When Project Management Methods Collide Conditions for agile and challenges in a large ICT firm using agile and traditional methods Master’s Thesis in the Master’s Program Quality and Operations Management SANNA DAHLQVIST MILDRED SKÖLD DEPARTMENT OF TECHNOLOGY MANAGEMENT AND ECONOMICS DIVISION OF INNOVATION AND R&D MANAGEMENT CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden 2022 www.chalmers.se Report No. E2022:057 Report No. E2022:057 When Project Management Methods Collide Conditions for agile and challenges in a large ICT firm using agile and traditional methods SANNA DAHLQVIST MILDRED SKÖLD Department of Technology Management and Economics Division of Innovation and R&D Management Chalmers University of Technology Gothenburg, Sweden 2022 When Project Management Methods Collide Conditions for agile and challenges in a large ICT firm using agile and traditional methods SANNA DAHLQVIST MILDRED SKÖLD © SANNA DAHLQVIST, 2022 © MILDRED SKÖLD, 2022 Report No. E2022:057 Department of Technology Management and Economics Chalmers University of Technology SE-412 96 Göteborg Sweden Telephone + 46 (0)31-772 1000 Gothenburg, Sweden 2022 When Project Management Methods Collide Conditions for agile and challenges in a large ICT firm using agile and traditional methods SANNA DAHLQVIST MILDRED SKÖLD Department of Technology Management and Economics Chalmers University of Technology Abstract The need for organizations to perform projects well has increased in today’s inter- national, demanding, and ever-changing market. To manage projects, organizations can utilize different project management methods, where recently there has been a transition from traditional to agile project management methods (APM). It has however come to light that the use of APM has not always been successful and that there are situations where traditional methods would have been more favorable. Therefore, for organizations to remain competitive, a need to apply the most favor- able method depending on the specific situation has arisen which has resulted in the coexistence of both agile and traditional methods within larger organizations, conducting projects in parallel. However, due to the different and oftentimes con- trasting principles of different methods, i.e., – agile & traditional – challenges are likely to arise within such environments. The purpose of this thesis is thus to under- stand when APM is more favorable compared to traditional methods. In addition, the thesis aims to identify potential challenges that could arise when agile and tra- ditional methods are employed by different teams within an organization. A case study at Ericsson was conducted to discover challenges that can arise due to the coexistence of agile and traditional methods. In parallel, a literature study was performed to identify factors that indicate when agile was more favorable to apply. In the study, 21 factors divided into four categories – project, team, organizational & external – were identified to influence the favorableness of APM. Furthermore, the case study identified five challenges that could arise within an organization when different methods coexist. Conjointly, these challenges pose issues within the whole organization. In its entirety, this thesis provides practitioners with an overview of factors deemed to influence when APM is favorable along with outlining potential challenges that can occur in an environment where different project management methods are used in parallel and where these reside within an organization. All in all, this thesis facilitates a more thorough understanding of when APM is favorable and what challenges organizations that wish to apply different project management methods can anticipate while providing a basis for future research. Keywords: agile, agile favorability, agile project management, challenges when meth- ods collide, coordination challenges, traditional project management. v Acknowledgements First of all, we would like to express our gratitude to everyone from Ericsson that were willing to dedicate their time to our thesis. Your knowledge, experience, and insights are the foundation of our master thesis. Without your contribution, our thesis would not have been completed. Secondly, we would like to direct a big thank you to our supervisors at Ericsson, Vandana Yadav, and Anders Hänström. Vandana, thanks for your encouragement and support in our work and for always being available to answer questions dur- ing the process. Anders, thanks for enabling a collaboration between Ericsson and Chalmers, without you our thesis would not have been possible. Also, a big thanks for your guidance in providing relevant interviewees for our thesis. Lastly, we would like to express our gratitude to our supervisor at Chalmers Uni- versity of Technology, Constantin Bremer. Thanks for your continuous interest in our progress and for taking the time to provide valuable feedback to ensure that our thesis was as good as possible. Sanna Dahlqvist & Mildred Sköld, Gothenburg, May, 2022 vii Contents List of Figures xi List of Tables xiii Terminology xv 1 Introduction 1 1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Purpose and Research Questions . . . . . . . . . . . . . . . . . . . . . 2 1.2.1 Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.2.2 Research Questions . . . . . . . . . . . . . . . . . . . . . . . . 3 1.3 Delimitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.4 Report Disposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2 Theory 5 2.1 Project Management Methods . . . . . . . . . . . . . . . . . . . . . . 5 2.1.1 Traditional Project Management . . . . . . . . . . . . . . . . 5 2.1.2 Agile Project Management . . . . . . . . . . . . . . . . . . . . 7 2.1.3 Hybrid Project Management . . . . . . . . . . . . . . . . . . . 9 2.2 Project Management Success . . . . . . . . . . . . . . . . . . . . . . . 10 2.3 Inter-team Collaboration & Coordination Challenges . . . . . . . . . 12 2.3.1 Prevalent Challenges from the Literature . . . . . . . . . . . . 12 2.3.2 Synthesis of Challenges from the Literature . . . . . . . . . . 17 3 Method 19 3.1 Research Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 3.2 Research Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 3.2.1 Empirical Context . . . . . . . . . . . . . . . . . . . . . . . . 20 3.3 Research Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 3.3.1 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . 22 3.3.2 Interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 3.3.3 Other Empirical Data . . . . . . . . . . . . . . . . . . . . . . 27 3.4 Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 3.5 Research Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 3.6 Ethical Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . 29 3.7 Method Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 ix Contents 4 Findings & Discussion 33 4.1 RQ1: Characteristics Determining when APM is Favorable . . . . . . 33 4.1.1 Project Characteristics . . . . . . . . . . . . . . . . . . . . . . 34 4.1.2 Team Characteristics . . . . . . . . . . . . . . . . . . . . . . . 39 4.1.3 Organizational Characteristics . . . . . . . . . . . . . . . . . . 43 4.1.4 External Characteristics . . . . . . . . . . . . . . . . . . . . . 46 4.1.5 Synthesis of Factors for Agile Favorability . . . . . . . . . . . 49 4.2 RQ2: Challenges due to Coexistence of APM & TPM . . . . . . . . . 51 4.2.1 ID1: Lack of Understanding of Other Methods . . . . . . . . . 52 4.2.2 ID2: Methodological Differences . . . . . . . . . . . . . . . . . 54 4.2.3 ID3: Mismatch in Organizational Characteristics . . . . . . . 57 4.2.4 ID4: Synchronization Issues & Lack of Adjustment . . . . . . 60 4.2.5 ID5: Ambiguity in Individuals’ Roles & Responsibilities . . . . 62 4.2.6 Synthesis of Challenges Identified . . . . . . . . . . . . . . . . 66 5 Conclusion 69 6 Implications 73 6.1 Practical Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 6.2 Implications for Research . . . . . . . . . . . . . . . . . . . . . . . . . 77 References 81 Appendices Appendix A – Interview Guide . . . . . . . . . . . . . . . . . . . . . . . . I x List of Figures 2.1 Illustration of the span of hybrid project management . . . . . . . . . 10 3.1 Illustration of the research process applied in this study based on the general process for qualitative research by Bell et al. (2019) . . . . . . 19 3.2 Illustration of Ericsson’s organizational structure . . . . . . . . . . . 21 3.3 Illustration of literature review process applied in this study based on the general process by Bell et al. (2019) . . . . . . . . . . . . . . . . . 23 3.4 Illustration of roles interviewed at Ericsson . . . . . . . . . . . . . . . 25 4.1 Illustration of the Cynefin Framework developed by Snowden and Boone (2007) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 xi List of Figures xii List of Tables 2.1 Traditional vs. agile project management approaches as per Dybå and Dingsøyr (2008) . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 2.2 The twelve principles behind the Agile Manifesto as outlined by Beck et al. (2001) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.3 Inter-team coordination and collaboration challenges deduced from literature and segmented into categories . . . . . . . . . . . . . . . . 17 3.1 Table of interviews conducted in this study . . . . . . . . . . . . . . 26 4.1 Factors for when agile project management is favorable from the lit- erature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 4.2 Relevant challenges identified during interviews at studied firm . . . . 52 4.3 Challenges identified at the studied firm with its respective issues . . 67 5.1 A summary of challenges identified in the studied firm . . . . . . . . 69 xiii List of Tables xiv Terminology APM – Agile project management, an iterative approach to managing projects. Cadence – Describes a sequence of events, equalized to the duration of a sprint in Agile or a team’s development cycle. GDPR – General Data Protection Regulation, a regulation in EU law aimed to protect personal information. Is applicable for all entities operating within the EU as well as entities outside who wish to conduct business in the EU. NVivo – Computer-assisted qualitative data analysis program used in the study to code gathered data. PROPS – A traditional project management method developed by Ericsson later sold to SEMCON, comparable to a Stage-Gate process. TPM – Traditional project management, a linear approach to managing projects. XLPM – A new version of PROPS, currently applied within Ericsson, comparable to a traditional Stage-Gate process. xv List of Tables xvi 1 Introduction This chapter provides a background that highlights the rationale behind the research conducted in this thesis along with the thesis’s purpose and research questions. 