“Jack,” Draper asked, “could you put a message at the bottom of everybody’s screen?”
“Oh, come on, we don’t want to do that!” Bhatia blurted out.
“But can you technically do it?” Draper asked.
“Of course we can technically do it,” Smith said.
“Oh, great,” Draper said. “And it can persist, right? You can put it on one message, and if he sends an email to somebody else, you can put it on that one, too, right?”
“Yeah, yeah,” Smith said, not convinced.
“So put ‘P.S.: I love you. Get your free email at Hotmail’ at the bottom.”
Bhatia and Smith communicated through pained expressions. “Oh, no,” they seemed to be saying. Draper had seen that look before. Of all the investors in the world, why did we end up with this idiot? Frankly, he didn’t care what they thought. This just felt right.
“Wait a second, guys, don’t you get it?” Draper asked. A tagline at the bottom of each message would act as free advertising. “I can send you an email and you can send it to all your friends and they get it and they can sign up and send it to their friends and pretty soon it takes off.”
Smith said, “I don’t think…”
Bhatia interrupted, “Let’s move on to other business.”
Draper agreed to table the discussion for now, but he had no intention of letting it go. He vowed he would keep pounding until they listened.
They launched HoTMaiL on Independence Day 1996. Not only did they like the symbolism—they viewed webmail as a populist tool because any user could log in from anywhere in the world—Smith had long promised the service would be ready by then. After turning on the registration function and hitting the switch in the early afternoon, Smith accompanied his tiny technical staff to Chili’s Grill & Bar in San Jose to celebrate. To keep track of sign-ups he brought along a laptop with an attached radio modem receiver on the back, the antenna sticking up like a divining rod. Over quesadillas Smith counted one hundred registrations in the first hour. After lunch they went to the movies, and by the time the summer blockbuster Independence Day began to roll, he tallied a total of two hundred sign-ups. Upon exiting the cinema, Smith logged in again to find that fifty more had joined HoTMaiL. They were finding the site via word of mouth and word of mouse. People were talking about it, and letting their friends and family in on the deal via email, using the Hotmail message as a proof of concept: 80 percent of those who signed up said that they learned about it from a friend.
Growth was robust but not staggering for the week. At the next meeting at DFJ Tim Draper once again pushed the two young entrepreneurs to insert a tagline into each message. Bhatia and Smith were adamant about not adulterating email. It just wasn’t done. They would feel like they were polluting emails with advertising, and what about privacy issues? If someone was adding a tagline, what else were they doing? A user would wonder what else they had access to, and they were also fairly certain it was unethical. But Draper wouldn’t let it go. The benefits, he contended, far outweighed the risks. If they were predicating their entire business on the size of their user base, they should be doing everything in their power to increase it as fast as possible. “P.S. I love you. Get your free email at HoTMaiL.” The more he said it, the more he liked it.
The next day Bhatia phoned Draper with the news that they agreed to do it, but without the “P.S. I love you” part. The impact was almost instantaneous. Within hours Hotmail’s growth took on the shape of a classic hockey stick curve. They started averaging 3,000 users a day, compounded daily. By Labor Day they registered 750,000 users and within six months they were up to 1 million. Five weeks after that they hit the 2 million user mark, adding more than 20,000 sign-ups a day, with Smith desperately trying to keep the servers up and running. At times, the site became sluggish and suffered major outages. But through it all, Smith, using little more than virtual spit and glue, kept Hotmail (they had dropped the awkward capitalization by this point) afloat.
The tagline with the clickable URL that Draper insisted be inserted in every outbound message served as a promotional pitch for the company. Simply by using the product, every customer became an involuntary salesperson. This implied endorsement from a friend or peer made it more powerful, and more far-reaching, than traditional advertising. The receiver of a Hotmail message could see that (1) his friend was a user, (2) it worked, and (3) it was free. Successful consumer branding is often based on user affiliation. (The cool kids wear low-cut jeans, so I will, too.) This plays to our tribal instinct, and resulted in clusters of users. Bhatia sent a message to a friend in India and within three weeks Hotmail registered 100,000 users there. It also became the largest email provider in Sweden without spending a nickel on advertising there. In contrast, Juno blew through $20 million in marketing and advertising, while Hotmail gained three times as many users in half the time.
