CHAPTER 6

Google, Mountain View, 2004

The internet was just reaching the masses when Sergey Brin and Larry Page walked onto the palm-lined campus of Stanford University as graduate students in computer science. They developed an algorithm called PageRank to search the rapidly growing internet and released it into the wilds of Stanford’s network in 1996. Two years later they incorporated Google to market the search engine they’d developed. By 2003, the company they founded was on the cusp of going public and making them both billionaires.

Google wasn’t the first search engine, but five years after its founding, it was quickly becoming what the Wall Street Journal dubbed the “front door to the web.” It had set out to challenge Yahoo! for search supremacy at a time when many realized that search would grow into a big business. Yahoo! had become one of the first of the era’s tech darlings when it sold shares in April 1996.

Google grew rapidly, both in a business sense and in the number of its shareholders of record. The latter was a metric that the SEC focused on, requiring companies with more than five hundred shareholders to file a document—a Form 10—with financial information for public investors. As Google neared that threshold, Brin, Page, and CEO Eric Schmidt began to think about selling shares to the public. To do so, they needed more experts in finance. Former Treasury Department official Sheryl Sandberg had been acting as the company’s senior finance official, but as a senior executive in the sales operation she was already spread thin. The company hired George Reyes, a longtime executive at Schmidt’s former firm, Sun Microsystems, to be CFO, and looked for other talent.

In March 2003, Sandberg hired a Harvard Business School graduate and Goldman Sachs investment banker named Kim Jabal as finance director in the online sales department. Raised in a Chicago suburb, Jabal had a round face framed by a head of tightly coiled blonde hair. She wasn’t particularly ambitious, but she was interested in math and engineering. As an undergraduate at the University of Illinois at Urbana-Champaign, Jabal had distinguished herself by being named an Edmund J. James Scholar, recognition awarded to students with “outstanding academic records, high aptitude, solid reputation, and self-discipline.” She joined Arthur Anderson after graduation and spent much of the 1990s planning large-scale information technology systems for clients.

When she grew tired of that world, Jabal headed east to Harvard and then back west for two years in Goldman Sachs’s technology, media, and telecommunications (TMT) practice. Older than many of her colleagues at Goldman, Jabal chafed at the male-dominated hierarchy and the always-on culture of the investment banking world. After one recruiting dinner in Boston shortly after Goldman gave BlackBerrys to junior bankers, she hopped into a cab with her colleagues, only to witness everyone go head-down, staring at their phones. Jabal was appalled. Goldman thought it was offering a great lifestyle perk, but she didn’t see it.

When Sandberg came calling in late 2002, Google was one of the hottest companies in town. Jabal had joined Google within months of the arrival of Buyer, the former Credit Suisse analyst, who had weathered the various investigations into Wall Street’s IPO activities with her reputation intact. She was hired to help Google prepare for the IPO. The two women worked in different parts of the company—Buyer reported to the CFO, while Jabal was in a business division. But they would overlap during IPO preparations.

Buyer hadn’t lost her fondness for evocative language or her aw-shucks personality. When Fortune caught up with her in May 2003, she told the magazine that she’d been hired to be Google’s “official lug-nut checker.” The comment gave cover to her real duties and downplayed her importance. Her office was on Salado Drive in Mountain View, in one of a collection of low-slung buildings just east of the Bayshore Freeway that had become Google’s home. Employees ate food prepared in a small mobile-home-type structure across the street.

The Silicon Valley that Google inhabited when Buyer joined the company was a very different world than the one she had known at Credit Suisse. The dot-com crash had thrown cold water on the over-the-top marketing and hype that she had described to Frontline the year before. The river of IPOs had dried to a trickle. In 1999, 476 companies had sold shares for the first time, while 380 firms entered the public markets the following year. By 2001, the tally had fallen to just 80, and activity was even lower in 2002, when 66 companies sold shares for the first time.

The dearth of technology competition meant that Google had its pick of talent. The company hired undergraduates out of Stanford to do customer service. Its market position also meant the company could think about doing something different when it came time to sell its shares to the public.

Buyer soon began working on “Project Denny’s,” so named because a small group of executives had begun holding covert meetings about Google’s planned listing at the chain restaurant close to the offices. Regularly in attendance were a core group that also included Reyes and two of Google’s lawyers, general counsel David Drummond and a more junior member of the team, Jeff Donovan. By then Brin, Page, and Schmidt had come to a belief that there ought to be a better way of going public.

At the time of Buyer’s arrival, the IPO market was quiet, with Silicon Valley still firmly hungover by the dot-com crash of 2000 and 2001. And yet the group was not content to simply restart the IPO market and bring heat back to a technology sector that was still licking its wounds. They set their sights on reimagining the decades-old process, bringing an engineer’s mindset to something that had long been the redoubt of Wall Street.

The founders had witnessed the excesses of the dot-com boom and bust, and as they began to think about their own IPO, myriad investigations into investment bankers’ allegedly illegal activity would occasionally make their way into the news headlines.

In meetings at Denny’s and in closed-door discussions at Google’s offices, consensus began to settle on holding some sort of auction. Google was auctioning off thousands of ads every day on its site. There may not have been a company with more auction experience on the planet. If structured right, an auction could also allow Google’s users, individual investors scattered across the world, to participate on the same terms as the largest investors.

Google’s executives “were engineers and computer scientists,” Buyer said much later. “And for engineers and computer scientists you can always architect a better way.” As they went looking for experts, the small group of employees found Notre Dame finance professor Ann Sherman, who had done a lot of work on auction theory, and Hal Varian, dean of the University of California–Berkeley’s school of information management and systems, who had started working with Google as a consultant in May 2002. They examined a couple of telecom IPOs in Asia that had used an auction process to sell shares. Singapore Telecommunications (Singtel), for example, offered three tranches of shares, including a retail portion that was priced at a discount to encourage small investors to buy into a national asset. Singapore’s government made it easy to place orders through automated teller machines.

Google’s rumored IPO sparked intense interest. At a May 2003 conference hosted by JPMorgan Chase in San Francisco, Schmidt brushed off IPO talk with a smile. In August, Brin told another conference that Google discussed a possible IPO with its board only once every two or three board meetings and that the company had more pressing issues at hand. Google didn’t need the money an IPO would provide, Brin told the audience, due to its profitability and annual revenues. He also cited distractions that he wasn’t eager to embrace, such as meetings with Wall Street analysts and investors, or disclosing the company’s financials in detail.

In Mountain View, Google executives began thinking about hiring banks. They wrote a request for proposal (RFP), a document that companies sometimes sent to investment banks ahead of a formal pitch process. Several questions in it were about auctions or hinted that Google might be leaning in that direction.

Google was obsessed with secrecy. The writers of the RFP, including Buyer, included slightly different questions in each so that if one of the questions was leaked to the press, they would know which bank had done it.