1.1 Background In an international and demanding market, performing projects well is a necessity for organizations to remain competitive. Simultaneously, managing projects is be- coming increasingly challenging due to today’s business environment characterized by high complexity and fast changes (Bergmann & Karwowski, 2019). To manage projects, multiple organizations have transitioned from plan-driven and linear meth- ods, collectively defined as traditional project management methods (TPM), due to their limitations in today’s ever-changing environment (Bergmann & Karwowski, 2019). As a substitute for TPM, agile project management methods (APM) has advanced as a new project management method. Since its introduction in the software indus- try, APM has gained popularity among practitioners worldwide due to its iterative nature and ability to adapt to changes throughout the course of a project (Fischer & Charef, 2021). Nowadays, the method is successfully adopted in various industries including automotive, health care, and telecommunications (Hohl et al., 2016; Paa- sivaara et al., 2018; Sindhwani et al., 2019). In fact, APM has become a necessity for many organizations due to its ability to handle the increased uncertainties in today’s unpredictable business environment (Fischer & Charef, 2021). In spite of the vast research highlighting the superiority of APM, mainly due to its ability to adapt to changes in customer requirements, it does not exist without challenges (Batra et al., 2010; Bergmann & Karwowski, 2019; Ciric et al., 2019). For instance, for APM to be favorable, a change in organizations’ culture is required which often is challenging (Thesing et al., 2021). Additionally, it has been found that the method is not always favorable (Denning, 2019) and that TPM, in certain envi- ronments, is deemed to be a more favorable project management method (Riesener et al., 2018). TPM is for instance more favorable than APM in environments where requirements are clearly defined and changes are minimal (Bianchi et al., 2020). However, the factors that influence if APM is favorable to use are currently vaguely defined, which has led researchers to call for future research to render when the 1 1. Introduction method is favored (Bianchi, Conforto, & Amaral, 2021; Bianchi, Conforto, Reben- tisch, et al., 2021; Conforto & Amaral, 2016). Factors suggested for future research include, but are not limited to team, project, and organizational characteristics such as team size, competency, and project complexity (Conforto & Amaral, 2016). Other environmental characteristics having a similar potential of influencing whether the adoption of APM is favorable are external and organizational factors (Bianchi, Conforto, Rebentisch, et al., 2021). Collectively, these characteristics make up the environment in which APM exists and thus influence whether the adoption is favor- able, making it of interest to study. Furthermore, since the importance of applying an appropriate method for managing projects has been declared by several authors (Noureddine et al., 2009; Špundak, 2014) and the favorableness of methods is dependent on the project environment, there is a need to apply both APM and TPM to manage projects within one and the same, often large and complex organization. Thus, for organizations that want to utilize the most appropriate method for their projects, different project management methods will coexist. However, due to the often different and sometimes even conflicting principles, cul- tures, or processes inherent in APM and TPM, challenges are likely to arise when these coexist (Theobald & Diebold, 2018). This has been highlighted by several authors who have outlined challenges of implementing APM in an organization that previously utilized TPM (e.g., Kasauli et al., 2020; Richter et al., 2016; Theobald & Diebold, 2018). Still, there are not that many articles published within the topic yet, which has led authors to call for future research to investigate what potential issues that can arise when APM and TPM are used in parallel within an organiza- tion (Batra et al., 2010; Theobald & Diebold, 2018). Consequently, these two areas call for future research and thus serve as the basis of this thesis that aims to provide answers to when APM is favorable to use compared to TPM and what the subsequent challenges are that might arise in large organizations that apply different methods, namely APM and TPM, to manage their projects. 1.2 Purpose and Research Questions In this section, the thesis’s purpose and aim are outlined, followed by a presentation of the research questions in connection with the relevant context. 1.2.1 Purpose The purpose of this thesis is to understand when it is favorable to use APM. In addition, the thesis aims to identify potential challenges that could arise when APM and TPM are employed by different teams within an organization. This aim is set in order to provide practitioners with guidance on when to adopt agile methods to manage projects within their organizations. 2 1. Introduction 1.2.2 Research Questions Since the wide adoption of agile in various industries and by different types of orga- nizations, there have been both successful and unsuccessful stories of its use. This has led multiple researchers to call for future research on when APM should be used and when more conventional methods are preferred instead. Thus, to fulfill the purpose of this study, one must investigate when APM is favorable to use. Research Question 1: When is it favorable to apply agile project management meth- ods? Albeit the recently revealed shift in application from traditional to more agile meth- ods, many and especially large organizations still use different project management methods. Thus, there exist large organizations having teams deploying agile and traditional methods to manage projects conducted in parallel. It is therefore of in- terest to investigate what potential challenges such organizations face. Research Question 2: What challenges can arise within an organization when agile and traditional project management methods coexist? 1.3 Delimitations In the investigation to answer the first research question, this thesis was delimited to examine the favorability of APM on a higher level of abstraction. Thus, the study does not delve into the favorability of distinct APM, such as Scrum or Extreme Programming. Furthermore, due to limited resources, the study is restricted to only investigating challenges that might arise when agile methods and traditional methods coexist within an organization. Thus, the researchers will not explore other areas such as potential benefits organizations that utlize both APM and TPM in parallel can experience. 1.4 Report Disposition The disposition of the report consists of six main chapters. The first chapter is an introduction, which provides the reader with a background of the subject, framing the purpose of the thesis and its corresponding research questions. Thereafter, a theoretical framework is outlined, which includes relevant literature from academia essential to fulfilling the purpose of the thesis. In the third chapter, the method of the study is elaborated upon including quality aspects and ethical considerations. In chapter four, the findings are presented and discussed, followed by a conclusion and implications in chapters five and six. 3 1. Introduction 4 2 Theory In this chapter, relevant literature is outlined to provide context to the study and provide guidance for the research questions. First, different project management methods are described. Secondly, prevalent inter-team collaboration and coordina- tion challenges based on previous research are provided. 2.1 Project Management Methods In order to manage projects, organizations tend to apply various techniques, proce- dures, and methods to accomplish a successful project. Together, these constitute a project management method (Gaborov et al., 2021; Špundak, 2014). For a project management method to be appropriate, it should provide guidance to the project manager throughout the process with the objective to increase the probability of the project being successful (Špundak, 2014). Nevertheless, managing projects is often not easy and the method must be able to address project-specific obstacles (Bergmann & Karwowski, 2019). Traditionally, project management mainly involved scope definition and thorough planning, before the plan was followed to achieve the established project outcome within the scope of time and budget (Nicholls et al., 2015). Nowadays, more mod- ern methods have evolved to support project management within the new business environment, characterized by its fast pace and uncertainties. One of these methods is agile project management. An overview of the differences between agile and traditional project management is provided by Dybå and Dingsøyr (2008) which can be viewed in Table 2.1. A further overview of the two project management methods is provided in the succeeding sections. Additionally, hybrid project management, which combines elements from the traditional and agile methods is described. 2.1.1 Traditional Project Management Traditional project management includes a set of various approaches that, for a long period of time, has dominated as the main project management practice (Bergmann & Karwowski, 2019). According to the authors, TPM is from the start depicted by a plan by which projects are executed sequentially with no going back to pre- vious project stages. Furthermore, the project outcomes are declared before the 5 2. Theory Table 2.1: Traditional vs. agile project management approaches as per Dybå and Dingsøyr (2008) Traditional Agile Fundamental assumptions Systems are fully specifi- able, predictable & can be built through meticulous and extensive planning High-quality, adaptive soft- ware can be developed by small teams using the princi- ples of continuous design im- provement & testing based on rapid feedback & change Control Process centric People centric Management style Command & control Leadership & collaboration Knowledge management Explicit Tacit Communication Formal Informal Customer’s role Important Critical Project cycle Guided by tasks & activities Guided by product features Development model Life cycle model (waterfall, spiral or some variation) The evolutionary-delivery model Desired organizational form/structure Mechanistic (bureaucratic with high formalization) Organic (flexible and partic- ipative encouraging coopera- tive social action) Technology No restriction Favors object-oriented tech- nology initiation of the project, including feasible objectives which will support the process (Behrens et al., 2021). The underlying thinking is that organizational mechanisms such as knowledge, and tools should be used so that desired project outcomes can be achieved (Salameh, 2014). Additionally, the method assumes all events which can affect the project are predictable (Ciric et al., 2019). Although there exist different approaches of TPM such as Waterfall and Stage-Gate, Salameh (2014) claim that it consistently comprises of the different stages initiating, planning, executing, monitoring, control, and project closure. In addition, Salameh (2014), describe that each project is executed by predetermined groups within the project team and that the output from each stage serves as the input to the next activity in line (Bergmann & Karwowski, 2019). Furthermore, each stage in TPM is performed from start to end without iterations or feedback between the stages (Fernandez & Fernandez, 2008) and when a stage is complete it is seldom revisited (Bergmann & Karwowski, 2019). Additionally, projects using TPM work with deliv- ering the final outcome. Collectively, this makes TPM especially suitable for larger projects under project conditions where the end-state is known and when none or 6 2. Theory minor project changes are anticipated (e.g., Batra et al., 2010; Boehm & Turner, 2003; Špundak, 2014), and for projects for which the company has established pro- cedures (Fernandez & Fernandez, 2008). As a consequence of TPM’s transparent structure, a shared understanding of the outcome and the process can be realized (Riesener et al., 2018). In detail, TPM can provide organizations with a stable structure for the project execution (Bergmann & Karwowski, 2019) which enables an organized division of responsibilities and roles amongst the project steps (Thesing et al., 2021). Further benefits of the method are that the project plan and corresponding resource and personnel allocation can be established beforehand which enables a smooth project execution (Fernandez & Fer- nandez, 2008; McCormick, 2012). Moreover, as project stages and team responsibil- ities are so well-defined in TPM, it facilitates a structure for handling dependencies between teams in a structured manner enabling international team collaboration (Papadopoulos, 2015). However, as a plan is established at the beginning of the project, TPM assumes that requirements are settled and as an implication, this leaves little room for changes in later stages (Bergmann & Karwowski, 2019). This is a major drawback of the method since requirements tend to change over time, as clients are not always able to declare all requirements at the beginning of the project (Ciric et al., 2019). Moreover, establishing a detailed plan in today’s ever-changing environment is difficult and according to Noteboom et al. (2021) and Špundak (2014), organizations should instead try to embrace change in order to succeed. For organizations applying TPM, the method’s lack of flexibility to changes can therefore result in costly endeavors such as increased development cost and longer time to market (Behrens et al., 2021; Fernandez & Fernandez, 2008; Špundak, 2014). Moreover, as customers are not able to further extend their requirements, the method can diminish the customer focus resulting in less customer satisfaction (Behrens et al., 2021; Fernandez & Fernandez, 2008). 