As Jurvetson related in a white paper, the Hotmail adoption pattern was similar to that of a virus “with spatial and network locality.” A person’s email address book is a type of virtual social network that is not encumbered by geography. A certain percentage of contacts will be friends, family, and colleagues who reside relatively nearby; others may be scattered throughout the world. A Hotmail message sent across the country might result in a new cluster of users. “We would notice the first user from a university town or from India, and then the number of subscribers from that region would rapidly proliferate,” he wrote. “From an epidemiological perspective, it was as if Zeus sneezed over the planet.”
Jurvetson noted a “mathematical elegance” to Hotmail’s “smooth exponential growth curves” in the company’s early days:
Cumulative users = (1 + fanout) cycles
Where “cumulative users” related to the number of Hotmail registered subscribers, “fanout” was the rate by which the product spread and “cycles” was the number of times the product was used in the time period since launch (or frequency multiplied by time). At the beginning, each Hotmail user, on average, brought in two new users each month. (In other words, the fanout equaled 2.) These two new Hotmail subscribers attracted two new users, and so on. That meant that one seed user equaled three users at the end of the first cycle, nine by the second, twenty-seven by the third, each cluster growing exponentially. Of course, this was a simplified model. There were other variables at play, like retention rates and “churn” ultimate saturation, when Hotmail reached so many users that its growth inevitably slowed, et al.
[ AN INSURMOUNTABLE LEAD ]
While companies like Hot or Not could toss more servers into the mix to keep up with increased bandwidth demand, Hotmail, which was suffering brownouts, system failures, and painfully slow download times, had to completely restructure its systems on the fly. This fell to cofounder Jack Smith, who had originally launched Hotmail on a nonscalable architecture because the company simply couldn’t afford anything else. He could only “silo” a limited number of webmail accounts on a given machine. “To add accounts we had to add machines,” Smith says, “but the registration system was a global concept. We had a hard time spanning the machines and we didn’t have a universal database.” He and his engineers were forced to manually customize machines to handle additional traffic, an inefficient, time-consuming process. “We were still growing like crazy, but some days we had such a hard day we had to shut off registration because we couldn’t accept more new users at that moment.”
As Hotmail was poised to register its millionth user, Smith designed a multitiered architecture to scale without limit. Instead of his original single-tier system, which stowed emails on one set of machines, the registration database on another, and the site on a third, he set it up so the load could be balanced across the entire network. “You had your database on one tier, the front end on another, and your mail spread over numerous machines instead of one per account,” he says. At the time there was no out-of-the-box server software with this capability, so Smith had to create new technology.
/> It took a month to design three major components—an email transporter, a back-end database, and a customer authentication database. Once it was in place, any technician could blindly add machines without reconfiguring anything if the site started getting slow. But when Smith cut over to the new architecture it immediately melted down, and he quickly switched back to the old system. It didn’t take long to diagnose the problem. Because of budgetary constraints, Smith had skimped on hardware. He went out and bought two dozen far more powerful motherboards, grabbed whomever he could find in the office, and brought them into their data center, where they swapped out the old ones. “The next night we cut over to the new architecture and it worked beautifully,” he says.
With a robust infrastructure Hotmail continued its frenetic growth, reaching 3 million users in two hundred countries by April, becoming the third largest email provider after AOL and CompuServe. Bhatia returned to the financing table for more money, this time securing $3 million from DFJ and Menlo Ventures at a more agreeable valuation. Hotmail had topped 5 million by its one-year anniversary, registering 60,000 new users a day, and on the way it had earned a reputation for speed and reliability. While messages could take a minute to transmit between two AOL accounts, and between fifteen minutes to several hours to travel between AOL and non-AOL accounts, Hotmail was significantly faster. And after a shaky start, it didn’t suffer the kinds of outages plaguing AOL, which at times simply couldn’t handle the flood of email.
