This study was primarily motivated by practical observations and reports from different software development organizations across multiple European coun-tries, which suggested that knowledge sharing does not work well for a wide variety of reasons that were not previously well described. Some previous re-search efforts focused on motivation to share knowledge or reporting knowledge sharing practices used in software development organizations. Given the men-tioned reports from practitioners, this study did not try to repeat the efforts of mapping practices and motivational factors, meaning “What works and why?”. In-stead, it looked at the matter from a different angle by asking: “Why knowledge sharing currently does not work and what prevents it from working well?”. The col-lected data and results indicate that this approach is something that has been missing and that has the potential to bring useful knowledge for both practition-ers and researchpractition-ers. As a pilot study, it will hopefully encourage more research-ers to focus on this topic for example by constructing applicable solutions and recommendations.
The results suggest that practitioners might be well aware of the importance and benefits of sharing knowledge with their colleagues, hence, possible expec-tation that the lack of knowledge sharing is caused by not seeing the importance or benefits might not be correct in all settings. Researchers should be aware that simply reminding the importance might not have a significant effect. However, the importance of knowledge sharing in software development teams might still need to be emphasized and explained to customers´ representatives without IT background, who might sometimes obstruct knowledge sharing efforts and prac-tices.
There are currently many practices for sharing knowledge. It can be argued that reporting them over and over or introducing new ones might be of a low value to practitioners unless what prevents the practices from being used is ad-dressed first. Discouraging factors should be considered while designing new
efforts and practices to support or encourage knowledge sharing. Potentially use-ful information for future researchers can be that the data suggested a demand especially for simple, fast, and easy-to-implement practices for effectively shar-ing one´s knowledge.
6 CONCLUSIONS
The main objective of this thesis originated from the surprising discrepancy be-tween the theoretically declared importance of knowledge sharing in software development teams and the lack of action and attention paid to it in practice. The primary goal of the thesis was to attempt to identify what is the reason for this discrepancy and how the gap between what is declared and how the reality looks like could be reduced or even eliminated. More specifically, the focus was placed on identifying the discouraging factors, which prevent or discourage software development professionals from effectively sharing their knowledge with their team colleagues.
The prior literature provided the base for the whole research effort with the primary focus being on knowledge, its coordination, and what are the known challenges and practices of managing and sharing knowledge in software devel-opment teams. Additionally, the attention was more closely paid to the distrib-uted software development, which is currently a common working arrangement in this global industry, and where different types of distances between colleagues make sharing knowledge much more challenging.
The empirical research aimed at understanding the phenomenon more in detail and did not pursue acquiring generalizable results, therefore, the interpre-tive case study method was selected, and qualitainterpre-tive empirical data were col-lected during semi-structured interviews. Study participants came from three different teams in one software development organization and held multiple dif-ferent roles that are typical in the field (software developer, application manage-ment specialist, infrastructure specialist, manager). Owing to this role coverage, the richness of the empirical data increased.
The results of the thesis confirm that there truly is a gap between how im-portant knowledge sharing is considered and how this importance can or cannot be seen in reality. A set of discouraging factors in knowledge sharing was iden-tified. Some of them did not seem to play a significant role in the case environ-ments, but most of them were considered to be quite strong. It is believed that the significance of individual factors may vary in other environments, hence, all the factors were reported.
The discouraging factors that were found to have little effect in the case en-vironments were: Importance, Motivation and benefits, and Willingness. Partici-pants claimed that, in their opinion, knowledge sharing is very important in the field of software development. Furthermore, they were able to reasonably ex-plain their motivation for sharing knowledge and describe its benefits for indi-viduals and the whole team. There was no strong evidence that practitioners would be unwilling to share their knowledge; however, some interesting reasons why individuals could be less proactive were discovered and described.
Based on the collected data, the most significant factors seem to be Per-ceived difficulty, Lack of attention, Insufficient or incorrect encouragement, and Missing systematic approach. Sharing knowledge was considered to be difficult
especially when there were no recommended ways how to do it and the expected level of quality of documentation was not agreed upon. Too often practitioners seemed to focus only on the primary work (writing code, solving service requests, etc.) and forget or ignore knowledge sharing needs, because those were consid-ered a side activity. Encouragement of sharing knowledge in the teams appeared to be insufficient, incorrectly targeted, or expressed in way too abstract state-ments. When not enough attention is paid to knowledge sharing, there might not be any process, model, or even simple team agreement constructed and all (if any) knowledge sharing activities and efforts are only ad-hoc. That might be a missed opportunity to learn from the past and reduce the influence of the other discour-aging factors.
The main contribution of this thesis can be seen in identifying and describ-ing the discouragdescrib-ing factors that hinder or prevent knowledge shardescrib-ing in soft-ware development teams. Only once these factors are known and understood, it is possible to address them. Some suggestions for limiting the effects of these fac-tors and for increasing the quantity and quality of knowledge sharing were also introduced. Other researchers can benefit from understanding the current needs of practitioners and challenges that they face in this area to better target their future research efforts.