2.1.2 Agile Project Management Agile project management refers to a set of diverse approaches utilized to man- age projects, such as Scrum and Extreme Programming (Papadakis & Tsironis, 2018). With roots in the software industry, APM has evolved to a commonly ap- plied method within multiple industries (Thesing et al., 2021). The nature of APM is formed on the basis of the four values of agile, presented in the Agile Manifesto (Beck et al., 2001): • To value individuals and interactions over processes and tools • To value working products over comprehensive documentation • To value customer collaboration over contract negotiation • To value responding to change over following a plan In addition to the four values presented, the Agile Manifesto includes twelve support- ing principles outlined in Table 2.2. Although the Agile Manifesto was developed 7 2. Theory with regards to software development projects, Špundak (2014) claims that the val- ues are applicable for APM as well. APM is characterized by enhanced collaboration with customers and the ability to adjust to changes. In addition, Bergmann and Karwowski (2019) emphasize the method’s incremental nature as a result of multiple planning and development cy- cles, which allow for continuous adjustment to new requirements and data, thus allowing for continuous improvement. Consequently, the method can be seen as heavily reliant on customer feedback to guide the project direction (Vinekar et al., 2006). According to Nerur et al. (2005), this allows APM to address uncertain- ties. However, McCormick (2012) highlights sufficient customer involvement as a risk of projects more easily going off track in situations where customers are faced with uncertainty and therefore can not convey what the desired project outcome is. Nevertheless, it is evident that APM is a people-oriented method to manag- ing projects (Bergmann & Karwowski, 2019) and that appropriate utilization of the method can result in multiple advantages for organizations including shortened delivery time, increased customer and stakeholder satisfaction, increased efficiency, and resource-savings (Ciric et al., 2019; Salameh, 2014; Thesing et al., 2021). Ad- ditionally, Thesing et al. (2021) highlight the advantage of identifying errors early compared to traditional methods. Although being described as a revolutionary method, some organizations using APM are facing difficulties. Ciric et al. (2019) claim that there exist organizations that are struggling to achieve agreements with stakeholders and to prioritize the needed work. Consequently, implementing APM into organizations is seen as one of the ma- jor challenges of the method. Moreover, there is consensus among multiple authors that one reason for APM implementation failures lies in the organization’s culture (Loiro et al., 2019; Papadakis & Tsironis, 2018; Thesing et al., 2021). APM requires a culture that embraces new challenges (Papadakis & Tsironis, 2018) and according to Noteboom et al. (2021) the desire for the method must be both a top-down and a bottom-up initiative. Hence, implementing APM often requires a change in culture and as well-established cultures are difficult and time-consuming to change, many organizations are struggling to implement APM (Boehm & Turner, 2003; Nerur et al., 2005; Papadopoulos, 2015). An additional reason why APM has resulted in several challenges is the remaining skepticism towards the method, which can result in a lack of management support and sufficient resources needed for the method to be prosperous (Ciric et al., 2019). According to Noteboom et al. (2021), lack of support from management is especially critical to address for the method’s success. Lastly, APM seems to be particularly challenging to implement in larger organiza- tions (Dumitriu et al., 2019; Sommer et al., 2015). According to Bergmann and Karwowski (2019), this can be due to the inflexibility and high degree of system- atization and formalization often exhibited by larger organizations. Another reason is, according to Dumitriu et al. (2019) that larger organizations often have multiple 8 2. Theory teams pursuing projects simultaneously which results in many interdependencies. Consequently, the authors declare, this can lead to challenges concerning misalign- ment and knowledge sharing between the teams. Table 2.2: The twelve principles behind the Agile Manifesto as outlined by Beck et al. (2001) Agile principles 1 Our highest priority is to satisfy the customer through early and continuous delivery of valuable software 2 Welcome changing requirements, even late in development. Agile processes harness change for the customer’s competitive advantage 3 Deliver working software frequently, from a couple of weeks to a couple of months, with a preference to the shorter timescale 4 Business people and developers must work together daily throughout the project 5 Build projects around motivated individuals. Give them the environment and support they need, and trust them to get the job done 6 The most efficient and effective method of conveying information to and within a development team is face-to-face conversation 7 Working software is the primary measure of progress 8 Agile processes promote sustainable development. The sponsors, developers, and users should be able to maintain a constant pace indefinitely 9 Continuous attention to technical excellence and good design enhances agility 10 Simplicity–the art of maximizing the amount of work not done–is essential 11 The best architectures, requirements, and designs emerge from self- organizing teams 12 At regular intervals, the team reflects on how to become more effective, then tunes and adjusts its behavior accordingly 2.1.3 Hybrid Project Management To harvest advantages from different project management methods, hybrid project management has evolved as a method. Hybrid refers to a combination of practices and tools from different methods (Bianchi, Conforto, & Amaral, 2021). Conse- quently, this implies that hybrid can refer to a mix of components from a vast amount of approaches, such as a combination of different agile approaches such as Scrum and Extreme programming (Papadakis & Tsironis, 2018). However, in the literature hybrid is often referred to as the combination of TPM and APM (Bick et al., 2017; Dumitriu et al., 2019; Vinekar et al., 2006), and it is this interpretation of hybrid that will be applied throughout this thesis. Nonetheless, hybrid should not be seen as a single method that organizations can apply, but rather as a method that incorporates elements from two approaches 9 2. Theory in a variety of combinations (Bick et al., 2017). The reason lies in the fact that there is no “one size that fits all” and the method should rather be developed for the specific situation (Bianchi, Conforto, Rebentisch, et al., 2021; Gaborov et al., 2021; Špundak, 2014). Hence, the combination of elements, whether more agile should be implemented compared to traditional or vice versa depends on the specific organization and the project environment (Bick et al., 2017). Moreover, the authors claim that hybrid should be viewed as the span between agile and traditional project management as illustrated in Figure 2.1. Hybrid project management Traditional project management Agile project management Figure 2.1: Illustration of the span of hybrid project management By utilizing hybrid, organizations can become more flexible to changes in require- ments while still having established a structure for the project process (Dumitriu et al., 2019). According to Batra et al. (2010), this is one of the major benefits of hybrid, as a lack of structure, which often APM implies, can create confusion within large and complex projects where it is often critical with stable planning and coor- dination. However, if the organization were to utilize a purely traditional approach, its flexibility to change would be limited. Therefore, balancing agility and stability is of value as it allows an organization to both adopt and optimize (Vinekar et al., 2006). Thus, an appropriate adoption of a hybrid approach can harness benefits from both methods and therefore be more favorable in comparison to APM or TPM (Bianchi, Conforto, Rebentisch, et al., 2021). However, combining elements from the two approaches that are often viewed as two opposite of extremes can be challenging (Batra et al., 2010). The rationale of the challenges lies in the different and conflicting assumptions of TPM and APM – see Table 2.1 – which needs to be combined in the hybrid approach (Batra et al., 2010; Dumitriu et al., 2019; Vinekar et al., 2006). Specifically, Dumitriu et al. (2019) highlight the challenge of establishing self-organized teams in a strict and controlling environment that characterize traditional projects. Therefore, the authors add, it is a possibility that hybrid should be viewed more as a challenge and not as the solution it is portrayed as by many authors. However, multiple authors argue that bipolar approaches can be combined and generate benefits if adopted in the right circumstance (Batra et al., 2010; Riesener et al., 2018). Furthermore, it is also argued that organizations often adopt hybrid due to the difficulties of implementing APM (Bianchi, Conforto, & Amaral, 2021; Bianchi, Conforto, Rebentisch, et al., 2021; Gemino et al., 2021). 2.2 Project Management Success When deliberating on project management success, one must first distinguish it from project success. The two concepts are strongly related to one another and in a 10 2. Theory study by Shokri-Ghasabeh and Kavoousi-Chabok (2009), 43% of the 65 respondents consisting of Ph.D. students within project management, project management pro- fessionals, and project managers, viewed the two as identical. The confusion around the concepts is not surprising as both consider the achievements of projects (Munns & Bjeirmi, 1996). This has resulted in multiple studies investigating how a distinc- tion can be made between them (Papke-Shields et al., 2010). From the research, it is detailed that project management success involves mea- surements throughout the project’s process and is evaluated at the end (Andersen et al., 2006; Cooke-Davies, 2002; Papke-Shields et al., 2010). Furthermore, multiple authors declare that the success of project management is typically measured with the criteria, time, cost, and quality, generally referred to as the iron triangle (e.g., Cooke-Davies, 2002; Pollack et al., 2018; Radujković & Sjekavica, 2017; Zid et al., 2020). In contrast, project success is viewed as a wider concept, evaluated against the project’s objectives (Gomes & Romão, 2016). Moreover, project success also incor- porates the long-term impact of the project throughout its life (Andersen et al., 2006; Munns & Bjeirmi, 1996). However, unlike project management success, there exist no general criteria of how project success is measured (Khan & Spang, 2011). Traditionally, project success was measured with the iron triangle (Bergmann & Kar- wowski, 2019) and some authors still see them as central when evaluating project success (Serrador & Pinto, 2015). However, most authors view the iron triangle as conventional and argue that project success includes many more aspects (e.g., Gomes & Romão, 2016; Irfan et al., 2019; Tam et al., 2020). For instance, Shokri-Ghasabeh and Kavoousi-Chabok (2009) and Noteboom et al. (2021) argue that project success should also include internal and external stakeholders’ satisfaction while Pollack et al. (2018) emphasize achieving strategic business objectives. Additionally, Andersen et al. (2006) highlight the importance of gaining knowledge from the project as this will aid future undertakings. The view of project success being a wider concept than project management success aligns with the proposed relationship between the concepts that project management success is a subset of project success (Andersen et al., 2006; Munns & Bjeirmi, 1996; Radujković & Sjekavica, 2017). This is reinforced by the fact that a project can be regarded as a failure although project management is deemed