iSixSigma

measuring process adoption

Six Sigma – iSixSigma Forums Old Forums Software/IT measuring process adoption

Viewing 12 posts - 1 through 12 (of 12 total)
  • Author
    Posts
  • #26020

    Khandekar
    Participant

    Is there a way to measure how well a company is adopting a new software development process.
    Our group is in the process of adoting a standard process towards our goal of CMM level 2. This standard process involved various phases for the development lifecycle. Currently we are tracking One major deliverable from each phase. So foe example for high level design phase, HLD document itself is a deliverable. Depending on how many deliverables are in a project , we have that many opportunities and we calculate DPMO based on that.
    Our current DPMO is at about 180000 and we are wondering if there is any industry standard/benchmark that we can set as a goal for ourselves.

    0
    #63212

    Ty
    Member

    Atul,
    I’m interested in doing the same, but haven’t gotten very far yet.
    May I ask how you calculated that your DPMO is 180,000 so far? You have obviously defined your opportunity and defect into a working definition to quantify this number. I would appreciate discussing this with you.
    Ty

    0
    #63214

    Gary A. Gack
    Participant

    hi Ty and Atul,
    Suggest you take a look at Dave Hallowell’s most recent article – http://software.isixsigma.com/library/content/c031029a.asp which deals with opportunities in software – you may want to contact Dave direct for more detail – [email protected]
     
    best regards

    0
    #63215

    Khandekar
    Participant

    Thanks you for your response.
    I looked at the article and it deals with Phase Containment  or defects per unit (FP or KLOC). This is good for measuring process effectiveness but I am looking more from a process adoption point of view. This will definetly be our next step but at this point we are too new into the process to measure it’s effectiveness.
    What we want at this point is a way to measure how the process itself is being adoptes across the organization. We have a standard , defined process and teams are supposed to adopt it into their life cycles. Some are doing a good job in adopting and others are not. But our visibility into this adoption is very subjective at this point. Our hope is to develop our current model of DPMOs into something that can be used succesfully to measute the adoption. And for that we are lookig to see if there is any benchmark we can use as a goal for us.

    0
    #63217

    Gary A. Gack
    Participant

    although it’s not exactly on your point, you may want to take a look at George Eckes’ book “Making Six Sigma Last” – he describes an approach to measuring the extent to which a Six Sigma culture has been adopted in an organization – not exactly what you seem to want, but perhaps a close analog. Regarding benchmarks, I’m not aware of anythig like that – if you hear of anything I would love to know about it
    best regards, Gary Gack

    0
    #63218

    Thomas Gorham
    Member

    You write:
    “I am looking more from a process adoption point of view”
    Ask yourself, “What are the goals of adoption?” (as opposed to the outcomes you hope to acheive by having the process adopted).  Some candidates might be:
    –  Project plans are developed based upon a tailored version of the standard process (assuming that not all projects have the same characteristics).
    –  Roles and responsibilities are assigned based upon the standard process.
    –  The project is being managed against a baseline of the tailored project plan.
    –  Project progress is determined by standard process exit and entry criteria being met
    –  My personal favorite…The business customer’s participation in the project is based upon the standard process and the customer understands their responsibilities at those touch points.  (If the customer thinks that they are playing a different game than the standard process, it clearly isn’t being adopted and likely will not be adopted.)
    These are some candidate behaviors that demonstrate adoption.  Failure to demonstrate the behavior can be considered a defect in adoption.  When you begin to do a root cause analysis on these defects you have entered the realm of OD/change management.    
    I hope this helps.
    Thomas Gorham
     
          

    0
    #63220

    John J. McDonough
    Participant

    Your SQA reviews should be able to measure process adoption.
    Indeed, the CMM pretty much says so:
    The purpose of Software Quality Assurance is to provide management with appropriate visibility into the process being used by the software project and of the products being built.
    You should frame your SQA process to develop reasonable metrics that can help you understand how you are progressing in adopting the organization’s software processes.  Simple metrics of how often SQA reviews get done on time can help you gain insight into how well you are executing that process … particularly if you have a well thought out SQA checklist.
    You will find it a little challenging at first, though.  If your SQA process it to have any hope of being effective, you are going to have to be pretty aggressive in improving the process, especially in the early going.  The odds of you having a really good SQA process out of the box are pretty slim.  Of course, a continually evolving SQA process is going to frustrate your efforts to use the results to gain insight into adopting the other processes.  But it’s the best tool you have, so you will just need to be careful in interpreting the data.
    –McD
     

    0
    #63224

    John J Noronha
    Participant

    Hi,
    Would it be fair to conclude then that the project must have a mature s/w process in place (equivalent to level 4 SEI-SW/CMM at minimum) before looking at sixsigma methodologies ?
    Regards
    John

    0
    #63225

    John J. McDonough
    Participant

    John
    I wouldn’t leap to that conclusion at all.  I would suggest, however, that it would be hard to get to level 4 without looking an awful lot like a Six Sigma organization.
    The SQA review is a KPA at level 2.  It can be a very key KPA in helping facilitate achieving the other KPAs.  Now certainly, a level 1 organization may well not have the sort of defined process that helps implement Six Sigma, but by level 2, you should be well on your way.
    I know from experience that Six Sigma can be a large help in getting to level 3, and I suspect that it can be helpful in getting to level 2.  But if your level 1 is truly “chaotic”, then maybe not.
    –McD
     

    0
    #63226

    John J Noronha
    Participant

    JohnIn theory I would agree with your view, as SEI requires implementation of “measurement and analysis” even with level 2 KPAs. However, in practice I feel one really must have well established level 2 management processes as well as well established OPF, OPD, ISM and SPE before a Six Sigma process can add value.In my view, a Six Sigma methodology would really add value in implementing QPM (Quantitative Process Management), as well as the other KPAs of level 4 and 5.If implemented at level 2 there is the possibility of the effort not gathering momemtum and fizziling out altogether after a short burst of initial enthusiasm !Would like to know how others feel too.RegardsJohn

    0
    #63230

    Gary A. Gack
    Participant

    Hi John,
    We have seen many instances of major gains in situations where CMM was not in play at all, and processes were at best level 1, maybe worse – so, it’s clear you don’t need to be anywhere on the CMM scale to get benefits from Six Sigma – management & customers care about business results ($), not maturity level
    best regards,
     

    0
    #63232

    Khandekar
    Participant

    Sorry guys for not responding for last 2 weeks. We had our formal CMM-2 assessment and I am happy to tell that we were assessed at level 2 on friday. Wew!! what a journey,,, This is a great relief as a member of SEPG and a major achievement.
    John; yes, we do have a functional but not so old SQA group and we do have regular project audits for process compliance. Your suggestion is a good one and we are now talking about adopting it going forward to measure process adoption. Our SQA audits on projects focus on KPAs and NCRs always correspond to one or more common features.
    We will most probebly use # common features in NCRS vs total Common features used in the audit as a measure of our DPMOs. I hope this is a better indicator of adoption. Atleast it will help us keep audited projects well in shape for our level 3 sample.
     
     

    0
Viewing 12 posts - 1 through 12 (of 12 total)

The forum ‘Software/IT’ is closed to new topics and replies.