Part I: An Interview with James Everingham, Head of Engineering at Instagram
By Pratap Ranade
Quantum Management, Wardialing and Finding Your Passion as a Programmer
Part I: Quantum Management
Enigma: We’re excited to speak with James Everingham today to learn a bit about what makes him tick. James, would you share with readers a brief overview of your background and your current role at Instagram?
James: I've spent over 35 years doing everything from engineering, managing large scale teams, and building my own companies. Prior to Instagram, I worked as the VP of Engineering at Yahoo, CEO and founder of Luminate (acquired by Yahoo), CTO and co-founder of LiveOps, Engineering Director at Tellme (acquired by Microsoft), Engineering Director at Netscape (responsible for the flagship Netscape browser), and engineering and management positions at Oracle, Borland, and Penn State University.
Enigma: Great. Let’s start with a topic you’ve written and spoken about—“Quantum Management.” Hicham (Enigma’s CEO & Co-founder) was particularly inspired by your talk. What do you mean by Quantum Management?
James: It was a metaphor for how management is changing. I wanted to describe what it means to manage for unimaginable outcomes. Quantum Management is an attempt to use quantum mechanics to explain how management should be applied when products are more invention-based than assembly line-based.
Management is a relatively new field—before the industrial revolution, there were no large workforces so you didn’t need management. The closest thing to a large workforce were armies and religions. As a result, most management originated in army theory.
Henry Ford established the assembly line (the “line”) in the 1900s to mass produce automobiles for the American consumer. To bring cost down and reliability up, he knew exactly what he wanted to build—it was an optimization process. This is where management began—the effort to get teams to work together cheaply and efficiently. This lasted through the mid-1900s.
Then Peter Drucker came along with a new theory of management as creative fields like advertising emerged.
Now, we’ve evolved into a state where innovation and invention is our product. Classic management doesn’t work well with that. How do you optimize for innovation? How do you optimize a process when you don’t know what you want to build?
Managing when you know what you want to build looks different than managing when you don’t.
Let’s take a real example—the Manhattan project. This was a research and development undertaking during World War II that produced the first nuclear weapons. It is horrible that this happened, but at the same time, it is the most amazing scientific project ever. It started with one guy—Oppenheimer—who had a theory. He knew we could create a big explosion if we split uranium, but no clue how to get it. A very different problem from Ford’s.
He scaled his project in less than 5 years from 1 to 40,000 people, beginning by hiring the smartest physicists. He focused the team on the outcome and then he got out of their way.
Kennedy did something similar with the space program. He didn’t know anything about space, but he said, “we will put a man on the moon in 10 years.” The first response he got was, “no.” But he asked why, repeatedly, and then just focused the team on the outcome and put a timeline in place.
Vastly different skillsets are required for the best results when we are asking teams to invent.
Vastly different skillsets are required for the best results when we are asking teams to invent. Simple things that we can do while interacting with them can dramatically change results. A good example is: 1) brainstorming your own ideas on whiteboard, and 2) prematurely trying to course correct and engineer that you think is going to fail. You have to be hyper aware of how to interact with the team.
Quantum mechanical principles are great metaphors for how you manage a team. In a classical management paradigm, like Ford’s, your parts are predictable. But with creative teams, like Kennedy’s, they are unpredictable. How do you get meaningful results from unpredictable parts and how do you manage the output?
You’re familiar with Schrödinger’s cat—the observer effect. [For those of you unfamiliar with it: Physicist Erwin Schrödinger proposed a gedanken (thought) experiment, in which you have a cat in a box. There is a vial of poisonous gas that can break depending on the decay of a radioactive isotope (a quantum mechanically uncertain process). At a given point in time, an observer does not know if the atom has decayed and therefore does not know if the cat is alive or dead. Once the observer opens the box, the cat is seen to be either alive or dead. But, due to the laws of quantum mechanics, until the box is open (an observation is made), the cat is in a “superposition of states”—it is both alive and dead.] Classic computing is binary. 0 and 1. On or off. The cat is either dead or alive. But, in a quantum world, things are uncertain—the act of observing a system changes its state. It’s a really good metaphor.
