Chapter 20
Where We Stand
USERS WERE COMPLAINING again. We heard a rising chorus of annoyance with pop-up ads appearing when people did Google searches. The ads opened new windows, cluttered users' desktops, and irritated the hell out of them. Either we stopped running pop-up ads, our users demanded, or they were prepared to stop using Google.
Google never did run pop-up ads. Others just made it look as if we did. A number of companies distributed free software for file sharing, media playing, and the adding of smiley faces to email. When users downloaded those programs, they unwittingly loaded their machines with code that launched pop-up ads or even collected credit card numbers and other personal information.
Some software surreptitiously took ownership of computers and made them "slavebots" that could be harnessed as part of a gigantic network to launch denial-of-service attacks or send spam. Techies called these parasitic programs "malware," "adware," or "spyware." We lumped them all together and called them "scumware."
Matt Cutts hated scumware. Intensely, personally hated it. His job focused on blocking people who tried to trick or "spam" Google into listing their sites higher in our results. He fought "black hats" every day, and scumware distributors were the worst of the worst. It sickened him that users thought we were the ones degrading their Google experience. In late 2001, he began monitoring the rise of scumware and pleading with any Googler who would listen to do something about it.
I was with Matt. I loathed seeing notes from users threatening to quit Google because of something over which we had no control. I worked with him to draft a lengthy email response, in which we explained what was happening and how to fix it. People replied with apologies and thanked us for alerting them to problems they hadn't been aware they had. Matt wanted to do more. He proposed that the Google toolbar include software that killed pop-ups and that we forward complaints we received to the FTC. He also suggested we post a note on our homepage explaining that Google was not at fault.
Wayne Rosing supported the idea of a "full-scale crusade-jihad" against those responsible, but others worried about the danger of declaring war. We suspected there were hundreds of scumware creators, and we knew they could be ... scummy. And spiteful. They didn't like to be thwarted, and they had no scruples about attacking those who tried to stop them. If we aggressively pursued them, they would target our site to make an example of us, causing even worse problems for users. Marissa suggested a compromise. If we could detect that a computer had been infected with specific scumware applications, we could show a message telling the user what to do about it. With that approach, we wouldn't confuse people who weren't experiencing problems and we wouldn't make ourselves too broad a target.
That solution turned out to be infeasible, so we fell back to posting a note linked from the homepage. We would tell users their pop-up ads weren't coming from us and casually mention that we did have our own, very discreet, very targeted, keyword-advertising program. Two marketing objectives satisfied simultaneously.
Larry didn't like the second part. Most of the people seeing our homepage would never advertise with us and might not even know we ran ads. Why disillusion them if they had no need to know? Larry never wanted to give people more information than he thought it was useful for them to have. I deleted references to AdWords.
In January 2002, we added a line to the homepage: "Google does not display pop-up advertising. Here's why." It was linked to a page that began, "Google does not allow pop-up ads of any kind on our site. We find them annoying." The response was immediate and positive, and I found it intoxicating. Google was becoming my own personal publishing platform. Mentalplex, the 9/11 news page, "Ten Things We've Found to Be True," and now "No Pop-Ups." The hits kept coming. We had built a global bully pulpit and my voice rolled forth from it. My thoughts, my ideas, my imprecations would be seen by more people than read the New York Times or watched a network newscast. I was the man behind the curtain giving voice to the all-knowing Oz. I tried to keep my ego in check.
The day after AdWords Select launched, the Associated Press ran a story about the service that said in part, "Online search engine maker Google Inc. is introducing a program that allows Web sites to be displayed more prominently if sponsors pay more money—an advertising-driven system derided by critics as an invitation to deceptive business practices." The article portrayed us as no better than Overture. It was flat-out wrong, yet major news outlets around the country ran it verbatim. The word "bastards" got a real workout in the Googleplex that day.
We had so carefully distinguished ourselves from the evil diminishers of search integrity, and all for naught. Cindy jumped on the AP to issue a correction, and they did, but she also reconsidered her original decision not to issue a press release about AdWords Select. She maintained a reporter-centric PR strategy of close communication with key journalists rather than "press-releasing" every burp, hiccup, and sneeze happening at the company. The strategy worked fantastically well most of the time, but when a reporter got a big story wrong, there was no official Google version to contradict it. Cindy and PR manager David Krane filed copy at two a.m. with the PR Newswire, saying, "Google's unbiased search results continue to be produced through a fully automated process and are unaffected by payment."
The AP story had been a fluke, an anomaly in a pattern of favorable press, but Cindy knew things would change. No one stayed beloved forever. Two days later she began formulating a "credibility campaign" to emphasize that not all search companies were created equal. We would use our own site to present our unfiltered messages in coordination with op-ed pieces in newspapers and executive speeches to select audiences.
"I want to kill the perception that we're selling our search results ASAP," she told us. "Our brand has been injured and we need to fix it. We're Google! Let's be outrageous and daring and have some fun with this."