At Morgan Stanley, the task of responding to the RFP fell to the technology investment banking team, led by Michael Grimes and a capital markets banker in his midthirties named Colin Stewart. Stewart was a Dartmouth College alum who was based in the firm’s Menlo Park office, a rising star who had recently returned from a post running equity capital markets in Asia. He reported to Jon Anda. Ted Pick, head of the bank’s syndicate desk and known for being surly, was also involved in the deal.

At Credit Suisse, some of the bankers decided to lean into Google’s questions, and thought about pitching a modified auction for a retail portion of the offering. The year before, Hambrecht co-managed an offering for Instinet, a trading system then owned by Reuters, using some elements of a Dutch auction. The Credit Suisse team thought it could do something similar.

Google’s executives read through the RFPs they’d received. They invited Morgan Stanley, Credit Suisse, Goldman Sachs, JPMorgan, and many other banks to an event that would be held at the Page Mill Road offices of Wilson Sonsini, Google’s law firm. The invitation came with a strict set of instructions. Bankers weren’t allowed to wear jackets and ties. Bank CEOs weren’t allowed, only those who would be working on the deal. When they arrived, the bankers were to avoid signing in under their name or their firm’s name to preserve secrecy.

As the date got closer, Goldman CEO Hank Paulson repeatedly called Google angling for a pass to attend. Paulson was an investment banker by training and enjoyed calling on powerful clients. Google executives weren’t swayed, and Paulson did not attend. In the years after the deal, Goldman executives would trade a story about Paulson just not being “Googley” enough.

On an autumn weekend in 2003, the bankers made their way to a large room on the second floor of a Wilson Sonsini office in central Palo Alto. John Hodge and Andy Fisher, bankers at Credit Suisse, were there. Eff Martin, Stephen Pierce, and George Lee led Goldman’s contingent in Paulson’s absence. Paulson’s former Goldman colleague Bob Rubin attended on behalf of Citigroup, where he served on the board. Nancy Peretsman, a banker with Allen & Co., brought her dog.

As they listened to the bankers, the Google team looked for substance. This was not a beauty pageant, and no bank was guaranteed a role in the transaction. JPMorgan made a cute video about Google, which failed to impress the Google team because it wasn’t substantive.

Goldman took another tack entirely. As the top IPO underwriter at the time, Goldman’s bankers arrived at the meeting floating on hubris. They told the assembled Google executives that an auction would be a mistake, and that if they pressed ahead it was going to be a disaster. They should let Goldman do it Goldman’s way.

At the conclusion of the pitches, some banks didn’t hear anything for many weeks. That didn’t stop some of them from talking to reporters, and the media hyped up the prospects for a future Google offering. In an October 2003 article, Reuters quoted an anonymous investment banker: “Search is very hot right now. They have great momentum. It’s a weak IPO market and there’s tremendous focus on Google.”

The upcoming offering, in other words, had done what Silicon Valley hadn’t been able to do for at least three years: capture the public’s imagination. In a Wall Street Journal article from November 4, under the headline “Our savior?,” writer Bret Swanson captured the mood:

The world’s expectations of Google, however, are grander still. In an IPO wasteland, where offerings this year hit a 29-year low, some in Silicon Valley view Google’s suspected 2004 public debut as a second coming. A $20 billion IPO could reinvigorate the Valley and part the waters for a new round of Internet capital. (Not to mention single-handedly salvage a nearly wiped-out $2 billion venture fund for top-flight Kleiner Perkins.) Still not satisfied, some are hoping Google uses its exalted platform to transform the entire hierarchy of the closed and clubby IPO system.

During the first week of November, Google sent a one-page document with five questions about auctions to Credit Suisse and asked for a meeting. Andy Fisher was playing golf with some East Coast friends to celebrate their birthdays at Pebble Beach. He had no choice but to cut his trip short and drive north.

The search company’s executives had asked the bankers to keep their identities secret, so when the Credit Suisse team arrived they signed in as “AF & Co,” Fisher’s initials. The meeting was held in a nondescript conference room with Brin, Page, Buyer, Reyes, and Drummond.

A month or more later, Buyer called Credit Suisse again. Google was ready to make its selection. But first, Schmidt wanted to meet one of the firm’s bankers and look him in the eye. “Eric wants to have this meeting, and we want one person from Credit Suisse to come and meet with him,” Buyer said.

Since Credit Suisse figured the discussion was likely to deal with the mechanics of an auction, they chose to send Fisher, one of their ECM bankers, to meet with Schmidt.

Fisher crammed for the meeting. “I prepared like nobody’s business, because I figured it was probably the most important meeting of my career.” It would be in Mountain View.

Fisher met with Schmidt in a large auditorium. The two men found a couple of chairs in the huge space and sat down for a discussion. It was just the two of them. Fisher was primed to talk about auctions. “Can you please tell me by what the demand for IPOs is overstated in an IPO?” Schmidt asked him. He was referring to percentage or amount: in other words, how much of the demand for IPOs was bullshit.

Fisher hadn’t prepared to answer that question. “I teach a class at Stanford, and there was a guy in my class, and he worked for a money manager,” Schmidt continued. “He said it was his job just to lie as much as he possibly could about how much demand he had, so he could get as much stock as he possibly could. So how much is the demand overstated by?”

Fisher thought there was no hiding anything from this guy. “Well, these things are twenty times oversubscribed,” he answered. And then he walked Schmidt through the math. “It’s twenty times oversubscribed because all these people give you orders for 10 percent of the deal. They don’t really want 10 percent of the deal—they don’t expect to get 10 percent of the deal.” By giving a bigger number, the investor expected the bank to factor that into the allocation process. Hopefully, the investor’s outsized order would garner it a bigger percentage of the deal.

The meeting went on like that, with Schmidt asking questions about the inner workings of the IPO market and Fisher answering, for about forty-five minutes.

Fisher left thinking he’d performed pretty well. He’d answered Schmidt’s questions and given him the sense that Credit Suisse knew what it was doing. The following Friday, he got another call from Buyer. Google had hired Credit Suisse as one of the lead banks for its upcoming IPO. Fisher was elated. Who were the other bookrunners? he asked.

Morgan Stanley and Credit Suisse. That’s it, she said. Goldman would not lead the deal.

After hanging up, Fisher did what any self-respecting banker would do after getting a call like that. He screamed.

And then he took the afternoon off. He grabbed a young banker who was a decent golfer and headed to the public course at Crystal Springs, off Route 280.

With the banks hired, it was time for the organizational meeting to bring everyone up to speed.

It was again held at Wilson Sonsini’s offices. Drummond, Google’s general counsel, had been a lawyer at Wilson when he helped Page and Brin incorporate the business, and the ties between the two firms were strong.

Morgan Stanley and Credit Suisse hired Simpson Thacher & Bartlett LLP partner William Hinman. Hinman was already a well-respected lawyer, with an honors degree from Michigan State University and a law degree from Cornell University. He had been a member of the editorial board of the Cornell Law Review during his graduate schooling.

As bankers and lawyers gathered for the meeting, many of them still didn’t know how big Google was by revenue, or how fast it was growing. The Credit Suisse bankers had organized a betting pool among themselves, and its members made a wide range of guesses.

