More on this book
Kindle Notes & Highlights
They all support a fundamentally similar approach to getting work done while avoiding sclerosis: give teams a great deal of autonomy, use an alignment process to establish their goals and monitor progress toward them, and then get out of the way and let teams work with high autonomy.
much better strategy is accomplishing your objectives and hitting your key metrics. In a culture with a strong norm of ownership, doing those things will give you prestige and high rank in your work environment.
The dysfunctions and disappointments of the NPI process were instrumental in sparking a 180-degree shift at Amazon away from trying to define and manage dependencies and toward trying to flat out eliminate them. This was a shift away from a culture in which management exercised a great deal of control, toward a culture in which management encouraged high levels of ownership. It led to two-pizza teams, single-threaded leaders, and many other organizational innovations. But before that shift could begin in earnest, Amazon needed to rewrite just about all of its software in a way that had never
...more
So Bezos set out to do something very different. A CEO who, according to former Amazon engineer Steve Yegge, “makes ordinary control freaks look like stoned hippies,” mandated that his company would build a software architecture that devolved control to teams throughout the company, allowing them to access data and software without having to ask anyone for permission or support (and without crashing the entire company). Yegge wrote that this mandate “was so out there, so huge and eye-bulgingly ponderous, that it made all of his other mandates look like unsolicited peer bonuses.” It took
...more
Nadella writes of Microsoft’s senior leadership team (SLT), a small group of the most senior executives in the company, that “debate and argument are essential. Improving upon each other’s ideas is crucial. I wanted people to speak up. ‘Oh, here’s a customer
As he writes, “I found that the key was agility, agility, agility. We needed to develop speed, nimbleness, and athleticism to get the consumer experience right, not just once but daily. We needed to set and repeatedly meet short-term goals, shipping code at a more modern, fast-paced cadence.” He stressed to me that “speed of iteration matters. So one of the fundamental things we do is in any project review, we look at not just what was the input and the output but also the speed of the input-to-output iteration.”
Gervais had a simple but savvy technique for starting to break down this deep defensiveness: he simply got people to talk about themselves—to, as we say, open up. Nadella writes that at the meeting we shared our personal passions and philosophies. We were asked to reflect on who we are, both in our home lives and at work. How do we connect our work persona with our life persona? People talked about spirituality, their Catholic roots, their study of Confucian teachings, they shared their struggles as parents and their unending dedication to making products that people love to use for work and
...more
What is fascinating about this is that is also the roots of failure and having the therapy take over. You need a culture of excellence enabled by vulnerability.
One important move toward one Microsoft was to eliminate existing profit centers (for the Windows operating system, Office application suite, and so on) and replace them with a single corporate profit and loss (P&L) statement.
Nadella’s second major move to restore alignment was to take away from individual groups ownership of important assets like code, data, and data centers. As we’ve discussed, owners of such assets have a strong tendency to gain status by acting as gatekeepers.
She stopped me mid-riff with “You’re missing it, they are actually hungry to do more, but things keep getting in their way.”
actors. It’s instead a fight against a badly configured environment—one that lets us Homo ultrasocialis gain status in ways that aren’t tightly aligned with the goals of the organization.
Fundamental attribution error
As Ford and Sterman looked at projects across a wide range of companies and industries, they kept seeing the same phenomena: late surprises, high uncertainty, and badly missed due dates. A manager they interviewed summarized the grim situation: “The average time to develop a product is 225 percent of the projected time, with a standard deviation of 85 percent. We can tell you how long it will take, plus or minus a project.”
Everyone withheld knowledge that their subsystem was behind schedule. Members of the liar’s club hoped someone else would be forced to admit problems first, forcing the schedule to slip and letting them escape responsibility for their own tardiness. Everyone in the liar’s club knew that everyone was concealing rework requirements and everyone knew that those best able to hide their problems could escape responsibility for the project failing to meet its targets.
We’ll call it the “Are you on time?” (AYOT?) game. A critical feature of this game is that observability is low for most of a project’s timeline; it’s hard to know from the outside whether a given team member is falling behind.
After a lot of discussion, the group settled on a name for this declaration: it would be the Manifesto for Agile Software Development. It was posted online by attendee Ward Cunningham in 2001.4
How to Make (Almost) Anything, which was started in 1998 by Media Lab professor Neil Gershenfeld.
Gershenfeld’s “lectures” are a fast-flowing stream of mini explanations and demos; in other words, a bunch of models. Students quickly start making use of those models. They’re expected to make something every week and be ready to show off their creation in class and talk about how they built it.
