analysis we provide innovative and

modern ways of visualizing business data. Our system, besides presenting the data in numerical

form will also provide appealing graphics and charts to help managers detect critical performance

issues. This includes visualizing the costs of business processes, business activities, resources

and resource pools. Most of these graphs will support drill-down features to help managers and

decision-makers understand the structure of costs. Implementing the charts library needed to

provide the visualization features was the responsibility of a fellow master student from the same

research group. The charts library used the data available via the JSON API implemented within

our tool and described above. This allowed us to test and verify the API created to export informa-

tion generated by our costing service.

7. Automatic Pre-configuration: As we stated before, one of the reasons that make SMEs resort

to consulting services when deploying costing software is the cumbersome and sometimes very

complex method of configuring Business Processes, Activities and Time-Equations inside the cho-

sen tool together with the lacking of the needed expertise to do the job. This is one of the reasons

leading organizations to opt for a one-time big cost analysis project. Besides increasing the costs

of cost analysis, lacking the skills needed to update the costing software to reflect new or updated

Business Processes leads to incorrect and inaccurate results when performing the analysis.

Organizations are in constant change and Business Process and Activities are constantly reflect-

ing those changes. So that our tool may address this problem, we developed a feature that makes

our tool capable of “inferring” the Business Processes, Activities and the relationship between both

of them to make the process of configuring the tool less complex. By analysing the transactions

uploaded to the costing tool, it can gather the needed information to construct the models that are

needed to match an organization workflow. These models can be later modified and updated man-

ually if needed. Therefore, this feature can be very useful to help bootstrap a cost analysis since

most of the cumbersome and tedious work would be done automatically and without demanding

specific knowledge.

Offering a costing service in a cloud environment helped us achieve the needed technological cost re-

duction. Current solutions require local software installations that raise the costs of the service because,

in addition to compel the purchase of the technological equipment needed, also involves operational

costs such as maintenance and initial configuration. Those tools are also very complex and require ex-

pertise whenever modifications to the model are needed. This issue prevent managers from performing

cost analysis as an ongoing process.

27

As for the costing methodology, we adopted TDABC for the reasons stated in the Related Work chapter

(Chapter 3). TDABC is an accurate costing methodology that solves the problems identified in previous

methodologies, that is simple to understand and implement, providing quick benefits for those who

adopt it (Pernot et al., 2007). TDABC also has clear connections with BPM that helped us connect it with

Business Process Costing Templates.

Regarding Business Process Costing Templates, we chose to use them within our service because they

provide a way of creating cost templates to some industry and distribute them for all the organizations

that operate within that industry. These templates can be created and modified by an organization or

by a cost analysis expert and be later included within our tool. Providing cost templates to more than

one company leads to cost reduction regarding the needed expertise and consultancy services, since

the cost of a template can be distributed by more than a single organization. These templates can be

later improved and adapted to the reality of the organization deploying the template. Even though the

organization may incur in a cost by doing this, it will be a lower cost when compared to the cost of a

complete analysis.

Another interesting feature of Business Process Costing Templates is that they allow benchmarking

between organizations within the same industry since they provide a core set of resources, business

activities and business processes that, although can be modified by each company, will still retain most

of the core components that can easily be used to perform cross benchmarking.

Finally, What-if Analysis, Data Integration and Data Visualization, are meant to provide means of assess-

ing the organization’s performance. Although these features are not directly related to cost reduction of

the cost analysis, which is the main problem we want to solve, they are required to comply with the

guidelines proposed by TDABC. Still, they add value to the tool since they provide the manager the

needed capabilities to perform cost analysis.

We believe that the implemented tool, which comply with all the characteristics stated above, will provide

cost analysis as a more affordable service to SMEs than current solutions.

5.2 Costing Service - Analysis Process

We will now present a typical use of our costing service. We will describe the process that a user must

follow in order to complete a TDABC analysis using our costing service.

Figure 5.1 shows the process of performing a cost analysis using the costing service. Users should

start by configuring resources and resource pools and importing transactions. The order in which these

activities are performed is irrelevant. If the costing service is being used by more than one user, each

user may execute or be responsible for one of the activities. It is also important to note that by configuring

we mean define and make the relevant associations. In the case of the resources, users should define,

28

for every resource belonging to the organization, the name of the resource, its monthly capacity and

the cost of providing such capacity. On the other hand, in the case of the resource pools, users should

define the name of the resource pool and its classification, whether it is a support resource pool or a

functional resource pool. Afterwards, users should configure the resource pool structure, that is, which

functional resource pools are charged with support resource pools.

Next, users must associate resources to the resource pools defined earlier. Resources can either be

associated to support or functional resource pools. After completing these associations, resource pools

will have their cost calculated in real-time so that the users can know the costs of their resources struc-

ture. This helps users validate the calculated values in order to prevent typing errors and other types of

configuration issues.

The next step is the most important step of the analysis. Users should first decide if they want to

automatically configure business processes and activities or if they want to manually specify them. The

main difference is closely related to the quality of the data imported to the costing service. If users

know that their data is valid and corresponds to the processes of the organization, they can let the

tool automatically configure business processes and activities. On the other hand, if users already have

some sort of ”optimized” business process template, they should manually configure the tool. Users may

also let the costing tool infer business processes and activities and then fine-tune them. We encourage

users to perform an automatic configuration since this simplifies the process of analysis even further.

Finally, users should associate the functional resource pools to the business processes that those func-

tional resource pools are accountable for and then order the software to compute the analysis. Running

the analysis finishes the process of cost analysis. However, users may change resource cost values,

fine-tune activities and business processes or change associations and then compute again the analy-

sis.

Figure 5.1: Costing Service - Analysis Process

5.3 Costing Service Tool

To start the development of the tool that instantiates the costing service, we created a data model that

respected the rules of both TDABC and Business Process Costing Templates as stated in Related Work

29

chapter (Chapter 3). Afterwards, we started the development of the costing tool needed to instantiate the

costing service. We developed our costing tool according to the guidelines defined by both TDABC and

Business Process Costing Templates meaning that they represented our requirements document. Our

artefact was developed using open-source technologies, being the core of the costing tool developed

using the Ruby on Rails web framework. The data model can be graphically seen in Figure 5.2

Figure 5.2: Data Model of the Costing Tool

To validate the data model and the developed costing tool we searched for field studies that had informa-

tion on practical applications of TDABC so that we could map the data collected by those studies, as well

as the results obtained, against our data model and costing tool. We found a set of papers describing

field studies (Adeoti & Valverde, 2012, Granof et al., 2000, Pernot et al., 2007) that were suitable for

our purpose although they provided very few detailed information. We ended up choosing a field study

related to the internet service industry that described the application of TDABC in an IT Service Provider

(Adeoti & Valverde, 2012). We then matched the information present in that field study to our data model

and loaded our costing tool with it to compare the end results of the analysis, provided by the costing

tool, against the manual analysis done by the authors of the field study.

