Learning From Failure #2

A little over a year ago, I published my first post on this topic.

This second post was triggered by the following First Round Review article, covering the work of Dave Zwieback:

This is How Effective Leaders Move Beyond Blame

The article itself is a mixed bag. Some parts cover topics that are fairly trivial, but one part stood out clearly above the rest, and was good enough in and of itself to justify a post. It’s the part that outlines the core principles behind Dave’s approach to learning from failures and conducting “learning reviews” (which is a much better name for postmortems):

  1. The purpose of the learning review is to learn so that we can improve our systems and organizations. No one will be blamed, shamed, demoted, fired, or punished in any way for providing a full account of what happened. Going beyond blame and punishment is the only way to gather full accounts of what happened—to fully hold people accountable.
  2. We’re likely working within complex, adaptive systems, and thus cannot apply the simplistic, linear, cause-and-effect models to investigating trouble within such systems. (See A Leader’s Framework for Decision Making by David J. Snowden and Mary E. Boone)
  3. Failure is a normal part of the functioning of complex systems. All systems fail—it’s just a matter of time. (See How Complex Systems Fail by Richard I. Cook, MD.)
  4. Weseek not only to understand the few things that go wrong, but also the many things that go right, in order to make our systems more resilient. (See From Safety-I to Safety-II: A White Paper by Erik Hollnagel, et al.)
  5. The root cause for both the functioning and malfunctions in all complex systems is impermanence (i.e., the fact that all systems are changeable by nature). Knowing the root cause, we no longer seek it, and instead look for the many conditions that allowed a particular situation to manifest. We accept that not all conditions are knowable or fixable.
  6. Human error is a symptom—never the cause—of trouble deeper within the system (e.g., the organization). We accept that no person wants to do a bad job, and we reject the “few bad apples” theory. We seek to understand why it made sense for people to do what they did, given the information they had at the time. (From The Field Guide to Understanding Human Error by Sidney Dekker)
  7. While conducting the learning review, we will fall under the influence of cognitive biases. The most common ones are hindsight, outcome, and availability biases; and fundamental attribution error. We may not notice that we’re under the influence, so we request help from participants in becoming aware of biases during the review. (Read Thinking, Fast and Slowby Daniel Kahneman)

The beauty of these principles is that they synthesize ideas from both systems thinking and behavioral psychology into a set of guidelines for how to operate in a complex-adaptive human system (aka “organization”). They are able to avoid the typical failure modes of traditional “root causes analyses” and “5 whys” exercises by calling out the root cause in advance and focusing the the exploration on understanding the conditions, and preventing “human error” from being anything but a symptom to a much deeper problem in the system.

Learning From Failure #2

Doing Things that REALLY Don’t Scale

In a previous post I suggested that I may develop some of my key take-aways from the WBW posts about Elon Musk and his companies into separate blog posts. This is one of them.

The credit for getting the term “do things that don’t scale” into the Silicon Valley jargon belongs to Paul Graham, the Y-Combinator co-founder, who popularized it in one of his most notable and frequently-referenced blog post.

Paul’s post focuses mostly on tactics. On doing things that don’t scale in the way that you do them, not in the purpose of the things themselves. It focuses on the “how” no the “what”: recruit users in a way that’s not scalable, over-weigh feedback and be extremely responsive, provide a level of customer service that will be impossible to maintain when you get bigger, do things manually that eventually will be impossible to execute without automation. Paul advocates for scrappiness, which is a critical mindset for an early stage company. He argues that doing the right thing is far more critical than doing that thing right in the early stages. And he is absolutely right.

But in the Tesla story, the WBW folks tell us about a different type of “doing things that don’t scale”, a type that pertains not to the “how” but to the “what”. Things that REALLY don’t scale. It is beautifully illustrated through what they refer to as the Hershey’s Kiss business model:


Step 1: High-priced, low-volume car for the super rich. Come out with the expensive first product, but make the car so fancy that it’s worth that price—i.e. just make it a legit Ferrari competitor and then it’s okay to charge over $100,000 for it.

Step 2: Mid-priced, mid-volume car for the pretty rich. Use the profits from Step 1 to develop the Step 2 car. It would still be expensive, but more like a $75,000 Mercedes or BMW competitor instead of Ferrari.

