Always Day One: How the Tech Titans Plan to Stay on Top Forever

PORTFOLIO / PENGUIN

An imprint of Penguin Random House LLC

penguinrandomhouse.com

Copyright © 2020 by Alex Kantrowitz

Penguin supports copyright. Copyright fuels creativity, encourages diverse voices, promotes free speech, and creates a vibrant culture. Thank you for buying an authorized edition of this book and for complying with copyright laws by not reproducing, scanning, or distributing any part of it in any form without permission. You are supporting writers and allowing Penguin to continue to publish books for every reader.

LIBRARY OF CONGRESS CATALOGING-IN-PUBLICATION DATA

Names: Kantrowitz, Alex, author.

Title: Always day one / Alex Kantrowitz.

Description: [New York] : Portfolio/Penguin, [2020] | Includes bibliographical references and index.

Identifiers: LCCN 2019054201 (print) | LCCN 2019054202 (ebook) | ISBN 9780593083482 (hardcover) | ISBN 9780593083499 (ebook)

Subjects: LCSH: High technology industries—Management—Case studies. | Internet industry—Management—Case studies.

Classification: LCC HD62.37 .K36 2020 (print) | LCC HD62.37 (ebook) | DDC 338.7/6004—dc23

LC record available at https://lccn.loc.gov/2019054201

LC ebook record available at https://lccn.loc.gov/2019054202

Penguin is committed to publishing works of quality and integrity. In that spirit, we are proud to offer this book to our readers; however, the story, the experiences, and the words are the author’s alone.

Jacket design: Christopher Sergio

pid_prh_5.5.0_c0_r0

CONTENTS

TITLE PAGE

COPYRIGHT

DEDICATION

PREFACE: The Zuckerberg Encounter

INTRODUCTION: Always Day One

Ideas vs. Execution

Miracles in Miami

The Engineer’s Mindset

When Things Speed Up

1. Inside Jeff Bezos’s Culture of Invention

Meet Amazon’s Science Fiction Writers

Jeff Bezos’s Robot Employees

Hands off the Wheel

Life After Yoda

Insist on the Highest Standards

That Article

Outputs

2. Inside Mark Zuckerberg’s Culture of Feedback

Facebook the Vulnerable

Building a Feedback Culture

Pathways

Facebook’s Day One

From Broadcast to Private

The Most Chinese Company in Silicon Valley

Enter the Machines

Robot Compensation

New Inputs

Facebook’s Next Reinvention

3. Inside Sundar Pichai’s Culture of Collaboration

The Hive Mind

The Toolbar Episode

The Path to Chrome

The Shift

The Split

AI Answers

The Great Reinvention

Uprising

Exodus

4. Tim Cook and the Apple Question

A Culture of Refinement

The Refiner’s Mindset

Silos and Secrecy

A Form That’s Right

The HomePod Debacle

Hands on the Wheel

The IS&T Problem

Face-off

A Drive down 280

5. Satya Nadella and the Microsoft Case Study

Nadella’s Day One

Democratic Invention

Constraint-free Hierarchy

Collaboration

Microsoft’s New Decade

6. A Look into the Black Mirror

Always Black Mirror

The Dystopia Is Now

The Erosion of Meaning

From Doomsday to Disneyland?

7. The Leader of the Future

Something New Wouldn’t Hurt

The New Education

Caring

Watching the AI

The Case for Thoughtful Invention

Onward

ACKNOWLEDGMENTS

NOTES

INDEX

ABOUT THE AUTHOR

To everyone out there trying to make it

PREFACE

THE ZUCKERBERG ENCOUNTER

In February 2017, Mark Zuckerberg summoned me to his Menlo Park, California, headquarters for a meeting. It was my first time sitting down with the Facebook CEO, and it didn’t go as anticipated.

His company, per usual, was enveloped in controversy. Pushing hard to grow its products but reluctant to moderate them, it had allowed them to fill with misinformation, sensationalism, and violent imagery. Zuckerberg seemed ready to talk about it, and I was eager to listen.

Facebook’s main office—a vast, open, concrete slab of a structure—is an unnerving place to walk into. It has nine lobbies, two layers of security to enter, and guards demanding you sign a nondisclosure agreement before taking another step. Once inside, I made my way to a glass-walled conference room, smack in the middle of everything, where Zuckerberg holds his meetings. And after he finished a conversation with his chief operating officer, Sheryl Sandberg, he ushered me in, along with my editor Mat Honan, for a chat in full view of anyone walking by.

