Swarm Management: Anybody who has led guilds or raids in World of Warcraft can learn how to lead a swarm. Or, for that matter, most entrepreneurs who have led small-scale teams dependent on trust. In essence, it’s the same social and psychological mechanisms.
Swarmwise chapters – one chapter per month
1. Understanding The Swarm
2. Launching Your Swarm
3. Getting Your Swarm Organized: Herding Cats
4. Control The Vision, But Never The Message
5. Keep Everybody’s Eyes On Target, And Paint It Red Daily (this chapter)
6. Screw Democracy, We’re On A Mission From God (Jul 1)
7. Surviving Growth Unlike Anything The MBAs Have Seen (Aug 1)
8. Using Social Dynamics To Their Potential (Sep 1)
9. Managing Oldmedia (Oct 1)
10. Beyond Success (Nov 1)
The actual book is expected to be available by July 1, 2013.
If I had to pick one skill that was crucial in allowing me to lead the Swedish Pirate Party on its journey from two lines in a chat channel to taking seats in the European Parliament, it would be skills and experience in project management.
This term, project management, is somewhat of a misnomer in this context. When we talk about management, we talk about appointed positions — Dilbertesque pointy-haired bosses, all too often. But good project management is not so much management as it is leadership. Leadership is not an appointed position, like management; leadership is a state of group psychology.
The first time I was trained in the enormous difference between these two concepts was in my officer’s training in the Swedish Army. (I hold the rank of second lieutenant.) Any dolt with pointy hair can be appointed to become a boss in the organizational chart, but in order to lead, you must deserve people’s confidence and trust.
An organization works at its best when these two roles coincide in the same person. When they don’t, the organization works terribly.
This boils down to a breakdown of the concept of responsibility. It consists of two equally important parts — accountability for a certain result, and authority to make that result happen. Accountability and authority must always follow one another as responsibility is delegated.
All too often, you will hear somebody being asked to “take responsibility” for a development gone bad, but what they’re really being asked is to take accountability for something without the corresponding authority. Unfortunately, taking accountability without such corresponding authority is the same thing as taking the blame for events that go wrong outside of your control. Only the most forward and simultaneously naïve people accept such accountability, and sadly they are all too often sacrificed as corporate scapegoats by those with more ruthless ambitions.
The reverse, authority without accountability, is equally bad. You can almost hear Stalin’s maniacal laughter in the background as Eastern Europe was being enslaved when somebody manages to get authority without the accompanying accountability.
The takeaway here is that authority and accountability must always follow each other in the concept of responsibility. Your swarm’s leaders will not have much of either, though, to be honest. They may get responsibility for a small budget as your swarm progresses, matures, and grows, but as we recall, they never get to tell anybody what to do — nobody does.
This is also why, as we discussed in chapter 3, the organizational chart of the swarm’s scaffolding should lag slightly behind the observed reality. You don’t appoint somebody to lead a function — you observe that somebody is already leading a function, voluntarily taking accountability for it, and ask politely whether they would mind if that fact were made formal in an announcement together with the corresponding authority (if any).
Along the same lines, the crucial project management skills that helped me lead the Swedish Pirate Party into the European Parliament were not the skills you’d learn in a project management class — things about gates, schedules, budgets, or stakeholders. It was much more the soft skills that come with experience: how to maintain a group’s motivation, focus, energy, and commitment to deliver.
Incidentally, these were skills I learned as an entrepreneur and a project manager during the dot-com boom of the late 1990s.
I founded my first company at age sixteen and had my first employee at age eighteen, so there was plenty of time to learn. But the environment in the dot-com era was something truly challenging, as people didn’t work for the money.
There was such a shortage of skilled coders, system architects, and designers everywhere you went that people could basically walk into any company and say, “Hello, I would like to work here.” The response from the company would be, “Yes, sir/ma’am — what salary would you like?”
In this environment, where people would literally have a new job before lunch if they felt like leaving their current one at the morning meeting, it was obvious that people didn’t work for the money. People invested their energy, focus, and commitment into changing the world for the better. Having rent and food taken care of was just a necessity ticked off the everyday checklist.
Thus, the psychology of this era — leading companies and projects during the dot-com boom — matches leading a swarm almost to the letter. In swarms, people don’t work for the pay, either (there isn’t any, to begin with), but they invest their energy, focus, and commitment to make the world a better place. Therefore, the leadership styles that work well are pretty much identical.
Of course, this also dispels the myth that you can’t lead a group of volunteers the way you would lead a company. Leadership is exactly the same in both cases. Leadership is psychology, and has very little to do with a paycheck and much more to do with deeply ingrained social wiring in human beings.
