Late and Overbudget -- what is it about?
How to be a cheerful, effective and successful project manager, even when things aren't going well.
I'm just reading Guy Kawasaki's book "APE". Which has made me think that I should have another go at saying what it is that I'm trying to do by writing this book that I've been calling "Late and Over-Budget".
Commitment and Consistency vs Planning Fallacy
The reason project management is such an emotional business is because of two fundamental principles of human psychology. + The principle of commitment and consistency + The planning fallacy
The principle of commitment and consistency is that people want to do what they say they're going to do, because people who don't do what they say they're going to do tend to be regarded as mad or bad.
The planning fallacy is the mistake that almost everybody makes when they're planning. We assume that things will go as well as they can do. We do this consistently, even when we've had lots of experience of things not going so well.
Combine these two things, a constant tendency to underestimate how long things take and a desperate desire to honour our commitments and do what we say we're going to do and the result is misery.
The more I think about it (and the more that I continue to work as a project manager) the more I see that Commitment Jujitsu is both the way most projects are managed and the main reason that so many people who work as project managers feel so miserable. I think there are some really interesting things that can be done to deal with these problems.
From the point of view of the planning fallacy, the naive optimistic "inside view" of a project can be informed by "The outside view" what actually happened on other projects.
From the point of view of commitment and consistency project managers can be careful that they commit to the right things and make sure that they get the right commitments from others.
Ideas about this come from Venkatesh Rao's mentioning of the importance of certain things being "illegible". He gets the idea for this from James "Scott's Seeing Like a State". My idea about this is that there is a certain level of clarity that is needed to allow the project to get "green lit" and funded. And a certain level of clarity that various levels of management can take about what is actually going on in the project. At the same time there is a different level of clarity that's needed inside the project in order to actually get things done.
Out of this kind of thinking comes a beginning of an understanding of what plans are for. Plans are there to hide detail, hide progress and hide the issues, so that the people who are funding a project can be given the impression that it's OK to green light a project, or continue funding a project that has started.
Manipulating clarity is a lot of what the people who have the title "project manager" actually do.
Coase's theory of the firm - As bad as it can be, as good as it gets
Companies and firms are more efficient than markets for a whole slew of risky endeavours, including a lot of software development. But within an organisation, this has the opposite effect from what you might think. An organisation which, taken as a whole, is very efficient, can, internally, be massively inefficient. This is why when you look at the software development that goes on inside organisations that have some kind of strong market position, a virtual monopoly, or a state monopoly, what you see internally is not efficiency, but massive amounts of inefficiency. This is why the software development that you see inside banks appears to have almost been designed to be as bad as it possibly could be.
As with clarity. It isn't enough to say that we should be honest and transparent, so with efficiency, it isn't enough to say that we should be "more efficient". The efficiency of the IT and software activities of a bank, or a government department is a function of the overall structure of the institution. All the inefficiencies are set by some magic hand at exactly the right level for the organisation not to collapse, whilst at the same time making sure that the absolute minimum gets done for the maximum amount of money.
Competence - Clueless, Losers and Sociopaths
The terminology is to negative. But the ideas are very interesting. The most important idea here, from the project management point of view is to challenge the idea that the problems that are presented in managing a project can be solved by increases in competence and because of the reasons we've outlined above around clarity and efficiency, this isn't going to work. Either getting your team to do things better, or communication more clearly exactly what your team are doing and what the issues are with what they're doing isn't going to make your life easier, it's most probably going to result in more trouble.
Here's an example. I was once the ScrumMaster on a team that had been going for a good long while and had really started to hit its stride. All the way through the process of the team hitting its stride, disappointment had been expressed by my boss and by the product owners at the speed of development, at how long things were taking. So I was very happy to come along to a progress meeting and state that we were making good progress, that we were even getting through the work faster than we'd anticipated.
Without missing a beat, my boss said "Oh -- that's great -- maybe then we can take some people off of your team and put them on this other team that's struggling." It didn't take a second. Success didn't breed success, it bred breakdown and hobbling to some "acceptable level" of failure.
Competence won't save you. Getting better at the job won't save you. As Lester says in rather brutally Freudian terms in "The Wire", "The Job won't save you, it won't fill your ass up." Learning to profit from success and from failure is the only thing that might save you. That's what the sociopaths come to understand.
My other big realisation. This was the big "Aha" moment of last summer, was that organisations would never progress if it were down to the "Losers" -- the people who put their faith in competence. In order for an organisation to progress, we need sociopaths, and their ability to just shrug their shoulders and say "fuck it". We need their willingness to create moral hazard, it is what makes the world go round. They can do this for three reasons.
- They know that they aren't competent and competence won't save them
- They know that they can blame failure on someone else (those dummies who are foolish enough to think that success and failure are down to competence).