At the meeting, Schmidt got up at the front of the room and began to present Google’s financials. In 2003, the company’s net revenue was close to $1 billion.

One lawyer in the room was stunned. “Oh,” he said.

It was simply orders of magnitude bigger than any other private tech company at the time. To put it in perspective, bankers had been taking companies public with $3 million to $5 million in quarterly revenue at the height of the dot-com boom in 1999 and 2000.

Most of the Credit Suisse bankers had guessed something considerably less. Only one got anywhere close to what Schmidt disclosed.

Once the organizational meeting was completed, Google and its advisors got down to designing the auction. Morgan Stanley and Credit Suisse were invited to a meeting at headquarters, convened in Google’s executive offices, in a glass-walled conference room. Once again, the number of bankers who were allowed to attend was restricted. Jon Anda, head of corporate finance for Morgan Stanley and the former equity capital markets chief, was there. Fisher was there for Credit Suisse, sitting across the table from Schmidt, Page, Brin, and Reyes.

Anda, Stewart, and others at Morgan Stanley were striving to come up with an auction process that could handle the huge influx of orders they expected from retail investors. It wasn’t an easy task for Wall Street information systems in 2004.

The bankers also spent a lot of time working through the ways in which they would accept the orders from other banks that would help underwrite the offering. Individual investors had never before put in an order with a price limit, which meant some of the banks that would be accepting those orders needed to add a field to the order-entry systems. And Morgan Stanley needed to come up with a way to accept the orders electronically from other banks and merge them with the other orders. Google wasn’t interested in having the orders sent by fax.

As they worked out the details, the Morgan Stanley bankers had to bring along the rest of their firm. A new approach meant more risk, and their bosses, Vikram Pandit and John Havens, wanted to know how the Google IPO could enhance the firm’s franchise, not destroy it.

Another idea would have been to have an auction for institutional investors and let retail investors tag along but not let them actually set the price. The bank even submitted a patent for an institutional-only auction. The Google executives didn’t like that idea.

To accomplish what it wanted, the bankers had to make some hard choices. A good deal of those choices had the effect of stripping elements of an auction out of the system.

When it came time for Anda to speak, he stood up and approached a whiteboard at the front of the room, where he scrawled a few of Morgan Stanley’s ideas for the structure of the auction. He proposed that investors could put in a bid and only change it once. When Anda was done walking the room through some of the bank’s suggestions, Schmidt rose and took his place at the whiteboard. With a marker that might as well have been red, he proceeded line by line to cross out the items Anda had listed.

Google wouldn’t be doing what Morgan Stanley had proposed, and the company wanted its bankers to know quite clearly that it knew it had veto power and intended to use it. Google wanted the auction to be as flexible as possible to encourage individual investors to get involved. The bankers at Morgan Stanley, working with Google engineers, weren’t sure their technology could handle it. They worried that they wouldn’t be able to sort through the thousands of orders that they were expecting, some in penny increments. Everything that Google wanted to do that would enhance flexibility for retail was something that made the software that much harder to write.

“Google was very much, ‘Let’s democratize this,’” said one of the lawyers involved. “The banks were like, ‘Let’s make sure we can make it work.’ And there was a little tension between those two ideas.”

Morgan Stanley ultimately built an auction system that could accept a billion bids.

With the broad outlines of the auction system coming into focus, Google’s team wrote the company’s prospectus. At the same time, its lawyers prepared to file a Form 10. Google had breached the threshold of five hundred shareholders. The document, governed by the Securities Exchange Act of 1934, served to register securities for potential trading on an exchange. According to SEC rules, the form needed to be filed no later than 120 days after the end of the fiscal year, meaning it had to be in by April 29. Google planned to surprise the market and release its IPO registration statement at the same time.

To do so, the company needed to submit its paperwork through an intermediary called a financial printer. A holdover from an earlier era when companies didn’t have the typesetting capabilities to create their own securities prospectuses, the financial printer played a role in helping them format documents. By 2004, the printer still played a pivotal role, taking the Form 10 or an IPO prospectus from the company, proofreading, formatting, and translating it, if necessary, and then submitting it through its system to the SEC.

As the deadline neared, the lawyers assembled at the offices of the printer, R. R. Donnelley, on California Ave. in Palo Alto, behind Wilson Sonsini’s offices. Like many of Palo Alto’s buildings, Donnelley’s office was clad in stucco, with a red tile roof, a nod to the region’s Spanish heritage. Visitors walked through several large arches before reaching an inner courtyard.

They set up a long table in one room. As April wound down, Google’s founders and top executives assembled around the table to do something they called page flips. Drafts of the prospectus were printed out from a kiosk and passed out to everyone to read from beginning to end. Then someone would lead the group, a conductor of sorts, through a page-by-page review. If someone had a change, the conductor made it on the master copy, finally handing that over to the customer service desk at Donnelley for the printer to log them in to the official version housed on their systems. At one point, Google sent over some English majors who were on staff to rewrite some of the legal jargon, and their edits had to be inputted in the same way. Because the process was so archaic, what should have taken minutes instead took hours.

Sometime before the countdown to the deadline, Thursday, April 29, a crowd assembled around the customer service desk at Donnelley. Brin, Page, Schmidt, and Drummond were there for Google, as were the company’s lawyers from Wilson Sonsini. Some of the investment bankers were there, and at least one of their lawyers.

Everyone was “holding their breath while last-minute changes were being forced through the document before you hit that magic button that says ‘Submit’ to the SEC,” according to one of the people there.

Finally, Donnelley’s people submitted the document to the agency.

As people celebrated, some of the lawyers continued to prevail upon the printer until the Form 10 was submitted.

Google’s prospectus became required reading from Silicon Valley to Wall Street. It was the first time the public was seeing the company’s financial results. It was also the first time the public got a look at the structure of its IPO, which the prospectus addressed in at least three sections.

Near the top of the document, Google explained the reasoning behind its choice of an auction-based IPO. “It is important to us to have a fair process for our IPO that is inclusive of both small and large investors,” the founders wrote, in a letter to investors they styled after Warren Buffett’s investor communications. “It is also crucial that we achieve a good outcome for Google and its current shareholders.”

They continued, “Many companies have suffered from unreasonable speculation, small initial share float, and boom-bust cycles that hurt them and their investors in the long run. We believe that an auction-based IPO will minimize these problems.

“An auction is an unusual process for an IPO in the United States,” they added. “Our experience with auction-based advertising systems has been surprisingly helpful in the auction design process for the IPO. As in the stock market, if people try to buy more stock than is available, the price will go up. And of course, the price will go down if there aren’t enough buyers. This is a simplification, but it captures the basic issues.”

The letter said that the founders were encouraging current shareholders to “consider selling some of their shares” as part of the process to ensure there would be enough supply to set a reasonable market. “The more shares current shareholders sell, the more likely it is that they believe the price is not unfairly low.

“We would like you to invest for the long term, and to do so only at or below what you determine to be a fair price. We encourage investors not to invest in Google at IPO or for some time after, if they believe the price is not sustainable over the long term,” the founders wrote, before moving on to sections about “Googlers,” or employees, and the company motto, Don’t Be Evil.

