I'm Feeling Lucky
I'm Feeling Lucky
The Confessions of Google Employee Number 59
Douglas Edwards
Table of Contents
Title Page
Table of Contents
...
Copyright
Dedication
Contents
Introduction
PART I
Chapter 1
Chapter 2
Chapter 3
Chapter 4
Chapter 5
Chapter 6
Chapter 7
Chapter 8
Chapter 9
Chapter 10
PART II
Chapter 11
Chapter 12
Chapter 13
Chapter 14
Chapter 15
Chapter 16
PART III
Chapter 17
Chapter 18
Chapter 19
Chapter 20
Chapter 21
Chapter 22
Chapter 23
Chapter 24
Chapter 25
PART IV
Chapter 26
Timeline of Google Events
Glossary
Acknowledgments
Footnotes
Houghton Mifflin Harcourt
Boston New York
2011
Copyright © 2011 by Douglas Edwards
All rights reserved
For information about permission to reproduce selections from this book,
write to Permissions, Houghton Mifflin Harcourt Publishing Company,
215 Park Avenue South, New York, New York 10003.
www.hmhbooks.com
Library of Congress Cataloging-in-Publication Data
Edwards, Douglas, date.
I'm feeling lucky : the confessions of Google employee number 59 / Douglas Edwards.
p. cm.
ISBN 978-0-547-41699-1
1. Google (Firm)—History 2. Internet industry—United States—History.
3. Corporate culture—United States—History. 4. Marketing—United States—History. I. Title.
HD9696.8.U64G56 2011
338.7'6102504—dc22
2010052588
Book design by Brian Moore
Printed in the United States of America
DOC 10 9 8 7 6 5 4 3 2 1
All of the author's profits from the sale of this book will be donated to charity.
Lyrics to Grateful Dead songs copyright © Ice Nine Publishing Company. Used with permission.
To Kristen, without whom the journey would have been
impossible and the destination meaningless.
Nothing to tell now. Let the words be yours, I'm done with mine.
—"CASSIDY" BY JOHN BARLOW
Contents
Introduction [>]
PART I: YOU ARE ONE OF US
1. From Whence I Came [>]
2. In the Beginning [>]
3. A World without Form [>]
4. Marketing without "Marketing" [>]
5. Giving Process Its Due [>]
6. Real Integrity and Thoughts about God [>]
7. A Healthy Appetite for Insecurity [>]
8. Cheap Bastards Who Can't Take a Joke [>]
9. Wang Dang Doodle—Good Enough
Is Good Enough [>]
10. Rugged Individualists with a Taste for Porn [>]
PART II: GOOGLE GROWS AND FINDS ITS VOICE
11. Liftoff [>]
12. Fun and Names [>]
13. Not the Usual Yada Yada [>]
14. Googlebombs and Mail Fail [>]
15. Managers in Hot Tubs and in Hot Water [>]
16. Is New York Alive? [>]
PART III: WHERE WE STAND
17. Two Speakers, One Voice [>]
18. Mail Enhancement and Speaking in Tongues [>]
19. The Sell of a New Machine [>]
20. Where We Stand [>]
21. Aloha AOL [>]
22. We Need Another Billion-Dollar Idea [>]
23. Froogle and Friction [>]
24. Don't Let Marketing Drive [>]
25. Mistakes Were Made [>]
PART IV: CAN THIS REALLY BE THE END?
26. S-1 for the Money [>]
Timeline of Google Events [>]
Glossary [>]
Acknowledgments [>]
Introduction
LARRY PAGE IS an intense guy. At least he was in 1999 when I first began working for the company he co-founded with Sergey Brin.
Whenever I found myself in a room with Larry, I felt an urgent need to do more, as though every second in which I wasn't communicating vital information was a waste of his bandwidth.
One day in 2002, I ended up alone with Larry in his office after a long and protracted battle over some policy or other. I had fought and I had lost, and I had come to opine on what I had learned and to extend an olive branch across what had been a turbulent time. Larry, dressed in casual shades of gray, peered intently at his screen. Or rather, at his two oversized adjacent monitors, filled with code and open web browser windows. Sergey, with whom he shared the office, was not on hand. Disassembled in-line skates, a crumpled hockey jersey, and a Japanese geisha doll kept watch over his empty chair.
"Larry," I began, "I know I haven't always agreed with the direction you and Sergey have set for us. But I've been thinking about it and I just wanted to tell you that, in looking back, I realize that more often than not you've been right about things. I feel like I'm learning a lot and I appreciate your patience as I go through that process."
I smiled inwardly. It was a well-framed corporate kiss-up. I'd humbled myself and given Larry an opportunity to analyze my strengths as a member of Google's management team and to reward me with comforting words and reassurances about the value I added. Now he would recount those occasions when my counsel had been sage and congratulate me on my perspicacity. I envisioned us engaging in the non-physical equivalent of a man hug before I trundled off to savor the moment with a freshly made cappuccino in the micro-kitchen. That's how you "manage up" in a large corporation.