Step 3: Low-priced, high-volume car for the masses. Use the profits from Step 2 to develop a $35,000-ish car that, after the government’s $7,500 EV tax credit and the savings on gas, would be affordable to the middle class.

Delivering on the company’s mission: “to accelerate the advent of sustainable transport by bringing compelling mass market electric cars to market as soon as possible.” will take years if not decades and a massive amount of funds. The only way to get there is through different types of products targeting different types of markets. The thing that doesn’t scale here is the strategy, not the tactics – it’s the business model, both in terms of the target market, and the product that’s being used to address its needs. But in the process, every phase provides enough revenue, and enough learning to drive the next step/phase in the evolution.

SpaceX is using the same approach: figure out how to put things into space ==> Revolutionize the cost of space travel ==> Colonize Mars

Uber is doing the same, but with slightly lower stakes and on a much accelerated timeline: Uber Black ==> Uber X ==> Uber POOL

I can imagine a similar trajectory for AltSchool: run our own private schools ==> provide services, software and expertise to other independent schools ==> transform public schools

Now THAT is doing things that REALLY don’t scale. Perhaps one of the only long-term strategies for driving change in industries where both a massive amount of time and a massive amount of money will be required.

Doing Things that REALLY Don’t Scale

Shorter Feedback is not Always Bett

If the only good thing that ever comes out of Facebook Notes, is more insight into Kent Beck‘s head – I’d still say that Facebook made a worthwhile investment.
It’s a short read that is best summarized by this quote:
Improvements in controlling systems have outstripped improvements in observation. Under the influence of “shorter is better”, the temptation is to continue to shorten the control loop, but that results in a loss of control.
At the macro level, many of us are working to change entrenched, slow-moving systems. We fight Inertia, our arch-enemy with experimentation, continuous improvement and iteration. But our ability to fight it effectively depends not only on how quickly we can make changes, but also on how quickly we can observe their impact. Ignoring the latter constraint can easily lead to chaos.
More specifically in the “people domain”, this observation highlights some fundamental challenges with the growing trend of abandoning traditional annual/bi-annual/quarterly “performance reviews” and replacing them with a “continuous feedback” scheme. It suggests that the latter may miss out on valuable patterns and issues that require longer observation and reflection time in order to detect.
Shorter Feedback is not Always Bett

From Purpose to Leadership and Back Again

It’s hard to ramp-up on Simon Wardley‘s thinking. I’m far from being an expert myself, but I’m slowly working my way towards that goal and find good value in the things I learn along the way.

At the core of Wardley’s thesis is a tool that he developed called a Wardley Map. But this is not the topic of this post. The topic of this post is a more recent piece by Wardley, putting the maps in the broader context of leadership and strategy:

From  Purpose to Leadership and Back Again

The following image outlines that content in the post:


  1. Purpose – your scope the moral imperative. The broader value chain that your organization contributes to. This not purely an input, as it may change dramatically based on the decisions you make
  2. Landscape (situational awareness) – understanding the landscape requires a map. Maps consist of three element: an anchor (chess board, compass), positions of the pieces  relative to the anchor (typically using some coordinate system), and movement (where pieces have been and where they can go). A Wardley map is used to understand a business landscape. User needs serve as the anchor; relative position is given through a chain of needs anchored to the user; and movement is determined through an evolution axis – another concept that’s worth further exploration.
  3. Climate – the rules of the game. Understanding the climate allows you to anticipate what will happen next: “pressing play” on the static map of how things are right now, and doing a reasonable job anticipating how it’ll dynamically change in the future. Wardley covers 13 such rules in the article.
  4. Doctrine – doctrine helps you decide how to organize around the map. This is meant to be context-free doctrine – one the applies in almost all situations regardless of context. Wardley covers 7 doctrine principles in the article.
  5. Leadership – The ability to anticipate movement on the map reveals multiple points of attack. Now context-specific gameplay comes into play and helps you choose the best option of this set of “where’s”.

That’s the key about strategy, the why is relative as in why here over there. You never start strategy with why, you always start with WHERE.

Mata-comment: this is the 2nd post in a row in which the authors referenced the work of John Boyd as an inspiration for their own thinking. Which probably means that reading his biography will move up in my reading queue.

From Purpose to Leadership and Back Again