Zuckerberg had been hard at work on his Manifesto, a fifty-seven-hundred-word exposition outlining Facebook’s response to the troubling content, and its role in its users’ lives more broadly. Coming in, I expected the typical CEO briefing: a lecture followed by limited time for questions. But after a brief overview, Zuckerberg started asking for feedback. What are the things we talked about that you feel didn’t come through in the text? he said. What’s missing?

As I answered, Zuckerberg listened intently. His posture didn’t shift. His focus didn’t break. And his reactions—first a gentle dispute of my suggestion that Facebook talk more about its power, and then an acknowledgment—made clear he wasn’t asking just for show. I had never seen a CEO do this before, let alone one known for his obstinacy. It felt different, and worthy of investigation.

After our meeting, I asked everyone I could about Zuckerberg’s unusual appeal for feedback. Is this normal? Has he ever asked you? Following several inquiries, I had my answer: the request was simply a glimpse into the way he runs Facebook. Zuckerberg has built feedback into Facebook’s very fiber. Major meetings end with requests for it. Posters around Facebook’s offices say FEEDBACK IS A GIFT. And nobody in the company is above it, not even Zuckerberg himself.

•   •   •

As a tech reporter in Silicon Valley, I’ve watched the tech giants’ unconventional run of dominance from the front row. Instead of following the typical corporate life cycle—grow, slow, stumble, and ossify—Apple, Amazon, Facebook, Google, and Microsoft have only become more powerful as they’ve aged. And perhaps except for Apple (more on that later), they’re showing little sign of letting up.

Amid this run, I’ve been struck by these companies’ uncommon internal practices. After countless interviews with chief executives, for instance, I had been convinced the world’s top CEOs were natural sellers, people who used the force of their personalities to rally others around their visions. But look at Zuckerberg and his counterparts—Jeff Bezos at Amazon, Sundar Pichai at Google, Satya Nadella at Microsoft—and you’ll see trained engineers more eager to facilitate than to dictate. Instead of answers, they have questions. Instead of pitching, they listen and learn.

Following that meeting in Menlo Park, I began digging into the tech giants’ inner workings more broadly—looking at their leadership practices, their cultures, their technology, and their processes—wondering if there was a link between their success and the unique way they operate. As common patterns emerged, that link became impossible to deny. And I became obsessed with uncovering what exactly they were doing differently, and why it was working. Two years and more than 130 interviews later, this book is the product of that journey.

What you’re about to read is the formula that’s allowed the tech giants to achieve their dominance and sustain it. This is a book about culture and leadership, but more broadly it’s about ideas and invention, and the path between. It’s about a new model for business in a time when companies can spin up new products in the blink of an eye, when challenge is constant and no advantage is safe. Leaning on an array of internal technology that’s enabled them to operate differently, much of which they’ve built themselves, the tech giants have uncovered this new formula early. The time has come to release it to everyone.

The companies detailed in this book aren’t perfect—far from it. In their unrestrained quest for growth, they’ve worked their people to the bone, missed obvious abuses of their technology, and retaliated against earnest internal dissent. Such excesses have caused the US government to consider regulation, and politicians to call for their breakup. Largely with cause. So to state it for the record: this book is not about growth, or growth hacking, or beating down smaller companies. It’s about creating inventive cultures, which I believe everyone can learn from. And for those looking to rein these companies in, understanding how their internal systems work can be a strategic advantage. To effectively diagnose illness, it’s helpful not only to look at the symptoms, but to understand the physiology.

If the tech giants’ knowledge remained only in their hands, the broader business world—and the regulators examining them—would be at a disadvantage. In our hands, we have a chance to even the playing field.

INTRODUCTION

ALWAYS DAY ONE

At an Amazon all hands in March 2017, a trim, confident Jeff Bezos stood before thousands of his employees, looked down at a stack of note cards, and read a pre-submitted query with an expression of mild disappointment. Okay, I think this is a very important question, Bezos said. What does Day Two look like?

For the past twenty-five years, Bezos had urged his employees to work each day as if it were Amazon’s first. Now, with Amazon marching toward a trillion-dollar valuation, and growing by approximately a hundred thousand employees a year, a (perhaps hopeful) employee was asking Bezos to imagine Day Two.

What does Day Two look like? Bezos asked. Day Two is stasis, followed by irrelevance, followed by excruciating, painful decline, followed by death.