Larry looked at me with the same stare he had directed at the code on his screen, as if he were trying to decipher some undigested bit of an equation that refused to resolve itself.
"More often than not?" he asked me. "When were we ever wrong?"
He didn't smile as he asked his question or arch an eyebrow to signify annoyance. He simply wanted to know when he had been wrong so he could feed that information into the algorithm that ran his model of the universe. If he had made a mistake, he needed to know the specifics so he could factor that into the next iteration of the problem if it reappeared.
"Oh. That's right," I thought, awakening from my reverie. "I don't work at a large corporation anymore. I work at Google."
Operating Principles
You know Google.
At least, you know what Google does. It finds stuff on the Internet. That's as much as I knew when I joined the company in 1999. I didn't know what a web indexer, a pageranker, or a spidering robot was. I didn't know how dogmatic engineers could be. I didn't know how many Internet executives could squeeze into a hot tub or how it felt to "earn" more in one day than I had in thirty years of hard work. I didn't know then, but I do now.
True, my story is one of rare opportunity and fortuitous timing, but not entirely so.
This book tells how it felt to be subjected to the g-force of a corporate ascent without precedent, to find myself in an environment where old rules didn't apply and where relying on what I knew to be true almost got me fired. It's not a complete history of everything Google did between 1999 and 2005, nor a completely objective retelling of Google's greatest hits. I wrote the official history of Google d
uring that period and inscribed it on the company's website.* Most accounts since have merely embellished it, and I don't intend to cover all that old ground again. Instead I'll give my insider's view of how things worked (and didn't work) and how we changed as individuals and as a corporate entity.
This book won't delve deeply into Google's current imbroglios over censorship, regulation, and monopoly. I include only what happened between my first day in 1999 and the day I left in 2005. We weren't yet worried about network neutrality, street-view data gathering, or offshore wind farms. Our big issues barely grazed the electrified moral fence of our "Don't be evil" credo: develop the best search technology, sell lots of ads, avoid getting killed by Microsoft.
While this story is told from a marketer's perspective and my title came to encompass "consumer brand management," this book is not just about marketing. I don't claim to have "built" Google's brand. The brand was built on the product, and the product was built by engineers—computer scientists who constructed systems as complex as any that ever launched a rocket into space, but powering instead a small rectangular search box that now appears in every corner of the Internet.
I'll describe the work habits that enabled them to accomplish a great deal in a short time and the shortcomings that developed in a company where every problem was viewed as solvable and every situation as reducible to a set of data points; where knowing you were right meant nothing should, could, or would stand in your way.
And I'll show how a company with a vision of providing access to all the world's information sometimes mishandled its own relationship with openness, honesty, and disclosure in ways that arose organically and inevitably from the attitudes of those in charge.
To start, I'll give you the background I wish someone had given me on my first day at Google, so you can appreciate the chaos without getting lost in it yourself.
Let's begin with a quick sketch of the company's founding. Google started as a joint research project by Larry and Sergey in 1996, when they were graduate students at Stanford. They based their project on a new approach to search technology that Larry named PageRank in honor of himself and because, well, it ranked web pages according to their importance.* Their algorithm took into account all the hyperlinks pointing to a given web page from other websites, as if by linking to a page those other sites were declaring it worthy of attention. Most search engines just looked at the content of the pages themselves and based their results on how often the searched-for word appeared on them. It was the difference between judging a stranger by his looks and gathering opinions from everyone who knew him. Because Larry and Sergey's technology analyzed what was going on behind the scenes, they called their search engine "BackRub." For a while, a photo of Larry's hand caressing a bare shoulder was their logo, but even after they airbrushed out the dark hairs, it looked like a shot from a low- budget porn flick.
In 1997, they changed the name to Google, which played to their love of math and scale (a googol is 10100). They chose the variant spelling for two reasons: the googol.com web domain was taken, and Larry thought they wouldn't be able to trademark a number. Larry was a very shrewd businessman—but we'll get to that.
Within a year, Larry and Sergey had taken leave from Stanford and set up in the Menlo Park garage of Susan Wojcicki, the college roommate of Sergey's girlfriend. Google's traffic began climbing and the company began hiring. They incorporated in September 1998, and when they outgrew Susan's garage in early 1999, they moved to an office at 165 University Avenue in Palo Alto. Six months later, having talked two venture capital firms out of $25 million, they moved into an industrial park at 2400 Bayshore Parkway in Mountain View. That's where I joined the company, which at the time had about fifty employees and was doing almost seven million searches a day. Even though that was a seventy thousand percent increase over the year before, it barely registered as a blip on the radar of major players like Yahoo, AOL, and MSN, which were each delivering on the order of half a billion page views per day.