In addition to its growth, the company was generating advertising revenue. For the first few months of Hotmail’s existence, Bhatia posted ad banners for free, just to give the site the patina of e-commerce professionalism; then he began layering in paid advertising. Revenue for the fourth quarter of 1996 was $350,000. Bhatia quickly found that he had neither the time nor the inclination to sell ads, so he outsourced this function to another company in exchange for a percentage of the revenue. According to Bhatia, the deal involved a million-dollar-a-month minimum. While Hotmail was still hemorrhaging money, it showed signs of profitability, and one month it actually flirted with breaking even.
By September 1997, with Hotmail edging toward 7 million subscribers and adding more features—instant messaging, a spell-checker, email search, and the ability to attach multiple files to a message—potential suitors appeared. General Electric pledged to head another funding round and offered to hook up Hotmail with its back-end credit card service to distribute monthly bills to consumers. John Doerr of venture capital firm Kleiner Perkins also wanted to lead a financing round and take a seat on the board; this would put Hotmail into a position for an initial public offering. An investor in Excite, valued at some $400 million, was interested in negotiating a merger, which would give Hotmail 33 percent of Excite’s stock and the opportunity to handle the Web portal’s email. Yahoo also floated the idea of joining forces with Hotmail, and Microsoft, after broaching a partnership, made an offer to buy Hotmail outright.
Each came with pluses and minuses. GE had enviable connections with NBC and MSNBC, which could transform Hotmail from a simple emailer to a destination site providing news and entertainment, but it valued Hotmail at $105 million, $20 million less than Bhatia wanted. Kleiner Perkins was the gold standard in tech IPOs, but arrived at an even lower valuation for Hotmail than GE had. Both Bhatia and Smith believed the offers from Excite and Yahoo, which was willing to bid $200 million, were far too low, and their mantra became: if you come in with anything less than half a billion dollars, get lost! That left Microsoft or the option of going it alone, building the business, then looking to cash in on an IPO.
Jurvetson advised Bhatia to spurn Microsoft and march toward going public. Unlike Netscape, which despite its staggering IPO was vulnerable because Microsoft could require PC makers to install Internet Explorer as the default Web surfing software and separate it from its user base, Jurvetson believed Hotmail could weather anything Bill Gates threw at it. He had been following discussion among analysts who believed the Web challenged Microsoft’s chokehold on the industry, a view he shared. The Windows environment didn’t matter much if you could jump the on ramp to the Web from any computer at any time. Microsoft had created its own webmail service on the Microsoft network (MSN) but had been having tremendous scaling problems. And while Hotmail was coasting past 10 million sign-ups (Fig. 8), MSN barely broke 2 million.
FIG. 8. Hotmail subscriber growth, 1996–1997. (Draper Fisher Jurvetson)
It was, Jurvetson thought, an almost insurmountable lead, and the key was Hotmail’s viral marketing. Even if Microsoft could somehow buy its way out of MSN webmail’s lagging growth, its viral coefficient still would never match Hotmail’s. Assuming it doubled at the same rate as Hotmail, only time shifted, and this meant the delta between the two got even bigger. When Hotmail hit 20 million users, Microsoft would only be at 4 million; Hotmail at 40 million subscribers, Microsoft at 8 million; Hotmail at 80 million, Microsoft at 16 million. Then came ultimate saturation. The way Jurvetson saw it, Microsoft’s only chance was to sabotage Hotmail’s servers.
[ HAGGLING OVER PRICE ]
Bhatia, however, was pondering two potential endgames that had recently played out with other tech companies: The first involved AOL, which rejected an offer from Microsoft and went on to multibillion-dollar revenues. On the other end of the spectrum was another toast of the Valley, PointCast, which had rebuffed Rupert Murdoch and News Corp’s $400 million bid and then faded into oblivion. Bhatia decided to re-enter the bazaar and try his hand at haggling with a team of Microsoft negotiators, who in the fall of 1997 jumped a plane from Redmond in a group of six to sit at a table in Hotmail’s conference room. At first they talked of a partnership, with Hotmail providing webmail to MSN, but after learning that Hotmail planned to move into news and entertainment to become a portal like MSN they said they couldn’t partner with a potential competitor. Instead, had Bhatia considered acquisition? At the right price, Bhatia replied, he could think of anything.
