How They Started Page 12

  Investors took note. In July 2007, Twitter landed $5 million in venture-capital funding led by former Odeo investor Charles River Ventures. Other funders included Union Square Ventures, Netscape co-founder Marc Andreessen, Ron Conway, and Feedburner creator Dick Costolo, who’d worked at Google with Evan.

  A whale is born

  After SXSW, Twitter would experience the hockey-stick-shaped, straight-up growth curve that is the dream of every start-up entrepreneur. In March 2008, when Twitter hit 400,000 users, the company set out to completely rebuild Twitter’s technology to support the skyrocketing user base.

  It didn’t work. Twitter would be down, sometimes for three days straight. The company still had only two dozen employees. Any big pop-culture or technology event—say, Apple’s Steve Jobs speaking at a conference—could cause a huge traffic spike and crash the system. Jack, Twitter’s first CEO, ceded the post to Evan.

  “We weren’t ready for the number of people around the globe who would find Twitter so useful and so relevant to their daily lives,” Biz recalled in a PBS interview.

  Why couldn’t Twitter fix its technical problems? It had money to hire more staff, but few engineers had both the needed expertise in Ruby on Rails and a willingness to work in Twitter’s frantic start-up environment, Stirman recalls. To top it off, the founders were picky about whom they hired. As a result, Twitter’s hiring lagged far behind its staffing needs. And by June 2008, QuantCast estimated Twitter had 700,000 monthly users.

  At first, visitors attempting to access Twitter during outages saw the Twitter bird gazing sadly at a damaged robot. Biz wanted something more reassuring and purchased a graphic by Shanghai artist Yiying Lu of a whale being lifted out of the water by many birds, each pulling a rope.

  Quickly dubbed the “fail whale,” the image became both an object of derision and a cult hit. The fail whale inspired a fan club, T-shirts, and at least one tattoo. The chronic technical breakdowns did not discourage potential investors, either.

  “When your product is so popular your servers are crashing every day,” Stirman says, “that’s the sort of thing venture capitalists salivate over.”

  In May 2008, Twitter would raise another $15 million from Union Square, Amazon founder Jeff Bezos’s Bezos Expeditions, Spark Capital, Digg founder Kevin Rose, and social media author Timothy Ferriss.

  Businesses@Twitter

  Twitter’s lack of a business plan made it the butt of jokes around Silicon Valley. Entrepreneurs created and compared lists of possible Twitter monetization schemes.

  An April 2008 TechCrunch story on the sacking of two developers held responsible for the persistent outages mocked, “Mighdoll Out—Business Plan Still MIA.”

  Despite this, the founders resisted the idea of simply slapping ads on the site and searched for less-obnoxious revenue ideas. Meanwhile, big companies began experimenting with Twitter. In 2008, Dell reported it had made $1 million selling reconditioned computers from its outlet store by offering discount coupons on Twitter. Cable giant Comcast overhauled its poor customer-service reputation with its Twitter account @ComcastCares, which responded quickly to customer complaints.

  At the same time, other entrepreneurs were building businesses on Twitter’s back. The company’s open source platform let developers use Twitter’s code to create related services. Hundreds of thousands of Twitter apps would be built, including the Twitter scheduler Tweetdeck, which became so popular that Twitter would later acquire it for $40 million.

  Some users tried out new personas on Twitter. A few were so amusing, they ended up with publishing and public-speaking deals. Two standouts were Fake Steve Jobs and Fake AP Stylebook.

  Cable giant Comcast overhauled its poor reputation with its Twitter account @ComcastCares, which responded quickly to customer complaints.

  Rise of the Twitterati

  Aside from the obvious fakery of Fake Steve Jobs, however, Twitter had a big problem with covert imitators. As celebrities and prominent politicians began to show an interest in Twitter, other users were setting up accounts, pretending to be a popular film actor or rock musician, and sending out messages that embarrassed the star. To encourage celebrity participation and eliminate these spoofers, Twitter created a Verified Accounts program.

