Larry's refusal to engage the privacy discussion with the public always frustrated me. I remained convinced we could start with basic information and build an information center that would be clear and forthright about the tradeoffs users made when they entered their queries on Google or any other search engine. I didn't really believe many people would read all the pages, or particularly care what they said. In fact, I somewhat cynically counted on that. The mere fact that we had the explanation available would allay many of their concerns. Those who truly cared would see we were being transparent. Even if they didn't like our policies on data collection or retention, they would know what they were. If they went elsewhere to search, they would be taking a chance that our competitors' practices were far worse than ours.
To Larry the risks were just too high. Once we squeezed the toothpaste out of the tube, we would not be able to put it back. And he would never do anything that might cost us complete access to what Wayne Rosing called our "beautiful, irreplaceable data."
Can't We All Just Get Along?
Ka-chunk. Ka-chunk. Over the rest of 2003, the product-development group under Jonathan worked with engineering to stamp out a series of features, services, and infrastructure updates.
Ka-chunk. Ka-chunk. The business-development group punched out partnership deals for toolbar distribution and advertising syndication.
I too ratcheted up my productivity. Writing copy for products and promotions was my primary activity, but there were other items on a list that perpetually expanded.
Set branding guidelines for partners who kept trying to use our logo in unapproved ways.
Arrange product placements. For example, I used Google to track down an email address for Sex and the City's props guy on a German website and the brother of a West Wing producer in Los Angeles. Google appeared in both shows at no cost to us.
Negotiate co-branding deals. Among other partnerships, I worked with DK books in England to release a set of Google-branded reference works for kids, with Time magazine to do an Internet search guide for use in classrooms, and with the World Puzzle Federation to sponsor their international championship.
Develop product identities. To brand our AdSense ads, I proposed "Ads by GOOOOOOOGLE," an idea that was rejected at first but implemented the following year.
No matter how efficient and productive I was, however, I felt hampered by the one thing that seemed to be moving backward. While I was communicating well with others in product management, my relationship with Marissa became more strained with each new project we worked on together. And we worked on an increasing number of projects together.
We disagreed about how to promote the Webby award won by Google news, which I thought was nice but no more exceptional than the four other Webbys Google had won and never acknowledged. Marissa wanted—and got—a special logo, a dedicated thank-you page, and a team picture on the site. Cindy stopped her from including the Webby in our online FAQ as well.
We had different views on Project Miles, a frequent-searcher program Marissa developed with no marketing input because, she claimed, Larry didn't want to use any resources to make it happen. In a test, the program put counters under the Google search box for some users. The number the counter displayed increased each time the users searched with Google. When Project Miles launched, someone from EPIC, a privacy rights organization, asked a New York Times reporter, "Do users know that Google is spying on them?"* Given Larry's hatred for frequent-flyer programs and his fears about raising privacy concerns, the test program was quickly canceled.
I argued against her idea to have staffers give away five thousand Google pens on the streets of Tokyo, which was opposed by our Japanese office and members of her own team. The promotion moved ahead anyway. Most of the Japanese pedestrians refused to acknowledge the Googlers, other than a handful of women who seemed quite comfortable accepting gifts from strangers.
All these instances of ignored or overridden input were eroding my authority to make decisions about our marketing. Worse, I had become a Cassandra—I could see bad things on the road ahead but couldn't stop us from recklessly rolling over them. Instead of having more credibility after the Froogle fiasco, I seemed to have less. I finally went to Larry to ask how we in brand management could align our efforts more closely with the company's goals. Work more closely with product management, he advised me. I assured him I would redouble my efforts to do so.
Two weeks later, Marissa closed the best window I had into Larry and Sergey's thinking about our brand strategy and the products in the development pipeline. She announced that Larry's weekly product-review meeting henceforth would be limited to Larry and members of the product-development team.
Product review was the meeting at which I learned about major initiatives while they were still the equivalent of a tropical depression forming somewhere in the Atlantic. At that stage, redirecting branding efforts might have some effect. Under the new system, I would not have input into products until they were full-fledged hurricanes five miles off the coast. I'd be able to do little other than board up the windows, huddle for cover, and pray for the best.
I thanked Marissa for trying to save me from an unnecessary meeting, but assured her I preferred to know what was happening with product development. Cindy reminded Marissa that since our product was our marketing platform, marketing needed to be part of the discussion that happened in the product-review meetings. The debate went on for a month, but the die was cast, and marketing was cast out.
Marissa told me that it was actually Larry and Jonathan who had suggested constraining the meeting to those in the product group. That may have been true, but since the word always came through Marissa, it was hard to know if something was lost in translation. I knew Larry hated large crowds, and now that the product group worked with engineering, perhaps he felt brand management's presence was redundant. Whoever instigated the change, it made my job harder. And it did not improve my working relationship with Marissa, who was promoted to director of consumer web products in mid-July 2003. I congratulated her and asked to set up a regular touch-base meeting so we could coordinate our efforts. It never happened.
