The second fix the new system needed was to correct a big problem that GoTo faced. GoTo displayed the price bid by each advertiser, and advertisers kept lowering their bids because they could see they were paying more than they needed to. If the high bid was twenty-five cents per click and the next bid was twenty cents per click, the top bidder was paying four cents too much. A smart advertiser would lower his bid to twenty-one cents.
Salar was intent on getting advertisers to tell us right at the start the maximum amount they were willing to pay for a click. He wanted to charge that full amount, just as Overture did, but he didn't want the advertiser to keep rebidding to lower it.
Eric showed him a solution. "The amount that you bid shouldn't be the amount that you pay," he said. He envisioned an eBay-type auction where the advertiser would pay the minimum amount necessary to win a position in the rankings. Eric had never heard of William Vickrey, the Nobel laureate who had created a "second-price auction" model; he worked out the idea himself. It just made sense to him that instead of charging as much as an advertiser was willing to pay, we should automatically lower the cost to the minimum amount required. Then advertisers would have no incentive to lower their bids, but they would have an incentive to raise them when the bids below theirs increased.
At first, Salar resisted the idea of a second-price auction, because it would confuse advertisers. They would have to trust us to lower their bids, and Salar wasn't sure they would be willing to do that.
Eric saw a fundamental difference between his approach and Salar's. "We both contributed to the design," he said, "but Salar was always looking at the product from the point of view of the advertisers, who he knew were the customers. They wanted transparency. They wanted control. I was looking out for the users."
Eric believed we should only show advertising when it was useful. He refused to put anything into the product that would weaken that principle. He and Salar went back and forth on the auction model and the need for secrecy about ad scoring. Salar was persuasive, but Eric brought an advantage to the debate about how the systems would be configured. "I was in charge of actually building them," he told me, "so as it turns out, I won most of the arguments."
It took both Eric and Salar to convince Larry and Sergey, who still hadn't signed off on offering CPC pricing, let alone on a method for implementing it. Salar explained the advantages for syndication. Eric showed how the auction model fixed GoTo's bid-lowering problem and suggested the changes might protect Google from GoTo's patent claims. Salar argued that different keywords had different values, so a fixed-price model made no sense.
Larry remained unswayed. It was only after Eric pointed out that ad quality and search quality were related—that with the changes we could control the relevance of the ads we displayed and improve the user experience—that he finally consented. As Eric put it, "Larry had always been the biggest champion of getting the best results for users. He was fine with making things tough for advertisers when we needed to."
In the end, Larry and Sergey weren't totally convinced, but they agreed to move ahead. That was not unusual for them. If you brought enough passion and logic to the fight, they'd take a chance that you might be right. Eric assumed they would rake him over the coals if it didn't turn out well, but he was glad to get the chance to move ahead and wasted no time putting together his team. He knew GoTo had a head start and was gaining momentum.
A Friend Accepts an Enemy Overture
Google's first great challenge had been gaining primacy in search. Its second would be winning the war for advertising revenue. Eric and Salar were preparing the company for that battle, but those of us not in the ads group at the time were left in the dark. All we could see was GoTo scoring deals and racking up sales. Not only was GoTo generating revenue, it was running the table, locking up portal sites with multi-year contracts. When our salespeople pitched Google ads to GoTo customers, they came away empty-handed, shaking their heads.
To make it worse, GoTo's sales team kept sending us things, as if we were a potential client. Every week, a new "gift" would arrive. A notepad. A golden egg. A plastic goose that lit up. A hand-delivered pie. Omid would display the latest item at TGIF each week and then send it flying with a kick. Except the pie. I'm pretty sure he ate that.
GoTo's management continued to crow about their successes. In August 2001, one month after Eric's team was given the green light to build a CPC system, a GoTo executive appeared at a Jupiter Research conference and claimed that all search would be paid for in the future. "He said that if users don't like the results, they can keep clicking," a Googler who attended the conference reported back. "It's no problem, since they'll eventually find what they want."
