CONTROLLING CHAOS: Twitter's Wild Ride From Doodle To IPO
In the fall of 2010, Twitter was in chaos.
The micro-blogging site, which had been founded four years earlier, still wasn't generating much revenue despite having 145 million users. The site was still crashing constantly. The staff were at each other's throats.
Chaos, moreover, was not an unfamiliar state for Twitter to be in.
Twitter's first CEO, Jack Dorsey, had taken the company from idea to reality, but then made a mess of the place before getting fired. Twitter's current CEO, Evan "Ev" Williams, had overseen huge user growth but was now moving too slowly and indecisively.
Twitter's board decided it now had to fire Williams, too. But who would take over?
A former stand-up comic and entrepreneur named Dick Costolo was one possibility. He was Twitter's COO. But Costolo had never run a large company before, so he hardly seemed the guy to finally get the chaos under control.
The largest company Costolo had managed, in fact, was the one he had founded, Feedburner. A content syndication service, Feedburner had had fewer than 100 employees when Google acquired it in 2007. Twitter already had more than 300. Also, Costolo was 47. In Silicon Valley, that's old. Famous founders like Mark Zuckerberg and Drew Houston were half Costolo's age and more accomplished.
At wits end, the Twitter board reviewed its options.
It could make Costolo an interim CEO until it found someone more established. Twitter had worked frequently with an executive recruiter, Paul Daversa, to unearth talent, and the board had a few more serious-seeming candidates in mind.
Or the board could take a chance on Costolo, who had risen to the occasion once before. The Twitter COO job had been a leap for Costolo, and he had excelled at it. Costolo also already knew the company's inner workings, and he had the respect of Twitter's management team, partially because he had helped hire most of them. Fred Wilson, a Feedburner investor and partner in Union Square Ventures, and one of Twitter's board members, trusted him.
In typical chaotic Twitter fashion, the choice was ultimately made by the man who was getting sacked.
In addition to being Twitter's CEO, Evan Williams was its largest shareholder. Williams and Costolo had known each other for more than a decade, having been introduced by a mutual friend in the mid-1990's, Eric Lunt. To soften the blow of his ousting, Twitter's board told Williams he could be part of the CEO search process. Initially, Williams wasn't thrilled with the idea of Costolo taking his place. But he came around and eventually pushed for Costolo to permanently take his job.
On October 4 2010, Williams resigned as Twitter's CEO. Costolo took over.
With Twitter set to go public at a $15+ billion valuation next month, the company's success seems as though it has always been smooth sailing.
In fact, it has been anything but.
The whole Twitter story, for example, even starts with an argument: Who is really responsible for the company's success?
Does the credit belong to Jack Dorsey, who came up with the original idea? Does it belong to Evan Williams, who backed the concept and then took over as CEO? Should it go to Dick Costolo, who turned Twitter into a real business? Or do Twitter's investors or other lesser-known heroes deserve the nod?
The truth is that, more than any other rocketship Silicon Valley startup in memory, Twitter's success has been a product of collaboration, not one genius visionary. At different times in its history, different people have provided the leadership, vision, and skills that the company needed.
But this group collaboration has been the farthest thing from pretty.
When Costolo was handed Twitter three years ago, the company was a mess. Its four-year history was bloodied with founder feuds, failed products, office politics, and executive turnover.
"The act of getting from there to here," Benchmark Capital's Peter Fenton says, "was violent."
Why was Twitter's early history so chaotic? And how did it finally get control of the chaos?
Over the years, we've reported many episodes in the Twitter story. Recent interviews with a number of sources, as well as reporting by other tech journalists, have helped us fill in some gaps. So here is the wild ride...
Business Insider/Mike Nudelman
People close to Twitter during its first five years describe the company as "chaotic," "stressful," an "absolute mess" and "a zoo."
That's because founders establish a company's culture. And Twitters founders fought a lot.
They fought about what the company's product should be. They fought about who should run it. And, right from the beginning, they even fought about who had created it.
The Twitter founding story most people know includes three people: Jack Dorsey, Christopher "Biz" Stone, and Evan Williams.
It goes something like this:
Biz Stone and Evan Williams had worked together at Google, after Williams sold a company called Blogger to Google. The pair then started a company called Odeo, a podcasting company that soon ran aground. Jack Dorsey, one of Odeo's employees, came up with the idea for Twitter. Stone and Dorsey made the prototype, and Evan Williams then built a company around it.
That's the story Stone told Howard Stern in a 2011 interview and that Dorsey has since repeated. But it's a lot different than the story Williams told on the day Twitter launched publicly in 2006.
"[Twttr is] a 20% project, if you will, that we-esp. Noah, Jack, and Florian-have been working on on the side for a couple months," Williams wrote in July 2006. "It has nothing to do with audio or podcasting [like Odeo did], but we think it's kinda neat."
Who are Noah and Florian?
According to early Twitter employees, they're the company's forgotten founders. Particularly Noah Glass, who started Odeo in his apartment before he ever even met Evan Williams.
Glass and Williams had met four years before Twitter was founded, in 2002. They were neighbors in San Francisco's Noe Valley. Glass had recognized the Blogger logo on Williams' computer and later spotted Williams in a magazine as the blogging site's founder. That's when Glass mustered up the courage to introduce himself.
"He was working on this wacky idea he later pitched me, which let people call a phone number, record a message, and then publish it to the web (via Blogger) by pressing a button," Williams wrote of his first meeting with Glass.
That "wacky idea" became Odeo, a podcasting startup. And Odeo eventually morphed into Twitter.
At first, Williams was an investor and advisor to Odeo, not a founder. Then he took on a more active role, acquired co-founder status, and became the company's CEO.
Technically, Christopher "Biz" Stone was an Odeo co-founder, too. But he came into the picture later.
