by Adam Fisher
I’m Feeling Lucky
Google cracks the code
Google was a dot-com poster child, but unlike the others it managed to escape immolation in the dot bomb. Brin and Page had the good fortune to secure $25 million of venture money in 1999, just before the fever broke. They also had the good sense to bank most of those millions. It took them a few years to figure out how to make search profitable, but again luck was on their side. The money sustained the young company through the launch of AdWords, a self-service ad-buying tool, in 2001. Two years later, Google perfected their advertising system with AdSense, which turned the entire internet into a billboard that Google could then sell. The timing was, again, perfect. Google became enormously profitable at the very moment that the rest of the Valley was at its most desperate. Google was able to expand on the cheap. With ad money pouring in, Brin and Page went on a buying spree: sucking whole companies, buildings by the dozen, and thousands and thousands of PhDs into their ever-expanding headquarters: the Googleplex. The much-anticipated Google IPO in 2004 marks the start of what Silicon Valley calls Web 2.0—a reboot, a leveling up, a phase shift. Post-Google, the Valley began to abandon the notion of the internet as a free “cyberspace” that one could “surf,” and instead started to regard the web as a vast machine possessed of a native intelligence—which it could direct, program, even own.
Douglas Edwards: If you read Larry and Sergey’s original paper that they wrote at Stanford where they talked about creating a search engine, they specifically said that advertising was wrong and bad and it would inherently corrupt the search engine if you sold advertising. So they were adamantly opposed to the notion of having advertising on Google.
Ray Sidney: So when I originally met Larry and Sergey in 1999, the plan was to license search—no advertising. We were going to have the best search out there and portals and companies and whatnot, and we were going to license it.
Marissa Mayer: The first revenue came on June 24, 1999, and it was licensing. Netscape was basically paying Google to power their search.
Larry Page: OEM-ing a search: You can get revenue from that, but it is limited. People aren’t used to paying huge amounts of money for that.
Ray Sidney: Then people started reading about how much money was being brought in to various other companies by search advertising, and it was kind of decided that we were leaving money on the table.
Douglas Edwards: There was a lot of pressure to generate revenue, and so Larry and Sergey decided that advertising doesn’t have to be evil—if it’s actually useful and relevant.
Ray Sidney: We had this great website and that was really getting popular so maybe we should think about putting some ads up there? “Let’s try it out!” Thus was born the first Google ad server. They had to call companies: “Hey, would you like to advertise on Google?” It was a manual process. The first ads were July or August of ’99.
David Cheriton: The thing that many people don’t recall from that first web era was that a lot of the companies that started out as search companies introduced advertising and then compromised the search in favor of the advertising, so nobody trusted the search.
Paul Buchheit: Companies would just mix the ads in with the regular search results so people would think it was a search result. It’s kind of like fake news or something.
Brad Templeton: In most of the publishing business, there is a history of having a wall between editorial and advertising. Google really took that to heart. They tried to keep their wall between the advertisement and the actual editorial content pretty strictly divided. But all the other websites were basically tearing down the wall. Google said, “Let’s not do that.”
Douglas Edwards: That was something Larry and Sergey were adamantly opposed to. They felt that it was so corrupt and so disingenuous. It was like a religious thing with them, so they wanted to make sure that nobody would mistake Google ads for search results.
Paul Buchheit: Sometime in early 2000 there was a meeting to decide on the company’s values. They invited a collection of people who had been there for a while. I was sitting there trying to think of something that would be really different and not one of these usual “strive for excellence” sort of statements. I also wanted something that, once you get it in there, would be hard to take out.
Brad Templeton: “Don’t be evil” was the phrase.
Paul Buchheit: It just sort occurred to me.
Sergey Brin: We have tried to define precisely what it means to be a force for good—always do the right, ethical thing. Ultimately, “Don’t be evil” seems the easiest way to summarize it.
Paul Buchheit: It’s also a bit of a jab at the other companies, especially our competitors, who at the time were, in our opinion, kind of exploiting the users to some extent. They were tricking them by selling search results—which we considered a questionable thing to do, because people didn’t realize that they were ads.
Sergey Brin: We think that’s a slippery slope.
Brad Templeton: That’s what “being evil” would be, actually.
Scott Hassan: Then Larry came up with a great idea: having self-serve ads, where you could just go to a website and click a couple things, and then ads would be shown on the Google website, without talking to a salesperson at all.
Marissa Mayer: AdWords was “pay by credit card and we’re going to run your ad.”
Ray Sidney: That’s definitely important. You can call the big advertisers and say, “Hey, how would you like to start advertising with us?” But you can’t call up all the little people who are interested in spending a small amount of money to advertise whatever small things they’re selling—what you would call the long tail. So that was definitely key.
Douglas Edwards: One of the key innovations was that the ads were going to be created by the advertisers themselves and not in any way be vetted by Google. I thought that that was absolutely insane. I was sure that was suicide. And I told Larry as much. And so we did it anyway. We put it up on the home page as a link. And about two hours after we did that we got our first AdWords customer: Lively Lobsters in Kingston, Rhode Island. Ryan Bartholomew was the owner and sole employee.
