Proper and Improper Use of CMMI

Just a few thoughts on some questions to pose as a sort of “guide” for whether or not you might expect benefits and value from using CMMI.  These also have the benefit of helping CMMI be implemented in a more lean/agile approach.

When implementing CMMI, Are you seeking . . .

• Improvement or Compliance?
• Empowerment or Definition?
• Clarity & Awareness or Constraints & Rigidity?
• Bottom-up input or Top-down direction?
• To understand whether what you’re doing is working?  or Whether you’re doing what the process says?

In this case, we also value the things on the left more.

:-)

The things on the right are a longer road, with questionable benefits and many risks.  The things on the left get you to benefits and value sooner with less carnage and baggage.

Take your pick.

Love & Marriage: CMMI & Agile Need Each Other

An article in this month's CrossTalk periodical, is now out.

See it here.

Download it here .

Enjoy!

 

 

 

 

P.S.  There are other great articles in the issue as well.  I'm in great company with an article by my friend, colleague and client, Jeff Dutton.  And, don't miss out what's coming next in v1.3 from my buds Mike Philips and Sandy Schrum!

Getting Started with CMMI and/or Agile.

Why you're probably not equipped to be asking the the questions!

Starting a CMMI or Agile initiative is a culture-changing endeavor.  Don't underestimate the changes you'll undergo.

Both CMMI and Agile often require fundamental shifts in culture that usually results in making <gasp!> changes to how things are in your business!

This in-your-face provocative < 3min video is intended to alert executives interested in either CMMI or Agile (or both) that there are no easy answers as to the question often on their lips: “what does it take to do ____ ?”  And, that to do justice to their business, they’ll need to devote some time to understanding even the basic context in which to understand any answers they receive (let alone use) about deciding to move forward with an improvement effort.

After all, if everything were going perfectly now, executives wouldn’t be seeking changes to improve, so when it comes to making improvements, there’s going to have to be change -- so get ready!

(I’m assuming [big time] that moving forward with either CMMI or Agile is to achieve some improvement in something!)

In my next installation, I’ll talk about what to do when you really only need a CMMI ‘level’ and aren’t so much interested in any improvement.

Top 10 Clues You're Not Ready for a SCAMPI

10. Four months ago you couldn't spell "CMMI".

9. No one in your organization has ever had any training of any kind whatsoever in CMMI, appraisal planning, or process improvement.

8. You haven't worked with anyone (in-house or hired) who knows what they're doing with CMMI.

7. Price-shopping for lead appraisers seems like a good idea. Kind-a like price shopping for a heart-surgeon.

6. After months of work, you switch from one constellation to another and think your appraisal is still on schedule (see #9 and #8). Kind-a like switching your team from field hockey to ice hockey mid-season.

5. Your CEO is petulant about the delay of your appraisal yet has no idea that your actual process performance is in the toilet.

4. You've waited until it's time to start planning for your SCAMPI to start looking for a consultant to help you implement CMMI in an "agile" way.

3. The only people you're sending to Introduction to CMMI are the ones you plan to have on the appraisal. You have no back-up plan if they can't make it to training and/or to the appraisal. Class isn't for another month, and, they're the same ones who've been working on your processes for the past several months, but until now, see #9 and #8.

2. You haven't qualified the people in #3 with your lead appraiser (which you haven't hired yet + see #9 and #8), you haven't qualified the projects to be appraised with your lead appraiser (which you haven't hired yet + see #9 and #8), and nonetheless, you have established a level of effort for the appraisal despite all of the above.

And,

The Number One Clue You're NOT Ready for a SCAMPI:

1. After months of work, you still don't see there's a fundamental flaw in committing to or expecting others to commit to (including your appraiser no less!) a firm-fixed price contract without knowing the requirements.

I wish the above list of clues were tongue-in-cheek.

Sadly, it's not. There is, nonetheless, plenty of fiction in it:

• The order of the list is mostly arbitrary.
  
• The list is not scientific.
  
• The list really should be longer. A lot longer.
  
• These clues are just from my experience alone, and doesn't account for anyone else's, so it's pretty idiosyncratic.

While in many cases, any one of the items in the list of clues could easily sabotage an organization's process improvement effort (let alone an appraisal), one thing that makes it especially troubling is the preponderance with which I encounter a single organization exhibiting most or all of these clues! And yet, such organizations think they actually can dictate to their lead appraiser the terms and conditions and the readiness of their organization to be appraised!