Furthermore, we also conducted informal interviews with professionals from accounting, finance and

management to present the early stages of our costing tool and data model to gather feedback. In those

interviews the costing tool and the model were well accepted and we were able to gather feedback from

different viewpoints that helped us improve the first versions of the proposed artefact.

30

we will now provide an overview of the developed artefact (CaaSH) that implements the proposed costing

service. The overview will focus on the features that map those defined earlier in this chapter. We must

warn that all the monetary values presented in this document are merely indicative and do not represent

the reality. For confidentiality reasons we used Resource cost values that do not match reality. We will

further detail this aspect in the Demonstration chapter (Chapter 6).

Figure 5.3 shows the Dashboard of our tool. We opted to provide information regarding the number of

Transactions (8682) used to make the analysis, the number of Business Processes (77) identified in the

analysis and the list of top five most costly Resource Pools and Business Processes. Users can can

rapidly get more details of the displayed information in the Dashboard by following any of the hiperlinks

available.

Figure 5.3: CaaSH - Dashboard - Costing Tool

A user can view the transactions that were imported to the tool in the Transactions menu, as shown in

Figure 5.4. Users can create new transactions manually or import sets of data as described earlier in this

chapter. It’s also possible to perform an automatic configuration. As previously explained, this feature

will ”infer” from the transactions available the business processes and business activities configurations

and associations to simplify the process of bootstrapping the tool.

31

Figure 5.4: CaaSH - Transactions Sample - Costing Tool

Figure 5.5 shows the view from Business Processes menu. In this menu users have access to two

fundamental types of information: the average cost per execution of a business process and the

number of instances executed. The number of instances is an hyperlink to a page where users can

see detailed information about those instances, such as the real cost of each instance of a particular

business process, as shown in Figure 5.6. As stated before, this enables organizations to know the cost

of a particular instance of a business process. Organizations may also find erroneous paths or instances

that are significantly over or below the average value. Still, in the Business Processes menu, users are

able to search for specific information in the table and order its columns dynamically. These features are

common to all the tables available in the costing tool.

32

Figure 5.5: CaaSH - Business Processes Sample - Costing Tool

Figure 5.6: CaaSH - Business Processes Instances Samples - Costing Tool

Regarding resources and resource pools associations, Figure 5.7 shows how a user can associate

resources to resource pools. This association is made using a percentage value meaning that a resource

can belong to more than one resource pool. The tool automatically verifies if the sum of the values

33

associated from resources to resource pools is greater than 100 percent. If such a condition is verified,

the tool alerts the user preventing wrong values to be saved. This avoid distorted results after an analysis

is performed.

Figure 5.7: CaaSH - Resources and Resource Pools Associations - Costing Tool

Finally, Figure 5.8 shows an example of a chart generated within our costing tool. As we stated before,

the charts library was not made within the scope of this thesis but rather from the work of another master

thesis student from the same research group, under the theme ”Visualizing Business Process Costs”.

Our costing tool just exposed the API with the data needed to the graphical analysis. The data contained

in this visualization comes from the previously mentioned case study of TDABC in an IT Service Provider

(Adeoti & Valverde, 2012).

34

Figure 5.8: CaaSH - Chart Example - Costing Tool

5.4 Summary

This chapter presented the proposal to solve the problem of this research. This proposal consists on

an instantiation of a costing tool aimed to solve the problem of high costs of cost analysis through the

implementation of a cloud-based costing service. We aim to reduce the cost, the complexity and the

expertise needed to conduct a cost analysis inside an organization, typically a SME.

We presented the main features of our tool and how they contribute to solve the problem raised by this

research. We believe that our proposal and the demonstrations described in the next chapter have a

clear contribution to solve the identified problem.

35

36

Chapter 6

Demonstration

This section corresponds to the demonstration step of DSRM where the researcher must demonstrate

the proposal stated in the previous step.

According to the DSRM, our demonstration will consist in an instantiation, i.e., an implemented system

that address the problem stated earlier (Chapter 2). Therefore, in order to demonstrate our proposal we

developed a costing tool that supports the cloud-based costing service proposed, as stated in the previ-

ous chapter. In this chapter we will extensively describe our demonstrations that validate the proposed

costing service.

We demonstrated our proposal by instantiating our artefact (the costing service) in three real world or-

ganizations, namely ”Social Security IT Institute (Instituto de Informatica da Seguraca Social)”, ”Defence

Data Center (Centro de Dados da Defesa)” and “Card4B”.

The demonstrations consisted in instantiating the costing service to a particular organization, i.e., cre-

ating the cost template, within the costing tool, to the organization being tested. This includes the

definition of resources, resource pools, business activities, business processes and the relationships

between these entities to the particular environment of the organization being tested.

For each organization, we will start by explaining why we chose that organization to demonstrate our

proposal and what we expect to demonstrate with it. Next, we will explain the scope of the demonstration

that was defined with the organization and what was expected to be delivered from the demonstration.

Afterwards, we will explain how we obtained the needed data and how it mapped to the features and

concepts defined in earlier chapters. Finally, we will closely relate the data with the costing tool and

present the results obtained. We will do a more thorough and descriptive explanation of the demonstra-

tion in the first organization, so that all concepts become clear. We will then try to simplify the exhibition

of the remaining cases because the costing tool follows a generic approach to cost analysis being able

to support different types of organizations and businesses without modifications.

37

6.1 Social Security IT Institute

The Social Security IT Institute is public institute, integrated in the indirect state administration, with ad-

ministrative and financial autonomy and its own assets. It is an organization with nationwide intervention.

Although several state competences have been assigned to the Social Security IT Institute, we focused

our demonstration in the service desk competences of the institute.

As we stated before in the previous chapters, the quality of de cost analysis is highly dependent upon

the quality of the data available. In order to demonstrate or proposal, we needed an organization with

some degree of maturity regarding both the business processes and activities. We were looking for

organizations that had their business processes and business activities well defined and controlled so

that we could completely demonstrate our proposal. It was in this context that the Social Security IT

Institute appeared. After setting up the first details, it was agreed that the demonstration would focus on

the service desk operations of the institute.

The objectives definition was held in conjunction with the Institute. The organization stated that the

most interesting outcome they could get from a costing tool was the cost of an incident or a service

request. Being a service desk, incidents and service requests are the most performed activities. Even

though they are the most executed activities, the institute doesn’t have a clear knowledge about the costs

involved in each service request or incident. So it was agreed that the objective of our demonstration in

the Social Security IT Institute was to calculate the cost of each incident and service request. Has

an outcome, the organization wanted to know, for each incident or service request, how much does it

cost to execute the business process that accomplishes the service request or incident.

The Social Security IT Institute uses a software solution (EasyVista) for IT Service Management. This

software is responsible for gathering data about Incident Management, Problem Management, Request

Management, Change Management and Service & Operation Level Management. All the processes

and activities are configured and defined in this IT Service Management software, according to the best

practices defined internally. This software holds the service catalogue and all the workflows that control

how operations are managed and how incidents and service requests solutions are accomplished.

