Wrapped under the story of the Afghanistan war, the book ‘Team of Teams’ was a writing about 2 very different approaches of teams. One that strives for efficiency, and the other that strives for adaptability. As system problems become more unpredictable, the author finds himself limited by the ineffectiveness of the efficient structure that has worked for decades.
Despite equiped with the best resources, intelligence, communication, the US are losing against threats of dense interdependency that’s prominent in today’s world. They had to find ways to adapt in a chaotic environment, with speed and self-organization that wasn’t previously possible. His approach to the concept of ‘Team of Teams’ gave me new insights on reasons behind organization structures in the past, present, and how it could be in the future.
Huge respect for the author, not just for his ideas, but his reflection as a leader. I imagine that it was no easy feat to have the courage to push change on a military structure that has worked out for decades. Above being a great manager’s book I think it teaches a lot about cultivating the right culture, leadership, team member, foster collaboration, parenting, etc.
Summary & Thoughts
** Takeaway **: Shared consciousness, trust, and empowered execution is key to allowing a management style that mimics an organism, one that self-organize in a network stronger than sum of its parts.
** Complicated vs complexity notion **: Outside the classroom, complexity forces you to make trade-offs and adapt. The assumptions in our theory and models could be useless or a hindrance under complexity (unless you update them in perfect real time). Accurate prediction is impossible in this environment because even small uncertainties, can percolate into massive events through interconnections. What you want is to build resilience instead of defending against each symptoms (offence + resilience > defence) (reconfiguring > predicting).
** Tradeoff **: The book challenged me to discard mental models, but it made me ponder a lot about whether old top-down management and illusion of control was necessarily bad. It has brought industrialization a long way and surely many non-tech industries today still thrives on efficiency. After all, companies run under limited resources it’s difficult to scale and foster trust without having some sort of beaurocracy. As much as I like the world to move towards ‘team of teams’, it’s a reminder that Trust don’t come for free and this cost is essential in enabling the idea. Is the environment interdependent enough that one needs to cooperate to survive? That’s where team of teams will find its worth.
The Rising Tide Raises All Ships
In resrospective, I think my learning method often optimizes for efficiency (cut down time/energy). But collaboration builds something more adaptive. Perhaps it’s the same wisdom as ‘If you want to go fast, go alone. If you want to go far, go together" proverb.
But I do hope that just as the author envisioned, more problems will arise from the interdependency of the increasingly chaotic landscape we live in and we’re likely to see new solutions that embrace collaboration, instead of structurally isolated teams. We will live amongst the graph not a top-down tree.
Complexity: when things are no longer predictable
The book made a clear distinction between 'complicated’ vs ‘complex’ systems. ‘Complicated’ problems can be broken apart because they’re independent / loosely coupled. But being ‘complex’ is about the dense interconnection of parts that fluctuate and highly unpredictable. A good analogy of complexity: a break shot at billiards or the ‘butterfly effect’ concept. Unlike the ideal world where you can model a closed system, small deviations and interactions percolate in nonlinear ways and change the entire game. It’s a shape-shifting network of many direct and indirect links that changes the prediction, and its impossible to know the right lever to pull ahead of time.
Some properties of complex systems: - No perfect or permanent solution exist. - Planning is hard because env and solution keeps on changing. - Well-timed solutions > brilliant but expensive solution that makes assumption about future risks (adaptation > mitigation)
A digression from the book, the internet has brought me to a great framework idea called the Cynefin. One that classifies situations into 4 (Simple, Complicated, Complex, and Chaotic). And there are guidelines on how to respond to each of those situations.
Problems with efficiency (under complex env)
1. Hierarchy creates friction and limit adaptability
Although rules and constraints in the chain of commands had kept privacy and prevented errors, it also inhibits information exchanges and creativity that’s needed to fight complexity. Having a centralized person to predict, monitor, control, takes too much time in dynamic environments.
Dividing the task mutual exclusively (MECE structure) decouples the groups. It’s neat, but there’s lack of coherence with teams working in silos. They fight their own battles and requires an orchestrator. In soccer terms, the players are only protecting their patch of grass. Trust breaks down and egos arise (tribal competitiveness).