- They know that when things do succeed, they can take the credit for success.
And how do the sociopaths get their ridiculously risky ventures road? Ventures which no one in their right mind would ever fund or commit to? By clever manipulation of commitment and consistency (encouraging the Losers in their folly) and clarity (making things look like a much surer thing than they are -- this is what plans are for).
Calendar time versus event time
This is a hazy notion that I need to develop a lot more. And part of developing it would be coming up with a better vocabulary for talking about it.
Clock time ticks, and costs. Event time is annoyingly silent, but delivers value. A lot of project management is taken up with totally pointless arguments about clock time. Exactly how long is something going to take? Let's "move heaven and earth" (i.e. let's make a bunch of stupid decisions based on expedience that we'll regret later) to meet a deadline.
I got the idea for this section from reading A Geography of Time by Robert Levine.
I think this is really what's behind the whole Lean idea, put simply, don't do anything that doesn't produce value. And then attempt to reduce the amount of calendar time that it takes to do the thing that produces value.
There are a bunch of techniques that you can use to defend yourself against verbal attack. This is bloody hard, but it's worth mentioning the literature that's out there (there isn't much of it, and it's fairly vague) and it's also worth taking those books as inspiration and developing games and exercises that improve our facility in this area.
One way of thinking about this, is that conversational self-defence, especially keeping your cool when you're being attacked, is certainly part of what sociopaths are doing. As Suzette Haden-Elgin says in her book "Verbal Self-defence" one of the things that "men" often get in a work environment (but by men, I think she means "sociopathic men") is that it's just a game. They don't take it personally.
It's always useful to know that there are a bunch of alternatives to attack (which I suppose is what losing your temper constitutes). As Terry Dobson points out, you can block. You can run away! You can deceive and you can blend.
There are also some useful ideas in George J Thompson's book "Verbal Judo", but they're not fully developed. Not only would I love to develop a full two-day course in this, because I think it would be immensely useful to lots of different people, I'd love to be a real guru at the skills that such a course would teach. I think a lot of the time I'm at the "unconscious incompetence" stage with verbal self-defence. And it would even be a step to get to the "conscious incompetence" stage.
Confusion to our Enemies - the Buttonhook Turn and the OODA Loop
It's like the old joke. You don't have to "outrun the bear", you just have to outrun anybody else who's running from the bear. This is one of the problems with the concept of velocity in Agile methods. When senior managers hear that a team's velocity is 10, they don't say "great, now we finally have an idea about how long this project is going to take." They don't say "Hmm, I wonder how we could look at continuously improving the environment of the project so that we could make it 11 or 12." No. What they say is "why can't it be 20?" or "It has to double so that I can make this deadline." This is the idea of Boyd and the Buttonhook turn. What's important about a fighter plane is that it can turn inside the turning circle of other fighter planes. Actually, Boyd made this into a much more sophisticated concept involving energy and thermodynamics. What's important about a plane is that it can generate kinetic energy and then lose kinetic energy quicker than its opponents.
It's tempting to think there might be a lot of mileage in thinking about organisations thermodynamically. I can't help thinking that that's the problem with trying to "improve the efficiency" of a development team -- where does that free energy go? In a big organisation it just gets dissipated into the system as low-grade heat.
Crazy Idea number 1: Help from Improv
Some things that I've learnt from Improv.
Improv games are powerful tools, for teaching and for practice. Games demonstrate beliefs and behaviour. Games are an almost facile way of getting you to use a different part of your brain.
People don't want to be changed -- even if they say they do, and think they do consciously, their subconscious will do almost anything to prevent them from being changed.
Crazy Idea number 2: Inspiration from Magic.
This passage has been haunting me:
When magicians are good at their jobs it is because they anticipate the way an audience thinks. They are able to suggest a series of clues that guide the audience to the deception. Great magicians don't leave audience thought patterns to chance; they depend on the audience to bring something to the table -- preconceptions or assumptions that can be naturally exploited. That's why, despite what people think, children are often bad audiences for magic. They have little experience and make few assumptions. They might not take for granted that holding your hand in a certain way indicates that it's empty, or that walking around a table in a certain manner indicates that it's just a normal piece of furniture. A magic show is built on these tiny allowances -- children may not grant them Alternatively, anyone with a firm system of beliefs, anyone who has been forced to categorise or analyse information, is ripe for skilful deception. This is why there are famous examples of learned men of science being badly fooled by the simple tricks of fake physics.
What really bugs me about this that I know how "the audience" i.e. senior managers are going to react to a project. One of the clues I suppose is this notion of a "reaction". It's a reaction, it's not a considered response. So why aren't I better at supplying them with a better trick/performance?