Following the process described in Section 5.2, we started by what we consider the most difficult task

in the process of cost analysis: gather relevant data to feed the transactional data needed to perform

the analysis. We asked the Institute to provide us with a CSV or spreadsheet file containing the data

to be imported as transactional data. Since they use EasyVista for IT Service Management, the data

was easily exported to the format expected by the application. A sample of the data that was imported

to the costing tool can be seen in Figure 6.1. We had access to 8682 real transactions to perform the

analysis.

38

Figure 6.1: CaaSH - SS IT Institute - Transactions Sample

The next step in the process of cost analysis is to define and configure the resources and resource

pools that are responsible for the execution of the service desk business processes and activities. The

institute has one department (Technical Support) that is responsible for all the incidents and service

requests. Although this department is accountable for every business process regarding our analysis,

the department has other smaller departments that support several other activities that contribute for the

its own activity and function.

Regarding the resources, there are several resources involved in all the business processes and depart-

ments inside the organization. These resources are diversified and include technical and management

staff, electricity, rents, material and equipments. Although all these resources and resource pools (such

as the Technical Support department) are properly identified, the organization opted to avoid gathering

the unit costs of each resource and their contribution to the resource pools. This decision was justified

since the organization had been previously involved in an cost analysis project. From the results of this

project the organization knew the daily cost of the technical support staff from the technical support

department. However, even though they knew this cost, they were unable to link it to the execution of the

business processes, which is the main objective of our demonstration in the Social Security IT Institute.

Based on these limitations, we defined a resource and a resource pool that matched the daily value

supplied. We knew that the monthly time capacity of this resource was 8 hours/day and 22 days/month,

which was converted in minutes with a 10% waste on working hours, giving the final monthly time

capacity of 9504 minutes for each technical support worker. If we assume (since we cannot disclose

the real value) the cost of providing such capacity as 200e/day, the cost of providing 9504 minutes of

39

technical support labour would cost 4400e. Since there is no other support resources or resource pools,

this means that the CCR of this resource is 0,46e.

Following the costing service process of analysis, described in Section 5.2, we opted for the path that

led us to the automatic configuration of business processes and activities. We pursued this path since

there was no purpose in configuring the costing tool for business processes and activities that had no

match in the transactions that were supplied. Moreover, we knew that the organization was refactoring

and optimizing their service and business process catalogue, so we wanted to see if the automatic con-

figuration was able to detect strange patterns from the transactions used. Since the business processes

and activities matched those defined in the process catalogue, there was no need to fine-tune business

processes or activities. This is the expected behaviour because the transactions were generated by an

IT Service Management software (EasyVista) that controls the business processes execution according

to what was defined in the respective workflow.

Given that only exists one resource pool, the association between functional resource pools and busi-

ness processes it is always the same, i.e., the technical support resource pool is associated to every

business process configured within the costing tool.

From the 8682 transactions supplied, the automatic configuration of the costing tool was able to detect

77 business processes with 791 unique business processes instances. This means that the calculations

were performed using data from 791 complete executions of a business process (from the 77 identified).

Figure 6.2 shows a sample of the results obtained from the analysis. The samples shows, for each busi-

ness process identified, the average time and cost of execution as well as the number of instances that

were identified for that business process. The red and green rectangles also show another interesting

result from the cost analysis. The red rectangle shows a group of three business processes that were

identified by the costing tool and that correspond to the same business process, as we were able to

verify with the Social Security IT Institute management. This results from the wrong definition of work-

flows inside the IT Service Management software (EasyVista) which leads to wrong categorization of

incidents/service requests in the service desk. The green rectangle shows the same problem described

earlier but this time with an even minor difference (name pluralization).

Not only the costing tool delivered what was expected, i.e., the cost of executing the business process

that accomplishes the resolution of a service request or incident, but also provided valuable insights

to the organization regarding the workflows definition in the IT Service Management software. The

organization can easily know the average time and cost of executing a business process. Moreover, the

institute can further analyse the data and find the cost for every execution of every business process.

The Social Security IT Institute management members considered these results very useful since they

can now further analyse the different costs that the same business process generates. For instance, a

desktop related incident as an average cost of 23,22e but the minimum cost and the maximum cost of

such incident was, respectively, 2,31e and 70,83e. Having this information, the management may now

try to understand what motivated such difference and take measures to mitigate the cause.

40

Figure 6.2: CaaSH - SS IT Institute - Business Processes Sample

6.2 Defence Data Center

The Defence Data Center belongs to the General Secretariat of the Ministry of National Defence and

among its several competences they are also responsible for service desk activities. The interest on

demonstrating our proposal in the Defence Data Center arose when a member (IT Director) from the

organization saw one of our presentations describing the costing service. The presentation that gener-

ated the interest was focused not only in the costing service but also its instantiation, i.e., the costing

tool. This practical approach was very appreciated since we demonstrated that it is possible to deliver a

quick and accurate cost analysis using our costing service, as we demonstrated with the case study of

the Social Security IT Institute.

The Defence Data Center service desk uses EasyVista software for IT Service Management, i.e., the

same software used by the Social Security IT Institute. It is also configured according to the best prac-

tices defined internally (ITIL). This means that the demonstration was almost identical to the previous

case. Again, we followed the costing service process of analysis (Section 5.2), starting with the integra-

tion of the transactional data. The feed of transactional data was again done through a CSV file exported

from EasyVista, the IT Service Management software.

A sample of the data that was imported to the costing tool can be seen in Figure 6.3. We had access to

34917 real transactions to perform the analysis. This represented over four times the number of real

transactions available in the Social Security IT Institute. However, the costing tool is prepared to work

41

with a high volume of transactional data, so the performance of the solution was not affected and all the

features remained functional.

Figure 6.3: CaaSH - Defence Data Center - Transactions Sample

Having the transactional data imported, the next step was to define resources and resource pools. The

Defense Data Center has nine functional resource pools that are responsible for solving the incidents

and service requests from both the ”CDD” (Defence Data Center) and ”DSSI” (IT Services Direction).

The list below shows the functional resource pools that belong to each organizational unit.

CDD:

• ATASA - Application Systems Department

• ATAOS - Systems Operation Department

• ATAU - User Support Department

• ATACS - Communications and Security Department

DSSI:

• ATPCOIG - BI/BW Department

• ATDDM - Master Data Department

• ATIL - Logistics Department

• ATIRH - Human Resources Department

• ATIF - Financial Department

Each of these functional resource pools is responsible for the execution of a distinct set of business

processes. This means that each one of these resource pools was associated, within the costing tool,

to a different set of business processes. A business process can only be associated to one functional

resource pool, as we stated before in Chapter 3. Moreover, each of these resource pools has a set

of resources. In this particular case, all the resources that are directly linked to functional resource

42

pools are human resources. Again, each of these resources were configured within the costing tool and

associated to the respective resource pool. However, besides the human resources, both ”CDD” and

”DSSI” have other indirect costs, such as costs related to infrastructures, general expenses, material,

and more.

Regarding these indirect costs, they could not be directly linked to the resolution of incidents and service

requests because these are costs that support the overall activity of both the ”CDD” and the ”DSSI”.

