Lessons From CMMI (A Software Process Model) For BI Practitioners
Hexaware successfully completed the CMMI Level 5 re-certification recently with KPMG auditing and certifying the company’s software process to be in line with Version 1.2 of the model. This is the highest level in the Capability Maturity Model Integration model developed by Software Engineering Institute in collaboration with Carnegie Mellon. For the uninitiated, Capability Maturity Model Integration (CMMI) is a process improvement approach that provides organizations with essential elements of effective process.
Now, what has CMMI got to do with Business Intelligence?
I participated in the re-certification audit as one of the project managers and I learnt some lessons which I think would be useful for all of us as BI practitioners. The CMMI model has 22 different process areas covering close to 420 odd specific practices. Though the specifics are daunting, the ultimate goal of the model is simple to understand and there-in lies our lesson.
In the CMMI model, Maturity Levels 2 and 3 act as building blocks in creating the process infrastructure to ensure that the higher maturity levels are achievable and sustainable.The high-maturity practices (Levels 4 and 5) of the model focus on:
1) Establish Quantitative Goals in line with the business objectives
2) Measure the performance with respect to the goals using statistical tools
3) Take corrective action to bring the performance in line with the goals
4) Measure again to ensure that the action taken has contributed positively to performance improvement.
Key Lessons for BI Practitioners:
1) Single-minded focus to “close the loop” – CMMI model evaluates every project management action in the context of project goals and measures them quantitatively. Business Intelligence, ideally, should measure all actions in the context of business goals and provide the facility to compare metrics before and after the decision implementation.
2) Strong information infrastructure – Higher levels of maturity in CMMI are sustainable only if the lower maturity levels are strongly established. In the BI context, this translates to a robust architecture that makes measurements possible
3) Accuracy + Precision is the key – Controlling variation (sustainability) is as important as hitting your targets. BI in organizations is weak along the sustainability dimension. For instance, enterprises do have analytics around “How am I doing now” but not much on questions like a) How long will this growth continue? b) When will we get out of this declining trend? etc.
In a way, this post is related to one of my earlier blog on BI and Sig Sigma with the central idea being that, for enterprises to be analytics driven both numbers and processes behind those numbers are equally important. CMMI Model, in its simplest form, also has that as its core theme for achieving high process maturity in an organization.
Thanks for reading and please do share your thoughts.