I tried to limit the clues to only things you could pick-up from a conversation -- even if you know nothing about CMMI or the SCAMPI appraisal method. I tried to keep the clues to things that don't have to do with CMMI itself or process stuff as a general theory. If I had included such items, I would add:

• You don't know that to do a SCAMPI you actually have to have used the processes on actual projects.
  
• You don't realize that there's more to CMMI than the names of the process areas.
  
• You don't realize that the generic goals and practices aren't just "extra information".
  
• You don't know that it's still the lead appraiser's responsibility to approve the people and the projects to be in the appraisal.

If you look at some of the other behaviors of organizations not really ready for a SCAMPI, you'll find that they continue to:

• accept more work than they can handle,
  
• be unpredictable in what they will deliver and when,
  
• measure little other than billable hours, and
  
• have no insight into where their defects come from.
  

Despite the condition in which many organizations may have artifacts for an appraisal, they have seen no intrinsic benefits to their new set of processes. They've been just "chasing a level", which results in lots of work for no real benefit. You can guarantee these organizations will drop their processes shortly after the appraisal.

Really, what these clues point to is the dreadful lack of realization that first and foremost, process improvement requires the right culture for process excellence. The above list points to a fundamental absence of the culture for process excellence. What people don't realize (especially in the USA), is that process improvement is a total JOKE for companies with the right culture of process excellence. If people want the truly idiot-proof, guaranteed, easy path to just about any CMMI Maturity Level, they would need to go no further than to foster a culture of process excellence, then whatever they did from that point forward would likely cause just about every practice in CMMI to form on its own. In other words, these clues aren't about process, they're about culture and leadership. You don't have to know anything about CMMI to pick up clues about culture and leadership.

One of the failures of CMMI is that it fails to press the basic importance of culture and leadership for process improvement. It fails to communicate in no uncertain terms that an absence of the right culture (and what that looks like) and the absence of leadership (and how that shows up) will lead to a failure in process improvement -- CMMI or otherwise. I'm not blaming CMMI for not including this information in the text, because much of it is there. What I'm saying is that despite what is there about the importance of culture and leadership, most CMMI users fail to grasp these important points. The text, therefore, fails to communicate this in a way that people will pay attention.

CMMI should come with a warning similar to, "Don't Try This At Home!" or "Use Only as Directed!", or "Check with a Qualified Professional Before Beginning Any Process Improvement Program", or "You Must Be *This* Tall to Use this Book". Or simply, "Danger Ahead!". Something to get people's attention and direct them to some of the fundamentals of any improvement program.

This accusation is not just leveled at garden-variety CMMI adopters. Often, they're the hapless "stuckees" forced to "make CMMI happen" against all odds. At least there's ample reason to be sympathetic to their plight. What's inexcusable are the too-many consultants, instructors and appraisers who are willing to ignore these fundamental requirements-for-success and who are unwilling or unable to posture with prospects and clients in such a way as to impart the importance of culture and leadership to success with CMMI. So, that translates to a pathetic statement about the consulting abilities (and possibly the ethics) of too many people who take money to work with others on CMMI.

Sorry for the rant. But people need to be warned. Attempting a CMMI effort without the requisite culture and leadership attitude may yield a short-term appraisal rating, but will ultimately lead to medium and long-term failure. *THAT* I can guarantee.

This is why I'm such an advocate for agile methods. Agile methods impart some of the basic needs for long-term process success. And, it imparts them at a level of abstraction usable by people who aren't process experts, yet establishes many of the appropriate culture and leadership traits that so many CMMI-only efforts fail to recognize. Agile values, methods, and practices empower their teams, cause leadership to eliminate obstacles, value the input of customers and practitioners alike, values learning, develops multi-disciplined teams, and most importantly (as far as processes are concerned) promotes lean thinking. While Agile ideas may not change leadership and culture over night, they contain many of the right activities that can eventually win over those whose hearts and minds need winning over. And, with the appropriate use of CMMI, agile ideas can kick-start the motion and direction needed for long-term and ongoing improvement.

I'll be writing more about why CMMI and Agile need each other for a while.
Stay tuned.

Field notes from SEPG-NA 2009 - Thursday

San Jase, CA.  Actually there won't, unfortunately, be much to report today as I was side-tracked from all but one session.  The session I attended was an experience report from a lead appraiser working with a company whose total size was all of 12 individuals.  In fact, they were a distributed workforce with no central offices.  Everyone worked from home.  This was a report on how they achieved CMMI ML3.  The company itself had very impressive results made possible by very impressive people and attitudes.  Let's get one (or two) things straight immediately: (1) they really truly were ML3, no corners cut, they really truly did the work of defining, managing and using their processes, (2) they did NOT need CMMI to be disciplined -- to a person they were highly skilled, highly technical, supremely professional, absolutely committed their work and company, incredibly laid-back, and deadly serious about NOT causing themselves work that was not fun and benefit the work and company.