Eventually Odeo moved from Glass' apartment to Williams' apartment. Williams had used some of the money he got from selling Blogger to buy a house. His old apartment was a perfect headquarters for the startup.
In those days, Glass was Odeo's leader, not Williams. "I think [Odeo] was something Ev was interested in, but it was mostly Noah's thing," says former Odeo employee Ray McClure. Other early Odeo employees included Evan "Rabble" Henshaw-Plath and his wife Gabba. "Mostly it was the four of us working out of the apartment," McClure said.
Odeo also had an intern named Kevin Systrom, who would later go on to found Instagram. Systrom sat next to Dorsey and talked a lot about photos.
By June 2005, Odeo had a functioning podcasting product. But when it launched, people were skeptical of its potential.
"We apologize if the following profile lacks our usual excitement and enthusiasm," TechCrunch founder Michael Arrington wrote while profiling Odeo in 2005. "We just have a funny feeling that iTunes 4.9 is going to have a somewhat substantial impact on the podcasting portal market."
Arrington was right. Apple came out with an iTunes podcasting platform that it built into its iPods - and it then went on to sell 200 million of the devices. Also, Odeo employees found that they weren't even using their own product. "We built [Odeo], we tested it a lot, but we never used it," engineer Blaine Cook recalls.
In fall of 2005, Odeo investor George Zachary of Charles River Ventures says, the "shit hit the fan."
By that time, Odeo had 14 full-time employees. Biz Stone was one of them.
Williams and Glass still had money in the bank from the $5 million the company had raised. But the pair knew they either needed to shut down or pivot to a new idea. They told Odeo's employees to start coming up with new startup ideas that Odeo could try. "Hackathons" - entire days or nights dedicated to coding new products -became regular occurrences. Some employees broke off into groups.
Odeo co-founder Noah Glass gravitated toward Jack Dorsey, who Glass says was "one of the stars of the company."
Dorsey had an idea for a completely different product, "Stat.us." Stat.us would show people what their friends were doing at any given time.
The quiet designer had sketched out the idea for Stat.us on a piece of notebook paper years before joining Odeo. Now, he shared this idea with his Odeo co-workers over Mexican food on a playground near Odeo's office, South Park in San Francisco.
At first, Glass wasn't impressed. "He started talking to me about this idea of Stat.us," Glass recalls. "I was trying to figure out what it was he found compelling about it."
Suddenly, the brilliance of Stat.us struck Glass.
"We were sitting on Mission Street in the car in the rain," he says. "We were going out and I was dropping [Dorsey] off and having this conversation. It all fit together for me."
In February 2006, Glass, Dorsey, and a German contract-developer named Florian Weber presented Dorsey's idea to the rest of the company. Glass changed the name from "Stat.us" to "Twttr." With Twttr, a user could text a message to a phone number, and the message would be broadcast to all of the user's friends.
Text messages are limited to 160 characters. But each Twttr user needed a user name or handle attached to the message. To make room, each Twttr text was eventually limited to 140 characters.
Glass pushed for Odeo to drop other side projects and pursue Twitter. For that reason, other Odeo employees considered him Twitter's "spiritual leader." Williams green-lighted the project and put Glass in charge.
"In terms of our new projects, I feel most strongly about Twitter," Williams wrote Odeo executives in an email, excluding Dorsey. Williams was often indecisive and relied heavily on his gut when making decisions, as he did in this case. "We could have a lot more discussion, and I may change my mind, but I think I just need to make a call at this point, and my gut is pulling me to Twitter."
TwitterAn early sketch Biz Stone made of Twitter and its search functionality.Dorsey was still just an Odeo engineer then, but he held an important role in the development of Twitter. He and Florian Weber coded the service while Stone worked on designs. Glass worked on product development and came up with hallmark Twitter features such as time stamps on tweets.
At one point, the entire early Twitter service was running on Glass's laptop, an IBM Thinkpad, and a Verizon wireless card. "It was right there on my desk," says Glass. "I could just pick [up Twitter] and take it anywhere in the world. That was a really fun time."
Glass and Dorsey were more captivated by Twitter than anyone else at Odeo.
"There were two people who were really excited [about Twitter,]" says Odeo investor George Zachary. "Jack and Noah Glass. Noah was fanatically excited about Twitter. Fanatically! Evan and Biz weren't at that level. Not remotely."
Glass was excited about Twitter because of its ability to connect people in real time. In 2006, Glass especially craved emotional connections; his marriage was on the rocks.
"You know what's awesome about this thing?" Zachary remembers Glass saying. "It makes you feel like you're right with that person. It's a whole emotional impact. You feel like you're connected with that person."
His emotional investment in Twitter made it difficult for Glass to deal with what was about to happen.
Flickr via Jack Dorsey"On the phone with the ever-late Noah Glass," Jack Dorsey wrote on June 7, 2006. "The working ever late Noah Glass to you!" Glass responded.By March of 2006, Odeo had a working Twttr prototype. In July, TechCrunch covered Twttr for the first time. This time, Odeo's employees were actually using the product.
Odeo employees became obsessed with Twitter. In those days, Twitter messages were sent by SMS, a texting service that phone companies charged for. Odeo employees racked up SMS bills that totaled hundreds of dollars a month. The company agreed to pay their bills.
That August, a small earthquake shook San Francisco, and word about it quickly spread through Twitter. That was an early 'ah-ha!' moment for Twitter users and company-watchers alike. By that fall, Twitter had thousands of users.
That's when Odeo engineer Blaine Cook began to feel that there were "two companies" at Odeo - the one "Noah and Florian and Jack and Biz were working on" (Twitter) and the one everyone else was working on, Odeo.
Twitter, says Odeo employee Ray McClure, "was definitely the thing you wanted to be working on."
Not everyone saw Twitter's promise. At an Odeo board meeting that summer, Noah Glass presented Twitter to the directors. He and Stone even made a video about "Twttr 101." The board hardly blinked at it.
