Trouble Shooting For Productivity

Bruel & Kjaer configuration management plan has established the following goals for the SPI effort:

1. Better control the growth in software development expenses.
2. Standardize the development of software PARA3. Increase predictability in software development with regards to e.g. calendar time, resources, functionality, and quality.

The basic strategy of the SPI project in Bruel & Kjaer is to focus on the project managers. It is believed that gaining their support is vitally important for improvements to occur. The project was initiated in March 2007.

The SPI group soon realized that it lacked insight and information about software processes and problems. Relevant information was available from the PET and PRIDE projects, but these projects addressed only parts of the software process and were, therefore, not a suitable foundation for the formulation of an overall SPI strategy. Only the maturity profile and recommendations from the

Bootstrap assessments were available, but the SPI group found this insufficient as a foundation for planning and implementing specific improvements. The SPI group opted not to repeat the Bootstrap assessment as it feared that a new assessment, only six months after the previous one, would do more harm than good and destroy whatever was left of the interest in SPI.

At the same time, it became clear that the SPI initiative had to be based on the active cooperation and enthusiasm of the project managers and that they had to be involved in identifying problems and suggesting solutions. The SPI group therefore decided to develop and apply the diagnosing technique described here. This meant that the SPI group decided to:

•    conduct interviews with the seven project managers about their views of strengths and weaknesses in Bruel & Kjaer's software process,
•    Suggest improvement initiatives based on this information, and hence PARA engage the project managers in the implementation of initiatives.

The SPI group decided to use an interview guide approach where "topics and issues to be covered are specified in advance, in outline form, [but where] the interviewer decides sequence and wording of questions in the course of the interview". This approach ensures that the interviews cover important issues and topics, but it allows the interview to be matched to the topics and problems relevant to the particular project manager.

The interview guide approach puts a high demand on the interviewer's ability to phrase and rephrase questions on the fly and to follow-up on interesting themes without spending too much time on less relevant topics. The SPI group tried to reduce these risks by having more than one interviewer at each interview. All interviews were tape-recorded to support the writing of summaries.