Data collection and analysis

The interviews conducted were semi-structured. Interviews enable the persons' to reflect on their answers and provide explanations more than for example in surveys.

Participants can build their answers. Semi-structured interviews allow to find causal relationships between variables and probe the reasons behind the actions, decisions, and opinions that participants have taken. In critical realism, this is important to try to understand the underlying reasons for actions. A semi-structured interview also enables to lead the discussion to areas that have not been considered yet but are relevant.

(Saunders et al. 2016) This method suits the explorative nature of the study and enables to collect data for framework building and answering the research questions. The time reserved for the interviews was selected to be 60 minutes. This was selected due to the scope of the topic in the interview. This time limit was estimated to allow more in-depth conversation and creates a more pressure-free environment for the interviewees' when there is no hurry. (Saunders et al. 2016)

The structure and interview questions (Appendix E) were based on the findings from the literature review. The literature review was completed prior to the interviews to get a good understanding of the subject and to be able to better select appropriate questions and themes. Three themes were selected for the interview based on the findings and research questions: general FE process, the PM's tasks, and the role and the effect of the FE on project performance. The end of the interview consisted of overall level questions to collect development ideas. In addition, the case company's operation manual which describes the way to operate was studied prior to the interviews. This helped to get a good understanding of the designed process and responsibilities thus shaping the interviews. The structure of the interview is found in Appendix E.

The invitations were sent by e-mail and teams to the interviewees with a short description of the purpose and background. Interviewees from the sales organization answered the invitation within a couple of hours on average and all were excited to take part. The PM's answer to the invitation took on average a week and required reminders. The interviews were conducted in face-to-face meetings or Teams meetings, face-to-face being the preferred one. Before each interview, a short description of the purpose of the study and the themes in the interview were sent to each participant. In addition, the participants were asked to select one completed project and gather the information for that project to smoothen the interview event. Each participant was asked if the interview could be recorded. It was made clear that the recording or the interviews as such would not be published anywhere and would be used only for transcribing and research purposes. It was also mentioned that in the final report the data would be anonymized so that the interviewees would speak as freely as possible. These mentions help to reduce interviewee or response bias (Saunders et al. 2016). During the interview brief notes were taken of the key points. Taking notes during the interview can help to maintain concentration on the subject. The notes can also act as a backup if the recording does not work. (Saunders et al. 2016)

16 interviewees were selected from the case company from which 13 interviews were realized. The requirements were that they had comprehensive experience and had conducted different types of projects and attended FEs. It was decided that PMs, SMs, and their managers would be interviewed. PMs were selected as the study directly affects their work and they have experience and knowledge of the effects to the execution side. SMs on the other hand are the ones in charge of the FE phase, so it is natural to get their viewpoint on how the FE should proceed and what are the responsibilities for each party. This was considered to produce enough different viewpoints to give reliable results. The details for the interviews can be found in Table 3.

Table 3: Interview information.

Interview No/

interviewee No

Title Location Duration

min:sec

1 / I1 Project manager Meeting room 60:39

2 / I2 Project manager Meeting room 55:21

3 / I3 Manager, projects Teams 50:39

4 / I4 Senior sales manager Meeting room 73:21

5 / I4 Senior sales manager Teams 60:20

6 / I6 Senior sales manager Teams 54.23

7 / I7 Senior sales manager Teams 70:58

8 / I8 and I9 Director, Sales and Senior sales manager

Teams 58:14

9 / I10 Director, Sales Meeting room 49:04

10 / I11 Project manager Meeting room 43:59

11 / I12 Director, projects Teams 34:08

12 / I13 Project manager Meeting room 59:20

13 / I14 Project director Teams 65:01

It was assumed that selected interviewees had the best understanding and insight into the subject. This selection is also in line to the literature reviews project management and sales and marketing literature division. In total 13 interviews were conducted, in one of which two people attended at once. The managers for these parties were taken, to get an overall and more strategic take on the issue. In total 7 persons from project execution and 7 persons from the sales organization were interviewed. The average duration for the interview was 56 minutes and 54 seconds. The interviews were linked to 9 projects listed in Table 4.

