This is one of many entries in the Change Diary: a true story of success and failure, written as it happened.

Thursday

14 Mar, 2002

Today I learned that the organization boss has acceded to a real bullpen! The details aren't settled yet, but it's probably going to combine several small, related teams into one, and include QA and the project manager. Just what I've been wanting! Now I can emphasize the next change more: a bullpen doing mostly-XP with me as coach. The boss is going to want a detailed plan for migrating to the bullpen environment, so my next step will be to create that plan. It's going to be a big project, so I'll probably do it over my upcoming vacation. Hopefully they won't proceed while I'm gone.

So, why did this happen? I asked my champion the same question. (Uh... champion? It's a word I just sort-of invented: somebody I work with who has more clout than me and thinks along generally the same lines I do.) He said that he wasn't sure, but that the boss was pretty open to the idea. He thought it was just because lots of people were talking about the idea.

I can't prove it, and it's oh-so-unhumble of me to claim this, but I'm pretty sure that was my doing. I've been talking about bullpens a lot recently.

Hmm... a side note. When I say I've been talking about something a lot, I mean a few times a day. I actually don't talk about it that much, relative to the other stuff I do. The majority of my day is spent doing my actual work. Every so often, a legitimate opportunity to say something change-related comes up. By "talking about X a lot," I actually mean that the majority of my change-related comments are about X.

What does a legitimate opportunity to say something change-related look like? Here's some. I've put the ones that I think result in the most change at the top.

• Something goes wrong. I say, "Let's prevent this in the future by doing X." ("Next time, let's try X so that doesn't happen again."; "X is a good way to solve this problem; "We should do X so this problem doesn't come up next time"; ...)

• Something irritating or tedious happens. I say, "If we do X, we won't have so much tedious Y."

• Something vaguely related to process comes up. I say, "X is something related that we should try some time."

• I feel like ranting. I've tried to stop doing this, because I think it tends to hurt more often than it helps. I like to rant, which can be a problem.

After my intro, I say a few sentences about X and then let it rest.

Every so often, I'm asked a direct question about process. These occasions are rare but wonderful. I get to spew all kinds of stuff. Impact-wise, it's probably about the same level as "something vaguely related comes up." But since I get to explain my viewpoint in more detail, it probably raises my credibility a bit more than the few-sentence approach.

I also write papers. So far, I think I've averaged one every few weeks. The papers are targeted at people who are higher in the organization and generally have a direct call to action. So far, I've written them after I've already seeded the change pretty thoroughly via one of the above methods. No one has ever responded to one of the papers and said "Yes, let's do what's in this paper." But so far, the main actions recommended in all of my papers have been followed. I don't know if that's because of the word-of-mouth, the paper, a combination of the two, or just blind luck. So far I've written three papers:

• A process and rationale for evolutionary architecture on the organizational framework (5 Feb).

• A recommendation that we use agile development and collocation instead of building an up-front architecture (~15 Feb).

• A call to collocate developers, QA, and project manager in a bullpen (6 Mar).

Finally, I should emphasize that everything I do is based on moment-to-moment instincts. What I've written above isn't a set of rules I follow; it's an after-the-fact analysis of what I'm doing naturally. I think I'd come across as artificial and false if I tried to follow the above paragraphs by rote.

Years Later...

5 May, 2006

I find myself in a difficult position as I write these analyses. I don't know how to make an organization change. I don't really think it's possible.

What I have learned, through years of bloodied noses, is how I can change my behavior to make my ideas more acceptable. I've learned this through thoughtful observation of people's reactions to me. I've always been keenly attuned to people's reactions. My skill at influencing change, on the other hand, has had to be learned. It's been a struggle, and I haven't necessarily mastered it.

Since I don't know how to make an organization change, I'm uncomfortable giving advice to you, my readers, about how to do it. And yet this diary is all about organizational change, so I feel that I must say something. Besides, it's not entirely true. While I can't make an organization change, I certainly know ways to influence organizations so that they are more likely to change.

The only problem is that I don't know how to describe what I do. In these entries, in my halting way, I've tried to give it form. Compassion, respect, courage, trust... all the warm fuzzy words. Hot soup. (Can you believe this crap? It's the best I've got, though.)

Add "patience" to the list, and "integrity." Although I was barely patient. It had only been three months at this point, and already I was feeling burned out. I hid my impatience, but not perfectly, I suspect.

Add "credibility" as well. I've been blessed in that, for whatever reason, people typically have a first impression of me as a smart guy. I built on that strength by doing my work well. The essays probably helped, too. I'd bet that it at least caused some buzz. I had some small success in implementing my ideas locally, but not with much noticeable effect. Still, that probably helped spread the word.

Ultimately, I still don't know why the bullpen idea was picked up. I'm pretty confident it was the result of my efforts. It was the one thing I had decided would have the most benefit. Did anybody realize the idea was mine? It didn't seem like they did, which may mean that, despite my incipient burn-out, I was doing my change-agency job perfectly. Gerald Weinberg:

In order for a consultant to get credit, the client would have to admit there had been a solution. To admit there was a solution, the client would have to admit there was a problem, which is unthinkable. As a result, the only consultants who get invited back are those who never seem to accomplish anything.

Whether these consultants actually do accomplish anything is an unanswerable question. Whichever way it was answered, it would leave the consultant out of a job, so effective consultants make sure it is never asked. Unfortunately, so do ineffective consultants. The difference, however, is that when an effective consultant is present, the client solves problems.

Gerald Weinberg, The Secrets of Consulting

Ah, if only it was easy. I had succeeded in heating up the soup, but now it was out of my control... and I was about to be out of the office for three weeks straight.

Next: Week Fifteen