successful (Andersen et al., 2006). Contrariwise, a project can be successful by satisfying wider objectives, although the project exceeds its resource budget and thereby fail in regards to project management (Radujković & Sjekavica, 2017). Thus, achieving project management success can boost project success, but it is not the ultimate cause of it. 11 2. Theory 2.3 Inter-team Collaboration & Coordination Chal- lenges Within organizations, teams, regardless of method applied, often need to collaborate and coordinate with other teams in order to achieve desired outcomes. In detail, the need for coordination between teams may be due to changes in the process and interdependencies (Gustavsson, 2020; Rolland et al., 2016) that regard how activities, requirements, and resources affecting several teams are organized and orchestrated (Dingsøyr et al., 2018). Therefore, coordination is often described as the process of managing dependencies (Dingsøyr et al., 2018). Achieving inter-team collaboration effectively is important as organizations otherwise can face delays, frustrations, quality issues, and rework (Bjarnason et al., 2022). However, inter- team coordination is not effortless as it encompasses various challenges, such as lack of communication and alignment issues (Vlietland & van Vliet, 2015). Therefore, organizations need to mitigate these challenges to achieve effective and efficient inter- team collaboration and coordination (Bick et al., 2017). Furthermore, the emphasis on achieving effective inter-team collaboration is especially evident in environments where APM is used to manage projects (Bick et al., 2017; Uludag et al., 2018) as multiple self-organized agile teams often imply a great deal of dependencies between them (Bjarnason et al., 2022). In this section, the prevalent inter-team challenges identified in the literature are outlined and later synthesized. 2.3.1 Prevalent Challenges from the Literature A prominent challenge mentioned by several authors is that teams do not always know to which teams they are connected, meaning that a lack of awareness of de- pendencies exists which hinders effective collaboration (Bick et al., 2017; Stray et al., 2019; Vedal et al., 2021). This challenge is especially apparent between agile teams, but can according to Theobald and Diebold (2018) be even more complex in the interface between agile and traditional teams. Despite the project management method applied it is fundamental that teams must recognize having dependencies in order to handle them (Babinet & Ramanathan, 2008). After all, if teams lack knowledge of the activities and projects of other teams within their organization, it creates a situation of de facto not being able to manage dependencies (Stray et al., 2019). This can lead to situations where entire teams are held back from complet- ing their tasks which in turn can decrease teams’ efficiency, cause misalignment and cause several escalations for the project such as delays (Bick et al., 2017). Although known dependencies can restrict planning and be burdensome, unknown dependencies pose even greater coordination challenges for teams which increases the workload and creates delays (Evbota et al., 2016). Specifically, lack of awareness causes teams to be unaware of other teams’ activities (Bick et al., 2017) which in turn can cause misunderstandings and affect the project outcome (Rolland et al., 2016; Stray et al., 2019). In contrast, if teams instead were aware of dependencies, the coordination and timing of tasks between teams would be smoother (Cadiz et al., 1998; Espinosa et al., 2007). Moreover, it has been emphasized that recognizing 12 2. Theory dependencies at as early a stage as possible is essential to not have to re-do work from previous teams, and to reduce the integration risks with other teams (Martakis & Daneva, 2013). Nevertheless, being aware of dependencies is troublesome, and is a task that requires a thorough understanding and contextual knowledge providing further complexity to the coordination challenge caused by dependency unaware- ness (Cataldo et al., 2007). Moreover, a team’s and organizational setting’s size can further increase the issue of lack of awareness (Babinet & Ramanathan, 2008; Vedal et al., 2021). Hence, simply having an organization with a too great amount of dependencies might cause the organization and its teams to lose control over them, resulting in a lack of dependency awareness (Dikert et al., 2016). Additionally, be- ing aware of dependencies has proven to be especially challenging in large-scale agile organizations, where development is done concurrently (Kalenda et al., 2018; Stray et al., 2019) and where the scaling can cause teams to lose track of dependencies (Uludag et al., 2018). Thus, the issue is of greater importance in large organizations with multiple teams collaborating simultaneously. Another challenge that affects inter-team collaboration and coordination concerns a team’s understanding of other teams’ knowledge and corresponding capacity (Bjørn- son et al., 2018; Espinosa et al., 2007). Having reached a state of mutual under- standing of other teams, which for example involves being informed of other teams’ activities, and knowledgeable of others’ task domains is an important facilitator for efficient inter-team coordination (Espinosa et al., 2007). Not having the common ground of understanding described or if teams are not aware of each other’s work, cognitive distance in between teams can arise (Bjarnason et al., 2022; Bjørnson et al., 2018). Further, dissimilarities in knowledge regarding a system and lack of information-sharing between teams are examples of additional circumstances that can create cognitive distances (Bjarnason et al., 2022). In detail, lack of under- standing can cause teams to assume that what others are doing is complicated, which gives rise to a cognitive distance (Bjarnason et al., 2022). Besides decreas- ing the efficiency of inter-team collaboration, such distance can result in undesired outcomes and undesirable designs in development projects (Badampudi et al., 2013). Another issue included as a factor in lack of understanding concerns the knowledge of essential terminology in a project. When teams are proficient in the common terminology, this eases inter-team coordination and collaboration (Espinosa et al., 2010) which might be why cognitive distances decrease between teams that execute the same type of work (Bjarnason et al., 2022). Moreover, different understand- ings of concepts can besides misconceptions cause various interpretations of ways of working, project assignments, and working procedures. For instance, lack of under- standing can cause teams to employ frameworks differently in a company (Conboy & Carroll, 2019). As stated by a respondent in Conboy and Carroll (2019, p. 45) study: “In the absence of something clear and definitive, people just continue to do the same thing they always did”. This further signifies the inter-team challenge caused by a lack of understanding. Nonetheless, mitigating lack of understanding can be challenging as it might require 13 2. Theory teams to place additional effort to communicate more often to make sure enough knowledge is shared between them and a mutual understanding is instilled (Bjar- nason et al., 2022; Bjørnson et al., 2018). For instance, the authors state that this requires teams to be extensively aware of each other. Furthermore, teams might need to participate in other team’s meetings which is a time-consuming task but will allow them to be informed not only about the team’s undertaking but also their proceeding and prioritization in projects (Begel et al., 2009). However, if the knowl- edge is acquired, the interconnections of projects, teams, and overarching knowledge become clearer, which in turn increases the efficiency of inter-team coordination and collaboration (Damian et al., 2013). Studies also show that communication issues can hinder the coordination and col- laboration between teams. Several of these issues relate to whether teams are dis- tributed or not (Stray & Moe, 2020) as distance adds additional barriers between different teams that need to collaborate (Noll et al., 2010). For instance, the ab- sence of informal encounters to exchange information is seen as essential even though meetings in distributed settings can be scheduled in frequent intervals (Stray & Moe, 2020). Additionally, the absence of informal encounters and the distance can make it harder for team members to develop personal relationships (Noll et al., 2010), something that is often viewed as a necessity to coordinate work successfully be- tween teams (Bjarnason et al., 2022; Bjørnson et al., 2018; Evbota et al., 2016). A potential reason is the lack of face-to-face meetings, decreasing employees’ op- portunity to establish trust and form connections (Noll et al., 2010; Stray & Moe, 2020). The importance of personal relationships between employees as a means to facilitate coordination is further emphasized in a study by Begel et al. (2009), where approximately 87% of the respondents agreed that personal connections are helpful. Nonetheless, relying on personal relationships to solve the communication issues can be contradictory since these relationships are influenced by power relations and per- sonal interests which may be conflicting (Wohlrab et al., 2019). Another communication challenge, which is more likely to occur when teams are distributed globally, is delays in information (Badampudi et al., 2013; Stray & Moe, 2020). One reason is that distributed teams often communicate with each other less (Badampudi et al., 2013; Stray & Moe, 2020), while another reason can be that teams work in different time zones (Noll et al., 2010). For instance, if one team sends a request or question to another team in another time zone that arrives after working hours, this will not be answered until the next workday (Noll et al., 2010). Additionally, different time zones complicate the scheduling of meetings (Wohlrab et al., 2019), making it harder to collaborate efficiently. Thus, the collaboration between distributed teams places a larger emphasis on meetings to communicate, which can be especially challenging when teams are spread out (Stray & Moe, 2020). On the contrary, when teams are co-located in the same work environment the opportunity for communication is greater (Evbota et al., 2016). Despite this, the authors state that an open work environment, which fosters communication as teams are located in the same office space, can result in other challenges such as teams 14 2. Theory being distributed more easily. Moreover, although teams are able to communicate more easily, personal preferences of communication channels can cause inter-team challenges (Bjarnason et al., 2022). For instance, if one team prefers face-to-face communication, another team might prefer to communicate by email. Consequently, different preferences can result in a team not being aware of the interaction from another (Bjarnason et al., 2022). In close connection, Paasivaara and Lassenius (2011) explain that a challenge when teams communicate is the different perceptions among employees of how the com- munication channels should be utilized. This can cause both intra- and inter- communication issues (Paasivaara & Lassenius, 2011). Moreover, Evbota et al. (2016) state that another issue concerning communication channels is the lack of suitable ones. This aligns with the findings of Paasivaara and Lassenius (2011), which discovered that some team members express concerns regarding common meetings, claiming that these were not able to help them coordinate with other teams in a better way. Consequently, lack of suitable channels can hinder team members from sharing information with each other (Evbota et al., 2016; Nguyen- Duc et al., 2015). Another challenge identified as a communication issue is cultural differences (Bjar- nason et al., 2022; Nguyen-Duc et al., 2015; Noll et al., 2010). This issue can be a barrier and cause misunderstandings when teams communicate with each other (Nguyen-Duc et al., 2015). For instance, cultural differences can imply that the organization lacks a common language, which in turn makes communication harder (Noll et al., 2010). Additionally, cultural differences can result in challenges even if teams use a common language. As previously mentioned, cultural differences can cause a lack of understanding resulting in misinterpretations and different ways of solving problems, but it can also make it harder for teams to communicate with each other (Noll et al., 2010), hence resulting in a communication issue. Moreover, a strong culture within the team can create boundaries that decrease the efficiency of inter-team communication (Bjarnason et al., 2022). Further, the authors state that this is especially evident when APM is applied, as this method highly empha- sizes a strong intra-team collaboration culture. Finally, differences in language capability are also viewed as a communication is- sue affecting inter-team collaboration (Nguyen-Duc et al., 2015; Noll et al., 2010). For instance, this issue can make some team members more prone to use a specific communication channel (Nguyen-Duc et al., 2015). For instance, some members which have a lower level of proficiency in the common language may prefer text- based media such as email, rather than telephone communication as this will allow them more time to compose their questions and responses (Noll et al., 2010). As al- ready stated, this can result in issues due to different preferences of communication channels. Moreover, employees with stronger language skills can often seem more powerful and as a result, unintentionally hold back communication from those with fewer skills (Noll et al., 2010). This further aligns with the concern with relying on personal relations to solve communication issues. 