The meeting erupted in laughter. For the thousands of Amazonians in attendance, Bezos’s demolition of their unnamed colleague, who had ventured onto Amazon’s third rail, was a delight. As the crowd applauded, Bezos paused, broke into a half smile, and closed the meeting. And that is why it is always Day One, he said.

Day One is everywhere at Amazon. It’s the name of a key building, it’s the title of the company’s blog, and it’s a recurring theme in Bezos’s annual letter to shareholders. And though it’s tempting to read it as an order to work ceaselessly, particularly at the notoriously hard-charging Amazon, its meaning runs deeper.

Day One at Amazon is code for inventing like a startup, with little regard for legacy. It’s an acknowledgment that competitors today can create new products at record speeds—thanks to advances in artificial intelligence and cloud computing especially—so you might as well build for the future, even at the present’s expense. It’s a departure from how corporate giants like GM and Exxon once ruled our economy: by developing core advantages, hunkering down, and defending them at all costs. Getting fat on existing businesses is no longer an option. In the 1920s, the average life expectancy of a Fortune 500 company was sixty-seven years. By 2015, it was fifteen. What does Day Two look like? It looks a lot like death.

From its origins as an online bookseller, Amazon has lived its Day One mantra, inventing new businesses with abandon, with a near-complete disregard for how they might challenge its existing revenue streams. The company remains a bookseller, but it’s also a clearinghouse for almost every imaginable product, a thriving third-party marketplace, a world-class fulfillment operation, an Academy Award–winning movie studio, a grocer, a cloud services provider, a voice-computing operating system, a hardware manufacturer, and a robotics company. After each successful invention, Amazon returns to Day One and figures out what’s next.

I own a huge amount of Amazon stock, Mark Cuban told me in July 2019. Depending on what it did today, it could literally be a billion dollars’ worth of Amazon stock. And I own the stock because I see them as the world’s greatest startup.

Look around at the tech giants today, and you’ll see a similar path. Google started as a search website, but then invented a browser extension (Stay Tuned), a browser (Chrome), and a voice assistant (Google Assistant), and incubated a leading mobile operating system (Android). Each new Google product challenged the existing set. But by repeatedly returning to Day One, Google has remained on top.

Facebook has gone back to Day One multiple times. Starting as an online directory, the company reinvented itself with the News Feed, and it’s reinventing today by moving from broadcast sharing to intimate sharing: giving the News Feed over to Facebook Groups—a series of smaller networks—and treating messaging as a first-class citizen. In the most fickle of all industries, social media, Facebook still leads.

Until recently, it seemed like Microsoft’s inventing days were over. The company was so attached to Windows it almost let the future pass it by. But with a leadership change from Steve Ballmer to Satya Nadella, the company returned to Day One and embraced cloud computing, a threat to desktop operating systems like Windows, and became the world’s most valuable company once again.

Apple under Steve Jobs developed the iPhone, a device that rendered desktop computers like the Mac and portable music players like the iPod less relevant but also set the company up for years of success. Today, Apple is having its Windows moment. It must leave iPhone orthodoxy behind and reinvent itself again to compete in the age of voice computing.

At Amazon’s South Lake Union campus in Seattle, one of the newest buildings is called Reinvent. It’s an odd word coming from one of the earth’s most successful companies. But in today’s business world, where Day Two is death, it’s the key to survival.

Ideas vs. Execution

Operating an inventive company takes more than speeches and internal messaging. It takes a reimagination of the way you run a business, which is finally possible due to a revolution in the way we work.

There are two types of work: idea work and execution work.

Idea work is everything that goes into creating something new: dreaming up new things, figuring out how you’re going to make them, and going out and creating.

Execution work is everything that goes into supporting those things once they’re live: ordering products, inputting data, closing the books, maintenance.

In the industrial economy, almost all work was execution work. A company founder would come up with an idea (Let’s make widgets!) and then hire employees for execution purposes only (they’d be in the factory, making the widgets). Then, in the late 1930s, we started moving from an economy dominated by factories to one dominated by ideas—what we call the knowledge economy.

In today’s knowledge economy, ideas matter, but we still mostly spend our time on execution work. We develop a new product or service, and then spend our time supporting it instead of coming up with something else. If you sell dresses, for instance, supporting each design requires loads of execution work: pricing, sourcing, inventory management, sales, marketing, shipping, and returns. Additional support work props up these processes, including basic tasks in human resources, contracts, and accounting.