However, we could gather data from indirect costs that were specific for each of these departments.

In order to distribute such costs to the resolution of incidents and service requests, we followed the

method described in the Related Work (Chapter 3). We created three support resource pools, one for

each department (”CDD” and ”DSSI”), containing their specific indirect costs, such as general expenses

and material for that department, and another one, common to both of them, containing the common

expenses, such as electricity, water, infrastructures and more. The first ones were fully associated to

the belonging department while the last, the support resource pool that is common to both ”CCD” and

”DSSI”, was associated with a 50% weighing to both departments.

After configuring these resource pools associations, we had our cost structure finished. Afterwards,

following the costing service analysis process (Section 5.2), we let the costing tool automatically config-

ure business processes and activities and then we associated the business processes to the respective

functional resource pools listed above. We had to consult the service catalogue the perform such as-

sociation. Running the analysis led us to the final results while completing the costing service analysis

process.

From the 34917 transactions supplied, the automatic configuration of the costing tool was able to detect

453 business processes with 5340 unique business processes instances. This means that the calcula-

tions were performed using data from 5340 complete executions of a business process (from the 453

identified).

Figure 6.4 shows a sample of the results obtained from the analysis. The sample shows, for each

business process identified, the average time and cost of execution as well as the number of instances

that were identified for that business process. Again, the red rectangle shows a group of three business

processes that were identified by the costing tool and that correspond to the same business process. As

we stated in the previous demonstration, this results from the wrong definition of workflows inside the

IT Service Management software (EasyVista) which leads to wrong categorization of incidents/service

requests in the service desk. Once more, we assumed resource and indirect costs that do not match

the real values provided by the Defence Data Center.

43

Figure 6.4: CaaSH - Defence Data Center - Business Processes Sample

The costing tool delivered what was expected, i.e., the cost of executing the business process that

accomplishes the resolution of an incident or service request. It also provided valuable insights to

the organization regarding the workflows definition in the IT Service Management software. Now, the

organization can easily know the average time and cost of executing a given business process. The

organization may also know the time and cost of each instance of a given business process. This

information is valuable since an organization may try to further analyse why some executions are better

than others and take measures to mitigate potential problems that may detect.

6.3 Card4B

Card4B develops and operates integrated mobility solutions through interoperable contactless ticketing,

passenger information, embedded systems and smartphones, systems integration and business intel-

ligence. Presently, Card4B is developing a project, designated ecoDrive - Intelligent Eco Driving and

Fleet Management, which is a multidisciplinary project, targeting the public transportation network, in

which INOV is responsible for the identification of business processes (BPMN) and the cost analysis of

those business processes using a TDABC approach.

The costing service described in the proposal chapter (Chapter 5) was adopted as a solution for the

ecoDrive project since it delivered all the needed features to accomplish the objectives defined. However,

this project required that the identification of business processes and activities was done prior to the

44

system deployment, meaning that we would only have access to real transactional data after the system

enters in production, since there is no digital data from the past. Former demonstrations followed the

opposite approach. One of the major purposes of the ecoDrive project is to deliver a system capable of

monitoring and controlling the activity of a public transportation organization, so we needed to identify

and define the business processes and activities that this system should control and monitor in order

to generate the transactional data needed by our costing tool. Given that the system is being built

from scratch, there is no transactional data to let the costing service automatically configure business

processes and activities.

The business processes identification work in ecoDrive project is outside of the scope of this thesis, since

it was not directly related to the costing service proposed. Thus, we will not describe the process and the

method followed to identify the business processes and activities. However, we will describe examples

of business processes and activities that were identified as well as their relation to the functional and

support resource pools of the organization.

Starting with the business processes, we were able to identify 10 different business processes, each of

them with their distinct set of activities. For example, we identified a business process ”Occurrences”

that is related to the different events that may cause changes to the operational service of a bus. The

activity list of this business process includes ”Change Driver” and ”Change Vehicle”. Another example

was the business process ”Corrective Maintenance” that features activities such as ”Repair of damage”

and ”Damage Report”. These 10 business processes and respective activities are the ones that will

constitute the foundation for the transactional data to be exported from the software being developed by

Card4B for the public transportation industry.

Following the costing service analysis process, these business processes and activities were manually

configured within the costing tool so that the results could be obtained without needing to order the

costing tool to automatically configure them. This is the alternative path to the one followed in previous

demonstrations. An organization may either define their business processes manually or let the costing

tool automatically deduce them, as stated in the Proposal chapter (Chapter 5).

Finally, resources and resource pools were configured accordingly to what was specified by the orga-

nization deploying the system. Following the previous example, we configured a resource pool, Main-

tenance Management, that is responsible for everything related to maintenance. That is, Maintenance

Management is a functional resource pool, that is responsible for all the business processes related to

maintenance such as the ”Corrective Maintenance” business process described earlier. This functional

resource pool has a set of support resource pools such as IT Management, Financial Management and

Administration. These support resource pools have resources such as technical and management staff,

electricity, taxes, communications and more.

As we stated before, we didn’t had access to real transactional data since the system is not in production

yet. This means that we cannot show real results for this demonstration. However, the system is fully

45

tested and will be deployed to production right after the main system. In this particular case, our tool

will import transactional data from JSON web services rather than from a CSV file, enabling the costing

tool to get data in real time. As a result, after configured, the costing tool can pull data, in a given time

interval, so that it can produce updated metrics without user intervention. Although this feature is only

available to Card4B, this could be easily extended to other organizations because the expected format

is the same as the CSV format. Every organization could create and deploy a web service exposing the

data in the expected format and link it to the costing tool.

In this demonstration, we explained how the costing service (and the costing tool) was used as part of

a bigger and more complex project. Again, our costing service approach allows organizations to ac-

cess accurate costing methodologies, such as TDABC, to know their internal costs. This demonstration

helped us highlighting the generic characteristics of the proposed service. Applying our service to a

completely different industry shows how the costing service may be applied to other industries while

providing the same features and benefits.

46

Chapter 7

Evaluation

This section corresponds to the Evaluation phase of Design Science Research Methodology (DSRM).

The goal of this evaluation is to determine if the solution proposed in the Proposal (Chapter 5) solves

the problem stated in the Problem (Chapter 2).

Evaluation is considered one of the most important aspects in the Design Science Research Method-

ology due to the fact that the evaluation step validates the contribution of the solution to the identified

problem as well as their utility, quality and effectiveness (Hevner et al., 2004). The researcher should

develop the criteria and assess the artefact performance against those criteria (March & Smith, 1995).

The author also defends that the researcher should mention if it worked, how it worked and why.

The evaluation method will consist in the following steps:

1. Interviews and Questionnaires: Gather feedback from the proposal through the demonstration

and identify improvements;

2. Pries-Heje et al. Framework and Osterle et al. Principles: To formally evaluate the research;

3. Demonstration Critical Review: To critically evaluate the research, objectives fulfilment and the

demonstrations conducted;

7.1 Interviews and Questionnaires

In this section we will present the conclusions obtained from our interviews. We will start by describing

their structure and the profiles of the interviewees and then, in the next subsection, we will present the

results and our evaluation of that results considering the objectives defined.

