In a CMMI ML 4 process, organization comes up with Process Performance Baselines (PPBs), define Business Objectives (BO) and Process Performance Objectives (PPO), derive Process Performance Models (PPMs). Project uses these data and information to compose the project quantitatively. Project is managed quantitatively and statistically using PPMs.
In CMMI ML4 organization targets can be same as that of current baselines. There has to be a quantitative linkage between BO and PPO. A PPM is not at all necessary to substantiate the relationships between BO and PPO, simple mathematical expressions will do. For example, BO can be a compound growth percent of PPO itself for ‘n’ number of years.
PPOs could be measured at the time of project closure only. Then how the probability of success could be monitored intermittently ..? In order to monitor this we need to define some prediction models (PPMs) with dependent parameters as PPO and the parameters which affect PPOs as independent parameters.
These can be summarised as,
Now what happens if probability of success to get an improved targets are observed instead of Probability of success to get current performance… Definitely, probability should be less than 50 %, since with current processes, targets can not be achieved. This means that current process needs to be improved to achieve PPOs. Here comes the role of CMMI ML 5 .