The Smartest Stock Buyback In History
The board's lack of enthusiasm for Twitter, in fact, soon helped Williams make the most lucrative business decision of his life.
In September 2006, a couple of months after Twitter launched, Williams wrote a letter to Odeo's investors. In it, Williams told them Odeo was going nowhere and that he felt bad about that. He felt so bad, he said, that he wanted to buy back their shares so they wouldn't take a loss.
In this letter to Odeo's investors, Williams wrote briefly about Twitter.
By the way, Twitter (http://twitter.com), which you may have read about, is one of the pieces of value that I see in Odeo, but it's much too early to tell what's there. Almost two months after launch, Twitter has less than 5,000 registered users. I will continue to invest in Twitter, but it's hard to say it justifies the venture investment Odeo certainly holds -- especially since that investment was for a different market altogether.
Williams proposed buying back Odeo investors' stock with his own money. Eventually, the investors agreed. Multiple investors, who combined had put $5 million into Odeo, say that Williams made them whole.
In other words, all of Odeo's assets - including Twitter - were sold to Evan Williams for approximately $5 million. That's why Williams is the largest shareholder of Twitter today, owning 12% of the company.
By comparison, Jack Dorsey now owns less than 5% of Twitter. It's not clear how much of Twitter Biz Stone or Noah Glass own. They either sold their shares, or their percentages are so trivial that they're not mentioned in Twitter's financial filings.
One big question Odeo investors still have is this:
Did Williams have an inkling of how valuable Twitter might become when he offered to buy back Odeo? Was the buyback offer a Machiavellian ploy? Or did Williams really just feel bad about the failure of Odeo's original idea and want to make his investors whole?
"Could Evan have known this would be the world's best thing ever and hid it while re-capitalizing the company?" one investor asks. "If there's ever any litigious stuff in the air it will be: How much did Evan know about the user engagement and numbers of Twitter at the time of buying it out?"
Those are fair questions. Williams knew that some Twitter's early users were highly engaged because he was paying monthly SMS text bills - as high as $400 - for his employees. Glass remembers mobile carriers telling him they'd never seen so much SMS activity before Twitter.
"You're going to want to bump up to the unlimited txting [sic] plan," Williams had written when he announced Twitter on his blog in July 2006.
Most Odeo investors, with the benefit of hindsight, echo the sentiments of James Hong, the cofounder of HotOrNot.com and an Odeo angel investor.
"Obviously, I wish what happened hadn't happened," Hong says. "There was a dark period where I didn't want to hear about Twitter."
Another investor agrees. "I wish he had reached out to me," says Mitch Kapor, still an active and successful investor in the Valley. "I think he could have, but he didn't. And I'd say it's sort of a shared responsibility."
One reason people feel comfortable bringing up questions about Williams' motives is because the Twitter co-founder has a reputation for being shrewd. He was described by many Odeo investors and employees as "calculating." And he had taken some flak for his behavior at his earlier company, Blogger: His co-founder, Meg Hourihan, told The New York Times' Clair Cain Miller after the Google buyout, "I don't think [Williams] took care of the people who got him to where he was." The Times also quoted Williams as saying that "all successful businesspeople make enemies along the way."
One early Odeo employee thinks Williams knew exactly what he was doing:
"Ev decided there was something interesting enough in Twitter that he wanted to buy all the assets and buy everyone out," he says.
Another employee, however, gives Williams the benefit of the doubt.
Matt BiddulphEarly Odeo employee Blaine Cook would become Twitter's first CTOOdeo engineer Blaine Cook recalls a meeting during the summer of 2006, a couple of months before the buyback, about whether or not to shut down Twitter. Cook says it wasn't obvious that Twitter would a big hit until six months later, in the spring of 2007, when it captivated the South By Southwest Interactive conference in Austin, Texas.
Also, a few months after buying Odeo back from its investors, Williams offered to let a select few buy into Twitter at a $25 million valuation, including Charles River Ventures, Silicon Valley angel-investing godfather Ron Conway, and Conway's SV Angel fund.
Tim O'Reilly, an Odeo investor who once employed Williams and runs O'Reilly Media, reflects: "It's certainly possible that Ev is more Machiavellian than he appears. I don't know. I take it at face value that he was doing what he thought was best. It's very easy to look back and say, 'Wow, I'd like to have a bigger piece of that.' It's very easy to say that."
CNBCThe park in San Francisco where Jack Dorsey first mentioned Twitter to Glass, and Glass was fired by Williams.In any event, by the fall of 2006, Williams had successfully purchased Odeo's assets back from investors. He promptly changed the company's name to Obvious Corp.
Williams had also concluded that one key Odeo employee wouldn't be part of the new company: Odeo founder Noah Glass.
On July 26, 2006, shortly after Glass presented Twitter to Odeo's board, Williams had asked Glass to go on a walk. The pair wandered over to South Park, where Dorsey had first mentioned the idea for Twitter that spring.
There, on a bench, Williams fired Glass.
Glass was stunned.
Williams offered Glass six months of severance while his Odeo stock continued to vest, according to a comprehensive new book by Nick Bilton of the New York Times. If Glass didn't accept the offer, Williams said, he would be publicly fired.
Glass took two weeks off and then told fellow employees he'd be leaving. Everyone seemed shocked.
"I remember when Noah told me he wasn't going to back to Twitter," says former colleague Ray McClure. "[It was] hard to hear him say that. It kind of blew my mind because I felt like we all identified with this, and of course I was worried about the team."
It was heartbreaking to see someone who cared so much about Twitter forced to sever ties with his product.
"Noah got really into it," says early Odeo employee Rabble Henshaw-Plath. "Seriously obsessive. I-don't-care-if-my-marriage-dies-I'm-focused-on-this into it."
So why did Williams fire Glass?