Ryan Bartholomew: I didn’t even know what I was doing, I just thought it was cool to buy lobsters for five bucks off the boat and try to ship them to buyers. And I was just working around the clock trying to figure out ways to get my listing out there. This was back in the days of Yahoo and AltaVista and I don’t even remember the others. But Google was around and very new, and I was up at two in the morning playing around on Google looking for something or other.
David Cheriton: The web had grown from this very early stage where there was a small amount of content and you could sort of find things with the help of Yahoo, to the case where the content had grown so rapidly, and had attracted all of these compromisers and game players, that it became harder and harder to find things. And so that was the environment that Google was in.
Ryan Bartholomew: And there was this little box to the side that was looking for advertisers for Google. I had never seen that before and I ended up clicking on it and running an ad for “live lobsters.” I later found out that I was the first advertiser on Google, ever. It was a self-serve, all-text display ad. You bid on a keyword and you make the ad and you pay a certain amount for it.
Ray Sidney: So in a very real sense, every time you do a search query on Google, an auction happens behind the scenes, between the various advertisers that are interested in key words that are in the query.
Ryan Bartholomew: Because you are bidding competitively against other people for keywords. It was priced in CPM.
Brad Templeton: That’s cost per thousand views. Ten dollars is a very common CPM in the marketplace. If you prefer to think of it as a cost per view, you can think of that as a penny per view. But usually people use the cost per thousand. M is the Latin for “one thousand.”
Ryan Bartholomew: I spent eighty-three bucks, and as far as I could tell, I didn’t get a single order! I
was pretty pissed off. I ended up calling Google and I got somebody and I said, “This sucks!”
Brad Templeton: For many years nobody was able to really get advertisers to pay very much. So, basically, the philosophy of sites at the time was “Yes, what can we do for you, sir?” And so you had web pages and they’d have pop-ups and pop-unders and dancing bears and everything else. And whatever the advertisers would pay for, you would do.
Ryan Bartholomew: I asked them, “Why aren’t you doing pay-for-click? Why aren’t you charging per click rather than just on the number of impressions? I didn’t get anything for my ads!” So they took my feedback and that was that.
Brad Templeton: So “I want to pay per click.”
Ryan Bartholomew: And after a while they did move to a pay-per-click model.
Marissa Mayer: It was “we’re going to let you pay per click.”
Brad Templeton: And again the other sites are doing it.
Ryan Bartholomew: GoTo.com was the first pay-per-click engine, there was a company called FindWhat.com, there was this whole pay-per-click ad model around prior to Google, and I had already been doing pretty well with it.
Marissa Mayer: But we’re going to rank your ad based on a quality score and how much you are willing to pay.
Brad Templeton: The brilliant thing that Google said was, “Okay. Pay per click, but if people do not click on your ad, we’re not running it.” Google charges you a fee if someone clicks on your ad. But if your ad is not performing enough, they take you out of the way, so they end up getting a minimum for every thousand impressions they do. Therefore they won’t run your ad unless it’s paying them the CPM figure they want. That was the genius! Google found a way that the advertiser could think that they were paying per click—which is what advertisers like—but they, the publisher of the website, could get paid per thousand views, which is what publishers like. They found this marriage which gave them what they wanted, and gave the advertisers the illusion that they were getting what they wanted, and of course who wins? Google wins.
Ryan Bartholomew: It was called AdWords Select.
Brad Templeton: And this was—by that standard of that day—a complete shock. “What do you mean, you won’t take my money?” That was just a completely foreign idea to web advertising at the time—and it forced all the advertisers to write good ads that people would actually click on.
Douglas Edwards: And it wasn’t just advertising for lobsters. Ryan quickly understood the opportunity for arbitrage. Amazon had an affiliate program where if you sent someone to Amazon and they bought a book, Amazon would pay you. So Ryan began taking ads out on Google not just for lobsters, but for books that would link to his Amazon affiliate page, where he would collect commissions. And of course in the same way he was using Amazon to generate revenue, he began linking to adult sites that were paying for traffic.
Ryan Bartholomew: I just couldn’t make much money on the lobsters, so in 2001 I ended up selling that company and ended up focusing on porn basically. Google allowed you to bid on adult keywords and run traffic to the porn affiliate programs, which get people to give their credit card and sign up for a free trial and all that stuff. It was a really simple business model: I would bid on any sex-related keyword that I could get, traffic was cheap—especially for the obscure stuff—and I would send it to a page that was nothing but text links. It would break down whatever their interest was, any kind of fetish, and then they would click it and it would take them to a porn site targeted to that interest. At one point there must have been forty or fifty affiliate programs for everything under the sun.
Brad Templeton: I remember the first party they had after they moved to downtown Palo Alto where they had actually put a projector doing search queries onto the sidewalk of University Avenue and that’s how you knew where it was.
