Outline:
- Processes have an infinite potential for improvement
- The greatest gains are achieved through rapid iteration
- The best decisions come from people closest to the process
- The world is less knowable than we think
- A clear direction is better than a detailed plan
This document was originally written at the end of 2017 to help me process my thoughts after three years working as a corporate employee in Amazon's Product Imaging Studios. I was self employed for about 8 years, from my junior year in college through 2014 when I moved to the Louisville area to work for Amazon. The transition from running my own small business with a handful of interns and employees, to working for one of the world's largest corporations, was thrilling. It also helped me to discover a great deal about my own core beliefs and values. My manager at Amazon doubled as the studio's Kaizen Promotion Officer. She introduced me to the world of Continuous Improvement and arranged for me to get a certification in Lean/Six Sigma. I had the opportunity to originate and implement several projects that were projected to save the company more than $356,000 annually. These improvements resulted in an 84% increase in productivity. I had always been a tinkerer, but this was the first time I had had the opportunity to see, and measure with precision, the impact of small tweaks on a very large scale operation.
After two years at Amazon, several management changes were made. The new manager insisted that it was not our responsibility to continually tweak our processes. Instead, they suggested that additional productivity gains would come from "cleaning up around the edges" and explained that this meant everyone needed to be more careful about returning quickly from breaks and lunch, and to stay focused on just the tasks each employee was responsible for. The contrast was stark, and the experience helped cement in my mind one of my most firmly held core beliefs.
Taiichi Ohno, father of the Toyota Production System (from which we get Lean Manufacturing, from which we get Agile Development), said,
The Toyota style is not to create results by working hard. It is a system that says there is no limit to people's creativity. People don't go to Toyota to 'work' they go there to 'think'.
Of course we should work hard. Hard work is fun and rewarding! But just working harder has diminishing returns. The biggest gains are not to be had by "cleaning up around the edge", but instead by continually focusing the force of human creativity right through the heart of the process. This is what makes us different from machines. We think about what we're working on, and we get annoyed by the parts that are inefficient or wasteful. It's our limitless creativity that drives us to improve processes and products past what we ever could have imagined.
Jeff Bezos illustrates this well in his 2003 TED talk. Bezos compares the Internet to the early days of the electric industry and the golden age of appliances. The first electric washing machine, first sold around 1908, was typically kept on the porch since it was very messy. It was connected to the electric grid by unscrewing a light bulb inside the house and screwing the end of a long cord into the light socket. An electric motor powered an agitator in the main wash tub and rollers for wringing out the clothes. People were mained or seriously injured when hair and other body parts were caught in the exposed moving parts. It had no off switch, so if something went wrong it had to be unscrewed from the light socket in the house. Some people were even killed when unscrewing the power cord and touching the open electrical contacts inside. This was because houses were not originally wired for "electricity". They were wired for light. When Sears began selling home appliances, one of their marketing taglines read, "Use your electricity for more light!" Of course, today this just sounds ridiculous. However, if we think that 100 years from now, or even just a few years from now, we can describe how we do things today and not think it just sounds ridiculous, we're fooling ourselves. One of my favorite illustrations of this is the comparison of a Formula One pit stop in 1950 and 2013. "I do think there's more innovation ahead of us than there is behind us," Bezos says. Personally, I think this will always be true, because there is always an infinite capacity for improvement.
Rapid interaction with the real world over a beautifully constructed master plan
Amazon has 14 Leadership Principles, and I have a favorite. My favorite is...
Bias for Action
Speed matters in business. Many decisions and actions are reversible and do not need extensive study. We value calculated risk taking.
Amazon leaders refer to this as the two-way door. If a decision or action is easily reversible with minimal impact, it should be taken quickly. I love this principle because of its relationship to learning. We can learn on our way through the door in both directions. We can learn from our successes and our failures. I personally believe we learn more from the failures than the successes.
Failures are critical to progress. In fact, I believe learning cannot occur in the absence of two ingredients: 1) a large number of opportunities to make similar judgments in similar situations, and 2) access to valid feedback. Valid feedback must have the ability to prove you wrong and it must sometimes do so. If it's not possible to find out that you are wrong, then it's not possible to know that you are right.
Faster feedback is more valid, since causality is more likely. Faster feedback leads to faster learning and develops intuition. Compare parking a car to docking a large ferry. Learning to park a car is easy because every movement of the controls is met with instant feedback. Docking a very large boat is difficult, requires lots of practice and usually involves many more tiny adjustments. A driver can learn to park a car intuitively, while docking a large boat will always be effortful. This is because every input to the controls has a long delay before the operator knows whether that input was the right amount. This is why we should favor failing fast, since this narrows the gap between our efforts and the available feedback from those efforts. The speed of feedback is exponential since faster feedback makes each iteration more valuable and also allows for more iterations in the same amount of time.
If valid feedback creates expert intuition, then we should expect to find the most expert intuitions among people who are closest to the feedback. This idea comes from Taiichi Ohno as expressed by Jeffrey Liker in The Toyota Way to Lean Leadership, "The further from the gemba the decisions are made, the poorer the decisions will be." (The gemba is the actual location where the work is being done). Liker continues, "It's hard to believe that a person who is not doing a job on a regular basis can actually be an expert on that job."
This idea is illustrated in a story told by David Bayles and Ted Orland in their book Art and Fear.
The ceramics teacher announced on opening day that he was dividing the class into two groups. All those on the left side of the studio ... would be graded solely on the quantity of work they produced, all those on the right solely on its quality.
...on the final day of class he would ... weigh the work of the “quantity” group: fifty pounds of pots rated an “A”, forty pounds a “B”, and so on. Those being graded on “quality”, however, needed to produce only one pot – albeit a perfect one – to get an “A”.