One reason may be their clashing personalities. Glass is loud, and Williams is quiet.
"Noah, you can always hear him talking," says McClure. "Ev, you can always hear him thinking."
Twitter via @cyrstalA coworker remembers Glass reprimanding Crystal Taylor during a stressful day, which may have led to his firing.One Odeo employee speculates that Williams fired Glass because Glass could get too emotional. After all, he was dealing with a failing company paired with a failing marriage. The employee remembers a time when Glass was "a little hard" on girl named Crystal Taylor, who remained at Twitter for a long time and ran its customer service division. "I think it was a day that he was kind of stressed. He was a little volatile."
Glass felt his firing was the result of a power struggle. Glass thinks he might have expressed too much interest in running Twitter for Williams' liking. Early on, before Williams or Stone were believers in Twitter, Glass wanted to split the product off as its own company and become CEO.
"I actually had done all the paperwork and was ready to roll," says Glass. "[Twitter as a stand-alone company] was ready to go. That's probably part of the reason why I'm no longer involved with it...I told [Williams] I would do things differently. When you speak truth to power, the ramifications can go a lot of different ways."
Another reason for Glass' firing may have been because another Odeo employee - an employee Glass thought was his friend - wanted him out.
According to Bilton's book, Hatching Twitter: A True Story of Money, Power, Friendship and Betrayal, Jack Dorsey met with Williams a few weeks before Glass was fired and told Williams he'd leave if Glass wasn't canned.
At first, Glass walked away with no Twitter shares. Later he was given some, although the amount he'll make when Twitter goes public is hardly life-changing.
Florian Weber, the other early employee who never received much credit for Twitter's founding, got even less. In 2005, Weber was hired by Williams and Odeo as a contractor, not a full-time employee. To this day, he owns no Twitter stock.
"It's not something that I pushed terribly hard for," Weber recently told The New York Times' Nick Bilton and Vindu Goel. Weber says he has no regrets and that he's happy with his life.
For Glass, being ousted from Twitter has been hard to stomach.
"I felt betrayed by my friends, by my company, by these people around me I trusted and that I had worked hard to create something with," Glass says. "I was a little shell-shocked. I was like, 'Wait...what's the value in building these relationships if this is the result?' So I spent a lot of time by myself. And working on things alone."
It's important to note that while Williams may have fired Glass, he has mentioned Glass' involvement in Twitter multiple times since letting his former friend go.
"Some people have gotten credit, some people haven't. The reality is it was a group effort," says Glass. "I didn't create Twitter on my own. It came out of conversations. I do know that without me, Twitter wouldn't exist. In a huge way." When news of Twitter's impending IPO broke in September, Glass crafted a message on the network he founded for the first time in months.
In October 2006, after Williams bought back Odeo's assets, he announced Obvious Corp to the world. He said Obvious was a new company fully funded by him, and that he alone owned Odeo's assets, including Twitter.com.
"I believe there is a lot of value in what we've built-both Odeo and Twitter-but I did not believe the [old] structure was going to lead to the kind of success we wanted," Williams wrote then. "In the new company, with a new structure, and a new model, I think they are great investments."
By early 2007, Twitter began to take off. Williams says rapid growth kicked off around January 2007 after Twitter opened its application programming interface (API) to the world.
In March 2007, Twitter won a prize for best startup at the South by Southwest conference. Twitter received so much attention there that its site crashed for two days.
One month later, Twitter had become big enough to become its own company. It spun out of Obvious Corp and Jack Dorsey, who was then 30, was named CEO.
"Ev called Jack a 'genius' but the word we use more often in the office is 'mastermind,'" Biz Stone wrote when the news was announced.
These days, Dorsey is widely praised for founding Twitter and another highly successful company called Square. Now 36, Dorsey has already accomplished so much that he's occasionally compared to Steve Jobs. But in April 2006, when he became Twitter's CEO, Dorsey wasn't yet well known. He was also inexperienced, and leading Twitter proved difficult for him. It wasn't long before his relationship with Williams began to sour.
In the summer of 2007, an earthquake shook Mexico City. Seven people there gave updates on Twitter. Their messages spread before the U.S. Geological Survey reported the earthquake, and an hour before CNN caught wind of it. By then, it had become clear that Twitter was no longer just a tool for telling friends where you were at all times. It was a fast way for news to spread, from people who were experiencing events in real time.
That summer, Twitter closed its first round of funding at a ~ $25 million valuation. Some of Odeo's investors were invited to participate. New faces, such as Netscape and Andreessen Horowitz founder Marc Andreessen, also joined the group.
One of these new faces would prove to be particularly important.
Williams reached out to an old friend, Dick Costolo, about investing in Twitter. The paid had been introduced a few years back, and Costolo had just sold Feedburner to Google. Williams asked Costolo via email if he would consider investing either $25,000 or $100,000 in Twitter.
It only took three minutes for Costolo to reply, according to the New York Times's Bilton and Roel.
"I'm on the $25,000 bus," Costolo replied. "Thanks Ev, this will be a lot of fun."
With its fresh infusion of cash, Twitter's growth continued to soar. Users were getting more creative with their tweets and crafting more of them. In August 2007, the hashtag was used for the first time. By March 2008, Twitter reached 1.3 million users. That May, it closed another round of financing, this time $15 million, from top investors like Union Square Ventures, Jeff Bezos, Spark Capital, and others.
As Twitter's public profile grew, however, its foundation began to crumble.
Twitter under Jack Dorsey's leadership, sources say, was chaos.
"Jack is crazy-smart and a product visionary and all that," says one source. "But Jack wasn't ready to be CEO at that time."
Dorsey struggled with the transition from engineer to boss. He wasn't leading teams well, and employees weren't happy working for him. He wasn't hiring new people fast enough. The site constantly crashed, making the Twitter "fail whale," which appeared when the site went down, a more prominent mascot than Twitter's blue birds. In 2007, the year Dorsey became CEO, Twitter's total downtime was six days. The constant crashing was a source of stress.
