Expedition Behaviour and Reverse Pyramids ▾

Tuesday, December 3rd 2019

A Heads Up: this post is a hard-to-read jumble of thoughts. More a reference note to my future self, and/or a place where I can point friends to specific sections they may find helpful.’
(unattributed inline quotes are from Seeking Wisdom)

“We understand life backwards but live it forwards” — Kierkegaard

For the past few months I’ve been thinking on how to best think about leadership in startups…or more generally, leadership in any team-based situation where you can expect that ‘the going will get tough’. So I…


…read through The Elephant in the Brain to try and understand how and why people rationalize their behaviors. Specifically, rationalizing self-serving behaviors at the expense of group (global) well-being.

…I read Adam Grant’s Give & Take, after his podcast interview on The Knowledge Project. Hoping to learn how to better-recognize people, and their underlying behaviors.

…I read The Art of Strategy. It’s a Game Theory book which I hoped would help me learn to better design cooperative systems. That is, systems which align self-interest and other-interest without conflict between the two.

I was interested in the idea that these are not (necessarily) opposing ends of a spectrum, but can be indepenent and mutually reenforcing variables:

“This hybrid engine of self-interest and concern for others serves a much wider circle of people than can be reached by self-interest or caring alone.” — Bill Gates

I want to learn to design cooperative systems that may continue in the correct direction, even if/when bad-actors became temporarily involved.

…finally, I’ve been working through Seeking Wisdom for awhile. With the advice to read books at different speeds (per their merit), this book has been the one I’ve read most slowly, of any book. It includes a lot of relevant learning, for this.

…then, books aside, I also spent a good deal of time talking with friends, mentors, fellow-founders, and a CEO coach.

It’s been deep work on a single question.


My motivation in all of this was to: update my understanding and expectations of human self and other-oriented behaviors. With the practical purpose of designing better human systems, and choosing better leaders.

To build an unselfish company, that would teach itself…

“Be warned that if you wish, as I do, to build a society in which individuals cooperate generously and unselfishly towards a common good, you can expect little help from biological nature. Let us try to teach generosity and altruism, because we are born selfish.” — Richard Dawkins, The Selfish Gene

In the past I’ve made mistakes here. Designing systems that led to conflicts in self vs other-oriented behaviors which people exploited, choosing the wrong leaders, or both.

These mistakes hurt the team I led, and equally hurt the leaders whom I should not have put in leadership roles.

Why Deep Work on The Question

When I asked for guidance from people (friends, mentors) outside our company, they quickly put the problem in their mind’s own categories…

“We don’t like uncertainty or the unknown. We need to categorize, classify, organize, and structure the world. It simplifies life…Our brain is wired to perceive before it thinks - to use emotions before reason.” — Seeking Wisdom: From Darwin to Munger

They would either categorize to (1) fit one of their own experiences — “Oh! This is exactly like when our company…” which had little to do with our company. Or by (2) rapidly zooming-in on what they assumed root causes. Which was tough to feel, really covered things…

“We tend to be looking for the root cause of something, but in complexity, there’s no root cause.” — Jennifer Garvey Berger, of Simple Habits for Complex Times

When I looked at this guidance and asked “Could our explanation of the event help us predict future similar events?” the answer with these observers’ root cause explanations was always no.

So I couldn’t put much stock in them…they felt like a delusional sense of clarity, not reality…“Stories may be selected to prove something and may give us a delusional sense of clarity.”

Their root cause might’ve touched edges of truth, but each root cause (and each individual had decided upon a completely different root cause) was too over-simplified to be practical. — Sidenote: 💙 Chris Argyris.

Given these people had only 2nd-hand information about our company neither lead to much real learning. They (graciously) tried to help, but on the outside looking in, they always had too little context to be effective.

So I also tried asking the leaders who I had chosen in the past who didn’t work out. And they had explanations ready for me: “it’s because this…, and it’s because that…”

But in every case these explanations were externally-focused: i.e. “You didn’t do…” and “You did”…or “The team didn’t do…” and “The Team did…”

I never heard an explanation that incorporated any measure of the leaders’ own behaviors, own choices, or own responsibility. Not surprisingly, when the details were shared, team members consistently (strongly) disagreed with the leaders’ externally-focused explanations, targetting others.

Their explanations probably also touched edges of truth, but in the end these explanations were more likely self-defensive than not, to…“deny and distort reality to feel more comfortable, especially when reality threatens our self-interest.”

Somewhere, something was wrong with how I was choosing and/or managing leaders. Everyone’s guidance on this seemed to touch some truth but never wholeformed, solid truths. The feedback was a mess.

What’s worse, everyone’s advice changed over time. Before and after a leader did/didn’t work out, the same people gave completely different feedback.