15 2. Theory Another challenge that needs to be managed in order for teams to collaborate ef- ficiently is lack of alignment (e.g., Gustavsson, 2020; Vlietland & van Vliet, 2014; Wohlrab et al., 2019). This challenge concerns various areas where misalignment can be devastating for inter-team collaboration. For instance, for efficient inter-team collaboration to occur, teams must have a common vision and align their interests (Wohlrab et al., 2019). Alignment in these areas is even seen by authors as a critical aspect for a project to succeed (Vlietland & van Vliet, 2015). However, achieving a shared vision and aligned objectives is challenging due to the involvement of multi- ple stakeholders who need to agree with each other (Evbota et al., 2016). Furthermore, without aligned objectives between teams, it is hard to achieve align- ment in how teams prioritize (Vlietland & van Vliet, 2014) which is another area often mentioned in connection to misalignment (e.g., Babinet & Ramanathan, 2008; Bick et al., 2017; Vlietland & van Vliet, 2015). For instance, in agile teams, the prioritization of tasks is often determined by product owners that can have different objectives which can vary depending on which manager they belong to (Vlietland & van Vliet, 2015). Thus, if the objectives vary, there will likely be a misalignment in prioritization in how teams view tasks to another. Nonetheless, misalignment in prioritization can occur even if the alignment of a common objective exists between teams. For instance, if a task dependency exists between two teams, where one team is dependent on another to perform a specific task, these teams must agree on a common prioritization for efficient coordination to be possible (Babinet & Ramanathan, 2008). Still, if the teams have different views on the priority of the task this can result in increased time to market and frustration (Vlietland & van Vliet, 2014). Different views on prioritization can oc- cur when there is a lack of information, which means that teams prioritize tasks without having the full picture (Evbota et al., 2016). This is hard to mitigate, as it is impractical for teams to share all available information among employees in an organization while it is also hard for employees to create an understanding of all information that exists (Evbota et al., 2016). An additional reason for misalignment in prioritization between teams is due to the intra-team focus that often exists which means that teams often prioritize their own back-log (Badampudi et al., 2013; Vlietland & van Vliet, 2014) regardless of the rest of the organization. Moreover, misalignment can occur due to rapid changes in requirements or when requirements are unclear, which can make it hard to prioritize (Evbota et al., 2016). Additionally, dependencies can exist between several teams, which further enhances the challenge to prioritize requirements accordingly (Evbota et al., 2016). Although it is challenging, it is essential to keep the prioritization clear and updated to avoid conflicts (Paasivaara et al., 2012). Lastly, another alignment issue, not as widely discussed, but which yet can occur although objectives and prioritization between teams are clear is misalignment re- garding Definition of Done (Vlietland & van Vliet, 2014), which is a set of criteria 16 2. Theory that must be completed before a task is viewed as done (Silva et al., 2017). In a study by Vlietland and van Vliet (2014), it was discovered that teams often uti- lize different definitions. Consequently, misalignment can result in different routines and technical inconsistency between teams (Berntzen et al., 2021). The challenge of aligning the Definition of Done between teams is further highlighted by Berntzen et al. (2021) and Vlietland and van Vliet (2015). 2.3.2 Synthesis of Challenges from the Literature The prevalent challenges found in the literature have been synthesized in Table 2.3, where the respective challenge is illustrated further with a few examples. Table 2.3: Inter-team coordination and collaboration challenges deduced from literature and segmented into categories Challenges Examples Illustrating the Challenge Lack of Dependency Awareness Example 1: Teams do not always know to which teams they are connected which hinder effective collaboration (Bick et al., 2017; Stray et al., 2019; Vedal et al., 2021) Example 2: Being aware of dependencies is a task that requires a thorough understanding and contextual knowledge (Cataldo et al., 2007) Example 3: Organizations with a too great amount of dependencies might cause the organization and its teams to lose control over them (Dikert et al., 2016) Example 4: Not having the appropriate awareness of dependencies when dealing with complex projects can cause misunderstandings and thereby affect the project outcome (Rolland et al., 2016) Lack of Understanding Example 1: Lack of understanding can cause teams to assume that what others are doing is complex, giving rise to a cognitive distance which decrease the efficiency of inter-team collaboration (Bjarnason et al., 2022) Example 2: Besides misconceptions, different understandings of con- cepts can cause various interpretations of ways of working, project assignments, and working procedures (Conboy & Carroll, 2019) Example 3: Gaining knowledge of other teams is a time-consuming task (Begel et al., 2009) Communication Issues Example 1: Lack of informal encounters is a barrier for distributed teams as it hinders trustful relationships and often leads to less inter- team communication (Noll et al., 2010; Stray & Moe, 2020) Example 2: Cultural differences can cause misunderstandings and create boundaries between teams, hindering efficient collaboration (Nguyen-Duc et al., 2015; Noll et al., 2010) Table Continues on Next Page 17 2. Theory Continuation of Table 2.3 Challenges Examples Illustrating the Challenge Example 3: Personal preference of communication means, differ- ent levels of language proficiency, and lack of suitable communication channels hinders efficient inter-team communication (Bjarnason et al., 2022; Evbota et al., 2016; Noll et al., 2010) Misalignment Issues Example 1: Misaligned objectives and lack of information can cause teams to prioritize tasks differently, consequently resulting in frustra- tion and delays (Evbota et al., 2016; Vlietland & van Vliet, 2014) Example 2: Intra-team focus causes teams to prioritize their own back-log (Badampudi et al., 2013; Vlietland & van Vliet, 2014) re- gardless of the rest of the organization Example 3: Different definitions of Definition of Done can cause differences in routines and technical inconsistency (Berntzen et al., 2021; Vlietland & van Vliet, 2014) 18 3 Method This chapter introduces the research strategy, design, and methods used in the thesis and the reasoning behind them. In addition, the research quality and ethical considerations of the thesis are discussed. 3.1 Research Strategy According to Bell et al. (2019), a research strategy accounts for the approach by which a study is conducted. Generally, there are two main strategies associated with different research orientations, namely quantitative and qualitative (Bell et al., 2019). A quantitative research strategy, typically entailing a deductive relationship linking theory and reality, is a strategy aimed to test theories. Contrariwise, quali- tative research often takes an inductive approach to theory (Bell et al., 2019) and is especially suitable for research that will collect and analyze words to generate new theories (Kaplan & Maxwell, 2005), which aligns with the intention of this thesis. In more detail, the rationale behind the choice of a qualitative strategy was the thesis’s aim to, through the research questions posed, generate theory that provided answers to when agile is more favorable and to identify potential challenges that could arise as a result of utilizing different project management methods within an organization. Hence, a qualitative study with an inductive approach was appropri- ate since new theory was the outcome of the research, generated from collecting and analyzing non-numerical data as described by Bell et al. (2019). Additionally, a qualitative research strategy is described as flexible and iterative (Kaplan & Maxwell, 2005). This further strengthens the choice of a qualitative strategy, as the process followed in this thesis is not sequential but rather iterative based on the general process described by Bell et al. (2019) – see Figure 3.1. Formulation of research questions Selection of relevant sites & subjects Collection of relevant data Analysis & interpretation of data Write up of findings & conclusions Identification of gaps Figure 3.1: Illustration of the research process applied in this study based on the general process for qualitative research by Bell et al. (2019) 19 3. Method 3.2 Research Design A research design serves the function of guiding the data collection and thus the chosen research methods by providing a structure for it (Bell et al., 2019). By con- sidering the purpose of the study and the research questions, a literature study was chosen to answer the first research question, when agile is favorable to use. Hence, an answer to the question was derived by primarily analyzing several published ar- ticles on the topic. In parallel, a case study was chosen as it was deemed the most appropriate research design to answer the second research question, to identify challenges that could arise within an organization as a result of the usage of APM and TPM. In detail, a case study implies that a single case of interest for the research, such as a unit or organization, forms the basis of the study and is therefore appropriate when an organization is thoroughly studied (Gerring, 2004). Further influencing the choice of a case study for the second research question were practical considerations, as the thesis emerged from an exchange with the Telephone Stock Company of LM Ericsson, henceforth referred to as Ericsson. This implied that the study relied on Ericsson for resources, specifically the opportunity for data collection. Hence, Ericsson is the investigated case and thus formed the basis of the study. 