It’s been an unusual weekend. Spent most of it closeted away with a bunch of very talented people, at an event organised by the Trinity Forum, headlined When No One Sees: The Importance of Character in an Age of Image. It worked a bit like an unconference: a small group of attendees, a core agenda run workshop style, lightly moderated and completely dependent on a participative audience. The format was garnished by some excellent guest speakers at mealtimes, and the surroundings were superb. More of that later.
I was very taken by something said by one of the visiting speakers. Headmaster of one of the larger private schools, he described his job as “being responsible for 1250 teenage boys every Saturday night”.
We were looking at the role of education and educationalists in the formation of character; it was a fascinating debate, bloglike in its tangential nature. At some point in the discussion, he was describing aspects of the pupils’ engagement with theatre and drama, and he made the observation:
We don’t use prompters
I think this is key. A simple decision — doing away with prompting — had a worthwhile impact on the takent and character of the students. They changed the way they prepared; they changed the way they responded when facing a problem; they changed the way they stepped in to help when others faced problems.
We need to keep examining what we do: every time we promote an inspection/repair culture, we tend to implement safety nets; the safety nets encourage slipshod behaviour, and soon we find that all we are promoting is mediocrity.
If achieving mediocrity wasn’t bad enough, we tend to make it worse. Far too often, the mediocrity attracts another foul behaviour, an audit culture where the measurement process becomes more important than that which is being measured. How else can mediocrity rise?
Popularity: 1% [?]
Our global sourcing models fly directly in the face of this movement. We are arbitraging labor markets, sending work far away to teams who are disconnected from our corporate cultures and imperatives, and finding that we can’t rely on them to deliver high enough quality deliverables. so we take our smartest, most skilled practitioners and turn them into reviewers and auditors. I’ve been concerned about this for while because we are outsourcing the roles from which we grow the next generation of experts — so it doesn’t seem very sustainable. but now you’ve got me thinking about the corrosive effect of transforming folks from doers to auditors. ouch.
Bill’s commend of “we are outsourcing the roles from which we grow the next generation of experts” really hit home. I’ve been involved with “onshoring” for quite some time on the Evolving Excellence blog, even moderating a panel discussion on it at Kellogg last weekend. See
http://www.evolvingexcellence.com/blog/2007/05/a_tide_against_.html
Many reasons for not outsourcing were brought up… keeping knowledge in house, leveraging faster cycle times, etc. But not what Bill brought up. The generational, or knowledge generational, impact of outsourcing could be monumental.
Kevin
Intersting points indeed on the outsourcing. Not that I believe we should have no measures for what we do, they are important ways of tracking progress (or not), however to pick up on the point JP makes I believe we spend too much time gathering data and gernerating metrics to satisfy one audit/mgmt need or another. When is a metric not a metric? When it states something that can be easily seen without them. If we spent as much time fixing issues/bugs as we do proving with metrics that they exist I’m sure we’d get more done.
A basic set of metrics to track velocity and quality are really important. Without them you can’t tell how well you are performing, and you can’t assess the impact of changes you make to the team or the process. However, you are right, Andy, metrics are a slippery slope, and someone has to keep the organization’s never-ending hunger for new statistics in check or you’ll spend way too much of your capacity trying to measure your capacity in new and ever-finer-grained ways. :)
Also, just to be clear, I don’t have a problem at all with peer code reviews. You can get a tremendous amount of cross team learning by having other practitioners walk through selected areas of the system. (Although I still think Pair Programming, which is part a real-time code review, is infinitely more effective in this).
What I’m concerned about is that we are turning our practitioners into specifier/reviewers. Giving a coder some responsibility for reviewing the work of others doesn’t bother me; it probably gives everyone a better perspective on the work they are doing themselves. Giving their coding work away, and leaving them only with auditing responsibility — *that* gives me the shivers.