  One late-2007 adopter was then-presidential hopeful Barack Obama, who cannily used social media to rally voters. In 2009, Oprah, Lady Gaga and Ashton Kutcher joined, with Kutcher becoming the first person to gain a million Twitter followers. The celebrities were a promotional bonanza for Twitter—each drawing press coverage and masses of followers, many of whom joined to connect with their idol. Oprah created a 43 percent traffic spike when she joined live on her talk show in April 2009.

  The celebrity sparkle helped skyrocket Twitter’s audience to five million users by the end of 2008, and to over 71 million in 2009. With just 50 employees in 2008, Twitter struggled to keep up.

  “It’s like we’re on a rocket ship that we’re just painting and suddenly it took off and we’re holding on to the ship with our fingernails,” Biz told the New York Times.

  Institutional investors including T. Rowe Price and Morgan Stanley put another $135 million into the company in 2009, at a reported valuation of $1 billion. The financial support bought Twitter more time to explore non-intrusive ways of earning revenue from its mushrooming audience.

  The new emergency broadcast system

  Users delighted in announcing their trivial activities on Twitter. But the founders always imagined Twitter would have a higher purpose of connecting people around the world to promote good causes.

  In April 2008, Twitter fulfilled that purpose. When University of California at Berkeley student James Buckley was arrested while photographing protests in Egypt, his call for help was a single-word tweet: “Arrested.” His 48 followers quickly contacted the US Embassy and the press. Buckley was soon able to tweet, “Free.”

  In natural disasters, too, Twitter proved invaluable. When an earthquake hit China in May 2008, Twitter became the go-to news source for early disaster reports, as users on the scene tweeted information from cellphones. Reporters used the tweets in their stories, and charities joined in with Twitter-based fundraising appeals.

  Twitter became an essential tool for political activists. This was brought home during Iranian protests in June 2009, when the US State Department asked Twitter to postpone scheduled maintenance that might take the service offline.

  In late 2010, the Middle East began to erupt in anti-government protests that became known as “Arab Spring,” and as oppressive regimes in Tunisia and Egypt toppled, political observers credited Twitter with playing a supporting role. In perhaps the ultimate demonstration of Twitter’s importance, Egypt shut down Internet access in January 2011 to prevent protesters from using Twitter.

  A screenshot of a Twitter news feed.

  A business model is born

  As Twitter struggled to find a way to earn revenue, its leadership changed repeatedly. Evan had taken the reins from Jack in 2008, while Jack left to found the mobile-payments startup Square.

  In fall 2009 Evan brought on investor and friend Dick Costolo as chief operating officer. When Evan stepped down as CEO in October 2010, Costolo became the new CEO. In January 2010, Kleiner Perkins Caufield & Byers added a $200 million investment to Twitter’s pot.

  After some testing in late 2009 with half a dozen corporate advertisers, several revenue-generating programs officially debuted in summer 2010. Companies could buy “promoted” tweets—at $100,000 a day—and hashtagged “trending” topics. (If a promoted tweet isn’t clicked on, a “resonance algorithm” detects this and removes it from view.) Local ads were announced as a coming option, but at the end of 2011 had yet to materialize.

  Twitter reports that its promoted tweets get click rates of 3–5 percent, which is roughly 100 times better than typical click-through rates for online ads. In 2011, promoted tweets began appearing at the top of users’ search results.

  Where are t
hey now?

  In early 2011, Jack returned to Twitter as executive chairman while continuing as CEO at Square. Evan left Twitter, and Biz would follow Evan out the door to the re-formed Obvious Corp. For his part, Noah Glass’s Twitter bio reads, “i started this.”