3.2.1 Empirical Context This study took place at Ericsson which is a large firm in the information and communications technology (ICT) industry. Since its foundation in 1876, Ericsson has grown into a leading ICT provider with annual revenue of 232.4 billion SEK in 2020 and about 100 thousand employees globally, 14 thousand in Sweden (Ericsson, 2021). To fulfill their purpose, to create connections that make the unimaginable possible, Ericsson yearly pursues various types of projects to deliver high-technology products, software, and services to the market such as roll-outs of cellular towers and Internet of Things (IoT) solutions (Ericsson, 2021). Further, Ericsson has a matrix-like organizational structure that includes four business areas, five market areas, and numerous group functions (Ericsson, 2021). The organizational structure is illustrated in Figure 3.2. In short, the four business areas are responsible for pro- viding Ericsson with business solutions that can compete in the market (Ericsson, 2021). In turn, these solutions are to be sold and delivered by the five market areas. Lastly, the group functions provide support to the market and business areas by coordinating work, driving synergies, and the agenda of Ericsson. 20 3. Method CEO Business Areas Media Group Functions Market Areas Networks Digital Services Managed Services Technologies & New Businesses North America Europe & Latin America Middle East & Africa South East Asia, Oceania & India North East Asia Red Bee Media Customers Figure 3.2: Illustration of Ericsson’s organizational structure To manage its projects, Ericsson has long used traditional approaches such as PROPS, developed by the firm itself, as well as XLPM where both methods can be described as a traditional Stage-Gate process. Lately, Ericsson has implemented cross-functional teams that consist of employees with varying functional expertise, such as employees from differing organizations within the firm, for instance as Group IT and Group Supply. The teams were created to help develop Ericsson’s digital capabilities by addressing specific parts of the customer journey from when an order is placed to final delivery and after-sales services. More specifically, the teams were part of and pioneers in Ericsson’s initiative to shift towards agile across parts of the firm to keep up with today’s new business environment. Despite the adoption of agile to manage projects, there still are teams within Erics- son that pursue projects with a more traditional approach. Therefore, Ericsson can be viewed as a firm that applies different project management approaches depending on the specific situation, thus not having a standard project management method across the firm. Ericsson was viewed as an appropriate firm for this thesis as the purpose and aim was to understand when agile project management is more favor- able compared to a more traditional approach, and what potential challenges can arise when different project management methods coexist within an organization. In detail, several organizations within Ericsson make up the foundation of the case in this study. The reason why the case is not limited to one or two units is that the division of teams applying APM and TPM differs within Ericsson’s organizations. For instance, APM is the primary method for the cross-functional teams and within the organization Group Digital Transformation Office (DTO), where work to digi- tally transform processes is conducted, while TPM is more commonly used within the organization Group Supply, primarily responsible for planning and orchestration of the supply chain. Therefore, representatives from several organizations are in- cluded as the researchers deemed this would increase the reliability of the results. However, due to the limitations of the study in resources and time, as well as the size of Ericsson, not all areas can be covered. Moreover, as the study emerged from an exchange with the organization Group Supply at Ericsson, that is where the main 21 3. Method focus is due to practical considerations. Additionally, the focus is primarily on or- ganizations affiliated with Ericsson’s business areas, since these are responsible for developing products and solutions. Nonetheless, representatives from market areas are also included to cover the collaboration between these areas, i.e. – market areas and business areas. 3.3 Research Methods In this section, the research methods used in the thesis – literature review & empir- ical data collection methods – and the rationale behind the choices are described. The thesis was initiated with a literature review, for the researchers to gain an under- standing of the research area. Thereafter, the case study was initiated by collecting empirical data at the studied firm. In parallel, the literature study, consisting of an extensive literature review was performed in order to provide an answer to the first research question, when it is favorable to apply APM. 3.3.1 Literature Review A literature review refers to the process of gathering and synthesizing research avail- able and can be conducted for several reasons and in various ways (Bell et al., 2019; Snyder, 2019). In this thesis, the purpose of a literature review was to identify relevant data for the literature study aimed at answering the first research question. Additionally, it was used in the case study to establish an understanding of the re- search field and to facilitate the analysis of the gathered empirical data, supporting the answering of the second research question. Within the qualitative realm, narrative reviews are typically used due to their spe- cific suitability for qualitative research (Bell et al., 2019). Furthermore, it is regarded that such a review is custom for research areas that have been studied in various con- texts (Snyder, 2019), which aligns with the purposes of the literature review for this thesis. Moreover, a narrative review is advisable for studies with fewer restrictions regarding the boundaries of the review and thus allow for greater flexibility as these can be modified throughout the process (Bell et al., 2019). Therefore, this thesis used a narrative review where findings from the empirical data collection methods provided guidance to new relevant areas in both studies, i.e., – the literature study & the case study. To identify relevant literature, this thesis followed the process illustrated in Figure 3.3 which is based on Bell et al. (2019)’s general process. Thus, the literature re- view was initiated with articles recommended by the supervisor at Chalmers who possessed academic knowledge in the specific field. Hereafter, keywords relevant to the research questions were generated which allowed for further literature to be identified through Google Scholar and the databases available through Chalmers Li- brary. The literature search was continuously revised and narrowed as the literature and empirical findings provided more direction. Additionally, to further facilitate the identification of relevant literature, snowball sampling, which is a method that 22 3. Method through the reference list of already identified literature helps researchers discover further literature (Bell et al., 2019), was used. Generate keywords relevant to RQs Conduct search – using google scholar & databased available through Chalmers Library Read abstracts & retrieve relevant literature Keep notes using NVivo Write up of literature review Read recommend articles Keep notes using NVivo Use insights from literature & interviews Use snowballing Figure 3.3: Illustration of literature review process applied in this study based on the general process by Bell et al. (2019) 3.3.2 Interviews In qualitative research, there are various ways to collect empirical data, where one of the most common data collection methods is interviews (Bell et al., 2019). Inter- views are advantageous to use compared to other methods such as observations, as they allow a specific focus while simultaneously ensuring that the study has a wide coverage by interviewing diversified roles (Bell et al., 2019). For this reason, inter- views were chosen as the primary method to collect empirical data from the studied firm in the case study. Additionally, the interviews were conducted online due to practical implications, such as the fact that the researchers were located in another city and that a large part of the firm still works from home due to COVID-19. Thus, online interviews were considered the most appropriate way to collect data during the given circumstances. Normally, interviews in qualitative research are less structured to allow for more flex- ibility and to ensure essential thoughts and perspectives from the interviewees are captured (Bell et al., 2019; Jenner et al., 2004). Typically, qualitative interviews are divided into unstructured and semi-structured (Bell et al., 2019), where the latter is used in this case study. The rationale behind this decision is that semi-structured interviews, compared to unstructured, allow for some comparability between inter- views when they, at a later stage, are being analyzed while simultaneously ensuring relevant topics are touched upon in the interviews (Bell et al., 2019). Lastly, semi- structured interviews are preferable when the interview is conducted by more than one person (Bell et al., 2019), which was the case in this thesis, as one researcher was responsible for taking notes while the other researcher led the interview. Furthermore, when using semi-structured interviews it is recommended to have an interview guide that includes the main themes that the interview aims to capture Bell et al. (2019). Additionally, the authors suggest that the guide follows a few baselines such as developing the guide from the study’s research questions, ensuring that there is an order in the line of questioning, and a balance in the language. 23 3. Method Moreover, a guide should not include any leading questions and the question asked about the interviewee should be relevant (Bell et al., 2019). With these considera- tions in mind, an interview guide was developed – see Appendix A – and used during all conducted interviews. Additionally, it is of importance to consider the process of how interviewees are se- lected. To initiate the sampling process, the supervisor at the studied firm proposed a few potential interviewees. Thereafter, the study used purposive sampling, which is a non-probability sampling technique, typically used within qualitative research, where the selection is done with the aim of the research in mind (Bell et al., 2019). Purposive sampling is further regarded to increase the representativeness amongst interviewees chosen and can also help display eventual heterogeneity within groups (Maxwell, 2012), which strengthened the choice of a purposive sampling technique as various roles within the studied firm were desired to gain a more honest depiction. In detail, purposive sampling can be divided into various forms where primarily the- oretical sampling and snowball sampling was used in this study. Theoretical sam- pling can further be described as an iterative process where data is collected based on previous sampling (Bell et al., 2019) which aligns with this thesis’ aim, where interviewees were deliberately chosen based on their appropriateness. Additionally, theoretical sampling entails theoretical saturation, meaning that researchers are able to evaluate throughout the data collection process whether further data is needed (Bell et al., 2019). This further emphasized the choice of the sampling method, as the researchers concluded the interview process when they felt that data saturation was achieved, meaning that gathered data from following interviews did not indi- cate any new information critical to answering the research questions. Additionally, snowball sampling influenced the sampling process as it allowed potential intervie- wees to be identified from already conducted interviews (Bell et al., 2019). Noteworthily, the sampling process was also influenced by convenience sampling, which is a technique where interviewees are chosen based on their availability (Bell et al., 2019). This practice was applied due to the study’s time limitation, meaning that only employees who were available during the allotted time for the interview process were interviewed. At the end of the process, 26 semi-structured interviews with a length of one hour each had been conducted. The interviews were conducted using the video conferenc- ing tool Microsoft Teams, which allowed the interviews to be performed in a, for the interviewees, familiar setting as suggested by Bell et al. (2019). Moreover, all inter- views for which consent was given by the interviewee were audio-recorded to ensure details were captured in their entirety. Worth