The burden of execution work has made it nearly impossible for companies with one core business to develop and support another (Clayton Christensen calls this the innovator’s dilemma). Those who’ve tried have almost always pulled back, or found it impossible to sustain multiple businesses at once. GM historically made many things other than cars, Professor Ned Hill, an economist at Ohio State University, told me, citing refrigerators and locomotives. They were an octopus, and they couldn’t manage it.

Drowning in execution work, today’s companies thus devote themselves to refinement, not invention. Their leaders might desire to run inventive cultures, but they don’t have the bandwidth. So they deliver a limited set of ideas from the top, and everyone else executes and polishes.

But running a company with an inventive culture, rather than one of refinement, is now suddenly possible. Advances in artificial intelligence, cloud computing, and collaboration technology have made it feasible to support existing businesses with much less execution work, helping companies turn new, inventive ideas into reality—and sustain them. These tools are the next evolution of a workplace software explosion that’s made companies more efficient, and AI is putting it into overdrive. Experts say AI will free people up to do more creative or more human work. But more precisely, AI enables them to do more inventive work. This, I believe, is a critical factor behind the tech giants’ success.

Pushing a new wave of enabling technology forward, the tech giants have figured out how to minimize execution work. This creates room for new ideas, and they turn those ideas into reality. Their cultures therefore support invention, not refinement. They remove barriers that would prevent ideas from moving through the company, and bring the best of those ideas to life. Simple in theory, but complex in practice, this is what makes them tick.

For some time, I was convinced the tech giants would hold this advantage over the rest of us for many years. But then I took a trip to Miami.

Miracles in Miami

Cee Lo Green probably never aspired to life as a corporate performer, but the hefty high-pitched singer embraced the role in October 2018 as he stood before eleven hundred badge-wearing professionals making small talk, checking their phones, and attempting to network inside Miami Beach’s LIV nightclub.

As the badge wearers chowed down on sliced brisket, jalapeño mac and cheese, and blue crab risotto and indulged in an open bar, Green had some fun. He teased his top hit, a song by the name of Fuck You (or Forget You when it’s on the radio), and talked up their achievements. You’re celebrating success in your life, right? he said, moving about the stage in a white jumpsuit and sunglasses.

When the first few bars of Fuck You rang through LIV, the crowd went ballistic, and the wide-smiling Green fed off its energy. If you know something you need to say ‘fuck you’ to, now is the time! he yelled. A roar of fuck-yous rang back from those in attendance.

Green’s performance at LIV would’ve been relatively nondescript, if not for the fact that it kicked off a conference hosted by UiPath, a little-known company whose software can watch your screen as you work and, with some labeling, automate your tasks. UiPath and its counterparts are on track to automate work across millions of jobs in the coming years, making the chorus of F-yous a bit jarring.

Months before the show, I heard rumblings that UiPath had the potential to change the nature of corporate work, in a manner that might bring the broader business world closer to the tech giants’ way of working. And after investors handed the company $225 million earlier that fall, I decided to take my notebook to South Beach to see what it was all about.

UiPath, I learned, makes it simple to automate routine work performed on a computer. Its software can observe your mouse movements and clicks and, with some guidance, figure out how to perform your tasks. UiPath’s robots (which have no physical presence) can take on a seemingly unlimited array of execution work, including entering data, generating reports, filling out forms, composing formulaic documents, and emailing those documents to designated recipients. In human resources alone, these bots can write standard new-hire letters, register new employees in various benefits systems, and, when the time comes, write their termination letters too.

This type of execution work fills significant portions of millions of people’s workdays, and some of the world’s most recognizable employers—Walmart, Toyota, Wells Fargo, UnitedHealthcare, and Merck among them—made their way to Miami to compare notes on how they might automate it.

SMBC, a Japanese bank, said it had already deployed one thousand UiPath bots and planned to add a thousand more within the year. Anoop Prasanna, Walmart’s head of intelligent automation, praised UiPath’s ability to automate work, complaining only that he couldn’t roll out the technology fast enough. Holly Uhl, who runs automation efforts at State Auto, an insurance company, told me in a quiet moment that UiPath saved her company thirty-five thousand hours of previously human work over seventeen months, and would ramp up even further. It’s going to continue to grow, she said.

The biggest news at the conference was that UiPath’s process automation technology was going to