"Jack's role was that of a founder and a heavy influencer at the board level," one source says. "But he is not operational...he's not broadly effective."
"Jack is very smart, but he is not a leader," says another. "He can inspire investors more than employees."
Dorsey's interest in and facility with the financial side of the business left much to be desired. Dorsey, according to Nick Bilton's book Hatching Twitter, made mathematical errors while keeping track of Twitter's expenses. He set up partnerships with texting companies but SMS fees were so high, Twitter was spending nearly 6-figures per month.
Dorsey had also not yet focused entirely on Twitter: He wanted to be a fashion-designer, and he'd often leave work to take night classes. He also left early for art and yoga classes. His extracurriculars eventually became such a distraction, Bilton reported, that Williams eventually sat down with Dorsey and said, "You can either be a dressmaker or the CEO of Twitter. But you can't be both."
Sometime during 2008, Williams got fed up. He decided that he should be - and would be - the CEO of Twitter.
Williams had more money than Dorsey and three or four times as much Twitter stock. As Twitter's majority shareholder, when Williams made a decision, it was final. The Twitter board agreed that Dorsey wasn't working out, and they backed the decision to remove Dorsey.
When Williams came after Dorsey's position in the fall of 2008, it was an ugly "violent exchange," a source says.
"Jack wanted to be CEO, but it wasn't really the right call," this person said. "It was very clear Ev wanted to be CEO and it was pretty clear to people in the company that Ev was a better choice."
Unlike Dorsey, Williams was a proven entrepreneur. "Ev had more credibility than Jack," says the source. "He wanted the job. He got the job. That's why their relationship is a little touchy...He made it happen."
Publicly, Williams lay Dorsey's firing at the board's feet. And, ultimately, the board may have done the dirty work. Two of its members, Bijan Sabet and Fred Wilson, took Dorsey out to breakfast and delivered the blow, Bilton reports. They told Dorsey he'd receive $200,000 severance and become a silent chairman; Williams would be Twitter's new CEO.
On October 16 2008, the news was made public. "While the board of directors and the company have nothing but praise for where Jack has taken us, we also agree that the best way forward is for Jack to step into the role Chairman, and for me to become CEO," Williams wrote then.
"Twitter held all my desires in the world," he said.
Dorsey learned a common startup lesson the hard way: Coming up with an idea doesn't mean you own the business.
"I let myself be in a weird position because it always felt like [Evan Williams's] company," Dorsey later said. "He funded it. He was the chairman. And I was this new guy who was a programmer, who had a good idea. I would not be strong in my convictions, basically, because he was the older, wiser one."
One year later, in 2009, Dorsey founded another company, Square, with a new group of investors. It was a startup no one would be able to rip away from him. So far, it has been a phenomenal success.
Dorsey's distracted management style left Twitter in a bind internally. But, externally, the startup's profile was skyrocketing.
Dorsey had been what the company needed to get off the ground. When he became Twitter's CEO, the company only had a few thousand users. When Williams took over a year and a half later, that number exceeded 1 million, and millions of tweets were being sent per day.
Still, Twitter's next three years, from mid-2008 until mid-2011, were "crazy intense," a source says.
"The website was going down all the time while usage was going up. The valuation kept going up, and pressure kept going up," this person said. "It was never calm. Ever."
By his own admission, Williams made a lot of mistakes as CEO. "I've screwed up in many, many, many ways in terms of managing people and product decisions and business," he told the New York Times' Claire Cain Miller in 2010. But, more importantly, Williams oversaw spectacular user growth and kept the company headed in the right direction.
Williams had trouble making decisions, which frustrated his board and fellow executives. When it came to hiring, he moved slowly and preferred to pull in family and friends, who weren't always the best candidates.
Plus, the same growth problems that had plagued Dorsey also plagued Williams. While Twitter's site crashed less frequently under Williams, it was still an uncomfortably common occurrence.
Williams could have made all his Twitter problems go away quickly by selling the company. And he could have made a fortune doing it.
In the fall of 2008, Facebook came after Twitter. Facebook offered to buy Twitter for $500 million in a cash and stock deal. Twitter's previous round of financing had valued it at $98 million, so the offer was tempting. Williams alone would have made hundreds of millions of dollars.
Facebook and Twitter spent a few weeks discussing the deal. Facebook, although it initiated the talks, didn't love everything about Twitter. Twitter wasn't generating much revenue, and its SMS fees alone could have cost Facebook up to $75 million annually, AllThingsD's Kara Swisher reported then.
Twitter's team asked for a final night to sleep on the offer. When Twitter's board woke up, there was a page-long email from Williams in each member's inbox.
In the letter, Williams asked his board when it's the "right" time to sell a company. He then expressed why he didn't want to sell Twitter. Williams didn't need to convince his board. The email was more of a courtesy since Williams still owned a controlling stake in the company.
"He would have made hundreds of millions," says a source. "He could have said, 'This is Zuckerberg's problem now.' But he was a huge believer in the company."
Williams' decision not to sell Twitter to Facebook for $500 million was one of many smart decisions he made as Twitter's CEO.
Under Williams, Twitter began making strategic acquisitions. A Twitter search company called Summize was Twitter's first and arguably most important acquisition, because it helped solve Twitter's "fail whale" problem. The Summize deal was finalized under Dorsey's CEO term, but Williams led the project.
TwitterJack Dorsey signing the paperwork for the Summize acquisition, which was completed while he was CEO of Twitter"[Summize] was really important for Twitter for a whole bunch of reasons," a source says. "Twitter needed the technical capability to make the site work, and the Summize guys were instrumental. Their search engine was great, but the team [practically] saved Twitter."