mentioning is that a majority of the 26 interviews were held in Swedish as this is the native and preferred language of most interviewees. In Table 3.1 all conducted interviewees are listed with their respective roles and the organization within Ericsson they belong to. However, to mitigate the risk of interviewees being identified, any detailed role titles have been changed to general role titles. For instance, the role of a Manager is applied to describe 24 3. Method both Line Managers, Product Managers, and Process Managers. Additionally, only the overarching organization is represented in the table for the same confidentiality reasons. To distinguish interviewees who belong to the same overarching function, but different sub-organizations, a segment name has been added after the name of the organization to visualize this, i.e., – segment A, B, C. For an additional overview of the areas and roles covered in the studied firm during the interviews see Figure 3.4. From the Figure, it can be depicted that two out of four business areas, namely Networks and Managed Services, and two out of the five market areas, namely Europe & Latin America and North America, are represented in the study. Additionally, the Figure shows that six organizations in total, some with affiliated sub-organization, jointly measured to nine, are represented in the study. Noteworthy is that although an organization is affiliated with a specific market or business area, the scope of the organization can imply that its business extends to other areas. For instance, Group Supply formally belongs to the business area Networks. However, the group function is responsible for supply chain planning and orchestration all the way from new product introduction through manufacturing and distribution to finally preparing their customers’ supply chains across all of Ericsson’s business and market areas. Hence, although not all Ericsson’s business and market areas are formally covered, due to the limitation in time and scope of this study and the large size of Ericsson, the researchers have done their best to cover a wide range of areas and some of their associated sub-organizations to provide a foundation for the analysis. Group Digital Transformation Office Business Area Market Area Managed Services North AmericaEurope & Latin America Group Information Technology (IT) Group Supply Development Unit Networks Networks Network & MS Delivery Group Sales #12 #4 #1#1 #1 #3 #2 #1 #1 Senior Manager (2) Senior Manager ManagerManager (5) Manager Manager Manager (2) Manager Product Owner (2) Product Owner (2) Product Owner Change Lead (2) Agile Coach Agile Coach Business Analyst Business Analyst Enterprise Solution Architect Figure 3.4: Illustration of roles interviewed at Ericsson Additionally, interviews have been performed with multiple roles, three Senior Man- agers, 11 Managers, four Product Owners, three Change Leads, two Agile Coaches, two Business Analysts, and one Enterprise Solution Architect. Noteworthy is that although several interviewees are classified as Senior Managers and Managers, a majority of these have experiences from previous roles such as a Product Owner or a Team Member, as they have been at Ericsson for a long time. Their long ex- perience was also valuable as they could provide a perspective on what it was like before Ericsson adopted APM, and provide a comparison between then and now. Therefore, the division of roles interviewed can initially be considered unbalanced, 25 3. Method although the researchers have deliberately chosen to interview employees with long experience, as a limited number of interviews could only be conducted due to the study’s limitation in time and resources. Another thing worth mentioning again is the fact that the interviewees’ role titles have been adjusted to general roles, where many interviewees were given the title of Managers as they had responsibility for others, compared to for instance Business Analysts that are primarily viewed as team members. Consequently, the range of areas and roles combined – see Figure 3.4 – provide a substantial foundation for the analysis conducted in this study. Table 3.1: Table of interviews conducted in this study Role Affiliation Date 1 Manager Group Supply – Segment A 2022-02-15 2 Product Owner Group Supply – Segment A 2022-02-16 3 Enterprise Solution Architect Group IT – Segment A 2022-02-16 4 Change Lead Group Supply – Segment A 2022-02-17 5 Product Owner Group Supply – Segment B 2022-02-18 6 Manager Group Supply – Segment A 2022-02-21 7 Senior Manager Group Supply – Segment A 2022-02-22 8 Manager Group IT – Segment B 2022-02-22 9 Business Analyst Group IT – Segment C 2022-02-23 10 Product Owner Group Sales 2022-02-24 11 Senior Manager Group Supply – Segment A 2022-02-24 12 Manager Group IT – Segment C 2022-02-24 13 Manager Group Supply – Segment A 2022-02-24 14 Senior Manager Network & MS Delivery 2022-02-25 15 Product Owner Group Supply – Segment B 2022-02-25 16 Manager Group Supply – Segment A 2022-02-28 17 Manager Group Supply – Segment A 2022-03-02 18 Agile Coach Group DTO 1 2022-03-02 19 Business Analyst Group Supply – Segment A 2022-03-03 20 Manager Development Unit Networks 2022-03-03 21 Manager Group IT – Segment C 2022-03-04 22 Manager Group DTO 1 2022-03-07 23 Change Lead Group Supply – Segment A 2022-03-07 24 Agile Coach Group Supply – Segment B 2022-03-08 25 Manager Group Supply – Segment B 2022-03-14 26 Manager Group Supply – Segment A 2022-03-15 1: Digital Transformation Office 26 3. Method 3.3.3 Other Empirical Data To facilitate the understanding of the contextual situation, internal documents pro- vided by Ericsson regarding the firm were utilized. These documents include how the firm and its organizations deploys agile and traditional ways of working, and how Ericsson is structured, which provided the researchers with an understanding of the context of the case prior to the interviews. Additionally, documents regarding organizational structure were used to provide an understanding of which part of the organization the interviewees belongs to. This was used both to map already conducted interviewees and to find new potential participants which can potentially provide another perspective. 3.4 Data Analysis Analysis of qualitative data primarily involves the identification of categories and concepts in data, which often can be laborious due to its commonly unstructured nature (Bell et al., 2019; Vaismoradi et al., 2016). Although no general rules regard- ing qualitative data analysis exist, many researchers within the qualitative realm deploy coding as an approach for analyzing data (Vaismoradi et al., 2016). In this thesis, coding was deployed in parallel with the data collection to facilitate both the collection and the process of analyzing data (Bell et al., 2019). To analyze the data gathered from the case study, interview notes were initially scrutinized by both researchers to ensure the entirety of the interview was captured. Specifically, both researchers read through the notes and later discussed them to- gether to assure all details from the interview were captured in their entirety and with the specific formulation. In case of ambiguity, the audio recordings were re- played. Additionally, by scrutinizing the notes, the researchers increased their close- ness to the data, which enhanced their ability to make sense of it (Vaismoradi et al., 2016). This is of importance as the contribution to academia from qualitative re- search greatly depends on researchers’ interpretation (Bell et al., 2019). As a next step, open coding was conducted, which is a process where data is broken down into codes before they are grouped into concepts relevant to the study’s purpose (Bell et al., 2019). In detail, these concepts emerged from continuous reflection upon, comparison of, and analysis of codes which allowed them to be sorted and labeled into clusters of codes of similar meanings, known as concepts. Thereafter, a different level of coding was conducted with the aim to determine hierarchical connections and relationships between concepts to generate categories and sub-categories (Bell et al., 2019). Lastly, the codes that were in Swedish were translated into English. This was done after the coding process, so that the meaning of codes which are extracts of interviewees’ formulations, would not be affected by a translation that could not be completely equated with the original. To facilitate the coding process, NVivo was used. NVivo is a computer-assisted qualitative data analysis program especially useful when large datasets are being an- alyzed (Bell et al., 2019). NVivo allows for an easier process, as the software makes 27 3. Method it easier to systematically create and track codes which enhances the transparency of the process and consequently the research’s quality (Welsh, 2002). Additionally, literature identified in the literature review was utilized to facilitate the process as it provided guidance to the researchers during the analyzing phase. For instance, by using the knowledge concerning the research domain, gained from the literature review, the researchers could more easily identify relevant concepts and categories. For the analysis of the literature in the literature study, a similar approach was performed, where articles were coded in NVivo and later grouped into concepts relevant to the study. However, the coding did not have to be conducted on as many abstraction levels as the data gathered from interviews aimed to answer the second research question. Hence, a more facile coding process could be performed to generate theory to provide an answer to the first research question compared to the second one. 3.5 Research Quality To assess the quality of research it is relevant to use different evaluation criteria depending on whether the study is of quantitative or qualitative nature (Bell et al., 2019). The reasons for this, the authors explain, lie in the different objectives and approaches of the strategies. To evaluate the quality of this thesis, the criterion of trustworthiness is applied. According to Bell et al. (2019), trustworthiness is a commonly adopted criterion within qualitative research than can be split into four criteria: • Credibility – concerns whether the findings of the study are believable • Transferability – concerns whether the findings can be applied in similar con- texts • Dependability – concerns whether the findings are persistent within the context • Confirmability – concerns the degree to which the researchers’ values have in- fluenced the study To strengthen the thesis’s credibility, several actions have been taken. During the literature review, various databases were used to gather articles from numerous journals to increase the number of data sources, which according to Yilmaz (2013) enhances the credibility. Furthermore, the interviews conducted at the studied firm were performed with different roles and from different areas to allow various ac- counts of aspects – see Table 3.1 – something that can further enhance credibility (Bell et al., 2019). Moreover, as all but one of the interviewees provided consent to record the interviews, the researchers could ensure that the vast majority of all data was captured, thereby strengthening the study’s credibility. Moreover, the credi- bility of the study was strengthened with the use of respondent validation, which means that participants confirm the accuracy of the findings by providing com- ments on selected parts of the thesis (Bell et al., 2019). For instance, all presented quotes and their use in section 4 have been validated by the respective interviewees. This was deemed especially important for quotes that were translated from Swedish. 