After demonstrating our costing service, we conducted a small questionnaire to those involved so that

47

we could obtain a more structured feedback. The questionnaire was made to five interviewees with the

following business roles:

• Planning, Quality and Audit Manager

• Budget Management

• Planning and Control

• Accountable for Client Support

• IT Director

We carried out six questions to these professionals that helped us assess the artefact utility and the

fulfilment of the objectives defined earlier in Chapter 4. These questions were also defined having in

mind the needed information to formally evaluate the research. The six questions were:

1. Have you had previous knowledge of a cloud-based costing service using Time-Driven ABC method-

ology?

Yes or No Answer

2. Do you agree that the proposed service could lower the complexity, the costs and the knowledge

needed to perform cost analysis within your organization?

Scale from 1 to 10

3. Do you consider that the proposed tool delivered the functionality needed to fulfil the requirements

and characteristics of the proposal?

Scale from 1 to 10

4. Do you agree that the produced metrics improve the knowledge you have regarding business

process costs inside your organization?

Scale from 1 to 10

5. Do you agree that costing, when delivered as a cloud-based service, using Business Process Cost

Templates and Time-Driven ABC is easier to implement inside your organization?

Scale from 1 to 10

6. Would you use this tool to perform cost analysis inside your organization in an ongoing basis?

Yes or No Answer

We believe that these questions address the topics needed to correctly evaluate our research. The next

subsection will detail the results from this questionnaire.

48

7.1.1 Results

The questionnaire results were very positive and provided us feedback in different topics needed to

evaluate this research. This subsection will summarize the results and provide some insights that were

provided by the interviewees in an informal manner after responding to the questions.

The results from the questionnaires can be summarized as follows (and as shown in Figure 7.1).

1. Question 1: The participants answered unanimously ”No” to this question meaning that none of

the participants had previous knowledge of a cloud-based costing service such as proposed by

this research.

2. Question 2: The participants answered to this question with an average of 8,5. The response

range was [7,10]. We can conclude that the participants agree that our proposal could lower the

complexity, the costs and the knowledge or expertise needed to perform cost analysis within their

organizations.

3. Question 3: The participants answered to this question with an average of 7,8. The response

range was [7,9]. The practitioners considered that our tool delivered the functionality needed to

fulfil the requirements. However, some practitioners reported that the costing service should deliver

more advanced and even custom metrics. We explained that custom metrics were not part of the

proposal since our effort was to validate costing offered as a service.

4. Question 4: The participants answered to this question with an average of 8,4. The response

range was [8,9]. Again, practitioners considered that the metrics produced by the costing tool

improve the general knowledge regarding business process costs. However, they considered again

that custom metrics would be valuable to improve the tool.

5. Question 5: The participants answered to this question with an average of 6,6. The response

range was 3U[6,9]. The range of responses was higher in this case. Some practitioners reported

that although a cloud-based approach is easier to implement it brings some privacy concerns that

must be addressed. However, we explained that although our proposal is offered as a service, it

can be deployed internally, if the needed technological aspects are met. Yet, we consider that part

of the value of this proposal is the absence technological related issues and costs.

6. Question 6: The participants answered unanimously ”Yes” to this question meaning that every

participant would use our tool to perform cost analysis in their organization in an ongoing basis.

This validated our efforts to provide an easy-to-use tool that implemented the proposed costing

service.

The interviews provided us feedback that helped us evaluate our proposal. The next sections will use

the interviews results to justify some conclusions.

49

Figure 7.1: Questionnaire Results

7.2 Pries-Heje et al. Framework

Alongside with DSR guidelines and to better evaluate research artefacts, (Hevner et al., 2004) proposed

five different types of evaluation methods: Observational, Analytical, Experimental, Testing and Descrip-

tive. Even though the authors provided these evaluation paths, not much more guidance was given on

how to accomplish them.

Considering prior research done in the area of DSR evaluation, (Pries-Heje et al., 2008) developed a

framework to fill in this gap that could help researchers use and rigorously evaluate Design Science

Research and its artefacts. This framework consists on distinguishing evaluation in two separated di-

mensions: one related to the form of the evaluation, the other concerns the moment of the evaluation.

Regarding the form of the evaluation, it can either be artificial or naturalistic. In an artificial evaluation,

the solution is evaluated in a contrived and non-realistic way (using for example simulations, laboratory

experiments and mathematical proofs). In a naturalistic evaluation, a solution’s performance is tested in

real environments using real users, real systems and to solve real problems.

Concerning the moment of the evaluation, can be ex-ante or ex-post. Ex-ante means that the evaluation

takes place before the artefact is developed, meaning that it is not absolutely necessary to construct

an artefact to evaluate a theory. Ex-post means that when the evaluation is conducted the artefact is

already developed. To further clarify, the artefact can be a design product or a design process. Figure 7.2

provides a graphical view of the framework, resuming the two dimensions and the combinations possible.

In summary, there are three different aspects of a Design Science Research Evaluation (Pries-Heje et

al., 2008):

50

Figure 7.2: Strategies for DSR Evaluation (Pries-Heje et al., 2008)

1. What is going to be evaluated – Specifies if the artefact is a Design Product, meaning that it results

from a particular process and therefore is tangible, or a Design Process if it is a set of activities,

tools, methods and practices that can be used to guide the flow of production;

2. When is going to be evaluated – Specifies if the evaluation is done before the artefact (ex ante),

after (ex post) or both;

3. How is going to be evaluated – Specifies if the evaluation is performed in an artificial scenario

(artificial) or using real world systems (naturalistic).

7.2.1 Results

In the case of our research, our evaluations have always addressed the three questions (What, How and

When) the same way.

1. What is actually evaluated? The artefact evaluated is the costing service (a design product)

proposed in (Chapter 5). We evaluated the results achieved by creating the costing tool that

instantiates the costing service, the feedback collected among academics and practitioners and

the results of applying this system in practice, i.e., the three demonstrations performed;

2. How is it evaluated? The feedback gathered from experts and practitioners proved valuable to

evaluate our proposal. We also implemented our proposal in three Portuguese organizations that

represented a naturalistic evaluation since it was conducted using a real artefact, the costing tool.

3. When was it evaluated? In the case of our research the evaluation was made ex-post, i.e.,

after the development of the design product. We first developed the costing service tool and only

afterwards proceed to obtain feedback from experts and practitioners.

51

7.3 Osterle et al. Principles

Osterle et al. proposes an evaluation method based within four principles (Osterle et al., 2010). These

principles result from the memorandum (Osterle et al., 2010) mentioned in the DSRM section. It is worth

mention once more that these principles and the use of DSRM in Information Systems research have a

great support by the scientific community.

1. Abstraction: Each artefact must be applicable to a class of problems;

2. Originality: Each artefact must substantially contribute to the advancement of the body of knowl-

edge;

3. Justification: Each artefact must be justified in a comprehensible manner and must allow for its

validation;

4. Benefit: Each artefact must yield benefit either immediately or in the future for the respective

stakeholder groups.