Twitter's hyper-growth continued under Williams, in part because it began to be adopted by celebrities. Oprah Winfrey signed up for Twitter and received more than 100,000 followers within 24 hours. Ashton Kutcher signed up for the service in 2009 and encouraged other Hollywood stars to follow. Also, Twitter broke its first true news story.
On January 16, 2009, a US Airways jet made an emergency landing on the Hudson River. No one was killed, but all the passengers had to climb out of the plane and wait on its wings to be rescued. A Twitter user snapped a photo of the passengers and it spread quickly online. Soon, mainstream media picked up the Hudson plane crash story. Every outlet credited Twitter with breaking the news.
Janis Krums via TwitPicA passenger on a ferry going to rescue United Airlines passengers on the Hudson posted this picture to Twitter and it went viral.That was the moment that the ousted Jack Dorsey says he realized Twitter's massive potential. Twitter was more than the "what I'm doing" status updates that Dorsey had initially envisioned. It was the fastest way for important news to spread.
Williams also persuaded Twitter's executive team that the company needed to create its own mobile apps. Until 2010, Twitter had relied on third-party developers to develop excellent tools for Twitter. But Williams believed staying on that course would create too much brand confusion. And, eventually, Twitter itself would become irrelevant and die.
On April 10, 2010, at Twitter's first-ever "Chirp" conference, Williams announced the acquisition of the first of many application companies. Twitter bought the parent company of an app called "Tweetie," AteBits. At the time, Tweetie was the number one Twitter app in Apple's App Store.
Williams also radically expanded the size of Twitter's staff. He grew the team from 20 people under Dorsey to 300 by late 2010. He made subtle product innovations, too, like perfecting the retweet process, introducing a Twitter url shortener and fixing how @ reply messages worked.
Kevork Djansezian/Getty ImagesDick Costolo (right) was a friend of Evan Williams and was initially brought on as an independent board member.And he persuaded a small Twitter investor, his friend Dick Costolo, to join Twitter full-time.
By the fall of 2009, Dick Costolo was already a Twitter board member. Earlier when Twitter had been hunting for an independant board member, Costolo came to mind. He was a Twitter investor and a friend of Williams. He was also a former CEO the board already knew well.
Costolo took the board seat and began working closely with Williams. Everyone liked what he brought to the table and wanted more of Costolo, not less. Williams decided to change the nature of his meetings with Costolo. At a party in 2009, Williams offered Costolo the Chief Operating Officer position. Costolo accepted.
When Costolo joined Twitter full-time, he was one of the only business people on staff. There were only about 30 people at the company then.
Costolo set up the structure for Twitter's finance and sales divisions. He ran business development, began driving international strategy, and put an early ad strategy together.
By mid-2010, Twitter's traffic had grown to 160 million users. Instrumental executives like Ali Rowghani and Adam Bain had joined, Rowghani as CFO and Bain as the creatively titled "President of Revenue."
As the company grew, Williams seemed less and less interested in running Twitter. Gone were Twitter's startup days. In their place were piles of paperwork, HR headaches and performance reviews. The board and other Twitter executives noticed Williams' lack of enthusiasm. It also noticed the great job Costolo was doing as Twitter's COO.
It isn't clear if Williams agreed to resign, or if the board forced him out. The truth is probably somewhere in the middle. Regardless, according to Nick Bilton of The New York Times, Twitter advisor and honorary board member Bill Campbell sat down with Williams in 2010 and told him the board wanted to replace him as CEO with Dick Costolo. Costolo may not have been comfortable replacing the man who had brought him to Twitter. But eventually, with Williams' blessing, he took the job as interim CEO. Williams nudged the board to make the position permanent.
On October 4, 2010 Costolo was officially named Twitter's third CEO.
Everyone hoped the third CEO would be the charm.
People who know Costolo describe him as a "funny," "directed," "self-possessed guy" who "manages to be likable without alienating people."
He had to be someone with a positive outlook to wade through the muck that was Twitter.
"Dick certainly describes [his early days as Twitter's CEO] as a mess," a friend of Costolo's says.
The mess lasted well into 2011.
Costolo's first big mission was to clean up Twitter's board. It had somehow become full of uninvited bystanders, and too many Twitter stories were getting leaked to the press.
Twitter's actual board members then consisted of Spark Capital's Bijan Sabet, Union Square Ventures' Fred Wilson, Benchmark Capital's Peter Fenton, Williams, Dorsey, and Costolo. But these were not the only people who attended the board meetings.
"There were lots of observers and people dialing in," one source said. Kleiner Perkins' partner John Doerr, for example, floated around. Representatives from IVP and other institutions popped by, too. Executives at Twitter who had never been appointed to the board felt entitled to pull up chairs.
"There was a moment when the people in the board room became so many it was silly," one person said. "It was almost like crashing the board meeting with these people coming to the party. You had the IT guy coming to the board meetings. You had John Doerr who wasn't even an observer coming, I guess because billionaires can do whatever they want. Bill Campbell, who was an advisor, [was also] coming to the board meetings."
The messy board may have personally affected Costolo too. A report from Bilton's book Hatching Twitter suggests that in September 2010, when Costolo was still COO, honorary board member Bill Campbell nearly fired him in a fit of frustration. A source who has spoken with both Campbell and Costolo, however, says that never happened.
One by one, Costolo tracked the unofficial board members down and asked them not to come back. No one is quite sure how Costolo did this. All they know is that the offenders never showed up again.
As Costolo was cleaning up the board, two of Twitter's long-standing board members volunteered to step down. Spark Capital and Union Square Ventures each owned large portions of Twitter. These positions made their portfolios lopsided, and they were eager to sell some shares. But Fred Wilson and Bijan Sabet, the two board members, worried that selling shares while remaining board members would send mixed messages.