Well, come grading time and a curious fact emerged: the works of highest quality were all produced by the group being graded for quantity. It seems that while the “quantity” group was busily churning out piles of work – and learning from their mistakes – the “quality” group had sat theorizing about perfection, and in the end had little more to show for their efforts than grandiose theories and a pile of dead clay.
This story could have easily been included in the previous section about learning from mistakes. But here I use it to illustrate that those students who had spent the most time with their hands actually on the clay had gained the most expertise. Taiichi Ohno based his methodology on two pillars, Continuous Improvement and Respect for People. You cannot have one without the other. Why? Because if you don't respect your people, you won't listen to the front-line workers when they suggest ideas for improvement of the processes they use every day. This means underutilizing the most valuable expertise in your company.
Bottom-up over top-down
As companies grow and often lose that intense urgency of beating their competitors to market, they tend to slow down. As they slow down, standardization and bureaucracy grow. Like ice which forms more easily in slow moving water, processes are gradually locked into a static state. The thicker the "ice", the greater the pressure to have everything fit a standard of perfection delivered top-down. Just like the ceramic students, it's tempting to think that if there is just one chance to get it right, one perfect standard, then we must think very strategically about our one perfect way. So we think very hard about what's best, and then we hand it down to those who will follow that standard. I'm tempted to put something about agility vs standardization into a this-over-that format, but that wouldn't be quite right. Instead, I think Taiichi Ohno had it right. First he said, "Without standards, there can be no improvement." But then he added, "Standards should not be forced down from above but rather set by the production workers themselves."
Standards are not the enemy of agility, but the source of those standards can determine the quality of our processes and our ability to continuously improve them. This is a cultural quality that comes from deep respect for the highly refined expertise of those workers whose hands are on the "clay" day after day.
The world is very complex. That was an understatement. For every new element in the system, the number of interactions grows exponentially and so does the probability that we will misunderstand the system. As the number of interactions grows, the level of precision necessary when measuring each interaction also grows exponentially if we expect to make a meaningful prediction about an outcome. To illustrate this point, mathematical physicist Michael Berry performed calculations to predict the movements of billiard balls on a table as they hit each other in sequence. This example comes from Nassim Nicholas Taleb in The Black Swan.
If you know a set of basic parameters concerning the ball at rest, can compute the resistance of the table (quite elementary), and can gauge the strength of the impact, then it is rather easy to predict what would happen at the first hit. The second impact becomes more complicated, but possible; and more precision is called for. The problem is that to correctly compute the ninth impact, you need to take account of the gravitational pull of someone standing next to the table. And to compute the fifty-sixth impact, every single elementary particle in the universe needs to be present in your assumptions! An electron at the edge of the universe, separated from us by 10 billion light-years, must figure in the calculations, since it exerts a meaningful effect on the outcome.
And this is an example with only inanimate objects. Now consider a system with free will agents, such as the market economy. In his 2011 TED Talk Trial, Error and the God Complex, Tim Harford explains that in a typical Walmart, there are approximately 100,000 different products. That's just one store. In a major city like New York, Harford estimates that there are around 10 billion distinct products. That's just the products and doesn't consider the interactions between products or the free will customers.
Clearly, our brains cannot process, or even access, that kind of complexity. Yet somehow we still think to ourselves, "If I can figure out all the details beforehand, I think I can get it right the first time!" Why would we think that? To answer that question, Taleb draws on the work of Daniel Kahneman and Amos Tversky to explain how our brains cope in a world with so much complexity. Our brains store narratives, not data. Narratives are simplified versions of reality. First our brain reduces the complex multidimensional mesh of reality into a linear narrative (a single dimension), then it creates meaning by assigning an arrow of causality along that linear path. This is how our brain wants to store information. Taleb calls this the "disease of dimension reduction". That may be a bit too harsh. In many aspects of our lives these simplified versions of reality serve us quite well. The problem comes in what Taleb calls the narrative fallacy. The narrative fallacy is the belief that, after reducing the past to a simplified straight line narrative, we can then take that straight line and extend it into the future as a prediction of what will happen next. This may work in the more mundane activities of life, but in our most important endeavors we prove repeatedly that humans are terrible at predicting the future. "In spite of the empirical record," Taleb laments, "we continue to project into the future as if we were good at it." We continue to mistake the model in our heads for the actual system. The model is not the system. The map is not the terrain. It is just our best attempt so far to understand an incomprehensible system. "The world in our heads is not a precise replica of reality", asserts Kahneman.
Responding to change over following a plan
If the world is so unknowable, how should we expect to get anywhere? My belief is actually not that the world is unknowable. My belief is that the world is not knowable in advance, or at least not very far in advance. That means we should expect our plans to fail. There will always be flaws in our "pre-market" reasoning.
So how do we get anywhere? By starting to move, sooner rather than later. And by adjusting rapidly. The gains to be had by just starting are greater than those from our well laid plans. Our ability to adjust our plans is more important than the plan itself. This is not the same as being unprepared. Preparation is about understanding what you already know could happen, knowing what can be done about it, and then still expecting to encounter highly improbable and unpredictable events. Then, keep moving.
So what about strategy and long-term planning? I think Jack Welch says it best in his book Winning, "In real life strategy is actually very straightforward. You pick a general direction and implement like hell." Toyota calls this True North and it comes from the clear vision of company leadership. It's the one thing that should be top-down. Once you know where you're going, what matters next is assembling a winning team, one you know won't give up when plans fail, and then giving them the freedom and resources to exercise their zeal in a clearly communicated direction. This means that great leadership is about having a powerful vision and clear communication. Next it's about giving freedom to the individuals on your team, without strict control over how the vision is executed.
When leadership lacks vision and imposes strict control over the details, it's time to go.