7.3.1 Results

Our research met the four principles of Osterle. This evaluation is based on the feedback received from

practitioners, which were described in the previous sections.

1. Abstraction: The artefact we propose (the costing service) can be applied to all types of service-

oriented or process-based architecture organizations.

2. Originality: None of the interviewees had knowledge of any research or product similar to the

proposed artefact. Similar research was not found in the Related Work (Chapter 3) in respect to

the costing service.

3. Justification: Our artefact is justified by all the evaluations and positive feedback gathered during

this research, as we emphasized in this chapter. The Related Work (Chapter 3) contributed to the

relevance of this principle;

4. Benefit: According to the interviewees, at least in the industries consulted, there would be a

valuable benefit, since it would provide an easier, more affordable and faster way to implement a

costing methodology and conduct a cost analysis. We also confirmed this benefit by the results

obtained from our demonstrations, where in some cases were able to provide valuable insights to

an organization using the proposed costing service.

52

7.4 Demonstration Critical Review

We consider that our research was proper evaluated and tested with our demonstrations. We demon-

strated our proposal in three Portuguese organizations, from two distinct industries, with the objective of

providing those organizations with new ways to conduct simple and reliable cost analysis. The feedback

gathered from the questionnaires and the formal evaluations from the previous sections validated the

usefulness and the quality of the proposed costing service.

Our demonstrations targeted three organizations that are making clear efforts to consider business pro-

cess costing a priority. Two of the three organizations (the service desks), already have their services

oriented according to best practices such as ITIL and ISO 20000. That gave us an huge advantage re-

garding the costing service implementation. Although our service provides an easy way to start costing

business processes and services, we do realize that the organizations deploying such service should

have a considerable maturity in their architecture. Without such maturity the proposed service is much

more difficult to deploy because of the lack of transactional data. However, since many organizations

are already service-oriented, our costing service is a solution that delivers clear benefits to those orga-

nizations, as emphasized by the demonstration.

We were able to deliver average cost per business process and real cost per business process instance.

We consider the latter to be a great benefit to service-oriented organizations since it enables organiza-

tions to comprehend the average cost of supplying a service or executing a business process but also

the real cost for every business process instance executed. This helps organizations detect erroneous

paths and singular problems that occurred at a moment in the organization. With this knowledge orga-

nizations know their real business process costs and also know when and why some execution costed

them more. Afterwards, organizations may try to investigate the real cause of the problem, knowing

exactly which transaction and activity generated the issue. Therefore, our service can fulfil the objective

of providing insights about the flow of costs and execution performance metrics.

Finally, every interviewee that saw the costing service configuration and use from the beginning of the

process till it’s end, considered the process of conducting a cost analysis using our tool very easy and

understandable. This fact validated another objective aimed at providing a service that required low

levels of expertise with both the service and costing methodology.

53

54

Chapter 8

Lessons Learned

Our research raised several aspects worth mentioning. Some resulted from the Related Work and from

the design phase of our research while others derive from the experience and feedback obtained during

the demonstration and evaluation phases.

Regarding the lessons learned from the first stage of our work we observed that there is already a

clear effort to model business process costs and cost structures to easily provide some methods of

service chargeback and accurate cost allocation. We also noted that organizations are clearly trying to

standardize their business processes according to widely accepted best practice frameworks such as

ITIL/ISO 20000. This approach helped us note that inside the same industry some patterns emerge

when such standardization is implemented.

However, aside from these praiseworthy efforts, we were still confronted with missing communication

links inside an organization. We observed that management only wanted information that they could

understand while others were only concerned with modelling and monitoring business process data.

Almost completely unrelated with those described above, we have accountants calculating costs that

are not related or explicitly linked to business processes. Finally, IT is supposed to just provide the

needed data that feed the needs of those abovementioned.

Considering the demonstration step, the lessons learned were rather practical and closely related to the

costing service implementation described earlier in the Proposal chapter. We observed that costing is

a tough exercise. Since the accuracy given by the cost analysis is highly dependent on the quality of

the data available, it is possible to model how each condition affects or influences the cost of a business

process but it is only possible to evaluate them if they were somehow recorded. Data availability should

be considered when cost modelling starts. We realized that cost analysis and modelling should be an

incremental process since the analysis results encourage the organization to store and log more data

so that the model and the analysis can be fine-tuned according to new expectations and questions.

55

Targeting these problems, ERP’s, CRM’s and Service Desk tools are used to store data needed to

monitor and control business processes. However, we found cases where the information available was

incorrectly introduced or missing. Still, first run results motivated the demonstration participants to store

more accurate and broader data.

Finally, we also observed that organizations are not keen on providing data to the public domain, espe-

cially after the cost analysis is concluded. However, for research and development purposes the data

was provided as needed.

8.1 Objectives Accomplishment

In Chapter 4 we described a set of objectives that should be accomplished in order to costing service

proposed successfully solve the identified problem. As a main objective, we defined that our proposal

should present a costing service, targeted to SMEs, that should be more affordable and easier to use

than current solutions. Regarding the main objective, we consider that it was fulfilled since our costing

tool, that instances the costing service, is less demanding in terms of technology and expertise needed

to run the costing service.

We also settled other more specific objectives that we shall now explain how and why they were ac-

complished. The following list describes the objectives defined earlier and the reasons why they were

accomplished.

• The service should be easily available to organizations

– The costing service was created with the intent of being a cloud-based service. Organizations

can get access to the costing tool, that instances the costing service, using a SaaS approach.

Organizations may also deploy the service internally if they want to.

• The service should support any kind of organization or industry

– As we demonstrated, our proposal can be applied to different organizations and industries.

Although we only tested three organizations from two different industries, we believe that our

proposal can be applied to any kind of organization and industry since the models in which

our costing service is based are generic and are not associated with any type of organization

or activity.

• The service should not require specialized expertise upon the moment of application on an

organization

– The costing service has a simple application process as described in Section 5.2. The costing

service analysis process can be followed by any user without demanding significant or spe-

56

cialized expertise in costing. Moreover, the costing tool, which instances the costing service,

was developed to be simple and easy to use. Both the proposal and the evaluation showed

that the costing tool accomplishes the needed simplicity to be used by any organization or

user.

• The service should be more affordable than traditional methods from the beginning of the

cost analysis until its end, i.e., the moment when results are obtained

– Our costing service is less demanding regarding both human and capital costs. Demanding

less specialized expertise upon the moment of application on an organization lowers both

the human and the capital costs of the cost analysis. Furthermore, the costing service was

created with the intent of being a cloud-based service which lowers the costs related to the

technological aspects of the proposal. Since our costing service can instantiate existing cost-

ing templates for a given industry or organization, it will lower the particular or organization-

specific costs of consultancy services needed to construct the cost model of analysis. More-

over, if the organization already has their business processes defined according to a best

practice (such as ITIL), the costing tool can ”infer” the business processes and activities, so

that the cost analysis outputs match the business processes of the organization.