Feelings ran deep on the subject of paid placement. When the topic of Google's refusal to sell placement came up on the geek bulletin board Slashdot, the first posted response was "I swear I want to make love to this company."
A self-identified Overture employee didn't share those warm and fuzzy feelings. "As for the claim by Google that they are pure," he asked plaintively, "why are they getting into the ad search business?" His implication seemed to be that the whole business was tainted. I didn't think so. You could present useful ads, but you needed to make it clear they were ads. It wasn't hard if you were willing to give up the revenue derived from deceiving users.
Larry and Sergey took the long view. Overture and the portals were training users not to click on links, because when they did, they felt cheated. It was our goal to make ads so useful that people would actually go out of their way to click them, even knowing that they were ads and not search results. To our founders, not being evil equaled sound business strategy.
My first contribution to Cindy's credibility campaign explained that principle. "Why we sell advertising, not search results," I wrote on our homepage in March 2002. The link led to a page that began, "In a world where everything seems to be for sale, why can't advertisers buy better position in our search results? The answer is simple. We believe you should be able to trust what you find using Google."
It didn't generate as much interest as our "No Pop-Ups" message, but our sales team loved it. It gave their clients a rationale for our refusal to offer pay-for-placement and detailed why that made us more ethical and more effective as an advertising medium. Our business-development team, though, had qualms about the closing: "Other online services don't believe the distinction between results and advertising is all that important. We thought you might like to know that we do." What about our partners like Yahoo? Would they view this as a swipe at them? After all, they ran Overture's ads above our search results.
It didn't help that with our Yahoo contract up for renewal, Inktomi suddenly got aggressive in attempting to win back the business, asserting that users introduced to Google on Yahoo's site would j
ust search directly with Google in the future. Why would Yahoo let Google siphon off their audience? Inktomi even drove a mobile billboard around Yahoo's campus with the message "Do you, uh, Google? Google is stealing your users. A friendly reminder from Inktomi." We debated surreptitiously pasting the words "bringing you customers" over the phrase "stealing your users," but we decided not to legitimize our rival's feeble ploy.
Omid was not amused. He knew the perception was spreading that Google was not a friend of portal sites, especially since we were now openly seeking partners not just for search results but also for ad distribution. Sergey turned to marketing for data proving the perception false. We couldn't find any, though I spent months looking.
Our Earthlink win had cracked the icy stasis locking the search players in place. Suddenly conventional certainty was set adrift. Google had entered a new industry and won an account from a firmly ensconced leader. We had flawlessly implemented a substantial and complex advance in our back-end systems and transitioned from one economic model to another. Those accomplishments could easily have absorbed the full focus of a competent tech company for years. It was becoming clear that Google was more than just a competent tech company.
At the tail end of 2001, I had convened a group at Cindy's request to begin thinking about Google's evolving position in the marketplace. Since then, Susan, Sheryl Sandberg, Cindy, and a couple of other marketeers had gathered every few weeks to try and pin down Google's protean essence. We called our initiative "Baby Beagle," in homage to Darwin. Our corporate identity had morphed with our entry into ads syndication—but into what? We didn't want to be pegged as a portal, but we had outgrown the notion of being only a search engine.
Our group couldn't reach consensus. It was like the old story of sightless men describing an elephant by touching its leg, its trunk, its back. I needed to talk to someone who saw the whole picture. I needed to talk to Larry. The hour I spent with him and Sergey probing their vision for Google gave me my best look at their motivations and aspirations for the company. Cindy was the only other person in the room. It wasn't a press interview. They had no reason to shade their views or filter their thoughts. They expressed what they truly and deeply believed.
We spent the first fifteen minutes talking about what Google was not and what we would never do. Larry wanted Google to be "a force for good," which meant we would never conduct marketing stunts like sweepstakes, coupons, and contests, which only worked because people were stupid. Preying on people's stupidity, Larry declared, was evil.
We wouldn't mislead people like our partner Yahoo, which at the time was experimenting with a pay-for-inclusion program that sold placement in their results. Google wouldn't treat employees badly or sell products that worked poorly. We wouldn't waste people's time—a point Larry emphasized again and again.
We need to do good, he said. We need to do things that matter on a large scale. Things that are highly leveraged. When I asked for examples, he mentioned micro-credits in Bangladesh and the Rocky Mountain Institute and talked about changing business systems to make them environmentally friendly while saving money. He also talked about distributed computing, drug discovery, and making the Internet faster. And that wasn't all.
We should be known for making stuff that people can use, he said, not just for providing information. Information is too restrictive. In fact, we shouldn't be defined by a category, but by the fact that our products work—the way you know an Apple product will look nice and a Sony product will work better but cost more. We're a technology company. A Google product will work better.
We don't make promises and then break them.
If we did have a category, it would be personal information—handling information that is important to you. The places you've seen. Communications. We'll add personalization features to make Google more useful. People need to trust us with their personal information, because we have a huge amount of data now and will have much more soon.