In October 2001, GoTo signed a deal expanding their relationship with Ask Jeeves. They announced proudly that they had attained profitability and that they were changing their name to Overture.
Overture continued the boasting at the Direct Marketing Association conference in Chicago later that month. I was in our small booth off to one side of the large hall, which overflowed with direct-mail houses, list brokers, specialty printers, and foreign governments offering low taxes and cheap labor for call centers. We were handing out magnets and t-shirts and riding our Googlized scooter around the floor promoting a contest for free ads. We stood out from the conservative coat-and-tie crowd, members of which came by to tell us again and again, "I love Google!" The only booth generating more interest was staffed by Hooters waitresses handing out free chicken wings as the Bears game played on a big-screen TV.
Overture's chief operating officer, Jaynie Studenmund, gave a speech. She claimed Overture was having a "Barry Bonds year" and acknowledged that Google was "a fine company" and "absolutely in the business of making money off search listings too." She told the crowd that Google "has something called paid placement that's different from Overture's paid placement." She meant the clearly labeled ads next to our search results. I had to restrain myself from heckling her.
Later, Studenmund stopped by our booth with a half dozen of her colleagues in tow. She wanted to chat. Did we think the event was worthwhile? A success? I candidly gave her as much misinformation as I could, and, after helping themselves to our tchotchkes, she and her entourage floated off. They didn't seem too concerned about us. We were just a hungry stray fighting a pack of others to eat the scraps of their success.
Then, on November 13, 2001, Overture dropped a bomb on Google. They were signing a five-month contract with Yahoo to deliver ads targeted to search keywords. Yahoo would keep paying us for our search results, but those results would be monetized by another company. They would skim the cream after we fed the cow and milked it. Larry and Sergey bit their tongues, but they were furious at the setback. They couldn't take it out on Yahoo, which was still our biggest partner, and bad-mouthing Overture would imply that Yahoo had made a mistake. There was no immediate engineering fix to our problem, but that didn't mean we had no way to respond.
We Hold These Truths to Be Self-Evident
While we couldn't talk trash about our competitors, we could use our site to talk directly to our users about the principles that made us unique. I had some thoughts about how to express that.
"Never settle for the best," I wrote, and quoted Larry's vision of the perfect search engine, which would "understand exactly what you mean and give back exactly what you want." Only Google was engaged in an endless struggle for search perfection, I explained—a noble, perhaps quixotic goal that set us apart from portals laden with ancillary services, and from ad networks that put the interests of advertisers ahead of users.
To demonstrate Google's purity of heart, I compiled a list of "Ten Things We've Found to Be True." I tried to distill every interaction I'd had with our engineering team and with Larry and Sergey into discrete nodes that together mapped the attitudes of the company's creative force. I laid them out in priority order:
Focus on the user and all else will follow.
It's best to do on
e thing really, really well.
Fast is better than slow.
Open is better than closed.
Democracy on the web works.
You don't need to be at your desk to need an answer.
You can make money without doing evil.
There's always more information out there.
The need for information crosses all borders.
You don't need a suit to be serious.
Great just isn't good enough.
The fourth point stuck in Larry's craw. It described Google's preference for free, community-developed, open-source technology like Linux, which was increasingly viewed as a threat to the dominance of Microsoft's Windows operating system.
"Don't moon the giant," Larry admonished me. Larry, along with Netscapees like Omid, knew what came from waving red flags at the Beast of Redmond. Netscape had famously "mooned the giant" by boasting that their browser would turn Microsoft Windows into a "mundane collection of not entirely debugged device drivers." Microsoft had responded with their own version of "mooning." They bundled their Internet Explorer browser with Windows to turn Netscape into a barren, lifeless realm by "cutting off their air supply." Google's air supply was still coming through a rather narrow tube, and Microsoft could easily throw kinks into it by making changes to Internet Explorer or focusing on search themselves. I cut out point number four.