At the end of pages and pages of risk factors, Google listed six specific to its choice of an auction: the stock price could decline quickly; successful buyers might experience a winner’s curse, the term used to describe what happened when bidders won an auction, only to find little demand at higher prices than they paid; successful bidders looking to sell quickly by flipping the shares might be disappointed; successful bidders might receive the full amount of the shares they asked for; submitting a bid did not guarantee an allocation; the systems, procedures, and results of the auction might hurt Google’s brand.

Then, under the heading “Auction Process,” Google outlined five stages: qualification, bidding, closing, pricing, and allocation. The process wasn’t simple. Each buyer needed to obtain a unique bidder ID by clicking a hyperlink. This they could do once they’d met an underwriter group’s eligibility and suitability requirements—written to include all types of investors. Bidders would receive their IDs electronically once they’d provided certain information to the underwriter, and if they could show that they had electronically accessed both the prospectus and the transcript of a management presentation.

Once investors received IDs and the bidding started, they were to indicate the number of shares they wanted to buy and the price they were willing to pay. They could submit more than one bid. There would be a minimum bid size, though Google left that blank in the preliminary prospectus.

Bidders could withdraw at any time until the end of the auction process. If they submitted multiple successful bids at different prices, they should expect to buy all of the shares they had ordered. In other words, bids would be considered in total. Google and its underwriters reserved the right to reject bids that the company considered speculative or attempts to disrupt the bidding process.

The instructions continued. The “master order book” would be monitored to gauge demand, the prospectus stated, and the company provided warnings to investors that suggested Google was preparing for a frenzy of demand that might lead it to sell more shares in the offering, depressing the stock price. “We and the selling stockholders may decide to change the number of shares of Class A common stock offered through this prospectus,” the document read. “It is very likely that the number of shares offered by the selling stockholders will increase if the price range increases. In an auction process, this could result in downward pressure on the price. You should be aware that we have the ability to make multiple such revisions up until the closing of the auction and pricing of the offering.”

Before Google closed the auction, it would ask the SEC to declare the registration statement effective. The company could close the auction within hours of the offering prospectus’s being deemed effective. If bidders were asked to confirm their bids and didn’t, their bids would be rejected.

When it came time to decide on a value for the shares, Google said it would find the highest dollar figure at which all the shares could be sold, and then set the IPO price “near or equal to” it. The underwriters would then begin the allocation process. If the IPO price was at or close to the clearing price, bidders should expect to get roughly the number of shares they’d bid for. If the winning bids represented more shares than were on offer, Google would allocate them to winning bidders in one of two ways: pro rata, apportioning a percentage arrived at by dividing the number being sold by the total orders, or maximum share, based on an algorithm that favored smaller bids over larger ones. Tables in the document represented two hypothetical auctions to illustrate the two methods.

In either case, Google said that it would set the IPO price at a level at which bidders would receive at least 80 percent of what they’d asked for.

Having filed the prospectus with the SEC, Google began negotiations with regulators. The company intended to set up a website, Meet the Management, to inform investors about what it did and who the founders and senior executives were. It would serve as a central hub to keep everyone informed about the deal. The company’s lawyers submitted schematics of what each page of the site would say. In some cases, Google had ideas for educating the public that the SEC didn’t let them use. In one instance the company wanted to provide a Bidders Handbook that it had written in plain English as an appendix to the prospectus. In another, informed by their belief that investors should know something about Google, the founders wanted to set up a simple four-question test before buyers would be allowed to bid. The answers to the questions would be found in the prospectus or on the website. The SEC rejected both ideas. The agency also required the company to organize a process so that people without internet access could buy shares. The Google team set up a series of analog notifications it could send to people at each stage of the auction process.

The nuances of the auction and the SEC’s requirements bogged the process down and required Google’s team to submit amendment after amendment. All of their prospectuses were made public—unlike those issued for much later IPOs, which allowed companies to submit a draft registration statement confidentially and work out certain details with the SEC in private. Instead, the company and the agency hashed out the details of the auction over the course of nine amendments between the end of April and the IPO.

On May 21, the company put out an updated prospectus showing that it had hired thirty-one underwriters. Brin and Page were adamant that small-time investors have the same opportunities as the mutual funds and hedge funds, leading them to hire a group of minority-owned and regional brokerages for the deal in addition to other, more typical Wall Street underwriters.

Besides Credit Suisse and Morgan Stanley, who had been listed all along, Goldman Sachs, Citigroup, and WR Hambrecht + Co. received mention, as did Allen & Co., JPMorgan, and Thomas Weisel Partners. Farther down the list were Ameritrade Inc., M.R. Beal & Company, William Blair & Company LLC, Blaylock & Partners LP, E*TRADE Securities LLC, Muriel Siebert & Co. Inc., and others.

Thomas Weisel Partners, the last of the independent boutique investment banks focused on the technology sector, was a surprising choice. A successor to the Four Horsemen, Weisel was run by an investment banker who had founded Montgomery Securities. Since then, the firm had been overshadowed by much larger banks.

Its inclusion owed everything to Kleiner Perkins Caufield & Byers’ John Doerr, a Google board member who had a soft spot for Weisel as the last of the independent investment banking boutiques not to have been swallowed by larger rivals. He wanted to see Thomas Weisel Partners succeed, and he pushed for its hiring. Google promised the firm a very small fee.

Hambrecht, also on the deal, had used OpenIPO repeatedly by that time—for Andover.net and Peet’s Coffee & Tea Company, for instance. Though not all of his efforts had been successful, he was Silicon Valley’s foremost voice on using auctions to price IPOs. The Google team had spoken at length to Hambrecht about his process, though in the end Page and Brin decided on a compromise. There were aspects of Hambrecht’s system that they didn’t like, so they decided to create their own. Google wanted the offering to be accessible to individual investors, and they felt that Hambrecht’s OpenIPO didn’t do that in a way that they liked.

His presence on the transaction was a tip of the hat, and Google paid him a small fee from the IPO to recognize his pioneering approach.

The same day that Google announced Hambrecht’s involvement, an auction he was handling got pulled. Alibris Inc., the owner of a website that sold new and used books, had met with tepid demand for its shares. In a statement, the CEO said that he stood behind the auction process and predicted that it would become more common among big companies selling shares for the first time.

With such a large number of banks involved in the Google transaction, it was inevitable that many would find aspects of the auction design not to their liking. Some of the bankers were upset that Morgan Stanley and Credit Suisse were requiring them to sign nondisclosure agreements. Others didn’t know how many shares they would get in the IPO to pass along to clients, or how they would calculate the fees they should expect.

Compensation for typical IPOs is calculated under what’s known as fixed economics. The underwriting fee, calculated as a percentage of the deal or gross spread, is divided up into three sub-fees. There is a management fee, an underwriting fee, and a selling commission. The breakdown was often 20 percent, 20 percent, and 60 percent.

