Bulldozer00's Blog

Investment is to Goldman Sachs as management is to McKinsey & Co. These two prestigious institutions can do no wrong in the eyes of the rich and powerful. Elite investors and executives bow down and pay homage to Goldman McKinsey like indoctrinated North Koreans do to Kimbo Jongo Numero Uno.

As the following snippet from Art Kleiner’s “Who Really Matters” illustrates, McKinsey & Co, being chock full of MBAs from the most expensive and exclusionary business schools in the USA, is all about top-down management control systems:

…says McKinsey partner Richard Foster, author of Creative Destruction. If you ask companies how many control systems they have, they don’t know. If you ask them how much they’re spending on control, they say, ‘We don’t add it up like that.’ If you ask them to rank their control systems from most to least cost-effective, then cut out the twenty percent at the bottom, they can’t.” (And this from a partner at McKinsey, the firm whose advice has launched a thousand measurement and control systems.)

A dear reader recently clued BD00 into this papal release from a trio of McKinsey principals: “Enhancing the efficiency and effectiveness of application development”. BD00 doesn’t know fer sure (when does he ever?), but he’ll speculate (when does he never?) that none of the authors has ever been within binoculars distance of a software development project.

Yet, they laughingly introduce a…

..viable means of measuring the output of application-development projects.

Their highly recommended application development control system is based on, drum roll please…. “Use Cases” (UC) and “Use Case Points” (UCP).

Knowing that their elite, money-hoarding, efficiency-obsessed, readers most probably have no freakin’ idea what a UC is, they painstakingly spend two paragraphs explaining the twenty year old concept (easily looked up on the web); concluding that…

..both business leaders and application developers find UCs easy to understand.

Well, yeah. Done “right“, UCs can be a boon to development – just like doing “agile” right. But how often have you ever seen these formal atrocities ever done right? Oh, I forgot. All that’s needed is “training” in how to write high quality UCs. Bingo, problem solved – except that training costs money.

Next up, the authors introduce their crown jewel output measurement metric, the “UCP“:

UCP calculations represent a count of the number of transactions performed by an application and the number of actors that interact with the application in question. UCPs, because they are simple to calculate, can also be easily rolled out across an organization.

So, how is an easily rolled out UCP substantively different than the other well known metric: the “Function Point” (FP)?

Another approach that’s often talked about for measuring output is Function Points. I have a little more sympathy for them, but am still unconvinced. This hasn’t been helped by stories I’ve heard of that talk about a single system getting counts that varied by a factor of three from different function point counters using the same system. – Martin Fowler

I guess that UCPs are superior to FPs because it is implied that given X human UCP calculators, they’ll all tally the same result. Uh, OK.

Not content to simply define and describe how to employ the winning UC + UCP metrics pair to increase productivity, the McKinseyians go on to provide one source of confirmation that their earth-shattering, dual-metric, control system works. Via an impressive looking chart with 12 project data points from that one single source (perhaps a good ole boy McKinsey alum?), they confidently proclaim:

Analysis therefore supports the conclusion that UCPs’ have predictive power.

Ooh, the words “analysis” and “predictive” and “power” all in one sentence. Simply brilliant; spoken directly in the language that their elite target audience drools over.

The article gets even more laughable (cry-able?) as the authors go on to describe the linear, step-by-step “transformation” process required to put the winning UC + UCP system in place and how to overcome the resistance “from below” that will inevitably arise from such a large-scale change effort. Easy as pie, no problemo. Just follow their instructions and call them for a $$$$$$ consultation when obstacles emerge.

So, can someone tell BD00 how the McKinsey UC + UCP dynamic duo is any different than the “shall” + Function Point duo? Does it sound like the same old wine in a less old bottle to you too?