Misinformation between teams requires a long process to correct. Issues were flagged but not raised/addressed/prioritized because systems cut across teams, and no one knows how to handle it, take responsibility, or want to cause inefficiency (the value that feeds job security). There’s absence of the big picture vision shared between teams.
2. Planning / procedures can be fatal
Plans keep us efficient but under unfamiliar situations that could lead you down to ruin. Check-list/contigency-plans adds noise and paperwork to the process and only keeps on growing. Whereas there’s thousands of potential errors, so in emergency having one person keeping tab of everything or overemphasis of one procedure presents new blind spots. You can’t control chess pieces alone because in this game, the enemy could move multiple pieces simultaneously.
3. There’s no permanent solution
Adaptation > Mitigation. Mitigation tries to fix the root cause for future events (make assumptions), whereas adaptation is ability to respond to env. Well-timed solutions is better than brilliant expensive solutions that’s hard to reverse. So if anything, stick with simple solution.
Proposed solution: Team of Teams
1. Decentralized decision-making means speed and context
When things don’t go as planned, teams can adapt/communicate/react/improvise quickly in real-time. Just like ants, small interactions cause emergence of new behaviors without the need for direction/plan. The real strategy was in the structure and not in the plan.
2. Trust, Purpose & Context makes a strong team
Like a soccer team, every member should view the entire field (context), not just their close proximity, to make the right decision. There’s horizontal trust and everyone’s necessary for survival. Since tech outpaces the expert, there’s more need for team players. Juniors need to be more assertive and captains less forceful.
3. Team of teams: scaling trust and purpose
Interconnectedness allows small groups to thrive in complexity (i.e. start-up, terrorist, youtubers, chatrooms). The more you grow in size, the harder to adapt because the required links grow exponentially.
Unlike hierarchical siloed teams, a network is meant to be messy with some overlapping responsibilities and context of entire playing field. This is inefficient, big time waster, redundant, but that’s precisely how adaptive organisms work.
Thus, came in the big idea: “team of teams” (not one massive team). Everyone simply needs to know at least 1 person on every other team, to be aware and cooperate with other teams that are affected by the ‘butterfly-effect’ / ‘inter-dependencies (systems thinking)’ of the your team’s work.
** Education > Training **: Education is understanding relationships so you can be resilient against unexpected surprises, while training is memorizing singular actions.
4. Space for dialog
Office spaces with cubicals reflects a paradigm that reducing distraction and placing fear of being monitored = productive. However, this design of space not only reinforce this behaviorial paradigm, it kills idea sharing critical for innovation and creativity.
Modern firms are beginning to design spaces that intentionally allow people to bump each other. The lack of walls sparks collaboration, prompts new insights, and new work habits that proactively gets shit done. It’s a way of letting go and allowing information to flow out of your control (because you don’t know which connection/conversation has value).
Leadership is like gardening
There are situations when leaders don’t have the most context and is only an approval stamp that slows processes (in a world that’s faster). In soccer, it’s like asking the coach for permission to pass the ball. Leaders normally default to control whenever they can, but don’t provide much value (i.e. Historically navy has more power than army on land because of communication barrier). Instead, communicate the thought process and let anyone make the call. Let information flow down and decision made at the bottom (used to be reversed).
For e.g., rather than drilling requirements, Ritz. Carlton empower freedom for employees to spend $2k to satisfy guests at their discretion, and own their work / complaints. Psychologically it’s a different experience and you’re more invested seeing your decision’s impact.
Role of a leader is prioritization and coordination. Not puppet master but a empathetic crafter of culture. The gardener only fosters the environment, put the upfront work and maintenance, but letting each plants flourish on their own. Focus on shaping ecosystem rather than moving chess pieces.
On communication, most powerful instrument is one’s behavior and leading by example. Reflected on both thoughts and ‘voice’. Let their work and words matter. Openly admitting you don’t know earns respect than to pretend. Asking opinions/advice shows respect. Allow team members to help the leader help the team.