When I led the Swedish Pirate Party, I used the exact same skill set I had used as an entrepreneur. And it did take the swarm into the European Parliament, so it’s hard to argue with the results.
PROJECT MANAGEMENT AND SELF-ORGANIZATION
The first time you see self-organization happen, it feels like magic. After having communicated a vision, you can see how the people who listened to you start to self-organize to make your vision happen, without you needing to give directions — or, indeed, interfere.
The trick, then, is how to communicate the vision. If I had to give a quick answer to that question, it would be “with all the passion you can muster, from the depths of your heart, through the fire of your voice and the determination of the depth of your eyes.” You need to be positively radiant with your desire to change the world for the better, and, above all, communicate three values:
— We can do this.
— We are going to change the world for the better.
— This is going to be hard work for us, but totally worth it.
You will notice that we’re talking about changing the world in “we” form. This is crucial. There is never “I need someone to do X,” nor is there ever “You should be doing X.” There is just “We all need X to happen.” You don’t need to point at somebody, or even imply who should do it. Somebody will.
Let’s take these three values one at a time.
We can do this: Part of what energizes a swarm is the realization that the sheer number of activists can make a real difference in the world, and that the task would seem impossible, utterly unattainable, before you came on stage with this crazy idea. It could be changing worldwide policy on a small but important matter, it could be going to Mars, it could be dropkicking an entire archaic industry out of existence with a new, disruptive product or service, it could be solving world famine, illiteracy, or disease. Shoot for no less than the moon! Once you’ve run the numbers as we discussed in chapters 1 and 2, and communicated to the swarm that your insane idea is actually achievable, blue sparks of energy will jolt across the swarm with loud, crackling noises. People will look high from the excitement of being a part of this. Feel high, too.
(You can and should push it even further, by the way. After all, we’ve already been to the moon. Everybody knows that. So shoot for Mars instead! That project would energize people, electrify people. In contrast, you’ll never get a swarm energized around the idea of making the most professional tax audit.)
We are going to change the world for the better: Even if you could get a short-term swarm focus around hate and intolerance, this book is about good causes. Everything you focus on tends to grow, and all your values become organizational values. Therefore, a swarm built on distrust would quickly be devoured from within by its own negative feelings, and collapse, splinter, and fragment into irrelevance. In contrast, the swarms that are the subject of this book aim to go into the history books based on mutual trust to achieve the impossible. The people who devote themselves to the goal of your swarm do so to get a footprint in history. You should communicate that this is exactly what will happen, once the swarm succeeds. (And indeed, getting to Mars would get into the history books, as would eliminating illiteracy.)
This is going to be hard work: One key value you must never falter on is your honesty. You must always communicate the situation of the swarm and its place in the world exactly as you perceive it, even if that means telling people that the swarm has problems or isn’t gaining momentum. (However, you should always think of at least one way out of a bad situation, and communicate that, too — as in we can do this.) The key point here is that people should not think that changing the world for the better is going to be easy or come lightly. You said totally possible. You didn’t say easy.
Once you have communicated this to the swarm, you will start to see people thinking in terms of “how can I help make this happen?” When a couple of thousand activists think like this, magic happens.
Also, it is crucial that you allow the swarm’s scaffolding to keep growing organically. Train your closest officers in swarm methodology and techniques, as described in this book or remixed with your own flavors of style, and help them recruit new officers into the empty boxes that their own box connects to. Your swarm will always grow from the inside out — it can only grow on its edges, a concept we will return to in chapter 8.
This is part of the necessary scaling out.
DRAW THE TIMELINE FOR ALL TO SEE
A key tool in project management is the timeline. Between now and success, you will need to set subgoals to be met that are spaced about eight weeks apart. This may seem like a contradiction to self-organization, but it’s not: you’re telling the swarm the things that need to happen to get from point A to point B. You’re not saying who should be doing what and when.
There are many good reasons to do this. The first, of course, is to back up the initial energy with credibility in the swarm’s ability to deliver:
— Let’s go to Mars!
— Yeah…but, eh, how do you do that, again?
Setting subgoals, or milestones in project lingo, spaced about two months apart on the timeline communicates a path from now to success that not only helps people believe in the swarm, but also helps people choose to do things that are relevant for the current stage of the project. Each subgoal needs to be credible, relevant, achievable, and clearly contributing to the end success. It will also help jell the swarm into crack working teams that perform magic on shoestring budgets (or, more commonly, no budgets at all).