David Cheriton: And one of them showed me his laptop or some display, where you could see the searches that were being filtered out, and of course they were all sort of porn type searches. Without the filter, we would’ve all been arrested. And it just sort of struck me. I hadn’t really thought about the magnitude of the use of the search engine for searching for porn.
Ryan Bartholomew: The beauty of it is that it didn’t matter whether it was lobsters, porn, Aerosmith CDs, or snowblowers! It was all just arbitrage on the value of the traffic. This is what Google was built upon. This is ultimately what Google is. You take a general search and filter it by what people are interested in. The more accurately you can figure out what people are searching for, the more likely they are to buy that product or service. Amazon is kind of doing that, too; it’s a very sophisticated machine for the same idea—narrowing you down to exactly what you want.
Young Harvill: It’s still not really talked about a lot in Silicon Valley, but a lot of the killer app for some of the first technologies is porn.
Ryan Bartholomew: I remember spending $13,000 in one day on all adult-oriented traffic. And my margins on that traffic were something like 3X. So I must have made thirty or forty thousand dollars in one day. That was right at the peak.
Brad Templeton: This turned into Google’s great advertising success.
Ryan Bartholomew: And when the holidays rolled around I got a lava lamp from Google as a thank-you gift for sending them millions of dollars in business.
Scott Hassan: But AdWords was just about putting ads on the Google search engine. The real advance was AdSense in 2003.
Marissa Mayer: Which was “because you are paying per click or per lead, now we’re going to take your ad and put it on different websites.”
Paul Buchheit: AdSense, the content-targeted ads, was actually something that, if I recall, I did on a Friday. It was an idea that we had talked about for a long time, but there was this belief that somehow it wouldn’t work. But it seemed like an interesting problem, so one evening I implemented this content-targeting system, just as sort of a side project, not because I was supposed to. And it turned out to work.
Susan Wojcicki: It was a really novel idea at the time to serve ads that were targeted dynamically. People were saying, “This is a sports site, so we’ll serve a sports ad.” And we were saying, “No. We can actually look at the page in real time and figure out what this page is about.”
Paul Buchheit: What I wrote was just a throwaway prototype, but it got people thinking because it proved that it was possible, and that it wasn’t too hard because I was able to do it in less than a day. After that, other people took over and did all the hard work of making it into a real product.
Ryan Bartholomew: AdSense is just the inverse of AdWords. With AdWords you are paying for when the ad is clicked on, then you get that traffic and then it’s up to you how you make money on that traffic. AdSense is just the program where, if you are the publisher of a web page, you get a cut of what those AdWords buyers are spending. It’s like owning a billboard. You run AdSense if you owned a billboard and the person who is buying the ad on the billboard would be an AdWords customer—and Google is the middleman.
Douglas Edwards: Google manages the system: They collect both the advertisers and the sites that run the ads. And so there’s a viral effect that goes into play when you build a critical mass. The advertisers say, “Where are all the publishers?” “Well, they’re at Google.” And the publishers say, “Where are the most advertisers?” “Oh, they’re at Google.” So it feeds on itself until you become the predominant player in that space.
Ev Williams: That’s when Google came knocking. It was a really hard decision to sell Blogger to Google because Google was still private, no one knew what was going to happen. This was pre-Gmail, pre-anything but search and advertising. So that was confusing: “Why would they want to buy Blogger? We’re not a search engine.”
Douglas Edwards: It was really simple: Once we had AdSense, we wanted more places to put ads, and we thought Blogger was going to generate millions of pages of content on which we could put ads.
Scott Hassan: It expanded the reach that t
heir ad network had and allowed them to have a lot of inventory cheaply. It was a win-win for everybody.
Biz Stone: After Google acquired Blogger, I wrote an e-mail to Ev. And I said to him, “I feel like I’m the missing member of your team. And so, if you ever need to hire more people for the Blogger team—I’m your guy.” I was actually hired to be on the AdSense team—or whatever it was called in 2003—but I really worked on Blogger under Ev.
Ev Williams: And we had to go down to Mountain View, because that’s where the real Silicon Valley was.
Biz Stone: We were in “Building Pi,” of course. Nerd factory, right?
Brad Templeton: And by that point, they had become a pretty big company.
Larry Page: We finally moved into a big corporate park in Mountain View, where we are now.
Heather Cairns: We were moving into the old Silicon Graphics campus, where there’s still old Silicon Graphics people working there, and they weren’t too happy to see us.
Marissa Mayer: At that point SGI was doing poorly, so there were about fifty people on that whole campus.
Jim Clark: It was a sinking ship.
Heather Cairns: We are like, “Yay! Here are our pool tables and our candy! Yay! We’re Google!” And they’re looking out the windows at us playing volleyball and they’re like, “Fuck you!”
Jim Clark: They were upset that they had missed out on being part of Netscape.
Marissa Mayer: We’d been so disrespectful—loud and annoying.
Heather Cairns: We weren’t trying to be disrespectful. We were just stupid. We didn’t have a sensitivity to the fact that those people probably wouldn’t have jobs in a few months. They were clear on that. And they’re just watching the new blood come in—happy, eager, just bouncing off the walls.