All those websites are archived and searchable from the course’s home page. HTMAA students quickly learn that once they have an idea for their weekly project, the smart first step is to search the archive and see if anyone did something similar in the past. They almost always find several models this way, and are free to incorporate the best elements from them as they begin their own work for the week.
The story of Arthur Andersen brings up a fundamental riddle: How do companies lose their way so badly?
For example, partners during the firm’s final years were clearly listening to their internal press secretaries justifying the clean audits they were issuing for dirty clients. Even after a few close calls, overconfidence and self-justification were pervasive.
Toffler writes in Final Accounting (coauthored with reporter Jennifer Reingold), “As consulting became more important to the Firm, in which the only way you could win was by screwing your fellow partner,… Arthur Andersen became a place where people spent a great deal of time in bloody internecine battles that could have been better spent working with clients, pursuing new business, or mentoring young employees.”
The formal separation of Andersen Consulting (AC) and Arthur Andersen into separate business units in 1989 didn’t fix the problem. In fact, it probably deepened audit partners’ feelings of losing status and getting left behind.
The business geeks fight back against both by embracing a norm of openness, which we can define as sharing information and being receptive to arguments, reevaluations, and changes in direction. This norm is the opposite of the defensiveness, clinging to the status quo, and undiscussables chronicled by Argyris.
We’ve already seen several examples of how openness is practiced at geek companies. Let’s revisit some of them, and show how they combat the key elements of Model 1: assuming unilateral control, striving to win and minimizing losing or failing, suppressing negativity, and creating undiscussable topics.
Carl Bass told me that when he became CEO of the design software company Autodesk in 2006, “I still had this old-fashioned idea that being an executive leader meant that you make really important decisions. By the time I left, I made almost no important decisions. I mean that very seriously; I probably made one every two years. And to make them, I’d say to a roomful of people, ‘I’m going to make the call, but first I want to hear all your opinions, all right?’” At Google, Eric Schmidt also gave up even trying to appear that he was in unilateral control. “We had a list of running issues that
...more
Not suppressing negativity: Williams recalled that even though the change she was making was a two-way door, it was “probably one of the hardest things I’ve done in my career personally,” because it flew in the face of what she was used to: cross-functional processes, management review committees, elaborate decision loops, and other trappings of bureaucracy. Her
“radical transparency,”
one obvious problem is that this all falls apart at some point; and the emperor is still the emperor…
As this folktale shows, common knowledge has an odd, almost supernatural power to get a group of people to change their behavior in ways that appear coordinated but aren’t. My favorite demonstration of this phenomenon comes from a logic puzzle that exists in many versions dating back at least as far as the 1960s. One version is set on the fictional Bad Breath Island.16
a great example
Arthur Andersen’s defensive and unethical norms were distasteful to Toffler, but she generally followed them. As she writes, “One of the powerful personal lessons of my Arthur Andersen experience is that, despite my self-image as a debunker, my frequent battles with my bosses, and an occasional outbreak of ‘my way,’ I basically went along with the culture… If you hang around a place long enough, you inevitably start to act like most of the people around you.”
There are two big reasons why the business geeks don’t believe in immortal businesses. The first is that innovation and competition are a one-two punch that can knock out anyone.
That disruption becomes a lot easier if incumbents succumb to any of the classic organizational dysfunctions. The second reason that the business geeks don’t believe in corporate immortality, and I believe the bigger one, is that those dysfunctions just keep reappearing. They’re as hard to kill as the boogeyman terrorizing kids in a horror film.
Zuckerberg’s vision for the metaverse wasn’t shared by many of his employees.
Our companies often don’t want us to do these things, but “company” is a very remote and abstract concept to us Homo ultrasocialis. Self-image, social rank, prestige, and reputation, meanwhile, are as concrete and enticing to us as a delicious, calorie-rich meal. And status losses and social exclusion are as painful as a toothache or a broken heart.
“Cooperative enterprise” is a good synonym for “prosocial institution,” and a company is very much a cooperative enterprise. Henrich’s insight gives us reason to expect that all companies will in some way, at some point, experience the age-old tension between their goals and those of individuals and coalitions.
Cultural flashpoints change over time, but bureaucracy is ageless. It’s the oldest and most common way that “individuals and coalitions eventually figure out how to beat or manipulate the system to their own ends.” As we saw in chapter 5, bureaucracy causes companies to become sclerotic, uncompetitive messes. Today’s tech giants have solved many hard problems, but the gravity-like pull of bureaucratization still bedevils them.
to dismantle the bureaucracy that had been built up.
But I do know that they’re not going to be following the industrial-era business playbook, because that playbook just doesn’t work as well as the geek way.
Because of their belief in the importance of coordination, communication, cross-functional processes, and control, companies following the old playbook become more bureaucratic and sclerotic over time.
