Authors: Douglas Edwards
Our work lives were too full of threats and opportunities to waste resources on bureaucratic redundancy. During the dog days of summer in 2002, we were juggling a dozen chain saws.
The Chinese government unexpectedly blocked our search results, as they had done once before in the spring of 2001. We didn't know why, exactly, but suddenly we had thousands of emails in Mandarin asking about other ways to access Google. Then the block was dropped. No reason was given. Eric had discussed getting Al Gore to mediate with the Chinese, but I don't think that actually happened. At least I didn't hear him talking about it outside my office door.
We set up a program called Google Grants to provide free AdWords ads to nonprofits. I had proposed a simple public service advertising (PSA) program months before and worked on the user interface, but Sheryl Sandberg was the one who crystallized the idea and brought it to life.
We fought a trademark lawsuit over a company using our name for its online store. I was deposed. I had never given a deposition before and had to be coached on how to approach it. I spent days learning not to answer any question I wasn't asked and to ask for the definition of any term that was ambiguous. I perversely enjoyed my four hours of sweating in front of a video camera sparring with the opposing attorney.
"Did you communicate this verbally?" he asked.
Pause. Think. "Can you define what you mean by, 'verbally,' please?"
"Verbally! You know! Did you talk to him?" The attorney's frustration was audible.
"Well." Pause. "My understanding of, 'verbally' is that it means, 'expressed in words.'" Pause. "That can mean in writing. Or it can mean by speaking." Pause. "Did you mean, 'orally'?"
We won that case. Our attorney Kulpreet was so happy with my Forrest Gump impression that he recommended me for a peer bonus—a thousand-dollar award any employee could receive for helping another department in its hour of need.
As our network of ad-syndication sites grew from Earthlink and AOL to include Ask Jeeves and the
New York Times
website, we needed branding guidelines to tell each new partner how to display our search results and our ads and where to put our logo. Creating them turned out to be complicated and time consuming, especially after we won the ad contract for InfoSpace from Overture. InfoSpace ran a number of metasearch sites, which included Google search results alongside those of other search engines. The deal was a big bite out of Overture's revenue, but we didn't want our brand to become just another generic ingredient in a salmagundi of results. I wouldn't allow InfoSpace to put our logo in their ads with all the other search engines, even after they had paid stiff penalties to Overture to break the exclusivity clause in their advertising contract with them.
*
The sales team backed me up but made me go to dinner with "the client" so InfoSpace could tell me how wrong I was. I didn't eat much, and swallowed less.
But the main thing Sergey was focused on that summer was finding a billion dollars. CPC AdWords was an unqualified success. It gave every indication it would pump dollars into our bank account for many years to come. That worried Sergey. Someone, somewhere, was undoubtedly watching that cash flow and working to come up with a new development that would make our search ads obsolete. If that happened, Google would end up a bit player in the Internet economy. We needed, Sergey believed, to uncover another billion-dollar business idea and launch it quickly. Larry agreed wholeheartedly, his paranoia every bit as deep as his co-founder's.
It was Paul Bucheit who showed them where to look.
Paul liked to hack things together. In his free time, he was constructing an email system he called "Caribou"—named for a poorly managed project assigned to the comic strip character Dilbert. We would later call it "Gmail." He had been working on Caribou for a couple of years. Though dozens of Googlers, including Larry and Sergey, were using it, almost everyone he talked to at Google hated the idea of becoming a provider of email to the public. Some opposed it from a technical standpoint, some from a business point of view. But they all agreed on one thing. An email system should never display advertising based on the content of the email messages it sent or received. It was both technically infeasible and unacceptable from a product perspective.
"Before I was even working on Caribou," Paul told me, "Udi Manber
*
came in for a talk and I remember asking him, 'Do you think you could ever do content-targeted ads on email?' And he said, 'Oh no, we would never do that. That's a gross violation of whatever.' It was like, you should be fired for even asking that question."
Paul knew his mail system would never launch if it couldn't eventually pay for itself. He considered building in premium services or offering six-month free trials, but knew those models would inhibit widespread adoption. And Sanjeev Singh, a colleague helping out with the coding, kept bringing up content-targeted ads.
Paul liked to hack things together.
One night, he and Sanjeev went through his inbox one email at a time and tried to manually match each message to an ad already in the system. It wasn't that hard. Paul decided to put together a simple prototype that would do the matching automatically. He rummaged around his code files and came up with a classifier tool: software that could identify things that were related and group them. He had written it as part of Matt Cutts's porn filter project. Perfect for what he had in mind now. He reconfigured the porn classifier to match ads to the content of emails, flipped it on for all the users of Caribou at three a.m., and went home.
The next day, all the engineers using Caribou saw ads targeted to the content of the messages in their inboxes. They were not pleased. Paul recalls that Marissa, with whom he shared an office, was furious.
"We agreed not to do this," she insisted. "And then you went off and did it anyway."
"I don't remember ever agreeing to that," Paul replied. "Maybe you said not to, but I never agreed to anything. I'm not really that agreeable a person that I would ever agree not to do something."
Larry and Sergey weren't furious. They were impressed. The idea of content targeting had been around for a long time, but only as a theoretical concept. It was even on the engineering project list, but assigned a priority of one—the lowest—on a scale of one to five. Everyone assumed it would be hard to build and not terribly accurate. Yet Paul had built a working prototype in a couple of hours, basing it on a self-described "stupid hack" he had written for another project.