Yahoo was the Jabba the Hutt of the "search space" at the turn of the millennium, and it wasn't even a search engine. Yahoo was a "portal," a provider of mail and news and all kinds of services built around a hand-compiled directory of web pages arranged by categories. It had almost thirty million users, but it rented technology to power its search box from Inktomi, the leading provider of search to websites and corporate intranets.
Industry experts speculated, Would Google focus on growing its own site to compete with Yahoo, or would it become a technology supplier and compete with Inktomi? If we tried to do both—build a popular online search engine while providing search technology to other sites that hoped to do the same thing—we would end up competing with our customers. The question, however, betrayed ignorance of Larry and Sergey's aspirations and self-confidence. Why choose just to have cake when you could eat it too? Google would be both a supplier and a search site, because Larry and Sergey knew they were smart enough to isolate the part of the equation containing failure and work around it.
Their vision didn't end at winning the search wars. They would build a company to fix large-scale problems affecting millions of people and terraform the entire landscape of human knowledge. They would speed medical breakthroughs, accelerate the exploration of space, break down language barriers. Instead of putting a Band-Aid on global ignorance and confusion, they would clear the clogged arteries of the world's data systems and move information effortlessly to the point at which it was needed at exactly the time it was required. They would be, Larry believed, an information conglomerate on the scale of General Electric—the GE of IT. To do that, they would need better tools—starting with a search engine that actually delivered what people wanted to find.
Engineers rebel at inefficiency. Larry Page, more than anyone I ever met, hated systems that ate hours and produced suboptimal results. His burning passion was to help the world stop wasting his time.
That love of efficiency begat a fondness for frugality, because paying more than the bare minimum for something was by definition wasteful. Larry liked trimming unnecessary expenses, but it was Sergey who fully applied his razor-sharp intellect to cutting costs.
"That seems kind of expensive," Sergey said, looking at the hundred-dollar price for a cab from Malpensa airport to downtown Milan in January 2003. He, his girlfriend, and I had flown in for the opening of our new Italian office, and I was looking forward to traveling in style with the president of a booming Internet company. The dot-com era was over for everyone else, but Google's financials were deep in the black. Even though we'd flown coach, surely now we'd be kicking loose a little change to let the Old World know we had indeed arrived.
"Maybe we should take the bus," Sergey suggested, standing in the middle of the baggage claim area squinting at the signage. "It's less than five Euros a person." The bus? What? Were we college kids backpacking on spring break? Maybe we could just hitchhike into town. It was pouring out, and a cab would take us right to our door, not to some run-down depot a short walk from nowhere.
We compromised on the train, which ended up saving us fifty dollars, not counting the cost to my inflated sense of importance.
Efficiency. Frugality. And oh yes, integrity.
Larry and Sergey had an intuitive feel for presenting data in a way that improved the ratio of signal to noise. That means they didn't believe in adding unnecessary crap to the information you actually wanted to see. So, no blinking banner ads in Google search results. No links to every service Google offered pasted all over the Google.com homepage. And no intermingling of ads with actual search results as our competitor GoTo.com was doing. To corrupt a working system would be to profane perfection.
"We could try a loyalty program," I once suggested in a meeting about getting users to search more often, "like a frequent flyer program."
Larry raised his eyebrows the way he does when he considers an idea so blatantly ridiculous you should be ashamed of yourself for even thinking it.
"Frequen
t flyer programs are evil," he said.
"They are?" I didn't recall my Mileage Plus number ending in 666.
"They incentivize people to take flights that are not the most direct or the cheapest, just so they can earn points. Their employers end up paying more, and people lose time traveling."
Loyalty programs promoted loyalty above efficiency, and that was just wrong, wrong, wrong.
Efficiency, Frugality, Integrity. I suppose if you had stitched that onto a flag, most Googlers would have saluted. There were other operating principles I unearthed picking my way along through trial and error, but those three constitute the mother lode from which they were mined.
And while we're in the mines, let's explore exactly what my fifty-plus honest Google colleagues were toiling to accomplish so cheaply and efficiently.
You Don't Say
I was Google employee number fifty-nine, as near as I can tell, but I started the same week as other people, so my number might have been higher or lower. It doesn't matter. We each contributed according to our ability to improve information access for the betterment of all mankind. That the lowliest engineer's capacity exceeded mine by a bazillion percent made no difference in our status-blind environment.
Theoretically.
In reality, if you weren't an engineer, your first directive was to avoid impeding the progress of those who were. I'm not a technical guy. No one at Google ever said, "Hey, let's ask Doug!" when the flux capacitor hiccupped. But you couldn't work at Google without learning something new every day, even if you weren't trying to. Most engineers opened up about their work when I sat next to them at lunch, and generally they didn't mind using little baby-English words to explain things to me. Given the pressure, though, the engineers were biased toward being productive rather than talking about their productivity. It was a "Don't talk. Do." kind of culture, which made communication about our technical achievements erratic.