That brought the total number down to ten. I had intentionally included eleven because the last point went on to state that we would "always deliver more than expected." I tried to add another point to replace the deleted one, arguing that taking it to eleven would give the list a sort of postmodern, ironic hipness, but the executive staff said to let it go. They were a pretty literal bunch.
They had no objection to the rest, including my favorite, "You don't need a suit to be serious." I hated neckties. I hated ironing shirts.* I hated scratchy trousers and shoes that pinched. I wanted to elevate my sartorial disdain to the level of a corporate value and, God bless Google, they were willing to do it.
Not all items on the list would age equally well. In fact, Google amended it not long ago to make it clear that doing "one thing really, really well" could be extended to cover products like online chat and financial news. I knew that Google would ultimately outgrow search alone, but search was the battle we were fighting at the time and the way our users defined us. Who were we to reject the brand they had created in our name?
Only one point gave me pause: "You can make money without doing evil." It was at the heart of our response to Overture, but I felt uncomfortable talking about our desire not to prosper at the expense of our souls. I thought it could be perceived as naive and would invite criticism every time we didn't live up to our moral code. It also went against another core value, one I hadn't listed: "Underpromise and overdeliver." Up to that point, we had always been careful not to overstate our accomplishments or brag about what we had up our sleeves.
We never boasted about our willingness to sacrifice ourselves to make good things happen for partners, advertisers, or users, and we rarely talked about what we might do for them in the future, but we strove to exceed their expectations. Shutting down Google.com for Netscape had been a prime example. While Larry and Sergey would negotiate every penny out of a potential client's pocket, our goal was to give added value in return. That attitude was uncommon among tech companies. Microsoft, in particular, gained notoriety as the issuer of vast quantities of vaporware, products that they touted as industry standards, but which came to market much later than originally projected, if at all. It's easy to understand why companies did this—it was a way to lay claim to new markets and kept aspiring players from entering them.
We abhorred vaporware. Though recently Google has preannounced products and even whole industries it intends to revolutionize, in its early days the company kept launches secret and downplayed features. We wanted people to discover some things on their own. We'd launch a calculator and let users figure out it could tell them the number of hands in a fathom. We'd launch a spell checker that offered suggestions not just for common nouns but for the names of people in the news.
"Underpromise and overdeliver" became as important a mantra to us as "Don't be evil."
Eventually, users and analysts began speculating about what Google might develop next. Would it be a phone? A travel service? A carbonated beverage? The rumors were rife, and the lack of confirmation or denial from Google gave competitors pause. Google was no longer just a search engine. It was a technology company that solved hard problems. That made the addition of Google's name to any product plausible.
Understating our accomplishments could be frustrating at times, especially for salespeople. AdWords ads got up to fourteen times the clickthrough rate of untargeted banner display ads, but we only let sales reps tell clients they were three times better. We wanted to sound realistic to potential customers and keep our true performance hidden from competitors. If we delivered double or triple what we promised, who would complain?
But there was no way to overdeliver on not being evil—and there were an infinite number of ways to fall short of that goal. Evil, as an operating principle, was a common discussion topic around our cubicles. In late 2001, in Silicon Valley at least, many saw Microsoft as the primary practitioner of the dark arts in technology, using their monopoly power to corral innovative startups that might turn their Windows cash cow into hamburger helper.
The "Don't be evil" mantra had already taken root within Google when I composed my list of "Ten Things." Paul Bucheit came up with it in 2000 at a "core values" meeting held to codify the way Googlers should act toward one another. It was not intended to regulate our behavior toward non-Googlers, nor were the values supposed to be disseminated outside the company.
According to Amit Patel, Paul became disaffected with all the "corporate"-sounding suggestions his colleagues proposed—things like "Treat each other with respect," "Honor commitments," and "Don't be late for meetings." They were boring, and they were too specific. It was bad coding hygiene to build an itemized list if you could apply a general rule.
