Bob,

I agree that CMMI has evolved to handle the complexities of trying to standardize processes across a large enterprise.

Beyond a certain size, even *finding out* the current state of play is difficult, let alone trying to co-ordinate standards through a centralized body. Attempts to do this include Enterprise Architecture, CMMI, Six Sigma etc etc.

The problem is that "one size fits all" can easily turn into "one size fits none". Innovation is easily stifled when all changes have to go through layers of bureaucracy (which I presume CMMI imposes).

A thought: with the World Wide Web, we have seen extraordinary innovation combined with evolving standards for process standardization and interchange *without* a mandated policy. (W3C recommends, it doesn't mandate.)

So let's propose a continuum:

(1) At one end, we have small companies where policy should be constructed and rolled out centrally. This works because (a) all affected users can directly work with policy-makers to identify problems; and (b) solutions are comparatively cheap and quick to implement.

(2) At the other end, we have globalized phenomenons like the World Wide Web where standards and policies emerge organically because there is a mutual benefit for all parties to co-operate at a baseline level.

Corporations operate at a complexity which is somewhere in-between. So my question to you is this: what's the *smallest* level of intervention required in a corporation to maximise productivity through consistency?

Stephen, your last paragraph is completely on point. I have run into too many people in a variety of organizations who, while denying this vehemently, propose and champion a "one size fits all" strategy.

I have heard many opposing viewpoints, but none as succintly put as yours. "What's the *smallest* level of intervention required in a corporation to maximise productivity through consistency?"

May I quote you in the future????

Thanks

Hi Don,

Be my guest!

I'd love to know the *answer* to my question, though ...

Stephen wrote:
"The problem is that "one size fits all" can easily turn into "one size fits none". Innovation is easily stifled when all changes have to go through layers of bureaucracy (which I presume CMMI imposes)."

Your statement above is ~85% right. The first 85%.

CMMI does *not* impose layers of bureaucracy. How exactly do layers of bureaucracy improve productivity and performance and reduce waste?

Granted, layers of bureaucracy are exactly how way too many organizations get to a CMMI rating, but that's not the model's fault. That's a fault of making the rating the goal (at any cost) instead of honest improvement the goal (to lower cost).

Stephen wrote:
"The problem is that "one size fits all" can easily turn into "one size fits none". Innovation is easily stifled when all changes have to go through layers of bureaucracy (which I presume CMMI imposes)."
...and...
I'd love to know the *answer* to my question, though ...

=====>

I don't know the answer either but here's one observation about "standard processes" I've adopted. Processes that focus on "what" instead of "how" tend to be more useful and enduring.

Chris White

Hi,
Here's my two cents. I've been around in this field for over 20 years. I know big deal. I've seen a lot of things come and go. There are few things that happen.

1. Somebody has success with a project and tries to standardize the way they did it to be the way everyone should do it. After all, they did slice bread for the first time! Book deals have no influence.

2. Attempts are made to create or codify practices, in order to self-aggrandize, or put out a product that will help you...pay off your debts in 3 days...lose 200 pounds in a week and still eat everything you want to eat....get into a negatively amortized mortgage with a variable rate and never have to worry about losing your home...etc. etc. etc. Grand visions of embarassing profits from some such products or "education" programs are no influence. Not at all.

3. There is an attempt to institutionalize management skills, abilities, and methodologies by creating Best Practices, ITIL, CMMI, Six Sigma, PMP, certs, etc. Again book deals, possible software sales, and dues, I'm sure, have no influence.

I will say, having done some light reading in the third group that I have seen some valid ideas in them but again, can you make a silk purse out of a sow's ear. Can you create an IT Management degree program and crank out IT managers or is experience a better teacher? Can you cookie cutter the skills need to deal with all situations that come up in the business world? Can you legislate morality? Can you codify a person's dedication to quality, fairness, and thoroughness? I tend to doubt it and the state of many IT departments across corporate America bears me out.
For example, Are we really discussing whether disk-based backup is better that tape based? Really?? Sure do tape at the second stage so you can send off site. Do technical schools really think that their OS development courses are worthwhile? IBM couldn't displace Windows, even though they had a better more powerful product.

It oftentimes seems to me that the industry has progressed so little in these last 10-15 years. Perhaps I was spoiled in the 80's with things like client-server, RAD, 4th and 5th gen languages, SCSI, PCL, VLSI, super-servers, EEPROM, spreadsheets, word processors, portables, T1s, etc.
Now the only place I see progress on a similar level is in hard disk storage. My daughter's portable has 20 times the storage that I had at the first company I worked in IT for. There's terabyte externals! The sky is the limit there.

You have to do what I loved getting into this field for. Reading and research, expanding your knowledge and being discriminating in what you believe and/or use. Thank God Infoworld and PC Mag are still around. Everything else seems so vendor driven. They control the industry these days, once again. We thought it was the end of that when client server and PC networks started replacing large hosts and the monolithic companies behind them. In some ways I miss those days. At least the core excellence of companies like IBM kept everybody honest.

Geo