57

58

Chapter 9

Conclusion

Cost efficiency has always been a major concern to organizations from all industries around the world.

In recent years, economic crisis and increased competition in an increasingly global economy pushed

even further the need for cost efficiency and cost awareness. It became crucial to assess and bench-

mark an organization’s performance and to identify improvement opportunities across all sectors of the

organization and over the cost stream.

However, traditional cost accounting systems fail to provide the needed data to perform such analysis

because they are mainly focused on financial and tax accounting. New and innovative costing method-

ologies were developed to solve the problem but they are out of reach for the majority of the enterprises

because they are too expensive to implement.

We followed the guidelines proposed by the Design Science Research Methodology to help us conduct

this research. We started with the problem identification and motivation that led us to do research in

costing, costing methodologies, costing templates and cloud services. From this analysis we made the

choice of developing our artefact using concepts from accurate costing methodologies, cost templates

and cloud services.

Our artefact is a cloud-based costing service meant to provide costing as a service. This means that

our main objective was to develop a solution to reduce the costs associated with cost analysis so that

SMEs can reach the accurate costing methodologies needed to assess an organization’s cost efficiency

and performance. Our solution is more affordable and less complex than those available in the market

since we combine open source technologies, cloud services and costing templates. Respectively, this

will lower licensing costs, implementation and deployment costs, and expertise or consulting services

costs.

We validated our proposal in three Portuguese organizations belonging to two different industries: Ser-

vices industry (two Service Desks) and the Public Transportation industry. In all organizations we com-

59

pleted the demonstration by instantiating the proposed service using the developed costing tool. Follow-

ing these demonstrations we formally evaluated the proposal using the Pries-Heje et al. Framework and

the four principles proposed by Osterle et al. We also conducted interviews with experts, practitioners

and potential users. The feedback obtained was taken into account to improve the proposal and the im-

plemented service. Finally we communicated our results to the proper audiences through presentations

and demonstrations to practitioners.

In the next sections will further detail our conclusions by presenting which were the main contributions

of this proposal, the limitations we identified and remained unsolved and future work that can be used

to improve this research and that we believe would bring extra value to this proposal.

9.1 Main Contributions

We believe that our proposal brings a valuable contribution in the context of costing and cost analysis.

We also believe that our proposal and demonstration meet the requirements and objectives defined,

answering the research question we raised with this investigation. The resulting costing service allows

organizations to conduct a bigger share of the cost analysis without demanding for significant levels of

expertise or capital.

The proposed costing service not only provided a solution to the problem of high costs of cost analysis

but also delivered a tool capable of correctly and completely support TDABC and Business Process

Costs Templates. These methods combined with the capabilities of our tool produced well-defined steps

that act as a guideline for both analysts and managers looking for a cost analysis solution. The end result

is costing service capable of delivering the ability for an organization to do cost analysis with internal

resources and expertise, without needing substantial investment.

Both users and experts also evaluated the costing tool and the results obtained from the analysis using

the costing service in the demonstration step. Their feedback was highly appreciated and resulted in

several changes to the tool and it generally went accordingly to our best expectations. Interviewees

considered that the costing tool would be more affordable and easier to use than those available today,

making it a valuable contribution.

9.2 Limitations

The limitations regarding our proposal can be divided into two groups, technical limitations and concep-

tual limitations. The first group is closely related to the implemented costing tool while the second relates

to the lack of some conceptual aspects that complement those raised by this research.

60

Regarding the first group, technical limitations, although we consider our costing tool to be more than

just a prototype, we must state that it lacks some characteristics needed to be a full cloud-based costing

service, due to time restrictions. The developed costing tool does not support integration and importing

of tax and analytical accounting data. In order to avoid data input mistakes and to enable high volumes

of data integration this would be mandatory in a commercial approach. Although we did not feel the

need for this features when demonstrating our proposal, we do acknowledge the need for them in some

situations. We also consider that the developed costing tool, although producing some interesting and

useful metrics, lacks the ability to produce management reports.

As for the complementary conceptual aspects referred above, they are related to profitability, capital

and investment characteristics that are needed to correctly reflect all the costs within the analysis. The

research process and the interviews revealed that although costs and cost analysis are major concerns

to organizations, they are strongly tied to capital costs, working capitals, return on investments and

profitability. Although we excluded from the beginning of this research such concepts and metrics, we

must consider that the lack of such characteristics in our service is a limitation that must be addressed

in future research.

9.3 Communication

The last step of the DSRM is the communication of the artefact developed to the proper audience

highlighting its contributions. To do so, we decided to demonstrate our solution to practitioners and

academics.

We conducted several presentations of the proposed service and costing tool to experts in the area

and demonstrations to potential users. The audiences included the elements already mentioned in the

Evaluation Chapter (Chapter 7) and others such as IT Directors, Managers, Professors, Cost Accounting

Professionals and Cost Consultants. The presentations made to these professionals resulted in further

evaluation since we were always approached with more suggestions and improvement opportunities.

Some of these presentations were made outside the scope of the research demonstration in order to

show the value of our approach to several professionals interested in the area of costing. We would like

to emphasize the interest shown by every interviewee, since they considered that the proposed service

and costing tool deliver real benefits to SMEs.

9.4 Future Work

In this section we present a description of the possible future work to further complement the knowledge

presented in this dissertation.

61

Although we achieved three full demonstrations in two distinct industries, we consider that applying

the service to more complex and different industries would be desirable since that could validate even

further the proposed costing service. Our demonstrations were carried using a significant volume of

transactional data. However, high transactional volume industries, such as an organization from a man-

ufacturing and product market, would be advisable in order to further test the costing service.

Another interesting aspect would be to develop BPMN importing capabilities within our costing tool.

Since some organizations already have their business processes modelled in BPMN it would simplify

the process of cost analysis even further. However, this was just something to improve the definition

of business processes in case of some discrepancy between their definition and the transactional data,

since our costing tool is able to configure business processes and activities from the transactional data.

Regarding the resources and resource pools configuration, the tool should also have, as mentioned

before in the Limitations (Section 9.2), the ability to import tax and analytical accounting data. Again,

this would further simplify the process of cost analysis.

Finally, also related to the implemented service, the costing tool should be able to import more drivers

that are linked to the business processes and activities. This would enable the tool to produce more

business metrics than just the cost of business processes and activities. Since our effort was to develop

a way to reduce the cost of cost analysis, we opted to leave this feature as a future development be-

cause it would require transactional data to be much more specific than the datasets we had access to,

conditioning the costing service validation and demonstration.

62

Bibliography

Adeoti, A., & Valverde, R. (2012). A Time-Driven Activity Cost Approach for the Reduction of Cost of IT

Services: A Case Study in the Internet Service Industry.

Armbrust, M., Fox, A., Griffith, R., Joseph, A. D., Katz, R., Konwinski, A., . . . others (2010). A View of

Cloud Computing. Communications of the ACM, 53(4), 50–58.