The founders, according to Paul, "were more open-minded about it than most of the rest of the company." They immediately saw the potential for targeting ads, not just to email, but to the content of
any
web page that contained text. They pulled together a team in September 2002 to start working on a more robust content-targeting system under the code name "Conehead." A month later, Google content-targeted ads started showing up around the posts in the old Deja News Usenet pages that had been incorporated into Google groups. A month after that we paid for low-cost remnant ad space—impressions that had not been sold and would have displayed house ads or PSAs—on small test sites like the ever-agreeable
Wunderground.com
to see if we could earn more from the ads than it cost us to pay for the media. In March 2003, Google's content-targeted ads publicly launched, expanding the places our ads could potentially appear to include almost every web page on the Internet. Google shared revenue with sites (which we called "publishers") that ran our ads, so small businesses, blogs, and even larger content producers struggling to sell their own ads could now turn on the cash spigot by adding a few lines of Google's code. In June 2003, we even made the sign-up process for publishers self-service (like AdWords for advertisers), calling the new service "AdSense."
*
AdSense gave Google an incentive to help grow the amount of content online. The more content on the Internet, the more potential revenue for us. Larry's idea of digitizing all printed pages suddenly became a much higher priority. We launched Project Ocean in late 2002 to begin moving the "ocean" of offline content online. That initially meant sending Googler Cari Spivak to a giant used-book sale in Phoenix, where she bought twelve thousand volumes, loaded them in a rented truck, and shipped them back to Mountain View to be scanned. Soon Google would be signing agreements with university libraries to scan all their printed materials and battling publishers and authors in court over copyrights. And Google bought Pyra Labs, a small company offering tools for anyone who wanted to create and host a blog at its site
Blogger.com
.
*
Google's second billion-dollar idea had been much easier to implement than the first.
For Paul, the experience confirmed the power of prototyping to give definitive answers far more quickly than theoretical discussions. "Experiencing something is much more powerful than just talking about it," he reflected. "I didn't think content-targeted ads would work, really, but I thought it would be fun. I spent a few hours working on it. It wasn't that I believed in it that strongly, it's just that it was really easy." Once people saw the prototype in action they realized, whether they liked Paul's implementation or not, that content targeting could be done.
Google learned a lesson, too. Paul had put together a prototype not specifically called for by his project, just because he found it interesting. He admitted to being easily distracted and had gone off on tangents before, but usually his manager reeled him in. When he added a primitive spell checker to the product search he was building, Urs admonished him for having too many side projects and asked him to stay focused. That attitude changed after he developed the content-targeting prototype. Not just for Paul, but for the company.
"I feel like the concept of twenty-percent time came out of that," Paul told me. "I don't think it was ever specifically stated, but it was more officially endorsed after that." "Twenty-percent time" was a mandate that all those in engineering spend one day a week thinking about something other than their assigned projects. Ad targeting in email had not been assigned to anyone, and most who heard about it vehemently opposed it. Paul built it anyway, and the company's thinking shifted overnight. Other engineers had also done side projects that looked promising, like Krishna's Google news service. Larry and Sergey wanted to encourage that kind of behavior.
At first, Paul observed, APMs ambitiously tried to organize the twenty-percent time so it wouldn't be a wasted resource. "That was completely the wrong mindset," he said later. "'Oh, these engineers are working on random things. We need to coordinate them and manage it.' The real value is that people will do things that everyone thinks are a waste of time. That's where the big opportunities are. It's an opportunity because other people don't see it." Google itself was a canonical example. No other companies had thought search was important. If they had, Microsoft or Yahoo would have invested more heavily in technology and Google would never have gained such a big head start.
Overture observed our trials with content targeting. They didn't want us to have the "first mover" advantage in a new market, so in February 2003 they acted as if they had beaten us to the punch again—as they had with search-related advertising—by claiming that they would soon launch a "contextual advertising" service. A week later, our service was open and ready for business. As an astute poster on WebmasterWorld noted, "Overture announces first. Google launches first." Underpromise and overdeliver. I was gratified to see that people actually noticed. Google was again in front of the competition, and Larry and Sergey intended to stay there.
Overture's Content Match service didn't go live until the end of June. One reason for the lag may have been that Overture had planned to base their technology on content-targeting software from Applied Semantics, a company in Santa Monica. In April, Google bought Applied Semantics, bolstering our own capabilities while cutting the legs from under our competitor, which now had to look elsewhere for the technology or develop it themselves as we had done.
The firm was our largest acquisition to date, and we welcomed our Southern California colleagues at TGIF with beer and sushi served from atop a custom surfboard painted with a logo saying, "Google Santa Monica." The board was an apt metaphor. Content-targeted ads were the next big wave in online revenue, and we intended to ride it all the way to the beach.
M
Y THIRD YEAR
at Google was rushing to a close. Everything shifted as the company transformed itself from startup to global powerhouse. New patterns and rituals slid into place, locked down, and began to feel permanent.
In October 2002, our department took up residence in the Saladoplex building with the AdWords team, my seventh (or eighth?) move since joining the company. I now had a closet office of my own with a wall-sized whiteboard and a window overlooking the patio. There was a micro-kitchen twenty feet away, and the gym—a converted computer mainframe room with an elevated floor and sub-zero air conditioning—was just around the corner. I felt completely embedded in my cozy corner of the organization and amazed at how many Googlers swarmed around me. (Ants, too. They found a hole in my window's insulation and a path into my Google gumball dispenser.) There were so many Googlers that the company stopped giving out stock options to new hires. Or, rather, deferred them. The government said that if we had five hundred shareholders we would have to make our financial records available, as if we were a public company. Instead, we kept hiring, but held back stock grants. The SEC would have some questions about that when we filed for our IPO.