Instead, our communication channel kept degrading. One day I heard from someone in product management that Sergey was furious at marketing. He believed we were holding up the launch of a new Google toolbar until it included a way for users to clear their search history. One of the most common questions we received from users was about removing previous search queries from Microsoft's Internet Explorer browser. The question came up because when people started to type in a Google search, their previous search terms would appear below it on the screen. Users didn't want that information displayed. Even though Microsoft was responsible for storing that data, our toolbar engineers thought they could fix it so the previous searches didn't appear.
I thought that would be great, but neither I nor anyone else in corporate marketing ever asked for it to be included. The toolbar engineers liked the idea enough to work over the weekend coding it anyway. Someone informed Sergey that our group was not only holding up the launch for the history-clearing add-on but insisting that the new feature delete Google's cookie as well.
"Why the hell are you letting marketing drive your product development?" Sergey demanded of the associate product manager, who spent a half hour well after midnight trying to calm him down and explain what was actually going on.
"There were only three people at Google at that hour," the APM vented, "Larry and Sergey and you-know-who. She found out about it and then all hell broke loose."
The UI team also struggled to retain access to the product-development process but we found ourselves marginalized, especially on design issues around Google news, which had become an entity unto itself under Marissa's stewardship. One of Marissa's APMs sent out a drastically different design for news search results and requested UI feedback forty-eight hours before the "code freeze," after which programmers would implement the design. Such major changes would
normally be reviewed and tested for weeks, but Marissa informed us that Larry felt the project was behind schedule and so had asked her to form a task force to move it ahead.
The last straw for me was not about a major change, but about how a minor change was handled. Google's fifth anniversary as a company occurred in September 2003. Marissa reworked the wording of the alt-text of our Google birthday logo, the phrase that appeared when you rolled your mouse over the artwork Dennis had created for the homepage. I had written "Celebrating Google's 5th birthday" as the explanation for the cake and party hat decorating the logo. Marissa wanted "Happy 5th birthday, Google!" It seemed weird to me to congratulate ourselves on our own birthday, so I instructed Dennis not to change it. In the middle of the night, Marissa overruled me, claiming her wording was what we had always used. It wasn't. Mine was.
The next morning I drafted a note demanding that Larry weigh in once and for all about who controlled the copy on Google.com. He had already assured me privately that wording was my responsibility, but now I wanted it in writing and on the record. Cindy advised me not to send the note. I let her cooler head prevail. I don't know if Cindy said something to Larry, but after that things settled down.
Marissa and I still disagreed over wording on occasion, but, at least in the short term, we found ways to work out differences amicably. Maybe Marissa realized she had finally overstepped her bounds. Maybe I was adjusting to Google's new world order in which product management had a legitimate say in the brand messaging. Or more likely, keeping up with the overflowing load of projects kept us both too busy to waste energy debating passive vs. active voice and verb-subject agreement. By late 2003 I was getting more done, faster and more efficiently, than I had in any other job I'd ever held.
In Redmond, the Beast Awakes
One of my new responsibilities was running our weekly TGIF meetings. We had passed a thousand employees in April 2003 and our Friday get-togethers were now enormous gatherings held in a large open space on the first floor of the building next to the Saladoplex. Each Friday, a contractor named Michael "MLo" Lopez* helped me build an Apple Keynote presentation introducing just-hired Nooglers, highlighting department success stories, and providing insight into our financial health.
Larry and Sergey would barely glance at the script I gave them, then ad lib a comically surreal amble through the events of the previous week. "I can't really read this," Sergey might say, squinting at the revenue number on one of Omid's spreadsheet slides. "But you can tell it's big because it has a lot of pixels." I got my jollies by including horrible puns, juvenile animations, and absurd images in the slides projected behind them, which caused groans and laughs while they spoke.
Because information tended to trickle in throughout the afternoon from around the company, preparation took an entire day—my twenty-percent time. I did my best to keep the TGIFs entertaining, playing world music as people assembled to munch on snacks provided by Chef Charlie and giving Nooglers propeller beanie caps so they stood out in the post-presentation mingling. Some weeks we had skits, such as the interpretive dance, caped superhero, and flaming laptop that introduced changes to our Help Desk organization.
Fewer and fewer old-timers showed up. The engineers already knew what they needed to know and found TGIF a waste of time. The ratio of news to fluff was not sufficiently high to draw them in, which alarmed Larry and Sergey. TGIF was intended to bind the company together. Instead, the culture was separating like the layers in one of Charlie's parfaits.
The founders weren't above bribing senior engineers to attend. Each December, Larry and Sergey "surprised" the staff by handing out a year-end thousand-dollar cash bonus at TGIF. Three days before the 2003 distribution was to take place, they asked me for ideas about how to do the presentation. I suggested a casting tape for a (fictional) Superbowl TV spot. Given how often we derided the profligacy of dot-com companies and their mass-market advertising, few staffers would fall for it, but it would give us a framework.