As a tangible example, the first subgoal of the Swedish Pirate Party was registering the party with the Swedish Election Authority. When the party was founded on January 1, 2006, the deadline for registration was eight weeks out. We needed fifteen hundred signatures from identified citizens with voting rights in the imminent elections. This proved to be a perfect task to jell the geographic subgroups: it was a hurdle to clear, there was a deadline, it was doable, and it contributed in a very graspable way to the end success. We arranged a competition between the thirty initial geographic subgroups, where the winners in total count of signatures, as well as the winners in signature count relative to the size of their geography, both would get an original certificate of registration. A silly prize which we paid a small premium for — getting multiple originals of the certificate — but very, very symbolic and worthwhile when you’re building a movement that will change the world.
You will notice that I didn’t tell anybody how to collect those signatures. That’s where swarmthinking kicked in and everybody started sharing his or her experiences in a giant hivemind hellbent on success, not being the slightest bit afraid of learning by trial and error, as we discussed in the previous chapter. One of our best signature collectors at the end of the day was an activist named Christian Engström, who set the benchmark: it was possible to collect twenty signatures per hour if you were out on the streets in midshopping hours. That particular activist is now a Member of the European Parliament.
The second reason you need subgoals about eight weeks apart on a visible, published timeline is to create a sense of urgency. In general, if something is farther out than eight weeks, we don’t care about it at all, it’s just an arbitrary goal in the future. Your vision needs to be broken down into parts that are small enough that everybody can always see a closing goal on the near-term horizon.
I could mention many software projects here by name, projects that started out as two-year projects without such subdivision, and which, it was invariably realized, wouldn’t make it as the deadline approached. Even though it’s water under the bridge, I won’t name those projects by name here — mostly out of courtesy but possibly also due to nasty NDAs — but I’ll share this wisdom of project management:
— How does a project get to be a year late?
— One day at a time.
The key to shipping on schedule at the end of a project is to stay on schedule every day. This doesn’t mean that a failure to adhere to the schedule is a failure of the swarm; rather, you as a project manager should have anticipated possible deviations in both directions from the start and allowed for them in the plan. When making development plans, it is typically prudent to leave 10 percent of the time of every subgoal unallocated for unforeseen events. Only you can know how this translates to your swarm, but the key is to adjust the schedule and the plan every day to account for changes in a fluid reality. You can’t change the events of the past, but you can replan for the future to accommodate for what has already happened.
Every day, you need to make sure that everybody in the swarm can check how far the swarm as a whole has progressed toward the nearest subgoal and toward the end goal. Paint the targets bright red on a daily basis for everybody to see; make all the targets visible and show the progress toward them.
SETTING VISIBLE, ACTIVATING AND INCLUSIVE GOALS
Have you ever played World of Warcraft (or, for that matter, pretty much any modern game)? One thing that catches people’s sense of addiction is that there are always many different paths to choose from for getting a reward of some kind. Looking at World of Warcraft, you can level up (called “dinging” from the sound effect when that happens), you can learn skills, you can explore the map, you can get rich, etc. In Battlefield 3 and similar games, you can get all sorts of achievement awards based on how you play the game. There’s always something to strive for that suits your taste.
This phenomenon, that there’s always some visible, public reward to strive for, no matter your taste, is key to a successful swarm. A lot of this can be achieved by just measuring a lot of things visibly. Anything that you measure in public, people will strive and self-organize to improve without further interference from you.
Let’s take that again, because you probably skimmed it while speed-reading, and it is key to the whole swarm leadership concept:
Anything that you measure in public, people will strive and self-organize to improve.
It’s basically that simple, and that complex. The Swedish Pirate Party posts its liquidity, assets, debts, and donation summaries openly (as many political organizations do now, but not a lot did so in 2006). This leads to people wanting to break new donation records.
Same thing with membership numbers, and in particular their growth rate.
Same thing with response times to mail. Exposure events in oldmedia (TV, radio, newspapers). Mentions on blogs and Twitter. And so on.
(Some people refer to this as gamification, a term that can come across as unnecessarily derogatory. This is not about producing work of low quality because you somehow goof off and think you’re playing games while producing it; rather, it’s about finding ways to engage the reward mechanisms of the brain for doing brilliant work in the same ways that successful video games do.)
Three things emerge as important here. First, the conclusion that things that aren’t measured don’t get handled well, or indeed at all. This is partly true. Some things are fun to do anyway and will get done just because of that — this particularly involves social and creative activities. Routine activities that are the same from day to day require some kind of motivating visible mechanism, or, more efficiently, a competitive element.
Let’s take mail responses as an example. Responding to mail addressed to the swarm at some public request-for-information address is hardly a very visible task, nor is it a very creative one, and yet it is one of the more important ones. Quick response times with proper and correct responses can make or break your swarm once oldmedia decide to try you out. Therefore, this is something we need to pay attention to.