Costolo agreed and these two investors dropped off the board. Each sold a minority portion of their Twitter stock to Rizvi Traverse, a firm that quietly amassed the largest outside stake in Twitter - more than 15%. Evan Williams reportedly sold shares to Rizvi Traverse, too. Union Square Ventures and Spark Capital are the next largest outside shareholders. Each firm has more than the 6.7% stake owned by Benchmark Capital.
After Wilson and Sabet dropped off the board, seven true members remained: Costolo, Jack Dorsey, Evan Williams, Benchmark's Peter Fenton, Flipboard CEO Mike McCue, former DoubleClick executive David Rosenblatt, and Peter Currie of Currie Capital. McCue later left Twitter's board later when Flipboard became a conflict of interest.
With the board cleaned up, Costolo could begin running a tighter ship. But internally, the company was still in shambles. One person who was at Twitter during the Williams-Costolo transitional years describes that era as "a shitshow." "There were so many politics," this person said. "Like, a funny amount. I guess it was the culture."
Many executives were fired that year. Others quit. Twitter's VP of Engineering Mike Abott left in October 2011. VP of Product Satya Patel left in June 2012.
At Google and Facebook, some antics occur at the senior level. At Twitter, bickering trickled down to middle managers who tried to claw their ways ahead.
"If you look at Google, the same people have been running it [for 15 years]," the source said. "Once you see the top [people fighting] then people next to you start getting fired, you try to get leadership and [secure] your position. You do stuff you normally wouldn't [do]."
There would be a big road map meeting one day, this person says, then the people in that meeting would never be seen again.
There also wasn't a lot of trust among colleagues at Twitter. "A lot of people at Twitter told you what you wanted to hear," this person said. "Then thirty minutes later they'd be in a different meeting saying something different to someone else... [Some of the managers] tried to be okay with anything, but that's not a leader. The best leaders tell the same stories to everyone because [if you don't], you turn people against each other."
Some of Twitter's turnover and confusing culture problems were caused by growing pains.
"Startups are all messy inside," one person says of Twitter's tough years. "During those years, Twitter grew to hundreds of millions in revenue...I don't know how you go through that kind of growth in the public eye without it being crazy. It's like going through your teenage years and not being mad at your parents every once in a while. That's not dysfunctional, that's normal."
Slowly but surely, Costolo and a few of the executives he hired sucked the poison out of the company.
Even more impressively, they found a way for Twitter to start generating real revenue.
In April 2010, with the help of CFO Ali Rowghani, Twitter launched its first paid product: Promoted tweets. Advertising Age was the first sponsor.
Promoted Tweets show a sponsor's message to targeted users outside the sponsor's follower lists. Promoted Accounts, a second product, are used by brands or individuals who want to obtain more followers. A third product, Promoted Trends, stick a sponsored hashtag among other trending topics on Twitter. Promoted Tweets are now Twitter's most popular product.
Rowghani had worked with Steve Jobs at Pixar, and some of that management style rubbed of on him. He is widely respected at Twitter.
"Ali was my favorite," says a former employee who asked not to be named. "He was no bullshit. No politics. He would tell you what he thinks. He's very, very smart in the way he understands people and technology. And he knew how to grow the company from 200 to 2,000 people. He had good instincts. "
Rowghani is said to be the perfect complement to CEO Dick Costolo.
While Costolo, a quick decision-maker, can be "frenetic," Rowghani is "more deliberative and reflective."
They are a good team, one source says, because Costolo "needs to be out in the world glad-handing, and he feels he can do it because he trusts Rowghani to manage home base."
On December 19, 2012, Costolo made Rowghani Twitter's COO. Since then, Rowghani has developed Twitter's relationship with Apple (Apple now builds Twitter directly into its iPad and iPhone operating systems), and pushed Twitter into TV.
Like Rowghani, Adam Bain, the leader of Twitter's sales force, has excelled.
"The two key intellectual thought partners in terms of strategy [at Twitter] are Ali and Adam," says a source. "Adam runs the sales part. Adam is an effective leader of that group and they've done really well."
Bain's team performed so well last year that he maxed out his commission package and was listed the third highest-paid employee at Twitter.
Under Rowghani, Bain, and Costolo, Twitter's revenue has grown like this:
- 2009: $0
- 2010: $28 million
- 2011: $106 million
- 2012: $317 million
- 2013: $600+ million (estimated)
Two other people have helped Costolo excel as Twitter's CEO: Frenemy founders Jack Dorsey and Evan Williams.
While his tenure as Twitter's CEO wasn't perfect, Williams' work set Costolo up for success.
"Dick did a masterful job [as CEO] 50% because he's awesome and 50% because Ev did a fantastic job picking up all the pieces from Jack," says one source. "I think people have forgotten how critical Ev was. When he took over the company [from Dorsey] it wasn't a company. It was a product that was crashing. He did all these things that were critical to the company."
Dorsey, who briefly returned to an active role in Twitter in 2011, has found a way to help Costolo, too. He has become invaluable when it comes to spotting must-have technology and talent for Twitter.
For example, in April 2012, Dorsey led a hard push for Twitter to acquire Instagram.
Kevin Systrom, Instagram's co-founder, had worked with Dorsey and Williams as their intern at Odeo. Dorsey always supported Systrom. He personally invested in Instagram.
In 2012, at an Allen & Co. conference in Arizona, Dorsey and Ali Rowghani spent a lot of time schmoozing with Systrom. The three of them sat around a fire at the Ritz Carlton resort at Dove Mountain, where Rowghani and Dorsey made Systrom an acquisition offer, Kara Swisher reported for Vanity Fair. Twitter, they told Systrom, would be willing to buy Instagram for about $500 million in an all-stock deal.
This was a massive offer, given that Instagram was only a year and a half old. But Systrom had other options. Facebook CEO Mark Zuckerberg was also courting him. And Systrom wanted to run Instagram independently. Instagram soon accepted a $50 million investment from Sequoia Capital.