I drafted a script and gave it to Delicia Heywood, a marketing staffer, to shoot and produce. She came back forty-eight hours later with the tape we would use the following day. After a brief intro of scrolling text, we cut to a director's slate. Then Al Gore, giving an intense look of concern, asked with the emotional spark of soggy cardboard, "Are you searching for answers? Google.com can help you find them." A pause, and then Gore asked, "Was that too over the top?" Other Googlers appeared, including Chef Charlie and our sultry receptionist Megan, who leaned forward to whisper, "Looking for something? Need a good search?" as well as a guest appearance by the Pets.com sock puppet. When the video aired, the audience of Googlers went nuts. After it ended, Sergey announced that we had decided not to spend the money airing a spot after all, and instead would split the cash among the Google staff. That went over very well.
By the end of the next year, the amount of cash needed for the bonuses became unwieldy. Armored cars and shotgun-toting guards were needed to transport and watch over the funds until they could be distributed.
There was more good financial news in 2003. The slides at TGIF showing our corporate financial health kept pointing up and to the right. The board split our stock when it hit ten dollars a share in June.
Of course, we still faced competition, but we had momentum. Overture had bought the search engine company FAST in February to compete with us, but they were clearly too late to catch up on that front. We were not happy, but also not surprised, when Yahoo bought Overture in July for $1.3 billion. Now Yahoo owned the patents Overture had accused Google of infringing, which meant we would soon be in court with one of our biggest customers. The handwriting had been on the wall. Google had grown too big and threatening to be Yahoo's supplier. They would have to compete with us, and to do that they needed their own search-advertising technology.
More disconcerting was Microsoft's awakening to the power of search. "Google's a very nice system," Windows group VP Jim Allchin told the Seattle Times in February 2003, "but compared to my vision, it's pathetic."*
The threat to Google was real. If Allchin's vision included integrating search into the new version of Windows Microsoft was soon to release, it could eliminate the need to launch a browser and go to Google to search. Given the number of Windows users worldwide, our traffic could drop in a hurry. Without traffic, we would show fewer ads and make less money. Way less money.
"We will buy you," was Microsoft's ploy when it came to startups that threatened them, "or we will bury you." Google was not for sale. Nor would Google be foolish enough to partner with Microsoft, exposing our technology to them in ways that would let them "harvest" it and use it against us.† One business-development person warned me that Microsoft's MO as a company was to get close to startups, suck them dry, and then throw them away. Microsoft was methodical about it, giving generous terms to keep the startups alive, but essentially turning them into captive research-and-development centers. Microsoft would become the startups' biggest customer and thereby drive the direction of their development, perhaps offering to provide informal technical help, which necessitated a look at the startups' proprietary code.
When Microsoft turned its gaze to us, Larry and Sergey huddled with our board of directors and determined to stay the course. There was fear, but no panic. To survive—as companies like Intuit, AOL, and Oracle had in direct competition with the Redmond giant—we needed to focus on our core strengths: search quality, a comprehensive ad network, and content targeting. And the key tactic for implementing that strategy was one we had been employing from the very beginning: hire brilliant engineers to do brilliant things.
First we needed to prevent Microsoft from luring away any of the talent we had already found. That might be difficult as stock options vested for the earliest engineers and Google grew larger and more bureaucratic. On the plus side, most engineers would not view a move to Microsoft as an improvement in management streamlining. It was Cindy who raised the question to me of how we might help engineering find m
ore talent, perhaps in a way that would bolster our image with users as well. Ads aimed at engineers, but seen by consumers—consumers who would feel they were eavesdropping on a conversation intended for someone else. If users took joy from discovering Google on their own, such an indirect campaign would have an even deeper brand impact than one aimed directly at them. It was an inspired strategy, and I knew just the people to make it work for us.
A Brand-New Idea for Recruiting Engineers
Late in 2002, I received a poster in the mail. It was a montage of ads and PR stunts that had been put together to introduce the BMW Mini Cooper to the US market. Some of them made me laugh out loud. I dropped a note to the ad agency that had sent it, telling them we weren't in the market, but if they were ever in the Bay Area, I'd like to talk to them. Almost two years later, I convinced Larry and Sergey to put the agency, Crispin Porter+Bogusky, to work for us on my secret plan to do a consumer branding campaign in the guise of an effort to recruit engineers. I didn't tell Larry and Sergey the secret part. I just said we would hire an agency to help us find more technical talent, as they were desperate to do.
Advertising to engineers would be tricky. They don't like to be marketed to. I instructed the Crispin people not to tell anyone outside Google they were working on our behalf, and not to mention branding to anyone at Google. But I let them know I wanted a campaign to reinforce the positioning of Google as the world's best search engine because we had the best technology, created by the sharpest minds in the industry.
I'm Feeling Lucky: The Confessions of Google Employee Number 59 Page 41