A very working solution to this dilemma is to use internal competitions with silly prizes. (Tangible rewards should rarely be individual in a swarm — always foster teamwork.) Use divisions by geography or some other arbitrary line to create teams that compete against one another in providing helpful answers quickly.
This is the second observation that emerges. If measuring things gets them done (and indeed, there is no upper limit to how many metrics you can or should track publicly), measuring things in internal competitions gets them done even more. As I already mentioned, this is how we jelled the organization in the Swedish Pirate Party right after its founding when collecting signatures for the party’s formal registration. There is a social limit to how many competitions you can have working at a time, which is probably higher than one but lower than five — this is up to you and your swarm to find out.
The third observation is the crucial importance of measuring the right thing. There are many horror stories of people who measure the slightly wrong thing, and therefore end up with terrible results.
The takeaway for this third point is that some things can’t be measured directly, and so you have to find some other thing that you can measure that has an assumed or known correlation to the thing you want to actually measure. Take alcohol consumption, for example. You can’t measure alcohol consumption in a country directly, but you can measure alcohol sales. This was done in Sweden a couple of decades ago, and the authorities responsible for public health rejoiced as alcohol consumption — as it was assumed to be, and published as such — went down steadily, year after year.
Then, somebody in charge discovered that about one-third of what Swedes drink is moonshine vodka, au naturel or spiced into schnapps. (A proud handicraft of our people, I might add.) This was never sold in regulated stores, and therefore never measured. Bureaucrats who live for rules and regulations had been making false assumptions — that people cared in the slightest about what the law said in this aspect — and alcohol consumption had actually increased steadily, leading to bad conclusions and bad policy as a result of bad metrics.
In the software business, the examples of this are too numerous. People who are rewarded for finding bugs is a common example of such Heisenberg metrics.
(Werner Heisenberg was a physicist pioneering quantum mechanics. Quantum mechanics are mind-boggling, a study in masochism to learn, and fortunately quite beyond the scope of this book. The only relevant part here is that, at the quantum level, you can’t measure something without simultaneously changing it. This was not discovered by Heisenberg at all, but phenomena similar to this are named after him anyway as he was a famous quantum physicist who happened to discover something else entirely.)
When somebody is rewarded for finding bugs, then, by definition, you measure when they find bugs and probably make it public, in order to herald the best bug finders as an example to follow. However, the instant you measure this and reward people for it, a portion of the people tasked with finding bugs will split that reward with developers who introduce bugs and tell them where to look. Therefore, measuring the state of the swarm can change it completely in the measured aspect, if done wrong.
“I’m going to code me a new minivan this afternoon.” —Wally, from the “Dilbert” comic
This is a typical example of Heisenberg metrics. There was a similar effect with the site mp3.com, which was a pioneer in the music-in-the-cloud business. (They were so much a pioneer, in fact, that the copyright industry sued them out of existence, bought the remnants for scrap value, and closed them down.) They had this experiment in 2001 called pay-for-play where an artist would get a piece of the site’s revenue, shared between the artists on mp3.com according to how much they were played on the site.
Thus, a fixed portion of money was to be distributed to the artists of mp3.com, according to objective metrics of their popularity, as measured by number of track plays on the site for a particular artist.
Bzzzzt. Very bad idea. But thanks for playing.
What happened was utterly predictable — everybody wanted to support their favorite artist financially, and therefore set all computers they could access to play music from that particular artist from the site mp3.com, but with the volume turned off as to not disturb anybody. Some people coded playbots that would repeatedly stream an artist’s music to boost artificial numbers that translated into money. Heisenberg metrics.
At the end of the day, the conclusion here is that you not only need to visualize the progress toward the nearest subgoal and the end goal of the swarm, but many other metrics as well that indicate the overall health of the swarm’s performance. You should pay particular attention to the fact that as you increase the number of metrics visualized, the tasks that don’t get measured at all will get less priority. Some of them may be important.
DIFFERENT LEADERSHIP STYLES FOR DIFFERENT PHASES
Group psychology and individual proficiency of tasks mature as they gain experience. In different phases of group cohesion and individual proficiency, you need to lead in different ways.
Let’s look first at what it takes to train an individual in a new task. It can be something as everyday as handing out flyers, or it can be doing a live debate on CNN or al-Jazeera in front of several million people. The principles are the same, and people can sweat in anxiety before doing either one for the first time.
In general, I find that a model with four leadership styles works well.