Blake, M. B. (2007). Decomposing Composition: Service-oriented Software Engineers. Software, IEEE ,

24(6), 68–77.

Blocher, E. (2005). Cost Management: A Strategic Emphasis. McGraw-Hill College.

Columbus, L. (2013, March). IDC’s Top Ten Technology Predictions For 2014: Spending On Cloud Com-

puting Will Exceed $100B. Retrieved from http://www.forbes.com/sites/louiscolumbus/2013/

12/03/idcs-top-ten-technology-predictions-for-2014-cloud-spending-will-exceed-100b/

Commission, E., et al. (2005). The New SME Definition: User Guide and Model Declaration.

Cooper, R., & Kaplan, R. S. (1987). How Cost Accounting Systematically Distorts Product Costs.

Accounting and Management: Field Study Perspectives, 204–228.

Cooper, R., & Kaplan, R. S. (1991). Profit Priorities from Activity-Based Costing. Harvard Business

Review , 69(3), 130–135.

Dejnega, O. (2011). Method: Time Driven Activity Based Costing - Literature Review. Journal of Applied

Economic Sciences, 5(1(15)/ Spring 2011), 7-15.

Delloite. (2011). CIO Survey Report 2011 The Guerilla CIO Working Smarter to Add Value (Tech. Rep.).

Dillon, T., Wu, C., & Chang, E. (2010). Cloud Computing: Issues and Challenges. In Advanced

information networking and applications (aina), 2010 24th ieee international conference on (pp. 27–

33).

Ecorsys. (2012). Annual Report on Small and Medium-Sized Enterprises in the EU, 2011/12 (Tech.

Rep.).

Garg, S. K., Versteeg, S., & Buyya, R. (2013). A Framework for Ranking of Cloud Computing Services.

Future Generation Computer Systems, 29(4), 1012–1023.

63

Granof, M. H., Platt, D. E., & Vaysman, I. (2000). Using Activity-Based Costing to Manage more

Effectively.

Guerreiro, R., Bio, S. R., & Merschmann, E. V. V. (2008). Cost-to-serve Measurement and Customer

Profitability Analysis. International Journal of Logistics Management, The, 19(3), 389–407.

Hall, O. P., McPeak, C., et al. (2011). Are SMEs ready for ABC? Journal of Accounting and Finance,

11(4), 11–22.

Hevner, A. R., March, S. T., Park, J., & Ram, S. (2004). Design Science in Information Systems

Research. MIS quarterly , 28(1), 75–105.

Hicks, D. T., & Costing, A.-B. (2002). Making It Work for Small and Mid-Sized Companies. Wiley Cost

Management Series.

Kaplan, R. S., & Anderson, S. R. (2004). Time-Driven Activity-Based Costing. Harvard Business

Review .

Kaplan, R. S., & Anderson, S. R. (2007). Time-Driven Activity-Based Costing: A simpler and more

powerful path to higher profits. Harvard Business Press.

Lambert III, S., & Chen, K. H. (1996). Overhead Cost Pools. Internal Auditor , 53(5), 62.

Laudon, K. C., Laudon, J. P., & Brabston, M. E. (2012). Management Information Systems: Managing

the Digital Firm (Vol. 12). Pearson.

Lourenco, A. G. (2013). Analyzing Cost and Profitability using Process-based ABC.

March, S. T., & Smith, G. F. (1995). Design and Natural Science Research on Information Technology.

Decision Support Systems, 15(4), 251–266.

Marston, S., Li, Z., Bandyopadhyay, S., Zhang, J., & Ghalsasi, A. (2011). Cloud Computing - The

Business Perspective. Decision Support Systems, 51(1), 176–189.

Mell, P., & Grance, T. (2011, September). The NIST Definition of Cloud Computing (Tech. Rep. No.

800-145). National Institute of Standards and Technology (NIST).

Miller, J. G., & Vollmann, T. E. (1985). The Hidden Factory. Harvard Business Review , 63(5), 142–150.

Nandan, R. (2010). Management Accounting needs of SMEs and the Role of Professional Accountants:

A Renewed Research Agenda. Journal of Applied Management Accounting Research, 8(1), 65–78.

Narayanan, V., & Kaplan, R. S. (2001). Measuring and Managing Customer Profitability. Journal of Cost

Management , 15(5), 5–15.

OCDE. (2009). The Impact of the Global Crisis on SME and Entrepreneurship Financing and Policy

Responses (Tech. Rep.).

64

Osterle, H., Becker, J., Frank, U., Hess, T., Karagiannis, D., Krcmar, H., . . . Sinz, E. J. (2010). Memoran-

dum on Design-Oriented Information Systems Research. European Journal of Information Systems,

20(1), 7–10.

Papazoglou, M. P., Traverso, P., Dustdar, S., & Leymann, F. (2008). Service-oriented Computing: A

Research Roadmap. International Journal of Cooperative Information Systems, 17 (02), 223–255.

Peffers, K., Tuunanen, T., Rothenberger, M. A., & Chatterjee, S. (2007). A Design Science Research

Methodology for Information Systems Research. Journal of Management Information Systems, 24(3),

45–77.

Pernot, E., Roodhooft, F., & Van den Abbeele, A. (2007). Time-Driven Activity-Based Costing for Inter-

library Services: A Case Study in a University. The Journal of Academic Librarianship, 33(5), 551–

560.

Pries-Heje, J., Baskerville, R., & Venable, J. R. (2008). Strategies for Design Science Research Evalu-

ation.

Rafique, K., Tareen, A. W., Saeed, M., Wu, J., & Qureshi, S. S. (2011). Cloud Computing Economics -

Opportunities and Challenges. In Broadband network and multimedia technology (ic-bnmt), 2011 4th

ieee international conference on (pp. 401–406).

Sievanen, M., Suomala, P., & Paranko, J. (2004). Product profitability: Causes and effects. Industrial

Marketing Management , 33(5), 393 - 401.

Simon, H. A. (1996). The Sciences of the Artificial. MIT Press.

van Aken, J. E., & Romme, G. (2009). Reinventing the Future: Adding Design Science to the Repertoire

of Organization and Management Studies. Organization Management Journal , 6(1), 5–12.

Vanderbeck, E. J. (2012). Principles of Cost Accounting. Cengage Learning.

van Raaij, E. M. (2005). The Strategic Value of Customer Profitability Analysis. Marketing Intelligence

& Planning, 23(4), 372–381.

van Raaij, E. M., Vernooij, M. J., & van Triest, S. (2003). The Implementation of Customer Profitability

Analysis: A case study. Industrial Marketing Management , 32(7), 573–583.

Wei, Y., & Blake, M. B. (2010). Service-Oriented Computing and Cloud Computing: Challenges and

Opportunities. Internet Computing, IEEE , 14(6), 72–75.

Wileman, A. (2010). Driving Down Cost: How to Manage and Cut Costs-Intelligently. Nicholas Brealey

Pub.

Skoda, M. (2009). The Importance of ABC Models in Cost Management. Annals of the University of

Petrosani, Economics, 9(2), 263–274.

65

66