So in a traditional deal, the lead bookrunners, like Morgan Stanley and Credit Suisse, placed two-thirds of the stock with their institutional clients, and each could expect to take home 30–35 percent of the total fees. Auctions like those run by Hambrecht worked in much the same way, though Hambrecht charged a smaller fee than was common.

But in the Google deal, if all 60 percent of the selling concession went to other firms, Morgan Stanley and Credit Suisse would end up making very little. They worried that if the deal was as popular with retail investors as they expected, there was the possibility that despite having spent six months or more designing the auction, they wouldn’t make much money. So these two underwriters decided to do away with the selling concession and split the fees between just a management fee and an underwriting fee.

For decades, Wall Street salespeople were paid on IPOs by getting a commission based on how many of their clients put in orders for the stock. By the time of the Google offering, some had done away with those but still awarded sales credits that worked in the same way. It was a mechanism to reward salespeople for getting their clients involved in the deal.

By cutting the selling concession fee from Google’s deal, the lead underwriters took away any incentive that salespeople at the regional or smaller firms had to market Google’s deal or encourage their clients to buy shares. (Merrill Lynch & Co. dropped out of the syndicate, concerned about the fee structure and the large technology investment it would have to make.)

By late July 2004, some of the investment banks worried that Morgan Stanley and Credit Suisse’s insistence that they guarantee the orders from their clients, even if the clients couldn’t pay, might leave them on the hook to buy stock they couldn’t sell. The banks did not like the heavy involvement of retail traders.

Morgan Stanley wondered what might happen if one of those smaller brokerages had many clients who all put in high bids for the stock. If those bids helped make the deal frothy and pushed up the price, the bankers knew that a version of the winner’s curse might saddle investors with losses and that they might not have the money to pay for the stock.

The worst-case scenario would mean that investors wouldn’t be able to make good on orders they’d sent through the smaller brokerages, and Morgan Stanley and Credit Suisse would have to stand behind the deal. So the underwriters persuaded the SEC to agree that if more than 10 percent of the orders defaulted, the deal would be canceled. The default clause was often written into underwriting contracts, but the SEC had ruled in the past that IPOs meeting the default definition were still effective. The agency had insisted that those deals go through.

The lead underwriters also required that other members of the syndicate stand behind their clients’ orders, and they reserved the right to require letters of credit or other credit support if they felt they needed it. The duty of communicating the need for letters of credit fell to Credit Suisse’s Fisher, who held a call with the underwriters to tell them.

After the call, Goldman’s Stephen Pierce called one of the lead bankers. “We’re Goldman Sachs,” he said. “We’ve been in business for a hundred and some years, and we’ve never had to post a letter of credit for a deal like this and we’re not doing it.”

In the end, the lead underwriters required all bidders to have enough money in their account to cover the bid.

On July 26, Google disclosed for the first time the price at which it would sell its shares. After hearing an appeal from the company, the SEC allowed it to publish a wider range than usual. Google’s shares would be sold between $108 and $135 apiece. At the high end of the range, the company would raise nearly $4 billion. The high end of the range was higher than many Wall Street investors and analysts expected.

The following day, Clarium Capital’s Peter Thiel joined CNBC to talk about the Google IPO. Thiel had taken his payments company PayPal public on Valentine’s Day 2002, selling shares at $13, only to see them rise to more than $20. Though PayPal was sold to eBay later that year, the IPO experience led Thiel to believe that the company left too much money on the table. In December 2003, Thiel told Barron’s that he had considered selling shares directly to investors before giving in to some of his partners’ nervousness to go with a traditional offering, underwritten by Salomon Smith Barney. If he got a second opportunity, he said he’d like to try selling shares directly to investors.

Thiel also knew the Google founders, a fact that put him in an awkward position when CNBC hosts Larry Kudlow and Jim Cramer began to criticize the company’s offering. As veterans of the financial industry, both television personalities were known for being sympathetic to institutional investors who watched their network.

CNBC had also successfully built an audience with day traders during the internet boom, and both men spent time speaking on radio, giving them a perspective on what Main Street investors thought about Google. After bantering with Thiel about how Google’s valuation compared to Yahoo!’s, the hosts got right to the heart of their doubts about the company.

“Peter, let me tell you what I hear a lot of retail people say today,” said Cramer. “A hundred and thirty dollars, to heck with it. I’d rather go buy eBay, I’d rather go buy some of the other dot-coms, which then soar.” His comment hit on the widely held belief that investors buy IPOs for the first-day pop, the sugar rush of quick profits they had grown conditioned to expect in the go-go years of the late 1990s.

Google’s offering wasn’t intended to deliver that, Thiel replied. “Google is not going to get investors who are looking for a quick flip,” the PayPal founder said, making the additional point that Google was trying to discourage the hot money on Wall Street. “You shouldn’t expect a huge one-day pop. It’s going to be priced very efficiently.”

Kudlow jumped in to say that maybe the Google founders thought they were a little too smart for their own good. The market will make its own decision, he said.

“I think it’s absolutely stupid for these kids, whoever they are,” Kudlow said, “to think that that’s not going to happen.” He continued, coming back to Cramer’s remark about the expected price of the shares. “It’s a lot of money, one hundred to one hundred thirty bucks, or whatever the spread is anticipated. This is a gang that was going to democratize the share purchases. Now why not price the damn thing at fifteen dollars or twenty dollars?”

Thiel said that Google was looking for long-term investors. “They’re looking for people who are going to hold it in the long term. Now maybe there’s nobody like that left in this whole country, in which case they’re going to have a pretty big disaster when their IPO comes up, you’re right. But the assumption is that’s not the case.”

As the interview continued, Cramer got worked up. He could have been talking for any number of institutional investors. “If you had to ruin an IPO deal, I would do exactly what they did,” Cramer said. “They, first of all, didn’t court any of the institutions.” He was referring to large institutional investors. “The institutions are saying the heck with it. Secondly, I would price the deal in April or May, not in the dog days, but that’s because they dithered so long. Then finally, a hundred thirty dollars, the only people who really are going to do that are the institutions, but they’re already so upset about this that they’re going to go and buy eBay. I mean, who was in charge of this thing?”

Thiel tried to move the conversation with the strong-spirited hosts to a different place. “Let’s put this in a broader context,” he said. “You have to look at this as part of a long-standing struggle between the tech industry and the banking industry, and it goes to who captures the value. When you build a great company, do the people who build the company, the original investors, get it, or do the bankers on Wall Street? And a lot of what was going on here beneath the surface was a decision to try to cut back on Wall Street’s enormous profits and so on down the line.”

Kudlow softened his stance. The host said he liked the idea of teaching bankers a lesson and was in favor of the Dutch auction. Cramer countered by asking if Thiel had any involvement in the process. He didn’t.

As the segment wound to a close, Kudlow asked Thiel if he would buy the stock.

“Well, it depends on where the auction price is,” he said. “Probably not at one thirty-five. If it’s around a hundred, I might.”

In other words, anything higher than $100 was too much for the PayPal veteran to pay.