These four leadership styles are quite different, and you need to use all of them when leading a swarm, reading each situation and applying the corresponding style. A frequent comparison of these leadership styles is the progression of the narrative in the movie Karate Kid (the original, not the remake), and the combination of these styles and the ability to shift between them has been described as situational leadership.
It is a vital part of the leadership role to personally train those who regard you as their leader.
When somebody is entirely unskilled in an art, you need to give direct, specific, and explicit instructions. Hold his or her hand entirely. At this stage, you need to focus on the actions to take and how to do them properly, rather than explaining their purpose in the greater scheme of things.
— Paint fence. Up, down, up, down. Strong wrist.
— Wax on, wax off.
In a swarm scenario, we observe that direct instructions for donations yield much greater results than vague ones. The more decisions you leave up to the reader when doing a call for donations, the less money you’ll get. For example,
— We are out of flyers. It’s a luxury problem, as we are handing out more than we thought possible, but it is still a problem. Help us! Log onto your bank and transfer 25 euros into account 555-1337-31337 right now, exactly just right now!
will yield a result almost an order of magnitude stronger than this version:
— We’d appreciate if you’d help us fund our handout materials. Please donate any amount you would like to contribute to account 555-1337-31337 at any time in the near future.
The difference in results lies in the very specific instructions. Every degree of uncertainty leads to inaction at this stage. If you make people comfortable with acting, and lower the bar as far as you can for people to take action within their comfort zone, then things will happen just as you instruct.
Same thing with handing out flyers, as we discussed in the last chapter. You need to make sure that every flyer handout is preceded by a very direct and inclusive instruction detailing every part of handing out, like the instructions described in that chapter. This is direct leadership.
The next stage and type of leadership is applied when people have mastered the basic actions, but are getting frustrated over their lack of context. They don’t see the road ahead and don’t feel progress. At this stage, you need to drop the direct handholding leadership and encourage and explain why these actions lead to positive results.
— You’re not teaching me karate! You’re just using me to paint your fence and wax your car!
— Show me: wax on!
The third stage comes when somebody is proficient in the skills needed, but still not in his or her comfort zone. He or she has the skills and the ability to deliver, but just doesn’t know it yet. This makes for yet a third type of leadership, which basically is endless encouragement.
— I’m never going to be any good at this! (makes a backflip from standstill)
Finally, the fourth and final stage is when somebody is self-motivated and self-reliant. At that point, he or she has more or less ascended to be your equal and doesn’t require much in the way of maintenance. The only important thing is that you periodically recognize him or her when he or she walks an extra mile. In this scenario, the one thing to keep in mind is that you recognize an extra mile only when it really is an extra mile — nobody in the fourth stage wants to be commended for performing trivial and routine tasks well.
You need to assess every individual you work with here — you need to assess where everyone is on this scale in his or her specific context. He or she may be in several different places at once if he or she is working in multiple contexts.
To wrap this up, you also need to pay attention to how groups form and mature. Groups, too, will pass through stages.
When new people first meet in a working environment, you can observe them being very polite and friendly with one another. If somebody appears offended, apologies follow immediately. These are symptoms of a group that cannot yet deliver effectively. Politeness is a sign of an inefficient group that hasn’t learned how to work as a team; people are keeping distance.
Over time, as these individuals learn to work together, they also explore where their limits go, and these limits of people’s roles will start to collide and flow into one another. This is when they start fighting between themselves over rules and culture in the group. This is a significant step forward from overfriendly politeness and shows that the group is well on its way to becoming a well-functioning team.
Finally, in the third phase, you see nothing of the clearly marked distances that were there at the outset. A functioning team can be observed by everybody seeming to know what to do without anybody spelling it out; the group has learned how to work together.
(If new people are added to the mix, the group temporarily reverts into determining roles, culture, and boundaries.)
You need to be aware of these group phases in group psychology, and, in particular, you need to know that a small amount of conflict is actually a step of progress. A group that remains polite to each other has not learned to work well together.
We’ll take a closer look at group psychology and the inevitable conflict resolution in the next chapter, as we discuss how to make people feel included and constructive.
This is the fifth chapter of Swarmwise, a book arriving this summer. Did you like it? It’s going to be free to share (it, like this excerpt, is CC-BY-NC), but you can also buy it hardcover.
EXCERPT FROM UPCOMING BOOK
This is a part of the upcoming book Swarmwise, due in the summer of 2013. It is an instruction manual for recruiting and leading tens of thousands of activists on a mission to change the world for the better, without having access to money, resources, or fame. The book is based on Falkvinge’s experiences in leading the Swedish Pirate Party into the European Parliament, starting from nothing, and covers all aspects of leading a swarm of activists into mainstream success.