In October 1997, Bhatia, Smith, and Hotmail’s vice president of marketing traveled to Redmond and, after a tour of the campus, met with Bill Gates and two senior executives in his office. No small talk, right down to business. Initially intimidated by Gates, who was renowned for tearing apart weak arguments and weaker minds, Bhatia quickly recovered his bearings and the meeting lasted an hour and a half. Afterward Bhatia and Smith sat at an enormous table, facing a team of fifteen Microsoft negotiators—the chief financial officer, lawyers, business development staff, and accountants—who told them they wanted to buy the company for $160 million.
Bhatia thanked them for the offer and said he had to discuss it with his board.
“C’mon,” Gregory Maffei, the CFO, said, “is that in the ballpark?”
Bhatia refused to negotiate with so many people, knowing it put him at a distinct disadvantage. Back home, he polled his investors, half-joking he could get a billion for the start-up. The lowest figure came from Doug Carlisle at Menlo Ventures, who promised Bhatia that if he could get $200 million, he would commission a bronze sculpture in his likeness and install it in his lobby.
Returning to the bargaining table, Bhatia informed Microsoft’s swarm of negotiators that he wanted $700 million. This prompted the predictable storm of invective: “You’re crazy,” “Out of your mind,” “You’ve blown it,” amidst a stream of obscenities. Bhatia recognized the bluster for what it was: the American version of the Indian bazaar. Perhaps he should turn out his pockets and offer them a few dollars. You must be poor. He consulted his beeper and recited Hotmail’s latest subscription numbers, which were growing by the second. At that rate, by the end of the year they would have more than 12 million users and be, by far, the number-one email provider in the world. Microsoft upped its bid to $200 million and Carlisle crowed, “It’s statue time,” but Bhatia stayed the course.
The weeks dragged on and the pressure mounted. Several times Microsoft team members stormed out, but they always returned. Microsoft raised its o
ffer to $300 million and negotiators hinted it could always buy RocketMail, a much smaller competitor. Bhatia didn’t waver. Jurvetson backed him. “You don’t have to sell now,” he said. “Why don’t you wait until you’re big enough to buy Microsoft, rather than them buying you?” But his own executives, with dreams of a quick cash-out on their minds, urged him to take the money. Even his parents got into the act. He told his father that he had turned down $325 million. His father mumbled, “Um, okay,” and they hung up. A minute later the phone rang. It was his mother. “Sabeer, listen to me. You call that Mr. Gates back right now and you take his offer! Do you know how much that is in rupees?”
After Microsoft placed $350 million on the table, Hotmail management took a straw poll and voted to accept it. Bhatia overruled them, although he would later admit it was the scariest thing he had ever done. He heard more than once that it would be on his head if he screwed things up. Finally, on December 31, 1997, a few days after his twenty-ninth birthday, Bhatia closed the deal and Microsoft transferred ownership of 2,769,148 shares of stock, worth almost $400 million.
Less than two years earlier, Bhatia had entered the offices of Draper Fisher Jurvetson with little more than an idea—no prototype, not even a mock-up—which he sketched on Jurvetson’s whiteboard. Now he and Smith had created a viral loop business that had grown faster than any company in history and in the process had made each of them $75 million richer. But Carlisle never commissioned that statue. Bhatia’s mother told her son it was bad luck to have one modeled on a living person. Instead, Hotmail set the stage for other viral companies to follow suit: The first was ICQ (short for “I seek you”), an instant-messaging program launched by five Israelis four months after Hotmail. Within six months it amassed 850,000 users, doubly impressive since users had to install software.
Viral Loop Page 11