“Observing” is like talking with your team while they are in the box trying to solve a creative problem. “Observing” can kill our metaphorical cat. For example, imagine a manager walks into a team room and asks the team, “where are we with our problem?” The team might say, “we have tried X and Y, but we haven’t made headway yet.” Then a manager might suggest an idea, trying to help the team. But, there is an implied hierarchy here. If you are the manager and you suggest an idea, you have tainted the experiment. The team will now execute that idea. You’ve killed the creativity—you’ve killed the cat.
So, instead of trying to reduce the number of states by sharing an opinion, your goal as a manager should be to ask probing questions and instead try to expand the number of states. Get the team to consider other ways in which they might win. How is time spent? What is the number of users? Develop more and more measures of success. I’m expanding the set of states. It’s like a qubit—it’s neither zero nor one, but it’s in between. This might be failure, or it might be success. As a manager, your job is to give teams more creative fuel—to expand the number of states. This is a challenge for managers.
Consider how most companies work. You have frontline coders—they work on operational tasks. Then you have frontline managers—they define tactics. Finally you have senior managers who set strategy. You see suboptimal results when senior managers ‘sine wave’ down through the layers. They reach down and observe a system, incepting ideas and reducing creativity.
Building great teams for a quantum paradigm
Enigma: For this to work it seems like you’re assuming a few things about the people in these teams. What do you look for in the people you seek to hire?
James: Of course. The first thing you need to understand before you hire people is your values. You have to hire people and hold a high bar for a consistent set of values. At Facebook we are very good at values-based hiring. We look for people who are:
Reflective - open to learning
If I reach out to someone at Facebook, even though they may be an expert on a topic (and they likely are), I know that they will be friendly, that they will follow up, that they will be invested in my success and that they will truly want to work with me. At other companies, approaching an expert or someone you don’t know can be scary. You don’t know what you will get. So values-based hiring is always the first priority. Hiring people who share key values facilitates communication, reduces fear and increases collective intelligence—collective intelligence is super key.
For best results, you want highly collaborative people who want to work as a team. When we
talk, we get better ideas than when we are alone.
For best results, you want highly collaborative people who want to work as a team. When we talk, we get better ideas than when we are alone.
Creativity is also important. I look for people that have solved problems in different ways. What I don’t look for is expertise. This is key. You may want some expertise, but it is not always an asset. There are lots of studies that show that as we become an expert, we see less. On the other hand, people with less expertise but with sheer IQ—they are the ones who innovate.
Enigma: What are the best people looking for in their ideal employer? Are there common themes you’ve seen?
James: They want to learn. They ask about opportunities for learning, not opportunities to climb the corporate ladder. Good people will always push themselves to be a bit uncomfortable. You want people who are looking to stretch themselves. Curious people will ask lots of deep, insightful questions. That’s why it’s always important to leave plenty of time for candidates to ask questions at the end. At a lot of companies, multiple interviewers will say, “he asked the same questions to me… he only had one set of questions prepared.” You have to be careful not to mistake that for a tactic. Don’t forget that they are interviewing you too—they want to find out that everyone has the same answer, they are testing your team for consistency and cohesion.
Enigma: How do you think about entanglement within your organization?
James: People with better relationships with one another perform better. Creating camaraderie is important. Here at Facebook we encourage people to go and do things offsite and do things together personally—it leads to better results. Empathy is an important entanglement. This helps us perform better and build better software.
Enigma: Have you found a way to build it without forcing it?
James: Co-location is important—teams sit together. It is common sense. But most companies will do stuff like make the engineering and design teams sit separately from each other. You shouldn’t have functionally separated teams.
We try to keep the teams that need to work together close together. We also have “WFH Wednesdays,” to let people work from home, but emphasize that on other days we get great benefits by being together. The deeper you dig, the more you realize that people need to work together, physically.