Facebook's Mark Zuckerberg, unlike Twitter, didn't take no for an answer. He spent a weekend convincing Systrom to reconsider - and he succeeded. Ultimately, Facebook acquired Instagram in a $736.5 million deal, $300 million of which was delivered in cash.
The missed opportunity was a blow to Twitter and Dorsey. Dorsey's instincts about Instagram had been right. The failure to win the deal motivated him to try harder on the next one.
A few months after Facebook bought Instagram, in the summer of 2012, Dorsey set his sights on a new startup, Vine.
Vine was a video app that hadn't launched yet. It was built by three New Yorkers, Rus Yusupov, Dom Hofmann and Colin Kroll. People who saw the early prototype fell in love with it.
"Everyone on the app knew this thing was special," a Vine investor says. "It was such an awesome product. It wasn't even fully built and yet it was still special."
Investors could hardly contain their excitement. One, in fact, did not contain it.
SV Angel's David Lee, who is an investor in Dorsey's other company, Square, showed Dorsey the video app over dinner in California.
Dorsey instantly felt Twitter needed to own Vine.
After that first sighting, things moved quickly. Dorsey flew to New York and told Vine's founders they had to join Twitter.
When Vine's founders met Dorsey, they felt he was smart but intimidating.
Dorsey calmed their nerves. He told them Twitter wouldn't re-brand or shut down Vine. Instead, Vine would be part of Twitter's long-term vision. Dorsey also made the founders an offer few first-time founders could refuse.
Twitter offered to acquire Vine in a deal valued at tens of millions of dollars. It was a great deal for Vine's founders, who would all be joining the company and getting a lot of stock. It wasn't a great offer for Vine's investors, who wanted to see the app through to launch.
One source says Twitter offered Vine a cash and stock deal "magnitudes" higher than $10 million. Another confirms the buyout offer was higher than $10 million but less than $100 million - again, not bad for first-time founders and a startup that had yet to launch.
Some Vine investors tried to fight the acquisition. There was talk of making a counter-offer in a new round of funding that would have let Vine's founders personally pocket millions. But this time, Dorsey won. Vine's founders took Twitter's offer. Hofmann is now Twitter's General Manager of Vine. Kroll is the app's CTO. Yusupov is Vine's Creative Director.
Today, Vine has more than 40 million users and Twitter has added to the Vine team in New York.
"The brilliance was that Jack immediately recognized how important [Vine] could be and how good the team was," a source with knowledge of the deal says. "Jack met with [Vine], recognized immediately that Twitter needed to get the thing done, and Dick made it happen. They worked well together on that."
Over the past year and a half, the drama has died down at Twitter. Jack Dorsey and Evan Williams are even speaking to each other again. The pair broke bread over dinner at a Moroccan restaurant in San Francisco, Aziza, last May.
This year, Twitter has been aligned around one common goal: going public. But as people familiar with the company readily admit, Twitter still has a lot to prove.
Twitter will generate more than $500 million this year, but the company is not yet profitable. It lost $69 million on $254 million revenue during the first half of 2013.
Also, Twitter is struggling to attract more mainstream users. In the second quarter, its growth in the United States slowed to just 1 million new unique users. Even with its 250 million total active accounts, Twitter is one-quarter the size of Facebook.
Some Twitter shareholders worry what will happen once the company goes public.
"Twitter has tried a lot of things that have really failed miserably," one person says. "All they have [working] is the original [140-character] concept. Twitter hasn't innovated in the way you would expect a tech company to."
Twitter is a simple product, and for that reason it's also easy for competitors to duplicate, although the network effect that goes with Twitter's massive user base is hard to build from scratch. Instagram came out of nowhere and threatened Facebook. The same could happen to Twitter.
"One thing no one talks about is it can all go away," a Twitter shareholder who asked not to be named says. "All it takes is one company doing something a little bit better. Twitter doesn't do anything that incredible that a new startup can not do better, and if that happens, then its $15 billion valuation becomes worth much less very quickly."
But if anything, Twitter's rocky road to success proves the company is relentless.
Paul Graham, the founder of high profile startup accelerator Y Combinator, often says startups should be more like cockroaches. Ugly and tough. In many ways, that's Twitter.
"Apparently the most likely animals to be left alive after a nuclear war are cockroaches, because they're so hard to kill," Graham wrote in April 2007. "That's what you want to be as a startup, initially. Instead of a beautiful but fragile flower that needs to have its stem in a plastic tube to support itself, better to be small, ugly, and indestructible."
Some shareholders think Twitter's best days are still before it. Lately, instead of founder feuds, Twitters has become known as the ultimate source for breaking news. Twitter is where the world first learned of Whitney Houston's death, for example, as well as Osama Bin Laden's death and the Boston Marathon bombings.
Saudi billionaire Price Alwaleed and his investment arm, Kingdom Holding, invested $300 million in Twitter in 2011. Alwaleed doesn't plan to sell a single share when Twitter goes public.
"We believe that it is just beginning to touch the surface," Alwaleed recently told Reuters. "We will be selling zero, nothing, at the IPO."
Twitter's tumultuous story and ultimate success has been a true collaboration for which no one person can take credit. Dorsey's passion for real-time updates ignited Twitter. Glass ran with the idea and pushed Williams to pursue it. Williams funded the company when it needed a cash infusion, saw Twitter's importance went far beyond a vanity tool, and accelerated growth. Costolo cleaned up the company internally and monetized the already-blockbuster product.
Most impressively, Dorsey and Williams were each forced out as CEOs. But despite wounded pride and bitterness, they still returned to make Twitter better.
Bijan Sabet of Spark, who's been a part of Twitter's bumpy ride since 2008, is understandably thrilled with the outcome.
"For me personally," he tells NYT, "this is a once-in-a-decade or once-in-a-career kind of investment."