After each interview, the transcribing was done as soon as possible. Transcribing and making a full record of the interview as soon as possible helps to maintain detail and quality to the data. The longer transcribing takes, often the more detail is lost (Saunders et al. 2016). Contextual data was also included in the interview notes.

During the interviews, the interviewees were asked to discuss about a project. Ten interviewees represented nine projects (Table 4). All of the interviewees did not represent projects since they had not been that involved with specific projects for a long time, or they had not attended both sales and execution phases of a project, or they worked at managerial level and had more overall views on subjects. The interviewees were

recommended to select the most recent project in which the execution phase had already started and preferably also finished. This was done to get data on how the project had performed in the execution. Selecting recent projects also ensured that the project is fresh in the interviewees' minds. The projects were supposed to be from normal-sized to large and represent typical projects to the company.

Table 4: The example projects.

performance also varied from poor to excellent. Some of the projects were ongoing during the interviews. All in all, it was considered that these cases represent the typical projects in the case company. In addition to data from interviews, the case company's project management software was gone through to see the performance and financial results of the projects to get the most accurate data.

A thorough analysis was done on the data in Excel after all interviews were conducted.

The interview themes and questions were placed on rows and interviewees on columns.

Some additional rows were also added to get a more granular look into, for example, different tasks in the FE. Themes and categories for the analysis were selected based on the results of the literature review, theoretical framework, and the planned structure of the analysis. The data was picked from the transcribes to the table if it was considered to be related to the theme. The idea was to find reoccurring themes, differing opinions, and trends in the answers. The objective was to find answers to the research questions.

After the categorization in Excel, the analysis could be done quite easily since the data was well structured and unnecessary data was left out. The analysis structure was formed based on the research questions and also to be in line with the literature review structure.

The division between project management and sales organization was considered to be valuable since they had very different roles and views in the FE. Thus, in the analysis, it was shown from which viewpoint each comment came from. Also, the theory was divided this way. To visualize the answers the framework for the FE tasks was used. If the interviewee mentioned the tasks being part of the FE it was marked in Table 5. A similar approach was used in studying the PM's role in the FE in Table 6. To analyze the impact of the FE on the project the data from projects was collected. First in the interviews and then this was backed by data from the management systems. Again, putting this data into Table 4 shortly describing the project, FE, execution, and the performance of the project. This way the project data could quite easily be used to analyze the effect of the FE on the overall level. Conclusions and observations were made from the tables based on the number of the mentions and also the type of the mention. If the interviewee considered the aspect as a strength or important these were highlighted. In addition, the weaknesses that interviewees brought up were introduced. Observations, development ideas, weaknesses, et cetera, were picked based on how often they reoccurred in the interviews, which parties stated these, and how valid they were considered to be based on the interviewee's explanation and other evidence. If something occurred rarely it was also considered to be valuable to examine why it was stated rarely. Also, opposite

opinions for claims were tried to be found to cover all the aspects and be as objective as possible.

The findings were verified in a workshop. In addition, a separate meeting with a manager was arranged to further validate the findings. All the interviewees were invited to attend the meeting in which they could comment on the findings. Any material or preparation instructions were not given prior to the workshop. The workshop further validates that correct conclusions and interpretations have been drawn. The workshop was arranged as a Teams meeting and lasted 1 hour and 50 minutes. The workshop began with an introduction to the subject with the goals and objectives of the study. After that, the results were presented starting with the findings on the impact of the FE continuing to the FE tasks and ending with the PM's role. Lastly, improvement ideas were represented. The development ideas were prioritized in the workshop based on the opinions and comments. In addition, the ideas were evaluated based on the required effort and estimated impact. During the session, the discussion was encouraged and opinions were asked from the attendees about the represented findings.

In the discussion chapter, empirical findings are compared to the theoretical findings. For this, the framework from the literature review is used. The goal was to find what matches and what is different between these two approaches and try to find overall answers to the research questions. Each was critically analyzed in the context of research questions.

Concrete development suggestions for the FE of the case organization were given. Two requirements had to be fulfilled for suggestions. First of all the suggestion has to be supported by the empirical study and verified to be of value in the workshop. Secondly, the theory must have some support for the development idea. Especially If many people considered a thing as a weakness someone gave a development idea for it and it was supported the idea was considered valuable. In the last chapter conclusions with limitations, applicability, and theoretical contribution were given.