Key Attributes of a successful measurement program!
Software Quality Center LLC is a reputed partner of the CMMI Institute. We have been using CMMI® to help elevate performance for over 15 years and have seen the value of the models to deliver measurable results for our Clients/Organizations. We look forward to continuing to work alongside the CMMI Institute to extend the reach of the CMMI® frameworks to enable individuals and organizations to reach their goals.
We at SQC, have had many significant articles published in our newsletters over the past 15 years, these articles written by our Consultants and clients. We are pleased to share few of them with you, over the course of next few months, which you will enjoy reading and enjoy as a take-away for your everyday meetings and knowledge. It is also a great way to reinforce concepts of quality in your organization's staff in everyday meetings!
K. Ganesh Shenoy (KGS) - Senior Manager & Head of Quality, EmSyS, Larsen & Toubro Limited, Mysore. L& T Technology Service (Earlier known as Larsen & Toubro Limited, EmSyS) is one of our esteemed client and we associated with various business units of L&T from 2000 as a partner for their CMMI based process improvements journey. KGS was the Mentor for Process improvements program and our consultants learnt lot from him.
It will be our privilege and honor to carry a repeat of articles published in our 2004 newsletter written by KGS
Key Attributes of a successful measurement program
EmSyS is a business unit of Larsen & Toubro Limited located in Mysore. It is a design house providing software and hardware solutions to its global clients in the area of Embedded Systems & Software. EmSyS is one of the early adopters of CMMI where all the CMMI model components have been implemented i.e., Software engineering, Systems engineering, Integrated product & process development and Supplier sourcing.
Measurements form the backbone of process improvement in EmSyS. Based on our experience, without a strong metrics foundation, decisions are based on the perception of the practitioners and hence deployment of resources for improvement may not be very effective. Sometimes, these perceptions are heavily skewed by the instances in the recent past.
Interestingly, on one hand, metrics is a very useful & a powerful tool & on the other, it can be a double edged sword if precautions are not taken while defining metrics, ensuring the integrity of the collected data, considering the context of data & its analysis.
Success in a measurement program depends on how well the measures are linked to the organizational objectives/ information needs.
The key here is to identify the measures, which are critical to success, & not to identify measures just because there is a need to implement the measurements program. One of the best-known methods that could be used to define the metrics is adopting "Goal-Question- Metrics" method to arrive at the suitable measures for the organization considering the performance needs of the organization in each of the processes/ sub-processes. Setting the measures that help the organization in tracking its performance, taking preventive actions in advance & triggering the improvement activities, is very crucial here.
Defining the measure explicitly is very critical for the success of a Metrics program. For example, when we define productivity in Lines of code / day (LOC/day) it is important to define what does LOC include? Does it include commented lines? Does it include reusable modules? In case of maintenance projects - does it include added / deleted / modified LOC? Does it include autogenerated code? Without such clarity the interpretations as well as the outcome of analysis may be erroneous.
Effective usage of the metrics is possible only when the practitioners are familiar with metrics. Again a mere familiarity is not enough. The users should be aware of the importance & benefits of metrics usage. They should also be aware of the definitions, inclusion / exclusion criteria used for each of the measures. Nurturing such an environment is very crucial for the success of the metrics program.
Unless these definitions are explicit, there can be large gaps in understanding the measurements. This in turn can mislead the improvement initiatives & can upset the confidence of the practitioners in the measurement program. Finally, the interpretation of the metrics can be misleading if the contextual information is not taken into consideration.
From the point of view of sustaining metrics initiative in the long term, if any of the current measures are not contributing to further improvement either due to changed business conditions or due to changed priorities of the organizations or due to lack of controllable parameters then it is important replace them with more useful measures, facilitating the process of triggering performance enhancement initiatives in line with the organizational business objectives.