  In November 2011, Twitter had more than 100 million monthly users and 250 million tweets were posted daily. The company’s valuation was estimated at up to $10 billion. Hiring finally sped up in 2011, with 500 new hires added for 800 employees total (100 of them in ad sales). Most notably, 2,400 companies were advertising on Twitter. One success story: Paramount Pictures estimated that in a couple of hours on Twitter it sold $1.5 million in tickets for the opening day of its film Super 8.

  Internet-research firm eMarketer estimated Twitter would see $50 million in 2011 revenue and $250 million in 2012—substantial income to be sure, but not yet a figure that justifies the level of investment funding Twitter has accepted. Is there more potential ahead?

  “We’ve only achieved 1 percent of what Twitter can be,” Costolo says.

  Will Twitter go public or be sold? The huge amount of venture capital invested—including a mammoth $800 million funding round in 2011—means a cash-generating event will likely come soon. Both Facebook and Google are rumored to have made offers.

  TripAdvisor

  Discovering treasure on its travels

  Founder: Stephen Kaufer

  Age of founder: 38

  Background: Degree in computer science from Harvard University

  Founded in: 2000

  Headquarters: Newton, Massachusetts

  Business type: Travel reviews website

  TripAdvisor’s 60 million-plus user reviews are some of the most widely read hotel and restaurant assessments in the world. The site draws more than 50 million unique visitors each month and ranks among the world’s most trafficked websites. Just over a decade ago, however, TripAdvisor was on the verge of going bust.

  A true innovator, the company rode out the storm and led where others tried and failed, building a “movement”—a thriving community that has nearly eclipsed the company itself. Waving the banner of “citizen tourism,” the company attracts millions daily to its more than 50 million travel reviews. It boasts eight million photos from its members’ travels, and a staggering 98 percent of new topics in its lively forums are replied to within 24 hours. People cannot stay away.

  Going on vacation

  In 1999, Stephen Kaufer just wanted to take a vacation. He and his wife were trying to plan a trip to Mexico, and, on the recommendation of a local travel agency, they went online to check out a few resorts. It was easy enough to find the glossy brochures and guidebooks, he says, but what he really wanted were first-hand opinions from those who had been there. Instead, at every website they visited, they found that the photos and descriptions of each resort were the same.

  Stephen was frustrated. It was the height of the dot-com boom, but the principles of Web 2.0 had yet to take hold. Stephen just wanted an honest opinion of each resort, but all he could find in his fruitless Web searching was an endless loop of the same recycled promotional material. Eventually he found someone’s personal home page that featured a few candid photos from one of the resorts and a paragraph about its facilities. He then realized that if the reality of this resort was quite different from its brochure, then it was probably the same for many more.

  “Each time we were recommended a destination and hotel, I’d go online and search high and low to find more information, and each time, we’d find a problem—the hotel wasn’t up to par, the destination was unsafe, etc,” he says. “It took an enormous amount of time to research these places properly, and it was through these endless searches that the TripAdvisor concept was born—I wanted a single place where I could get the real scoop on a hotel, not just the official blurb from the property or a travel agent.”

  Stephen has always been an entrepreneur at heart. He co-founded his first business, a software development tool company, in 1984 while pursuing his computer science degree at Harvard. That company was eventually sold off in 1998, but, just a year later, Stephen’s experience trying to book a hotel had him again thinking entrepreneurially. He said to himself that there had to be a better way to plan a vacation online. Soon after, in February 2000, TripAdvisor appeared.

  “I was employed at the time, so I put the idea on hold for about a year and then started gathering a group of people who I thought would be interested in creating what would become TripAdvisor,” Stephen says. “In 2000 we were up and running, albeit in a different way than we do now.”

  Life, and near-death, as a search engine

  The company actually began life as a search engine. These were the days when sites like Lycos, Go and Excite were the most visited pages on the Web, and an upstart Google was making a fast name for itself with its authoritative search results. These were Stephen’s models, but approached from the opposite angle. Rather than trying to rank results based on their perceived worth, TripAdvisor simply provided unfiltered links to any travel reviews on the Web. Stephen’s aim was to unearth the raw, candid opinions of services and popular destinations in the travel industry—the opinions that lurked on personal Web pages in the furthest reaches of the Internet. Except it didn’t work.