Here his eyes took on a faraway look and his words came faster. Sensors are really cheap and getting cheaper. Storage is cheap. Cameras are cheap. People will generate enormous amounts of data.
Everything you've ever heard or seen or experienced will become searchable. Your whole life will be searchable.
Our conversation ended on that note. Not once did the subject of making money come up. Not once did he talk about advertising revenue or syndication or beating Overture or CPC or our new Search Appliance.
I was probably a naive middle-aged dreamer, because looking back at it now, I see there was nothing truly extraordinary about what Larry described. But when I walked out of his office I believed that for the first time in my life I had been in the presence of a true visionary. It wasn't just the specifics of what he saw, but the passion and conviction he conveyed that made you believe Larry would actually achieve what he described. And that when that day came, he would already be thinking another fifty years ahead. My respect for our two capricious, obstinate, provocative, and occasionally juvenile founders increased tenfold that day.
There were other glimpses of Larry's thinking. He and Eric shared a list of possible strategies that included Google as the publisher of all content, where users would pay us and we would reimburse the creators of everything from books to movies to music. Google as a provider of market research and business intelligence based on what we knew about the world. Google as an infrastructure platform and communications provider tying email and web data together. Google as the leader in machine intelligence backed by all the world's data and massive computing power that learned as it went along.
He had no small plans.
Eric, on the other hand, was the voice of corporate pragmatism. These grand schemes would have to be paid for somehow. "Any chart that goes up and to the right is good," he assured us. And, "I like to watch cash in the bank." I got the impression he shared my concern for all the things that could go wrong.
One fear I knew Eric had was of clowns. Specifically, the bozos who showed up at a company when it reached a certain size and bloated it with bureaucracy and bogged it down in mediocrity. Google's hiring guidelines explicitly stated we should only add people smarter than we were.
That's why we started running a line on the homepage that said, "You're brilliant. We're hiring." The engineers loved it.
I hated it. To me it reeked of arrogance and went counter to our "say little, do lots" brand strategy. I had opposed it when we ran it previously, but Marissa insisted the data showed it garnered more résumés than any of our other job-related lines. I got nowhere pointing out that a minuscule percentage of the people reading it on the homepage would be qualified to work at Google. Larry and Urs were willing to waste a few hundred million impressions to reach the dozen or so people they might consider hiring.
The page at the other end of the link had been written entirely by Jeff Dean. The word "exceptional" appeared three times in the first paragraph, and "problems" showed up four times in the next two sentences. I offered to smooth out the rough edges and nearly gave Wayne Rosing a heart attack.
"No!" he exclaimed. "Leave it alone! Please!" It was a page written for geeks, and if Jeff, our own über-geek, liked it, marketing's touch would only taint it. Cindy encouraged me not to let engineering roll over our department, so I sat down with Jeff and went through the copy line by line, making helpful suggestions. As I made the edits, Jeff said he liked most of them—then, as soon as I left, he undid them all. He knew what appealed to him and saw exactly how it would appeal to others like him.
It would be easy to assume from this anecdote that Jeff thought he was brilliant and was arrogant about it. That wasn't the case at all. In fact, Paul Bucheit told me, Jeff kept everyone humble. "You can't get up and be an asshole about being smart," Paul explained, "because Jeff's smarter than you and he's not an asshole." I think Jeff just looked at brilliance as a quantifiable asset. Since brilliance was a parameter for our search, it was best to specify that in a forthright manner.
; I knew it would be pointless to keep fighting, and the page went up as written. I was learning to pick my battles.
The Copyright Crusade
One battle picked us. The Church of Scientology filed a complaint under the Digital Millennium Copyright Act (DMCA) demanding we remove links to Operation Clambake (www.xenu.net) from our search results. Operation Clambake, based in Norway, sought to expose what it claimed to be unethical practices within the church. The DMCA was a federal law requiring companies to remove content that an owner asserted was protected. In this case, that content included some of the church's internal documents and photographs on the Xenu website. We had no doubt the church would sue Google if we did not comply with the letter of the law and remove the references to xenu.net from our search results for the term "Scientology."
Ironically, while we were intent on keeping Google's internal processes private, many on our staff supported the first amendment rights of church dissenters to expose Scientology's secrets. I heard grumbling in support of those threatening to boycott Google for kowtowing to "an oppressive and censorious organization." The law was the law, however, and the DMCA was the status quo. But at Google the status quo was nothing more than an inconvenience to be improved upon as time allowed. Evidently, the time for fixing DMCA removal requests was now.
Matt Cutts led the charge. He proposed we put on our game face and drop pointed hints about how far we would go to defend our results, implying that we were committed to a more combative stance than we were actually prepared to adopt. He drafted a polite letter to his contact at the church (from whom we had received previous complaints) in which he laid out a number of paths Google might take, from publicizing that we had eliminated results at their request to letting the courts settle the issue.
I'm Feeling Lucky: The Confessions of Google Employee Number 59 Page 34