Lesson 1: Except in rare circumstances, don't tell management that they need to adopt systems thinking. Most senior managers are eminent pragmatists, focused on their goals (or the goals they've been given). Systems thinking, as good as it is, is just a means to an end for them, and they'll be more likely to hear you if you help them achieve their goals than if you ask them to adopt your tools.

Listen to them, and discover what their problems and goals are in their words. Ask questions. Be curious. Don't fake being curious so that they'll open up; really be curious (people can often tell the difference, and it makes a difference in how much you'll learn, too). When they've stated a clear problem that systems thinking can help with, ask them if they'd be interested in a solution. Then show them the way forward, perhaps without ever mentioning the words "systems thinking."

Think of your trips to get your car serviced. The service people don't tell you about all the great, computer-aided equipment they have. They simply ask you what's wrong with your car, they fix it, and then they ask you if it meets your satisfaction. (Well, they also ask for some money, but that's a different story.)

You may work for an innovative manager who sees the strategic advantage in simulation, and you may be tempted to start talking about systems thinking again. Careful! I had a manager like that when I was starting in organizational simulation, and it seemed great. What it allowed me to forget was that the managers I was serving, his peers, didn't share his enthusiasm, and I needed to be solving their problems, not talking about my technology.

Lesson 2: Don't do your work in a vacuum. When I first started out, I'd go ask managers what their biggest problem was. They'd tell me, and then I'd head back to my desk and computer to start working on it. A week or so later, I'd drop by the manager's desk again to get more data, only to hear that that wasn't a problem anymore; they had solved it two days ago! When I'd ask what they did, it was clear they had found a good but often quite sub-optimum bandage to apply. That is, they fixed the problem well enough so that something else was crying out for help, but they perhaps hadn't fixed it for good.

It took me a few loops through this process before I discovered I could always be working on a problem someone had declared a week ago to be their biggest problem and that they had "solved" two days ago -- not a way to feel good about my contribution.

If a manager offers you a problem, work with them to solve it. Solicit the information you need while you're sitting with them, and capture it on paper in front of them. Scribble down statements, data, and fragments of stock and flow diagrams. Accept their feedback. If it's the sort of issue and situation where it's appropriate to pull together a group, do so, and use any of the facilitation techniques created to help with such work.

You'll probably need to do some of the more detailed modeling on your own, but don't stray long from involving them in giving you data or reviewing and guiding your progress. You'll have to judge how long you can stay apart, but, in most cases, you should be interacting multiple times a week.

To maintain that openness and that pace, you'll need to be good at modeling. If you don't feel comfortable working in front of your managers or internal customers, and if you have to spend more time studying than doing, get yourself some support, whether it means taking a course, bringing in outside consultants to help, or allying yourself with others in your company who can help you deliver the services your internal customer expects. You might find a consultant who will collaborate with you so that the two of you together deliver the value and you simultaneously increase your skills.

Lesson 3: Develop a knack for seeing patterns and recognizing likely underlying structures. One of the key mantras of systems thinking is that events are part of patterns and patterns are created by structures. Most people look at events and see events. When you see a notable event, see if it's part of a pattern. If it is, think about the type of structure that might create such a pattern, and look at the organizational system to see if it has such a structure. Then think about modifications to the structure that might fix the recurring events. Finally, test your hypothesis with simulation.

That approach will help you offer effective services faster, and your managers will appreciate that you can help them solve their problems well and quickly.

Of course, you get better at recognizing the structures that create specific patterns by doing lots of simulation.

Some years ago, I watched a manager talk about bouts of overspending followed by bouts of under spending. To him, it was an event of overspending that lasted several months, then a pause, then another event of several months of overspending. I saw this as a type of simple oscillation, and I began to look for a structure that could generate such oscillations.

Knowing what to look for helped me find a structure in the organization that could generate such a pattern, and that greatly sped the process. It wasn't too far from discovering the structural problem to proposing the cure, and then it was simply time for testing, approval, and implementation.

Incidentally, out-of-control spending dropped by 95% when we installed the new process in the real world.