1.2 Emphasize Knowledge Creation
In essence, improvement is knowledge creation. SPI is driven by knowledge about practices and perceived needs, insights gained during the improvement process, software industry standards, and state-of-the-art methodologies and tools. SPI efforts also depend on the implicit, individual knowledge of participants. However, the general idea is to make knowledge explicit and to share knowledge.
To create useful knowledge, you must be observant and systematic. Some knowledge will, of course, remain tacit in individual skills and organizational capabilities. You should, however, make an effort to learn from practice, to make the implicit explicit, and to build widely shared knowledge about software development and SPI. Also, your SPI group should understand knowledge itself in a broad sense, ranging from experience to general, established theory. Finally, your SPI effort's knowledge creation process should be cyclical in nature, as Figure 1.2 shows.
Figure 1.2 SPI should emphasize knowledge creation
Your SPI group's knowledge creation process must be deliberately designed and nurtured. We recommend that you pay particular attention to the following questions:
- How do you capture and evaluate your experiences?
- How do you combine them with other experiences and with your underlying theories?
- How is your thinking influenced by knowledge from outside the organization?
- What is the quality of your knowledge?
- How does your knowledge feed back into the SPI efforts?
Examples from Practice
As part of its SPI initiative, Danske Data established the Project Management Competence Center (see Chapter 5). The PMC Center's purpose was to be a meeting point where project managers could exchange ideas and solutions and discuss problems, issues, and challenges. Thus, through the PMC Center, Danske Data could better organize and use its project-management knowledge and bring in outside information to inspire project activities. However, the main motivation for establishing the PMC Center was to harness competence already present in the organization, much of which resided with a few knowledgeable and powerful project managers. PMC Center activities would help explicate this knowledge and make it available to all project managers. Danske Data also developed a project-manager training program in which experienced project managers led many of the key sessions.
Systematic was slower to realize the importance of the knowledge creation principle in its SPI effort (see Chapter 4). Several years ago, Systematic established a quality management system and was quickly certified as compliant with the ISO 9000 standards. When Systematic started its SPI effort, Systematic managers believed that the organization could reach CMM level 2 and even level 3 just as easily. It took a while for the SPI group to realize that this was not the case. Gradually, they concluded that the new change process was quite different. To succeed, common software processes had to be found, described, and institutionalized, and that would require a tremendous amount of work. This understanding emerged slowly as old knowledge was forced to give way to new experience. The company's experience in launching the ISO certification effort required a lot of process description and some management pressure. However, the knowledge explicated in the descriptions was at that time largely shared. With the SPI effort, processes were not in place in advance and the knowledge did not even exist. Thus, new knowledge had to be introduced and shared across the organization.
Two factors are key to facilitating a knowledge creation approach to SPI: systematic evaluation and state-of-the-art knowledge. When you systematically evaluate software practices ("diagnose" in the IDEAL model) and specific SPI initiatives ("learn" in the IDEAL model), you create a foundation for learning from experience. This learning in turn can lead to increased understanding of what it will take to improve your organization's software practices.
You can also import state-of-the-art theories and techniques from outside the organization. Such external knowledge can provoke your organization to change and can introduce what industry leaders consider common knowledge. The purpose is to transcend your organization's existing software practice when internal knowledge building is insufficient (for more on this, see Chapter 12).
Knowledge creation is undermined by myths and the "not-invented-here" syndrome. Building local knowledge is important, particularly when it is done publicly. Half-baked stories of the successes and failures of various past efforts can, however, reinforce myths and ruin good initiatives. It is therefore important that you create knowledge openly and that conclusions are tested in public. Otherwise, all knowledge will be equally important, there will be no sense of knowledge quality, and, ultimately, all knowledge will be equally meaningless and bound to remain private or localized in subcultures.
Most software practitioners are proud of their practices and results, and they often invent new approaches to deal with challenges they face. All this is positive, but if they value only in-house solutionsand thus the not-invented-here syndrome dominatespractitioners considerably reduce their ability to learn from state-of-the-art theories and techniques and thus limit improvement possibilities.
The value of knowledge creation is clearly expressed in the SPI literature. Humphrey (1989) argues that SPI initiatives must be guided by two types of knowledge: normative models (understanding the map) and systematic assessments (understanding the landscape). Assessments are particularly important here because they provide insight into current software practices; for examples see McFeeley (1996) and Chapters 7, 8, 9, and 11.
Checkland distinguishes between the perceived world and the ideas and concepts relevant to appreciating it (Checkland and Scholes 1990). People create the perceived world through interpretations based on ideas and concepts, and their experiences of the perceived world in turn yield new ideas and concepts. According to Checkland, intellectual work is not simply a matter of making sense of the perceived world; sense-making is a purposeful activity that brings experience and theory together in a framework. Individuals conceptualize frameworks and use them to support reflection and thinking, something that leads to action.
Nonaka and Takeuchi (1995) focus on the nature of human and organizational knowledge and explain how knowledge is created in daily organizational life. Their primary distinction is between tacit knowledge and explicit knowledge. Based on this distinction, they discuss four fundamental knowledge creation processes:
Socialization occurs when one person's tacit knowledge is directly adopted by others as tacit knowledge.
Externalization occurs when an organization explicates tacit knowledge as concepts and models.
Combination occurs when an organization brings together different sources of explicit knowledge to create new forms of explicit knowledge.
Internalization occurs when individuals adopt explicit knowledge in practice, thereby making it part of their tacit knowledge.
To create and manage knowledge successfully in your SPI efforts, you must ensure that all these processes are working at the individual and organizational level (see Chapter 14).