Gardening maintenance - A good visit to a subordinate accomplishes understanding, communicate guidance, and lead/inspire. For example, having agenda and thoughtful questions to fill in the gaps in mental model, or to ignite long term thinking (i.e. what would you do differently if …?, what are you thinking of X?). However, a bad visit leaves confusion and demoralization (i.e. inattentive from fatigue, cancelled visits).
Enforcing updates: Context must be provided back to the leader. If the operation reports are transparent, leaders don’t have to cut it open and intervene (eyes on hands off).
Trust & collaboration between teams:
Ideally, in an urgent moment, we want one side to be able to urge another ‘trust me’, while the others will make decisions on the fly based on passed information and ground-level interactions without hesitation. Supervisor spectates.
Prisonner’s Dillema - Everyone in the tables came with history and a particular way of viewing the problem, each had value but none could succeed in isolation. It’s not just an exchange, but a ‘relationship’. Both prisoners must have holistic view of each cards, for them to know that they’re on the same boat. You’re more likely to share assets knowing how it’s being used.
Some strategies: 1. Shared consciousness - townhalls or info summary anyone can subscribe to. 2. Exchange program - place a representative (superstar, like an embassy) to another team for 6 months to see war from the other side, smooth out communication, learn each team’s strengths, and expand comradery & trust naturally. 3. Empowered execution - decentralizing decision making
Trade-off: Efficiency vs adaptability in teams
Efficiency is undeniably important. It saves time in life/death situations, soldiers trained to be replaceable part of the system, placing predictability and order into a chaotic environment. The notion of reductionism is the need to optimize few parameters in order to shave off time, effort, resources (increase output, precision and productivity with less). To hire more untrained workers fast. To easily promote alignment and cooperation if everyone put their trust in the system.
However, while breaking it down and optimizing parts to cut inefficiencies, you turn the job into mindless cogs in an assembly line. You replace skilled creative workers with thoughtless unquestioning machines. Only the manager think, plan, coordinate, report, etc. They are the strategist of the game overseeing all movements. This works in ‘complicated’ problems, but will be a constant struggle in ‘complex’ domains. Protocols, policies, rules, paperwork, becomes increasingly intricate as to try cover all likelihoods.
In an environment where workers need to constantly take new roles, use unfamiliar tools, this can mean death when the system can’t adapt to new realities (i.e. failure of the Maginot Line defense against enemies that don’t play by their rules). The ‘never again’ mentality often won’t work because it’s impossible to foresee what’s going to happen.
While our strive for efficiency and foresight kinda work, not only did it kill the muscle to self-organize (antifragility), they also invite more unpredictable complexity against the rising tide of more complex threats that came because of it (or around it). For example, tech can be seen as a tool we build to pursue efficiency; but that same improved connectivity also shapes new complexity, shorten the reinforcing feedback delays, rumors spread faster, harder to respond and prioritize events, and the entire game is less deterministic than ever. Perhaps this trade-off was more of human intervention vs nature. The more we try to control, the less we actually do.
The Double-edge sword - what is at risk:
Requirement for holistic view, means extreme transparency (default to open) which exposes vulnerability of sensitive data. - To the author, leaks are a risk they’re willing to take. Benefits (necessary for winning the war) outweighs the cost (leaks), even if cost is more visible. - Value and power of information grows the more it’s shared. - You can’t make assumption on who ‘needs to know’ the information. What seems irrelevant could be relevant. - Withholding information will push teams back to working in silos, as they can’t assemble the full picture. Everyone should commit and release information for other teams to subscribe to. “Share information until you’re afraid it’s illegal”
Lack of context: Sharing raw information before analysis could cause misinterpretations (although it allows less friction to fix/correct things). Furthermore sometimes inaction is preferable than naïve actions. Democracy only works when people have enough context to make good decisions, so education needs to come before it. Similarly, sharing context without empowerment disables them from acting on new insights.
The author’s heuristic was to stick with efficiency if cost of error > cost of delay. Otherwise, prefer team of teams, where it’s better to ship 70% now than 90% tomorrow. To his surprise, it could even get 90% done today as the quality of decisions improve.