Professional investors were coming to a similar conclusion as Google executives fanned out across the country for a series of meetings to drum up demand for their shares. The roadshow required executives and their bankers to fly to various cities for often large meetings with investors, conducted in an auditorium or ballroom.

In late July, Google executives visited New York’s Waldorf Astoria for a luncheon in the hotel’s ground-floor ballroom. Only those investors who were clients of Morgan Stanley and Credit Suisse were allowed into the room, and only after showing identification. Hundreds of investors and analysts were there, giving the event a crowded and chaotic feel.

But when it came time for presentations from Brin, Page, Schmidt, and Reyes and a Q&A, the executives came off as less than polished. They spoke off the cuff, answering questions without consulting their presentation. At one point, one of the executives tried to tell a joke and it fell flat. “I certainly don’t think it was our best showing,” recalled Buyer, who was sitting in the audience. “It was a little more ad-libbed than would have been ideal. That’s the way that Sergey, Larry, and Eric wanted to do it.”

Google didn’t offer many new details about its competitive position to rivals or about how the auction would work, leaving some investors wanting for information. Writing about it at the time, the Wall Street Journal said that “when the presentation, and question-and-answer period that followed, failed to give the investors the insight they came for, there was some disenchantment.”

On August 2, Brin, Schmidt, Reyes, and Buyer appeared at a hotel on Boston’s waterfront for a presentation in front of a smaller crowd of fifty or sixty investors. As the investors found their seats at tables set with silverware and plates, they found a small menu that introduced the Google team. The Google team was dressed informally, without ties, in keeping with a decision to ignore some of Wall Street’s long-standing conventions.

“I was in Boston for another reason and conveniently managed to wrangle an invitation,” said Barry Randall, then the lead manager of the First American Technology Fund, part of US Bancorp Asset Management, based in the Twin Cities. Randall, used to evaluating thirty to forty IPOs a year in those days, took two pages of notes as he listened to the Google executives speak. “The first thing that stands out about it is nothing stood out about it. I mean, by this time everyone knew about the auction.”

Google’s management warned investors that the business was seasonal, with the two middle quarters of the year tending to be weaker than the first or the last. The discussion also touched on Google’s relatively new email product, Gmail; the fact that Google would have multiple share classes that gave the founders complete control; and traffic acquisition costs and other financial minutiae. Not one investor asked about the auction, according to Randall’s recollections and notes.

“It was shockingly tame and uncontroversial,” Randall said. His notes from the meeting were shorter than usual.

As Google set out to create the auction process, there were hundreds of tiny decisions they had to make. Many of those had to be run by the SEC, which was particularly sensitive about company executives talking up an upcoming offering outside of the filings they submitted to the agency.

From the moment companies filed a prospectus with the SEC, securities laws required them to observe a quiet period that prevented them from communicating publicly about the deal other than through use of the prospectus. Earlier that year, the SEC had slapped the wrists of cloud-computing pioneer Salesforce.com after the New York Times published an article that quoted CEO Marc Benioff while the company was going public. Four days later, the SEC forced the company to delay its offering and enter into a “cooling off” period that would let the effect of the article dissipate. The incident showed that the SEC was prepared to enforce a strict interpretation of what it considered an offer of securities.

During the week of August 9, as Google was preparing to close the bidding registration period, the company’s PR team received word from Playboy magazine that the publication would be accelerating the release of a piece based on an interview a contributing editor had conducted with the founders in April, before they’d filed their prospectus with the SEC. It would be in the September issue, available to the public that week. Now Google was quite possibly in the same boat as Salesforce.com. An offering that had already been delayed because of regulatory negotiations and trying to create something from scratch was now in danger of being delayed even further. There was a very real chance that the SEC would consider the piece a violation of the Securities Act of 1933, which forbids companies from hyping their offering during the registration period.

One Wilson Sonsini lawyer, after being told by Google executives about the forthcoming article, thought to himself, I either use my wastepaper basket to throw up in, or just to pack up whatever I could grab in my office, and go home and say, Okay, I’m done with being a lawyer at Wilson Sonsini. The lawyer called lead partner Larry Sonsini to let him know what was going on.

Before the lawyers could formulate a response, they needed to know what the article said. On Wednesday, August 11, Google and its team got copies of the magazine and someone took a pile over to Donnelley, the printer. By this point, food wrappers and old papers littered the room. As the stack of Playboy magazines sat on the table, lawyers and accountants helped themselves to a magazine from the pile and began reading the interview.

The lawyers knew they had to contact the SEC. The Wilson lawyers, including David Segre, one of the lead attorneys, were already connected to officials at the SEC’s Division of Corporation Finance, including Paula Dubberly and Barb Jacobs, assistant directors of the division. When the lawyers finally reached the officials by phone, the SEC drew a hard line.

“We can’t treat you better, or differently, than we treated Salesforce,” one participant remembers an official saying. “So, sorry, guys, we’re going to cool you off.”

Then Sonsini spoke up. Already an eminence in Silicon Valley thanks to his early embrace of emerging technology companies, he asked that the agency try to keep an open mind. That he was speaking to two female SEC officials about an unauthorized article in a men’s magazine wasn’t lost on at least one participant in the call. “Can you just let us at least try and convince you,” Sonsini said, according to this person’s recollection, “that that’s not the right remedy for the situation in which we find ourselves?”

Dubberly was noncommittal. “You can try, but we don’t see any way out of this.”

The team of lawyers got to work drafting a memo that would expand beyond nine pages to explain to the SEC why Google’s IPO could move forward. The work would take them late into the night and into the early hours of the next morning. The lawyers made several suggestions to make it easier for the SEC to rule in Google’s favor. One was that with most of the world anticipating the IPO, an interview in Playboy was effectively white noise. It wasn’t adding much hype to an already hyped-up transaction. The founders had already been on the cover of Newsweek and no one read Playboy for investment advice.

Simpson Thacher attorney William Hinman, who represented the underwriters, made the point that investors planning to participate in the IPO had to be registered by the time the auction started, on Friday, August 13. The magazine wouldn’t be on most newsstands by then. If the SEC required Google to enter a cooling-off period, the company would have to reopen the bidding registration, and then, they could be assured, everyone would have read the magazine. It would throw sand in the cogs of an already complex offering and potentially disadvantage some investors.

They scheduled a follow-up call with the SEC for the following morning.

As they were drafting their memo to the SEC, Hinman foresaw an issue. He was scheduled to be on a plane to Oregon to attend an important family commitment. He wouldn’t be able to make the call. But things should be in good shape, he told the founders. The SEC was open to innovation and wanted the deal to go well because Google had democratized the offering process and had broken up the old-boy network of allocations. He wasn’t too worried.

Brin wasn’t having it. “Oh no, you have to be on the call,” he said. “We want you on the call.”

Hinman didn’t think he could change his flight.

“Well, we’ll get you a private plane with a phone,” Brin answered. He made a call and spoke to someone.

