Simpler isn't actually better
One of the most challenging aspects of technical leadership, or any leadership for that matter, is managing complexity. People, plans, markets, and software systems, just to name a few, are all wildly complex things. Finding ways to separate the signal from the noise, set reasonable expectations, and observe results can be daunting.
We’re admonished to “keep it simple, stupid,” and given heaps of tools that promise to reduce any complexity to its essential, manageable bits. Tracking systems, dashboards, dependency maps, OKRs, and box-and-whisker plots abound.
Yet history has taught us that seeking simplicity, while often well-intentioned, can be disastrous. To be the nimble and effective leader who can manage through complexity with ease, you’ll need to do one thing that many leaders never learn to do: let go.
There is no better example of the danger of abstraction than the 2008 U.S. financial crisis. In the early 2000s, major financial institutions began adopting what is called a “Gaussian copula function” developed by a mathematics whiz named David X. Li1 to roll up the complexity of mortgage-backed securities into a single risk value.
As we now know, that risk assessment was fatally flawed. It worked well enough for bankers to get wildly rich selling these new securities they created, but it abstracted away the mercurial nature of the financial correlations the ratings were based on.
Ultimately, those in-built assumptions were upended by reality, the U.S. housing market collapsed, millions of people lost their jobs, defaulted on their mortgages, and the country as a whole suffered the effects for years.
What seemed like a useful abstraction turned out to be an Achilles’ heel of the entire United States banking system.
We humans love to reduce things to abstractions. We do it without even thinking about it. Professor Max Bazerman from the Harvard Business School shared this story in his book “Complicit: How We Enable the Unethical and How to Stop”:
I was teaching this group of executives online. I asked the class, “What caused the massive fraud at Theranos?” Each executive was asked to enter their answer in the chat function. The time allowed was generous.
Clearly, there were multiple causes of fraud at Theranos, but when I looked at the class’s answers, I found they tended to be simple and singular. 62 of the 70 responses offered a single cause. 56 of those 62 were a simple description of Elizabeth Holmes, such as her ego or lack of integrity.
None of those single answers is wrong by itself. You probably wouldn’t be wrong to characterize the whole scandal as having been caused by any one of those things. Where we go wrong is in adopting such a characterization as our sole view of reality.
I call it “The Dashboard Effect.” Once you have a credible aggregation or trend in front of you, it’s very easy to overlook its derivation, and act as though that one view of the world is the only view of the world.
We’re automatically driven to this by the fact that the world is complex, and its complexity is often unmanageable in raw form. Humans seek abstractions; we seek ways to comprehend what we’re looking at.
As a leader, one of your chief exports is clarity. What will it require of you to acknowledge, or even to embrace, uncertainty while shipping assuredness and poise to your team every day?
In complex situations, the people who are best equipped to navigate unpredictability and nuance are the ones who are closest to the action. In software development, that’s the people writing the code and the people talking to the customer.
Empower those people to navigate the situation by explicitly not managing it yourself. Provide a crystal clear intent for the action of the team. Provide answers to questions like “What does success look like?”, “How will we recognize it?”
In the book “The Art of Action,” Stephen Bungay examines how the Prussian army originally invented, and successfully implemented, this more “hands-off” approach to leadership in the 1700s, and how it won them numerous wars. It is now in use by militaries around the world.
One reason that this way of leading through complexity and uncertainty isn’t more widely adopted is that leaders find it impossibly hard to stop themselves from getting directly involved in any problem. The less things are going to plan, the more leaders try to intervene, measure, and control.
The key takeaway here is that we can’t predict complex markets, interpersonal situations, or team dynamics. The best we can do is to relentlessly, and as clearly as possible, describe the outcomes we want and give people agency and trust to act based on the local terrain they encounter.
Listen, and ask more questions
It has been shown that the most effective leaders listen at least 80% of the time. The most successful founders and CEOs in the world would tell you that they don’t know all the answers, but they know how to ask the right questions.
When you hear your ego whispering “I can solve this,” or “I know what to do here,” take a pause. You know a lot, and your desire to help is valuable! Ask some questions first, to make sure you really understand what’s going on. Don’t be a “seagull manager” who swoops in, shits all over everything, and flies away.
Get curious about what your teammate is trying to achieve. What seems to be the problem? What would a clean resolution look like? What have you tried so far? How does resolving this issue align with the intent we’ve previously laid out for the team? What are a couple of experiments you’d like to try? How would you assess the outcome of those experiments?
Every single one of those questions hands the full responsibility and agency over the situation to your team, and demonstrates your trust in them to find the right way forward.
Only in the most extreme cases should you lean in to create process, controls, measures, or take actions directly. Regard your direct intervention in the team’s operation as a vaguely insulting last resort.
Some questions for you
What aggregation or abstraction can you choose to let go of this week?
How do you communicate your clear intent to your team?
When something isn’t going right, what will it take for you to stay curious for a bit longer?
Was this useful? Want more? You know what to do!subscribe
What would you create, if you knew you couldn't fail? I help engineering leaders achieve their impossible dreams. Learn more here.