"Aren't all of these covered by, 'Don't be evil'?" Paul asked.
No one took him seriously. The meeting concluded with a list of eleven core values, which HR asked me to help wordsmith. "Don't be evil" wasn't one of them. The meeting left Amit unsatisfied, and he took it upon himself to proselytize the Word of Paul. Soon, "Don't be evil" began blemishing every markable surface like brown spots on ripening bananas. I had a rolling whiteboard in my cubicle, and one day when I came back from lunch, "Don't be evil" was neatly printed in one of its corners. I saw the phrase scrawled on conference room walls and twirling across laptop screensavers. Others saw it too. I had to assure job applicants, vendors, and visitors that it didn't mean the company was fighting Satanic urges.
It was intimidating to have a corporate commandment stare down at you wherever you went—a dry-erase Jiminy Cricket looking over your shoulder, passing judgment on your every action. That was Amit's intent. Its very simplicity made the phrase unforgettable and gave it the force of an irrevocable law.
"'Don't be evil,'" Paul explained, "is about not taking advantage of people or deceiving them. Anything deceptive is evil. So if we put up search results, move them higher because someone paid us, that's deceptive, that's abusing trust." Paul wanted Google to be the anti-evil company. Amit's marketing campaign sold the staff on formalizing the credo. Once it became a cultural meme, it was impossible to uproot. The effect was as if Amit had been scribbling with a permanent marker directly into our collective consciousness.
"I also thought it would be a good value because it would be difficult to remove once it was in," Paul admitted. "It wouldn't look too good to get rid of, 'Don't be evil.' Besides, Microsoft had a monopoly on evil. We didn't really want to compete."
The idea of not doing evil seeped into conversations as a criterion for evaluating products, services, and life decisions. People brought their own spin
to interpreting what it meant:
"If the ads looked more like search results they would generate more revenue. But wouldn't that be evil?"
"Resist evil. Don't make the toolbar less functional to appease Microsoft."
"You took the last éclair and didn't finish it? You are evil incarnate."
"I've noticed that people have this strange definition of evil," Paul observed, "which is, 'Anything I don't like.' In my mind, I can not like something, but it can still not be evil." But even strict adherence to Paul's original concern about selling placement in search results put us at odds with our industry. If pay-for-placement was evil, the market was in league with the devil. Overture was growing at an enormous clip and building a sprawling advertising network of sites running their ads.*
When we posted "Ten Things We've Found to Be True" on our website, we earned a handful of kudos from users for our stand in favor of integrity. But that did nothing to slow the growth of Overture, a juggernaut that now threatened to lock up all the advertising dollars flowing to search.
Some Positive Results
Overture's deal with Yahoo seemed to put them out of any competitor's reach. In December 2001, they signed a three-year agreement with Germany's biggest ISP. In January 2002, MSN announced they were testing Overture ads. They were everywhere and they were unstoppable.
Except. Except that by January 2002, Googlers had already been banging away in the Googleplex for almost two months on a working prototype of our new ads system. Just before midnight on November 15—two days after Overture's Yahoo announcement—Eric Veach had flipped the switch. The prototype had multiple bugs and lacked key features, but it could serve ads sold on a CPC basis through a real-time auction.
While Salar and Berkeley economist Hal Varian refined the bidding system, I thought about how we would sell the system itself. I compiled a spreadsheet comparing it to the original AdWords and to Overture. It looked very compelling on paper, but in reality it wasn't yet ready for public exposure. The gears and wires still showed through some parts of the interface, and some of the steps required to create an ad seemed counterintuitive. Birthed in the middle of the night, the newborn product could not yet survive in the cold competitive world into which it would soon be thrust. It needed to be wrapped in a user-friendly UI and given a name.
I'm Feeling Lucky: The Confessions of Google Employee Number 59 Page 32