  “I wanted a single place where I could get the real scoop on a hotel, not just the official blurb from the property or a travel agent.”

  Stephen says he had a site he liked, but he wasn’t able to figure out how it could make money. “When we first started out, our business model was entirely different to what it is today,” he says. “We’d planned to sell TripAdvisor as a rich database to travel portals, online travel businesses, etc. We’d hoped to offer such high quality content that simply being in the travel business necessitated access to the TripAdvisor database, which we would then license out and/or get a share of the revenue generated on the page views from that. All the major players would have it, and no one would try to build it themselves because we’d be so far ahead of the game—that was our planned business model.”

  But after a year, he had only one licensing deal, and Stephen says he often joked that the quarterly revenue check he received from this wouldn’t even cover the weekly free lunch he offered to his employees. “We found ourselves in the midst of a pretty fundamental problem—we wanted to be paid to have our content featured, but everyone else wanted us to pay them to have our content featured,” Stephen says.

  “We came across more trouble when 9/11 hit,” he adds. “It was a hugely traumatic time, particularly for the travel industry. Everything we had in the pipeline was stalled and we were struggling to move ahead and looking at going out of business within the year.”

  In the post-9/11 days, TripAdvisor wasn’t making any money at all, and Stephen was running through cash quickly. To avoid going bust, he slimmed down the team, took a pay cut, and carried on as best he could in the difficult conditions.

  “Getting further investment was a tough sell—we were asking for more money, but couldn’t show how we’d ever make any,” Stephen explains. “We did manage to get a bit more funding, and it was about this time that we took on Expedia as a client and changed our game plan—we started targeting our content at the end user, rather than travel businesses, and making money through cost-per-click (CPC) advertising.”

  Voyage of discovery

  Discovering a new business model was quite serendipitous for him, as Stephen admits that by fall 2001 the company was just months away from folding. Luckily, around this time Stephen noticed something interesting in his traffic figures: when TripAdvisor placed more relevant ads next to its search results, click rates on the ads soared to 15 percent—way more than the industry average of 0.2 percent. As this trend continued, Stephen found his elusive money-making potential and adopted a strategy of providing other companies within the travel industry access to its growing number of members via these targeted ads.

  When users searched for localiz
ed travel information, TripAdvisor began selling keyword-based text ads for companies that offered services within those markets. In essence, the company used the gravitas of its user-generated reviews as a lead through which it fed readers on to other industry sites to complete their transactions.

  “It proved very effective—we abandoned some of our original ideas and saw great success starting with Expedia, and then moving on to other big players and maintaining our growth to become what we are today,” he says. “Finding a way to make this content profitable wasn’t the easiest ride, though, and our business model underwent a few changes, with our biggest early success coming when we discovered TripAdvisor’s unique capabilities in CPC advertising.”

  When he first started out, Stephen never intended to appeal to the end user, but instead planned on selling TripAdvisor’s content to travel companies. When he realized the test site was getting 5,000 hits a day without any effort, he started looking for ways to monetize this traffic. When banner ads didn’t prove too successful, he focused on sending extremely qualified traffic to hotels and online travel agents, and then charging them for this traffic.

  With users searching for specific destinations on the website, they qualified themselves as interested in those specific destinations, and subsequently hotels in those places. Stephen saw there was huge potential here—if a user was looking at a hotel in Paris, and TripAdvisor surfaced a deep link that took that user straight into an online tour agent’s booking pages for that hotel, the company could make some money.

  “We approached Expedia and told them we’d like to advertise their hotels on our website in front of a highly targeted audience,” Stephen recalls. “We’d only charge them for the traffic we sent them, and we explained how highly qualified our traffic was. At the time, Expedia had no idea who we were, but we’d piqued their interest and they granted us a trial period.