‘Everything seems stupid when it fails.’ In hindsight, eveything seems obvious. But we should look at earlier decisions in the context of their own time…sometimes we have a hard time separating a true story from a false one. After an event a story is created so that the event makes sense.” — Seeking Wisdom: From Darwin to Munger

They would interview these potential leaders, alongside me. Advocate for them, judging them good leaders. Then, once failed, would forget (or downplay) their earlier advocacy…once reality was clear.

I could get plenty of post-hoc feedback. But I didn’t feel like this guidance was helping me get any better at the future.


The above was all a rambling way of saying that I couldn’t seem to learn much, by simply polling those around me for feedback. People with different interests, different backgrounds, or different contexts gave completely differenty and contradictory feedback.

But I really wanted to actually get better: at (1) choosing future leaders and (2) designing human systems for cooperation (≠ selfishness).

And I wanted to do it without lazily falling back on cynicism, and/or comfortable (but incorrect) explanations. I hoped to maintain an optimistic (but perhaps, more rationally-optimistic) view of human nature. Where the scars of bad experiences would not turn me cynical, but neither would I repeat those experiences.

“We should be careful to get out of an experience only the wisdom in it, and stop there, lest we be like the cat that sits down on a hot stove-lid. She will not sit down on a hot stove-lid again — but also she will never sit down on a cold one anymore.” — Mark Twain

I needed to truly learn something, that would improve actions and decisions in the future. But simply averaging out the conflicting poll-data wouldn’t do this.

So, the only answer was to do the work. To learn what I didn’t know that I didn’t know. So I could pick better leaders, and build better human systems.

⇢ tl;dr on outcome?

First, in the past I believe I put too much emphasis on potential leaders’ words and too little emphasis on their past actions:

“Consider people’s actual accomplishments and past behavior over a long period of time rather than first impressions. Since people leave track records in life, an individual’s paper record if often predictive of future performance and behavior.” — Seeking Wisdom: From Darwin to Munger

I repeatedly chose leaders who said they wanted to be a leader, and who said they wanted to face the challenges, etc. but who really didn’t have any ‘when the going gets tough’-type teamwork history in their past.

In hindsight, that’s become obvious.

I would then support them for quite a long time (these leaders told me as much at the time). But, eventually, after months of supporting this I would lose faith and patience in the leader. Then, faith-lost, slip into Setup to Fail Syndrome.

So things went from starting very well, to ending badly.

What I should’ve done is been more clear with myself that I was propping up a leader who shouldn’t have been there. That would’ve been better for everyone, including the leader who should not have been there.

But instead, the propping-up which was followed by an eventual lost-faith downward spiral, made things worse for all.

…again, anyway…

What follows is not a summary of what I learned. It’s just two isolated ideas that I want to write out from the process, so I can look on them later and sometimes share them with friends.

Expedition Behavior

The first is a new mental model, called Expedition Behavior. I believe this is by far, the best mental model I’ve come across to describe the responsibilities, relationships, and personalities of good leaders and teams in startups.

Most people say “it’s like a marriage” — but I now think that saying “it’s like an expedition” is probably much closer to the truth.

Neil Schulman on The Clymb wrote the best summary of Expedition Behavior I’ve found:

  • Support the Group’s Goals (i.e.; Group-Interest > Self-Interest)
  • Take Care of Yourself (i.e.; Mental/Physical Health, Performance-Shape)
  • Pitch In (i.e.; Do Your Work, Be Responsible)
  • Help Others but Don’t Do Their Work For Them (i.e.; Avoid Over-Dependence)
  • Moderate Potentially Annoying Behaviors (i.e.; duh…)
  • Admit Shortcomings and Correct Them (i.e.; No Rationalizing-Away Responsibility)
  • Support Everyone’s Growth (i.e.; Build Everyone’s Skills)
  • Be “Cow-Like” (i.e.; Don’t Get Too Emotional)
  • Relaxed Awareness (i.e.; Pay Attention, Be Responsible)
  • Be Funny (i.e.; Help Each Other Enjoy It)

Expedition Behavior is about process skills not technical skill. But it is technical skills that I/we too heavily-emphasize when choosing leaders for difficult expeditions (or startups):

“Technical skills are often emphasized at the expense of group relations, particularly in expeditions when the goal is clear and a high level of skill is required. Technical skills, while not enough to ensure safety or goal attainment, are essential and usually determine one’s assignment on an expedition, such as lead climber or medical personnel. Interestingly, it is process skills which emerge as being critical to success as time proceeds and conditions change during a trip.” — Christine M. Cashel, EdD., Group dynamics: implications for successful expeditions

I never ignored the need for leaders’ process skills. It’s just, I did not do a good job of judging them, because I judged their process skills through conversations.