Early Thursday morning, with the sun rising over the Palo Alto office park where they had been furiously drafting, the team of lawyers faxed the memo to the SEC. And Hinman headed to the San Jose airport. When he arrived, the lawyer asked the pilots of the private plane he would be taking if the phone on board worked well. What phone? they said. They hadn’t gotten that detail. Hinman asked them if he could rent a satellite phone. Yes, they said. “Two thousand dollars.” A satellite phone it would be.

In the end, the SEC checked in early and Hinman was able to take the call in the airport’s waiting area. He and the other lawyers explained their case, and the SEC hung up. By the time he took off in Google’s private jet, Hinman didn’t know if the SEC would let the company go.

In the end, the agency agreed that it wouldn’t force Google to delay the IPO, but the company needed to include the entire text of the Playboy interview in an amendment to the company’s prospectus. Doing so would ensure that investors reading the prospectus would know what the founders had said, and it would also mean that the company and its underwriters were taking legal liability for the article. If investors later found that they were misled by comments made in the interview, they would have grounds to take legal action.

On Friday morning, August 13, Google put out amendment No. 7, with the Playboy article listed as Appendix B.

By the time the article was widely available on newsstands that weekend, all the bidders had been registered with one of the underwriters or brokerages involved in the IPO. The auction had finally begun. Over the next several days, the bankers running the process realized that something was wrong. The individual investors that Google had worked so hard to court weren’t showing up to bid to the degree the company had expected. Many of the large institutions, the mutual funds, weren’t putting in bids for prices as high as the low end of the price range that was in the prospectus.

As Cramer and Kudlow had diagnosed, some of those investors didn’t like the auction process. It required not just an estimate of Google’s value but an understanding of auction bidding strategies. Legg Mason portfolio manager Bill Miller, for example, felt the need to seek out two auction experts for help in thinking through his bidding strategy.

Others were turned off by what they perceived to be the arrogance of the company’s founders and executives. At the Waldorf Astoria luncheon, Google’s executives had declined to provide details that investors were seeking. To other meetings, Brin wore a t-shirt, blue jeans, and sneakers with a thick heel. It wasn’t the kind of attire the professional investing class, most of whom wore suits, expected to see from technology founders trying to impress them.

Google had recently updated its financials to show that growth had decelerated in the second quarter. The company wasn’t offering any projections to Wall Street analysts to model their future performance. So as Wall Street tried to get its collective head around Google’s business over the next few years, estimates diverged widely. Some even thought that the company’s performance had peaked.

Google thought it didn’t need to do anything more to help investors. The IPO had been hyped by the financial media, and during months of preparation, the company had meeting after meeting about how to handle the large flood of orders it anticipated. Now that it was time to collect the orders and they weren’t coming in as quickly or at as high a price as the company and its bankers had hoped, Google had to take care of yet another headache. It had disclosed in an earlier prospectus that in issuing shares and options to some employees without registering them with securities regulators it may have violated state and federal laws. The infraction might lead it to have to pay $26 million to buy the stock and options back from shareholders. Now the company learned that the SEC had started a probe of its practices. It needed to alert shareholders. On Monday, August 16, Google disclosed that the agency had “initiated an informal inquiry into this matter and certain state regulators, including California, have requested additional information.” The company also disclosed that it had asked the SEC to declare its registration statement to be effective by 4:00 p.m. the following day. The offering was likely to close within hours of the statement’s being declared effective.

Tuesday came and went without any public word from Google. Brin, Page, and Schmidt were locked in a tense debate with their bankers. They had orders for roughly 20 million shares at around $85. That was more than 5 million shares less than they planned to sell and at a price that was 30 percent below the middle of the range written in the prospectus.

Members of the company’s board who were planning to sell shares on behalf of their venture capital firms—Doerr, for Kleiner Perkins, and Michael Moritz, for Sequoia—decided that if Google’s shares were going to be auctioned at such a low level, they didn’t want to sell. Between them, Kleiner Perkins and Sequoia had planned to sell 4.5 million shares. Now they wouldn’t. Google didn’t immediately alert other investors, including Ron Conway, a venture capitalist who had originally introduced the Google founders to Sequoia, about Kleiner Perkins’ and Sequoia’s change in plans.

“We were calling each one and saying… are you in at eighty-five dollars?” recalled one of the members of the Google team. “We were hoping everybody was, and I’m sure that we were pushing people to stay in, but because we needed that number.”

The SEC would need to know about Google’s lowered range. As its attorneys at Wilson readied another amendment for release on the morning of Wednesday, August 18, Segre prepared to make a case to the SEC that the lowered range wasn’t material. If the SEC didn’t agree, Google would have to submit another prospectus for review. That would likely delay the offering by twenty-four or forty-eight hours.

Early on the morning of August 18, Segre called a senior SEC official at her home in the Washington, DC, area. The lawyer explained why Google didn’t think the lowered price was material—Google didn’t need the money it was raising—and begged the official for a favorable ruling. After some back-and-forth, she relented.

In a statement on the web page it had set up for the IPO, Google told shareholders it would soon file an amendment to its securities prospectus stating that the price range had been lowered. In the amendment that came that morning—its ninth—Google said it was now seeking to sell shares in the range of $85 to $95.

The number of shares to be sold in the offering was also reduced. Google would continue to sell the same number of shares, 14.1 million, while the selling shareholders would only sell 5.5 million shares, not the 11.6 million they’d planned to.

Certain they could get the deal done at that size, Google’s bankers called investors to finalize their bids. Due to the auction structure, only investors who had already bid could participate in the auction at the lower price—investors who had sat out the first round because the price was too high couldn’t now put in bids at the lower price.

As the bankers made their calls, they were subject to Google’s restrictions on telling investors what other investors may have bid. It was common practice for bankers to share updates on investor order sizes or bid prices with other investors, but Google had decided it didn’t want its bankers influencing bids in any way.

No sooner had the bankers alerted investors to the lower price than one New York investor who had put in an order for a couple million shares at a price higher than $85 lowered his price to around $80. The bankers now had to persuade him to change his mind or find orders to take his place. Credit Suisse bankers called the investor and told him, effectively, that the company hadn’t lowered the range for nothing. There was a reason. Wink, wink. In so many words, they tried to tell him that they had enough orders to get the deal done at $85. He was going to miss out if he didn’t increase his bid. The investor didn’t change his mind.

Later that day, when it came time to price the deal, Credit Suisse’s bankers persuaded Brin and Page to leave the now untidy conditions at Donnelley and move to Fisher’s office around the corner, on Hanover Street. Others were connected by conference call. The bankers told the Google executives that at $85 they could allocate each investor about two-thirds of what they’d asked for. Increasing the price to $86 meant that Google would lose a few orders from investors unwilling to pay that price, leaving just enough to buy all the stock. Those investors would thus get 100 percent of what they wanted.

Though opinions differed, enough people advocated for setting the price at $85 and leaving some investors unsatisfied that that’s what happened. They would, the theory went, buy the stock once it started trading on the NASDAQ, putting a floor under the price. The Google executives agreed and the price was set.