The moral of this story is that the primary driver of improvement (of any flavor) is first and foremost attitude and culture.

Moving on.  I stopped by David Anderson's last talk of the conference on Metrics and Agile, and once again, it was close-the-doors-room-is-full SRO.  While other sessions were letting out early and people were streaming out right at the stop time (time for lunch!), David's session was full until kicked out.  Once again, put CMMI, Agile, and Maturity in the same space and sparks fly.  Dave impressed several very important people.

That's all I have to report today about the conference.  Stay tuned for SEPG-NA 2010.  I'm contributing to the program committee and know there are some really great things in store.  Planning for it started today and I've got action items due.  :-)

Field notes from SEPG-NA 2009 - Monday

San Jose, CA.  I'm at (no surprise) SEI's annual big deal conference, SEPG-NA.  As might be expected, attendance is way down due to the economy.  SEI had to scale back a lot of the more splashy touches -- no-frills tote bag, nixed VIP socials mixers bare bones staff.

I arrived in time to teach a CMMI-SVC Supplement course for the SEI on Sunday -- scheduled to coincide with SEPG for the convenience of travel -- that evening I shared conversation and a bottle of really nice California Merlot with Alistair Cockburn, Tami Zemel and Steve Masters.  Earlier in the day Alistair listened in on my class from the corridor and over cheese and fruit bluntly reported that the content made his ears bleed.  Unfortunately, he's right.  Despite the mostly very positive feedback, there's only so much charisma can do for certain SEI materials.

Alistair challenged me to explain CMMI to him in 5 minutes or less or he'd fall asleep.  I believe I succeeded.  He Tweeted as much at least.  As it turns out, not to either of our surprise, whether using agile terms or traditional terms, if you're working to improve the experience and situation of "development", you have the same goals and face the same challenges.  With that settled we called it a night and met this morning over breakfast to joke about travel anecdotes and strategize our individual plans for the day.

With other obligations on my plate for this week, this morning I only sat in on half of a half-day tutorial this morning on the excellent topic of The Role of Organizational Culture in Process Improvement.  Rather than a bunch of finger-wagging (which, from other presenters, such a topic title might devolve into), anthropologists, Palma Buttles and Fred Valdez, and process improvement uber-guru Judah Mogilensky gave a very well-informed, thoroughly enjoyable, interactive and insightful tutorial on several very specific attributes of culture that affect how to introduce, address and implement process improvement, and the challenges faced by consultants, appraisers and users alike due to culture.  Concepts on the perception of time, surface or hidden emotion/expression, stated vs. rewarded values, and so on.  During this session, David Anderson arrived.  We commiserated over the registration statistics and what it may imply for other large-scale conferences like Agile2009.

To round-out the day's sessions I attended Corey Ladas' mini-tutorial, Launching a Kanban System for Software Engineering.  He put up a slide depicting a "waterfall" life cycle which included a "stabilization" phase-gate to which he said, "I don't think I'm saying anything anyone doesn't already know will fail."  Someone in the audience stopped him to ask (with incredulous tone in her voice), "Are you trying to say that this approach doesn't work?"  <<Snicker.>>

After the tutorial, I headed off to the exhibit area for the "grand opening" of the exhibit hall.  As part of the fanfare, a troupe was hired to march around the exhibit hall in oriental dragon costumes accompanied by drums and cymbals.  It was festive and lively.  Though it would have been more appropriate had they been asked to start things off, lead everyone into the hall, do a circuit around the hall, then be done.  Instead, they continued to perform for a lot longer than needed.  In addition to causing traffic problems (which wasn't really a huge issue), they made it hard to speak while nearby.  That was an oversight.  After a break, they returned to continue, only playing softer.  Still, their initial display was too long and they didn't have to come back.  It wasn't that it was bad, it was merely unnecessary.  As for the exhibit hall... so sad... so many fewer, and each booth featured fewer people.  The student posters, were a refreshing new feature this year.  I was impressed with their efforts, both in terms of research and commitment.  First person I ran into was from, of all places, UMBC.  Yup, home turf.

Afterwards, David Anderson and his gf joined several of us for a wind-down at the Marriott's concierge lounge.  Well, as I should expect, my increased visibility within SEI and within the CMMI-oriented market has also resulted in never having to sit alone if I didn't want to.  Even then, I didn't always succeed in getting long stretches of time on my own.