How effective these changes are and how much thought has gone into these changes is anyone's guess.
The key point, however, is that the CMMI V2.0 is being projected as a bettered version of the CMMI suite, which again is open to debate.
CMMI V2.0 or, for that matter, any other model or framework are essentially benchmarking tools and are as good as the business value they create for an organization.
CMMI V2.0 has laid strong emphasis on business objectives and business performance, and rightly so.
So, changes and improvements that can further improve and deepen the alignment of CMMI V2.0 with business priorities would make it even better.
This post dwells upon two such aspects that can help in doing that.
CMMI Model Online Portal
The first aspect is related to the CMMI model online portal.
One of the key things in CMMI V2.0 is that the CMMI model is available as an online portal which is probably getting updated very frequently.
This is being cited as a strength.
Unfortunately, that view is not logically sound.
Any model or framework serves as a reference and can be used to benchmark its usage in a specific situation against the accepted base which is the model or the framework itself.
Models can change and do change but the changes should get introduced in a thoughtful manner and only at periodic intervals separated by a reasonable time gap.
The new version of a model should come out only as per the above principle unless there is some problem or "bug" in a particular version of that model, making an immediate "rectification" release necessary.
Frequent changes lead to a situation of trying to hit a moving duck, which is never a good idea as far as benchmarking is concerned.
Interestingly, one of the types of CMMI appraisal is now called as "benchmarking" appraisal.
So, in a way, the above is apparently a crucial structural challenge with CMMI V2.0 appraisal.
An organization starts its CMMI V2.0 implementation with CMMI model around a certain time duration.
When it is time for appraisal, let us say after 2 years, the extant copy of CMMI model might be different.
Incidentally, the CMMI model portal has the provision to download a pdf copy so it is possible to store a copy of CMMI model, as it was, at a given point in time.
The question that is but natural to be asked is:
"Which copy of CMMI model should be used for appraisal - the one from two years back or the one which is there now?
In case the changes introduced are significant, then ensuring all of them are adequately incorporated will be a logistics and coordination challenge for the organization.
But in case the changes are not significant, then this is not an issue per se.
That, then, brings up another point.
If the changes are not significant, why to change frequently?
Would it not be better to accumulate the change requests and improvements from actual experiences shared by the users and other sources and bring a new version?
And do that as a strategic event and not as a routine procedure.
The earlier arrangement, where a model version used to get released at a certain time interval was a better one for the model users and implementers.
The online portal would still be useful for the model developers and model maintainers as they can use it as a work in progress copy of the next version of the model.
Change requests and improvements can get constantly incorporated by the model developers and model maintainers.
And just like code-freeze in case of software development, there could be a content-freeze in the model development cycle, where the next version is baselined and formally released.
CMMI Appraisal Sampling
The second aspect is related to CMMI appraisal sampling.
The sampling now is done using a method that is called as a randomly generated sample (RGS).
In this case, any project can ger selected.
In CMMI V2.0 there is heavy emphasis on business objectives and business performance.
Given that, sampling should also reflect the same line of thinking.
What should happen ideally is that those projects which are critical to business should be a part of the appraisal.
The factors that typically govern “critical to business” are:
In case the above factors are not being considered in sampling, then the final sample would not be aligned with business priorities.
But in case the above factors are already being considered in sampling, then this is not an issue per se.
And the final sample would still be the same.
That, then, brings up another point.
If the final sample would still be the same in both cases, why to have this new method?
Would it not be better to let the sample be aligned with business priorities and in fact, bring in, or further improve the mechanism to ensure projects that get selected are critical to business?
CMMI Model Online Portal
The first aspect is related to the CMMI model online portal.
One of the key things in CMMI V2.0 is that the CMMI model is available as an online portal which is probably getting updated very frequently.
This is being cited as a strength.
Unfortunately, that view is not logically sound.
Any model or framework serves as a reference and can be used to benchmark its usage in a specific situation against the accepted base which is the model or the framework itself.
Models can change and do change but the changes should get introduced in a thoughtful manner and only at periodic intervals separated by a reasonable time gap.
The new version of a model should come out only as per the above principle unless there is some problem or "bug" in a particular version of that model, making an immediate "rectification" release necessary.
Frequent changes lead to a situation of trying to hit a moving duck, which is never a good idea as far as benchmarking is concerned.
Interestingly, one of the types of CMMI appraisal is now called as "benchmarking" appraisal.
So, in a way, the above is apparently a crucial structural challenge with CMMI V2.0 appraisal.
An organization starts its CMMI V2.0 implementation with CMMI model around a certain time duration.
When it is time for appraisal, let us say after 2 years, the extant copy of CMMI model might be different.
Incidentally, the CMMI model portal has the provision to download a pdf copy so it is possible to store a copy of CMMI model, as it was, at a given point in time.
The question that is but natural to be asked is:
"Which copy of CMMI model should be used for appraisal - the one from two years back or the one which is there now?
In case the changes introduced are significant, then ensuring all of them are adequately incorporated will be a logistics and coordination challenge for the organization.
But in case the changes are not significant, then this is not an issue per se.
That, then, brings up another point.
If the changes are not significant, why to change frequently?
Would it not be better to accumulate the change requests and improvements from actual experiences shared by the users and other sources and bring a new version?
And do that as a strategic event and not as a routine procedure.
The earlier arrangement, where a model version used to get released at a certain time interval was a better one for the model users and implementers.
The online portal would still be useful for the model developers and model maintainers as they can use it as a work in progress copy of the next version of the model.
Change requests and improvements can get constantly incorporated by the model developers and model maintainers.
And just like code-freeze in case of software development, there could be a content-freeze in the model development cycle, where the next version is baselined and formally released.
CMMI Appraisal Sampling
The second aspect is related to CMMI appraisal sampling.
The sampling now is done using a method that is called as a randomly generated sample (RGS).
In this case, any project can ger selected.
In CMMI V2.0 there is heavy emphasis on business objectives and business performance.
Given that, sampling should also reflect the same line of thinking.
What should happen ideally is that those projects which are critical to business should be a part of the appraisal.
The factors that typically govern “critical to business” are:
- Importance of the customer of the project - this is influenced by both current earned revenue and future projected revenue from that customer
- Size of the project - large projects are more complex and challenging and their success has a proportionately higher impact on the organization's success
- Project in a new vertical or a new domain - such projects can help create huge, sustainable revenue streams in the future
- Strategic value of the project - this would be driven by the view of the executive management in terms of the long-term survivability of the organization
In case the above factors are not being considered in sampling, then the final sample would not be aligned with business priorities.
But in case the above factors are already being considered in sampling, then this is not an issue per se.
And the final sample would still be the same.
That, then, brings up another point.
If the final sample would still be the same in both cases, why to have this new method?
Would it not be better to let the sample be aligned with business priorities and in fact, bring in, or further improve the mechanism to ensure projects that get selected are critical to business?
