Adoption of PSP practices is not as straight forward as obtaining training for the engineers that will use the PSP. By "adoption" we mean cultural acceptance and use of the technology. When careful attention is paid to the process of technology insertion, organizations are successful in adopting PSP.

Successful adoption of PSP has two different but supportive perspectives: the organization and the individual. The objectives and thus the motivation of each party are different. The barriers to success and thus the adoption plans are also different. Organizations are motivated to attain their business goals, whether they are decreased time-to-market, higher quality products, better ability to meet schedule, or increased and more predictable profitability. At this stage of PSP maturity (late 1997), individuals interested in the PSP are motivated by the desire to do better work and to personally improve. PSP maturity has not reached the stage where fear is a motivational factor.

For PSP to be successful on the organizational level, the organization must point to PSP as a key enabler for the attainment of specific business goals. These goals are organizational specific. Example goals that have been used by diverse organizations for PSP adoption include decreased time-to-market, higher quality products, better ability to meet schedule, increased and more predictable profitability, and staff understanding of SPI (CMM level IV and V) principles.

Successful adoption of PSP at the organization level occurs when the organization's engineers use PSP in their day-to-day practices, when project management uses PSP approaches (e.g., statistical process control, roll-up of planning estimates) to run projects, and when senior management provides support for PSP. Any adoption approach that does not address all these issues will probably not succeed.

Our advice to organizations that are planning PSP adoption is to apply the principles of technology adoption to the PSP technology. We recommend a multifaceted approach to organizational PSP adoption. Build and maintain senior management commitment. Provide PSP training, mentoring, and support to both project management and engineering staff. Apply the new PSP knowledge to several PSP pilot projects where results are measured. Finally, carefully analyze the pilot results with the goal of making a rollout decision. Successful PSP adoptions have all addressed these issues whether by accident or by intention.

Organization PSP adoption projects have failed in the past for a number of reasons. When training is provided only to engineers, there is no mechanism to constructively use the engineer's PSP data to help achieve project success. Furthermore, there is no mechanism to address issues arising from project management's misuse of engineer's PSP data.

When senior level commitment is not obtained and maintained the PSP effort is often dropped before it can be validated. Other technology adoption experience indicates that even if the PSP pilots did finish that senior management will not find the results as believable as they do when they are involved from the beginning. When mentoring is not provided to engineers after PSP training they revert to old (non-PSP) practices. Senior management must provide the work environment where engineers can use PSP. This is a requirement (discussed below) for an individual engineer to personally adopt PSP.

When pilot projects are not begun immediately after training, then the effectiveness of the training is lost. PSP efforts with single pilot efforts are exposed to a large risk related to the project and not PSP. PSP pilots have not completed because of external cancellation of the project (unrelated to project performance). There is also a large risk, during piloting, that backup PSP trained personnel will be unavailable should there be a personnel change. Finally, one data point from one pilot is should not be enough to make a technology rollout decision.

Successful adoption of PSP at the individual level occurs if after the PSP instruction the individual has internalized the PSP lessons and can demonstrate PSP practice in their day-to-day work. For software development work this implies they are planning their work based on data from past experiences, practicing early defect removal rigorously, and searching for process improvement opportunities. Furthermore, they are using PSP continuous process improvement principles to guide their non-software development work.

A number of strategies for the individual adoption of PSP have been tried. These have included the original CMU one semester class and its academic semester long descendents, the SEI's industrial version of the CMU class (either as a two week intense format or as one day a week for about 15 weeks), self study, and courseware independently developed from Humphrey's book. Only the academic and industrial strategies have shown repeated success. Self study and independently developed courses have not worked. The requirements for success seem to be sufficient student time, substantial student effort, access to mentor/instructor knowledgeable and experienced in PSP, and a work environment that supports PSP.

When the class is made so easy that no effort other than showing up is required then the students are not forced to see the value of PSP approaches and they never begin to practice PSP. The effort, difficulty, and the student's focus must be at least on par with that found in their work or study environment. PSP is hard. There is no way around it. It involves the student adopting a work approach of continuous improvement while getting the job done versus just getting the job done. This change in focus will not happen overnight.

PSP training requires about 130 hours of student time. Shorter courses have not worked. The entire approach of collecting data, making process changes based on that data, and then measuring the effects of those changes takes considerable time. The student must invest this time to see the benefit of discipline.

Mentoring is absolutely essential to success. The mentoring helps in understanding difficult concepts, in providing constructive feedback, and in motivating the need for change. During the PSP learning period, it is typically the instructor who acts as the mentor. After the learning period, when PSP is in daily use, it is important that mentors be available. In organizations where PSP has been institutionalized, these mentors are peers and immediate supervisors. In organizations where PSP is not institutionalized, the mentors are often the instructors acting as consultants.

Finally, the student's work environment is critical to successful adoption of the PSP. During the learning PSP learning phase, the work environment must support the learning of PSP. This means that for the 130 hours needed to internalize PSP practices, PSP is the student's work assignment. After the PSP learning phase, the work environment must support PSP practices. Management and coworkers must have an expectation that PSP planning practices, defect removal practices, and process improvement will be performed.

Our advice to individuals wanting to learn PSP is fourfold. One: get the support of your supervisor. Two: make sure there will be others in your work environment using PSP when you finish your training. Three: identify an instructor who not only understands PSP but is committed to its precepts. Four: prepare for the hard work of changing the way you work.