28 3. Method Transferability is due to the uniqueness of qualitative case studies, agreed upon by multiple authors to invariably be a key issue (Bell et al., 2019). To strengthen this criterion, a thick description of the case was developed as recommended by Bell et al. (2019) and Yilmaz (2013). In addition, the literature review aimed to answer the first research question, is of a more general nature, which therefore enhances the transferability of this thesis. Dependability is further described as a criterion that assesses the consistency of the process with regard to time and replicability by other researchers (Yilmaz, 2013). However, the emphasis on replicability is typically lower for qualitative than for quantitative research as it is regarded as unachievable due to the many procedures that can not be standardized (Jenner et al., 2004). To allow for replicability, the procedures applied have been documented. In addition, the case study’s replicabil- ity can be strengthened by publishing the used interview guide (Bell et al., 2019) which is another measure taken in this study to enhance its quality – see Appendix A. Lastly, as NVivo increased the transparency of the analysis process, this also strengthened the replicability as per Bell et al. (2019). Confirmability is always a concern in qualitative research since it is conducted on the basis of researchers’ interpretations (Bell et al., 2019). Hence, it is impossi- ble to guarantee that no personal values on inclinations are included (Bell et al., 2019). However, even if confirmability cannot be fully achieved, the level at which the internal interpretations can influence the research can be limited. To enhance confirmability, the emerging analyzes and conclusions in the thesis were the result of both researchers. In addition, the thesis’s results were continuously examined by supervisors from both the case firm and Chalmers. Although these measures cannot guarantee that no interpretations have influenced the outcome of the thesis, they are deemed to have mitigated the risks. 3.6 Ethical Considerations Although achieving good research ethics can be bothersome, it is of essence to pay attention to the various ethical considerations that can arise when conducting research (Bell et al., 2019). To evaluate the research ethics in this thesis, the four ethical principles outlined by Bell et al. (2019) will be used. • Harm to participants – concerns the avoidance of both physical and physio- logical harm • Informed consent – concerns participants being well informed before making a decision • Invasion of privacy – concerns regarding the privacy of participants • Deception – concerns whether the intention of the study is portrayed differ- ently from reality These ethical considerations can arise in several stages throughout the process and need to be managed (Brinkmann & Kvale, 2018). 29 3. Method Harm to participants can arise as a result of having put a research participant in a situation that it regards as stressful or that has the potential of having negative effects on its profession or occupation (Bell et al., 2019). To mitigate the risk of participants being harmed, Brinkmann and Kvale (2018) suggest that the interview should be constructed so that the communication between researcher and intervie- wee does not create feelings of anxiety, unease, or hurt the interviewee’s self-esteem. When developing the interview guide, these considerations were taken into consid- eration as a means of reducing the risk of psychological harm – see Appendix A. Participants were also informed of the possibility to at any time opt-out or decline to answer specific questions without reasoning, as recommended by Bell et al. (2019). Moreover, the interviews and the identities of the interviewees are kept confidential and anonymous. Furthermore, as a means to increase anonymity, generic names of roles were utilized – see Table 3.1. However, complete anonymity of the participants was deemed to not always be possible to achieve. Therefore, participants were in- formed of this before the interview was initiated. The issue of informed consent involves providing interviewees with sufficient infor- mation about the study as a whole to facilitate a fair decision-making process for participants choosing whether to participate (Bell et al., 2019). To ensure informed consent was obtained, the invitation sent out to potential participants included a description of why the researchers requested to conduct interviews with the studied firm together with the thesis’s purpose and aim. In addition, the participants that had accepted the invitation were informed again, prior to the initiation of the in- terview, and were given the opportunity to ask questions. In this way, the authors ensured that the participants were well informed as recommended by Bell et al. (2019). Invasion of privacy of participants is closely linked to informed consent (Bell et al., 2019), which implies that this issue can be mitigated with similar procedures. In addition, the interview guide was developed so that questions, only relevant to the study, are included and thus no unnecessarily private questions are asked. Moreover, to further mitigate the issue, the researchers treated interviewees with respect and made sure not to interrupt or ask inappropriate follow-up questions. The issue of deception relates to being dishonest about the research and what it entails (Bell et al., 2019). Furthermore, the authors describe that deception can arise when the study is misrepresented or when participants lack information or are misinformed. In this study, there is no indication of deception as all informa- tion conveyed about the study was truthful and aligned with reality. Moreover, as already declared, information about the study was provided repetitively to ensure participants were not misinformed or lacked information. In addition to the presented ethical principles, other ethical and legal considerations were deemed necessary to take into consideration. One such consideration managed in this thesis is data protection and management which includes both legal and ethical aspects (Bell et al., 2019). The aspects of data protection were handled by 30 3. Method complying with GDPR in conjunction with adhering to policies provided by the studied firm. For instance, the recordings were stored safely and deleted after the completion of the study, which interviewees also were informed about. Lastly, ethical considerations can arise with the use of copyright materials (Bell et al., 2019). In this thesis, the issue of copyright has been managed by complying with Chalmers’s policies. 3.7 Method Discussion The research methods employed in this research were chosen to fulfill the thesis’s purpose and aim by providing answers to the research questions. The methods were selected in the interest of available resources, specifically time and number of re- searchers, and by taking the scope of the thesis into consideration. Consequently, with more resources available for conducting the thesis, the methods chosen could have been adjusted accordingly. Firstly, a greater number of interviews could have been conducted if more time had been available for the study. Increasing the num- ber of interviews could have given better coverage across the entire studied firm. Additionally, by having time for more interviews, the researchers could have inter- viewed a larger variety of roles. Lastly, if more time had been available, the length of the interviews could have been increased to enhance the details and concepts captured. Now, the interviews were limited to one hour due to the busy schedules of the interviewees. Therefore it was impractical for the researchers to pursue all leads in-depth, which impacted the possibility to gain more empirical data to support the answer to the second research question. Collectively, more resources available for the interview process could have increased the comprehensiveness of the study. Furthermore, if it had not been for the remote set-up of the thesis, on-sight inter- views could have been conducted. One of the downsides of not conducting face- to-face interviews is that it might impact the connection between the researchers and the interviewee, and the researchers therefore can miss potential data that can be obtained from the interviewee’s body language and voice (Opdenakker, 2006). However, as video conferencing tools such as Microsoft Teams have become more common, it has lately not been regarded to be of great concern among qualitative researchers (Bell et al., 2019). As working from home, having Microsoft Teams as the major communication platform, has been the new normal for employees at the studied firm since two years back, the interviewees were considered to be comfort- able using the tool, further lessening the concern. In addition to face-to-face interviews, had the set-up not been remote and more time had been available, ethnographic observations could have been conducted. Ob- servations could have been used to strengthen the answer to the second research question, concerning the identification of challenges that can arise when different methods coexist. Observations are advantageous since they, in addition to facilitat- ing researchers’ understanding of the contextual situation from the perspective of the interviewees, provide a more objective perspective by excluding the subjectivity from interviewees (Bell et al., 2019). Therefore, if more time had been given and 31 3. Method if the circumstances were not affected by COVID-19 making observations infeasible at the studied firm, it could have been used to triangulate the data gathered from interviews. Instead, interviews had to constitute the main data source for the second research question, where the researchers interviewed various employees from diverse roles and from different parts of the firm, to establish a substantial foundation for the study’s analysis. Lastly, the literature review provided a basis for understanding, facilitated the anal- ysis of empirical data, and provided answers to the thesis’s first research question. To discover when APM is more favorable to use, a narrative review was conducted. However, if more resources were provided, a systematic literature review could have been conducted instead. A systematic literature review enhances the quality of the literature study, as such a review is less likely to be affected by the researchers’ prej- udice since it follows well-defined steps with clear criteria (Tranfield et al., 2003). Despite this, such a review is generally very time-consuming and therefore was not considered appropriate due to the existing limitations of this study. 32 4 Findings & Discussion In this chapter, answers to the research questions are presented. In the first section, findings from the literature study are analyzed and discussed in order to provide an answer to the first research question, when it is favorable to use APM. In the second section, the findings identified in the case study are discussed and analyzed in relation to the literature, to provide an answer to the second research question, concerning what challenges can arise when APM and TPM coexist in an organiza- tion. 4.1 RQ1: Characteristics Determining when APM is Favorable Within the literature, multiple authors testify of the importance of applying the right method when managing projects, as this otherwise can result in avoidable challenges and impact the project success negatively (Špundak, 2014). Moreover, the choice of an appropriate project management method should also facilitate an effective project and be tailored to its needs (Noureddine et al., 2009). Nonetheless, identifying the most appropriate method can be challenging given the numerous project and environmental characteristics that influence the method se- lection (Bianchi, Conforto, Rebentisch, et al., 2021). Additionally, there is still con- fusion regarding which characteristics mostly influence the favorability of a project management approach. According to Gaborov et al. (2021), the selection must be based on the stakeholders’ demands and inherent risks, size, complexity, and cost of a project. In contrast, Boehm and Turner (2003) claim that the most important characteristics are concerning culture, the project’s size, criticality, and degree of dynamism, as well as the competence of the team and its size. Vinekar et al. (2006) acknowledge these characteristics and add the importance of the customers’ cul- ture. Špundak (2014) agrees with many of the authors and adds the importance of requirements’ flexibility, stakeholder location, and customer availability while Fer- nandez and Fernandez (2008) adds that the quality of the project and the desired process transparency can also impact the selection. Thus, there exists no general agreement among authors on which factor