On someone’s words from one (or usually, many) conversations, not by ‘actual accomplishments and past behavior over a long period of time’.

In hindsight, it’s clear that…

“…predicting who has “good” expedition behavior (or EB) can be hard. Expeditions have a way of bringing out both the best and the worst in people; and no matter how many questionnaires your team has filled out, how many pretrip meetings you’ve held, or how many years you’ve all known each other, sometimes things don’t turn out the way you expect them to.” — Dave Anderson and Molly Absolon, NOLS Expedition Planning

The Expedition Behaviour lens has been clarifying. What someone says about themselves in conversation, means very little about how they will act.

I would never appoint an expedition leader who was untested in practice. Just because they said they aspired to leadership? Just because they showed technical-expertise as a great climbers, in their local gym? Never.

But that’s exactly what I often did appointing leaders in startups.

I took an impression of their leadership, responsibility, determination, and other process skills in a conversation about what they thought they wanted to do.

Then I sent our expedition into the wild, with an untested leader at its head. Small wonder it often went poorly.

It’s when the ‘going gets tough’, when the weather turns, when the packs get heavy, when leaders are most needed and untested leaders are finally tested. Unfortunately it’s here, when it’s getting hard, that rationalizations often replace responsibility and leadership in practice, crumbles:

“Poor expedition behavior is a breakdown in human relations caused by selfishness, rationalization, ignorance of personal faults, dodging blame or responsibility, physical weakness, and in extreme cases, not being able to risk one’s own survival to insure that of a companion.” — Paul Petzoldt founder, National Outdoor Leadership School

Because it will be tough:

“…if everyone on an expedition felt as if he or she was doing 110 percent of the work, each was probably doing just enough. Expeditions require a lot of effort. There are logistical chores, long hard days moving from one camp to another, technical obstacles to overcome, camp tasks to perform, and sheer emotional endurance involved in going on a wilderness trip of any length of time.” — Dave Anderson and Molly Absolon, NOLS Expedition Planning

What I need to do as a CEO, is test for this prior to setting out on any expedition. Prior to choosing any leader. To verify process skills in practice rather than in conversation. Never to be satisfied with evidence of technical skills and only aspirations, for process skills. Instead, hard evidence of both in tough, team-based (not solo) situations.

I (think) I better understand this now. So, I hope I can choose (and test for) better leaders on both technical and process skill.

⇢ Next section is a mostly-unrelated thought.

Reverse Pyramids

The second idea is reverse pyramids.

Let’s start with normal pyramids. Those with the pointy-end up. The ones we are all familiar with. It seems we all share a mental model of leadership and hierarchy in organizations that looks something like this:

Typical Organization Pyramid

It’s pervasive and universal. It’s the common way we think and talk about organizations. In practice, a real pyramid (from 1917) might look like:

Organization Pyramid By Tabulating Machine Co., December 1917; Picture by Marcin Wichary - File:Tabulating Machine Co org chart.mw.jpg, CC BY 2.5

The problem is I think these pyramids are upside down. I think the pointy-end of the pyramid should be facing down, not up.

It’s a subtle distinction — but I think it’s important.

I feel like I’ve seen many would-be leaders who seem to view leadership from the perspective of ‘moving up the ladder.’ Who seem to think of it as gaining power, influence, freedom, …or as being recognized (and rewarded) for contributions on ‘lower-rungs’ and rising above their peers, etc.

It’s my perceptions of their actions, and perception of some entitlement. I don’t know that anyone would actually verbalize this. But…

When people think of leadership as ‘moving up’ the pyramid I think this mental model really hurts everyone including the would-be leader. It gives them the wrong expectations, and when reality hits it hurts.

I think reality looks a lot more like this:

Leadership PyramidIf my bad drawing is confusing — leaders have more people standing on their shoulders, as they take on larger leadership roles. It’s a pyramid, but it’s a reverse human pyramid, because leadership is heavy.

Leaders do not gain more power, they accept more responsibility. It’s not that a leader is ‘the best among her peers’ and therefore merit moves her up. It’s a completly different paradigm.

It’s a function of impact (for good or bad), and responsibility. You become a leader because you choose more responsibility for others, and move down the pyramid, putting more weight on your shoulders.

Phil LibinFrom How and Why to Start A Startup - Sam Altman & Dustin Moskovitz

So flip the pyramid upside down to get a better mental-model. It may be less glamorous to think of but it’s more real. It’s a more true reflection of the relationship between impact and responsibility in leadership.

It’s a function of impact (for good or bad), and responsibility. Like Ben said, ‘with great power (or impact) comes great responsibility’.

In the future I hope to do what I can to better-test leaders. I hope my expedition partners see pyramids with the pointy-end down.