Stock was then allocated to those investors who had put in orders at the lower price. Fidelity, Legg Mason’s Bill Miller, T. Rowe Price, Jennison Associates. An A-list of investors, although the shares were sold at a lower price than the company initially wanted. At $85, Google raised $1.7 billion, making it the largest internet IPO of all time. Brin and Page became paper billionaires.

Despite the superlatives of the Google offering, the company’s executives were angry with their bankers over the way the deal had gone. The bankers understood it was not the outcome anyone had expected. “You had this whole confluence of factors: the design of the auction, the fact that growth was decelerating, the fact that some institutions didn’t want to participate in the auction, the fact that the company was arrogant on the roadshow, etcetera,” one of them said. “All of which conspired to make the thing less successful.”

With the deal done, Page and Schmidt left to take a private plane to New York to celebrate the IPO. The lawyers and accountants stayed behind at Donnelley to proof the final prospectus, effectively cleaning up after the party. Sometime before midnight on the eighteenth, Brin showed up. He wanted to soak up the ambiance and the stale air of the printer’s office one last time. They stayed there chatting and watching the Summer Olympics, then taking place in Athens, on a large television until Brin left sometime in the early hours of the morning.

In the end, Morgan Stanley and Credit Suisse each received 5.3 million to allocate to their clients, representing more than half of the shares on offer. Goldman was third in number of shares. The underwriters received a fee of 2.8 percent, with both Morgan Stanley and Credit Suisse earning $23 million.

The next morning, as part of their trip through New York, Page and Schmidt visited Morgan Stanley’s Times Square trading floor to watch the stock open.

Fisher had delayed a family vacation to Montana and stayed overnight in an airport hotel so he could catch a flight the next day. Waiting for his connection in the airport, he stared up at one of the televisions hanging from the ceiling that was tuned to CNBC. The anchors were talking about what a disappointment it was for Google to sell at $85 a share. The stock was likely to fall further in price, they suggested.

For a couple of minutes, Fisher watched them drone on and on. Then the banker dismissed the blather. Having seen all the orders, he knew how wrong they were.

When Google’s stock opened later that day, the first trade crossed at $100.

By the end of the trading day, Google closed at $100.34. The stock had risen 18 percent, an ideal outcome after all that Google and its team had gone through. And it had done so on a day that the NASDAQ Composite Index fell 0.6 percent. The Wall Street Journal put it succinctly: “Investors who went through the cumbersome auction for Google Inc. shares were feeling particularly lucky yesterday. The countdown to Google’s initial public offering of stock was filled with confusion, frustration and miscues. The first day of trading was no different. But shares of the Internet search-engine company proved to be big winners, at least for a day.”

The stock performance helped to soften any irritation Google executives may have felt earlier in the process. Buyer remembers being happy with the result. “We were thrilled with it,” she said.

When the IPO was done, the underwriters got a second bite at the Google apple. The terms of the deal were such that they were permitted to exercise a greenshoe option, by which they could buy an additional 15 percent of the shares at the IPO price within thirty days of the offering. Google’s underwriters bought 2.9 million more shares.

Google paid the entire underwriting syndicate $53.7 million in fees. Almost $5 million of that was then used to pay for expenses such as legal and printing costs borne by the underwriters.

Even though Google may have ultimately been satisfied with the result, Wall Street was not. Helped in part by the Wall Street Journal, a narrative began to take hold that the offering had been a failure. The newspaper put out an article headlined “How Miscalculations and Hubris Hobbled Celebrated Google IPO.” It opened, “Google Inc. may have needed Wall Street after all.”

A Columbia University auction expert offered up a quote that the drop in price reflected a reality that an auction introduced too many uncertainties and led investors to back away from participating in the IPO. A hedge fund manager said that the auction “managed to tee off the broader constituency of Wall Street, and it’s obviously hurt them,” adding, “Wall Street wins again.”

The newspaper wrote that “the result yesterday was a far cry from the great expectations, built up over the months and weeks leading up to the IPO, that demand for Google’s shares would help reignite demand for new tech stocks and pioneer a way for new companies to do an end run around Wall Street’s old cozy ways.” The article put a fine point on the already fraught relationship between the investment banks of the East Coast and the innovation economy of the West, and made the case that Google’s actions had escalated the hostility between the two subcultures.

“The people at Goldman think we fucked up the deal,” one of the people involved in Google’s offering said in an interview. “That’s what they wanted to think.”

Nonetheless, the narrative that the Google offering was a failure would be used as a cudgel over the following years to ensure that other ambitious startup executives didn’t make another serious run at reforming the system. While NetSuite and Rackspace resurrected the auction format just before the climax of the 2008 financial crisis, it would be fourteen years from Google’s IPO before another company with the consumer recognition and vision of the search giant would try again.

In the months after Google’s offering, the firm’s stock price moved all over the place. It touched a high of $201.60 on November 3 after reporting better-than-expected results in its first quarter as a public company and fell when Google insiders started selling. By some measures, the volatility exposed Google’s continued frostiness with Wall Street. Executives remained secretive and circumspect of professional investors. They still didn’t offer financial projections, which analysts and investors had come to rely on to judge corporate growth rates. As Google had prepared to release its first earnings as a public company in October, the lack of projections led to unusually wide estimates in what analysts expected for the earnings report, from a low of $0.22 a share to $0.61. The company reported net income of $125 million, or earnings per share of $0.45, excluding some onetime costs.

In November 2005, Google executives took an initial step in improving their relationship with the financial world by hiring their first head of investor relations, Kim Jabal, the former Goldman Sachs banker who had been hired by Sheryl Sandberg in 2003. The appointment would give Google an executive whose sole job was to focus on telling the company’s story to Wall Street and answering questions from analysts and investors about its finances.

Buyer had taken over some of the duties in the months after the IPO, according to some accounts, but it was time to get someone focused on it fully. The Playboy interview and its aftermath had made the founders wary of speaking to the press, and Schmidt recognized that the company needed someone devoted to speaking with investors. Jabal had worked in the background of Google’s IPO, pulling data on her business for the offering materials, but otherwise playing a minor part, and now she took on a greater role at the company.

Soon after she started, Reyes, Google’s CFO, misspoke about the state of the company’s advertising business, and the stock tanked. Google had to file a document with the SEC clarifying the remark. Jabal would handle most of the communications with investors from then on.

Jabal took her job seriously, encouraging the CEO and Google’s founders to be more considerate of institutional investors. She held a series of investor meetings, including a large one in New York. She persuaded Google’s executives to start providing financial projections for the first time. The combined efforts helped earn a place for Google for the first time on a list of shareholder-friendly companies.

In 2006, her primary purpose of designing the IPO complete, Lise Buyer left Google. Inspired by the work she’d done for Google and the experience she’d picked up, the Buffalo native started her own consultancy to advise companies on the IPO process. She called it Class V Group, a nod to whitewater rapids that are so dangerous to navigate that the American Whitewater Association recommended hiring a guide. Buyer thought it was a reasonable analogy for the IPO process.

It wasn’t immediately obvious that she would have a lot of work. After 380 companies went public in 2000, only 282 companies went public in the next four years combined. Another 169 went public in 2005.