'I'm leaving the company in two weeks," Dick Costolo said abruptly, his
face stricken, his fingers banging the wood-slab table before him.
Costolo, the bald and lithe chief executive of Twitter, was sitting in
the Waterthrush conference room on the 11th floor of his company's
headquarters, in the old Western Furniture Exchange and Merchandise
Mart, in downtown San Francisco. As the words left his mouth, Costolo's
nine top lieutenants, his so-called Operating Committee, looked up from
their iPads and smartphones, stupefied. Off in the distance, the
F-Market tram could be heard screeching to a halt; Ubers were pulling
over to deposit product managers at Twitter's Art Deco lobby. In the
cavernous staff commissary, there was the familiar din of conversation
as programmers poured Bolivian coffee into black mugs with bluebirds
draped around the stem. But in Waterthrush—which, like every
conference room at Twitter, is named after a bird—there was only
silence.

Then Costolo proffered the second part of his bombshell: "Jack is
coming in as the interim C.E.O."

The Jack in question, as everyone in the room knew, was Jack Dorsey, the
co-founder and former C.E.O. of Twitter, who had been fired seven years
earlier for prioritizing the pleasures of running a start-up over many
of the rigors involved. In those days, Dorsey spent a considerable
amount of time going to hot-yoga classes and taking sewing lessons, for
instance, when he could have been fixing the nascent social-media
company's server outages. This behavior rankled his colleagues and
scared his investors, and led to Dorsey's becoming another example of a
founder who was expelled from the company he had helped start. In the
intervening years, however, Dorsey mounted a remarkable comeback. Most
notably, he founded a mobile-payments company, Square, which was
estimated to be worth around $5 billion, and sat, coincidentally, a
mere block away from Twitter on Market Street.

For any normal public company, the sudden exodus of a C.E.O. in favor of
a former chief would be a logic-defying or even dystopian turn of
events. But Twitter is not a normal public company. From the moment it
was born, 10 years ago, it has existed in a near-constant state of
chaos. While the average tenure of a Fortune 500 company C.E.O. is
around a decade, Twitter has had five leaders in that same period. Its
four co-founders have each pushed one another out. If you discounted
Costolo's five-year reign, Twitter averaged one new boss per year. In
fact, this would technically be Dorsey's third time at the helm.

Introducing the Hive, VF.com's New Tech, Business, and Politics Site

Costolo's abrupt exit wasn't supposed to be this way. He had been
brought in as an adult C.E.O. to tame Twitter's madness. And, by nearly
every metric, he had achieved remarkable success. Under his tenure, the
company had grown from 300 employees to around 4,100. It had increased
revenues from literally zero to around $2 billion a year. Twitter had
also completely rebuilt its Web site to eliminate fatal server crashes.
But Costolo did have one problem that he could not solve. Twitter, which
was once among the hottest companies in Silicon Valley, had stopped
being cool. It had lost out on news to Facebook; on millennials to
Snapchat; on China and India to WhatsApp. Instagram was eviscerating it
on images. In a remarkably short time, it had slid from the
second-largest social-media company to the ninth. And while Twitter
still had 300 million monthly active users—or M.A.U., as they are
known in Silicon Valley—it had stopped growing. And there is nothing
scarier in Silicon Valley than a company that has stopped growing. As a
result, Twitter's stock had been on a jagged decline for 18 months.

Costolo, meanwhile, was looking for his exit. In December 2014, he had
proposed to a Twitter board member and its general counsel that he
resign in about a year. The plan allowed time for Costolo to groom a
successor, or for the board to find a replacement. Either way, unlike
his predecessors' departures, it would afford him the opportunity to
exit gracefully. But at Twitter, nothing is ever that simple. And soon
after that private conversation, the company's stock continued its
precipitous decline, and calls for Costolo's job mounted across the tech
and financial press. Then, in early June, Chris Sacca, a voluble
investor in Twitter, published an 8,500-word stream-of-consciousness
epistle, agitating for a change at the company. Sacca, perhaps livid
that he might lose his membership in the so-called Three Comma Club (a
pejorative term for tech billionaires), followed up with a series of
interviews and tweets lambasting Costolo. Costolo decided that he had
had enough agony. He was out.

Still, this being Twitter, the chaos was really only beginning. At
first, the board asked Costolo to stay on for a few months to ease the
transition. But Costolo refused, telling it that he didn't want to be
perceived as a "lame-duck C.E.O.," who would spend that time as a
punching bag for the media. Faced with a time crunch, the Twitter board
considered as his replacement co-founders Evan Williams and Jack Dorsey,
both of whom had previously run Twitter (before pushing each other out),
and both of whom, it should be noted, still sat on the Twitter board.
The board sided with Dorsey, given his success at Square, and his
allegiance to certain board members.

Yet when the board offered Dorsey the job, asking him to leave Square
for Twitter, he flat out rejected it. "I'll do whatever you need me to
do to help the company," Dorsey said, "but I won't leave Square." So
after numerous conversations, the Twitter board was left with no choice
other than the unprecedented measure of naming Dorsey the interim C.E.O.
of Twitter while he simultaneously ran Square.

Given that the S.E.C. needed to be notified, Costolo and Dorsey had to
move quickly to tell their employees and managers. On that windy June
morning, Costolo walked into his office, sat on his couch, and hastily
sent an emergency e-mail from his iPhone asking his Operating Committee
to convene immediately in Waterthrush. Down the block, Dorsey was
concurrently informing key employees at Square that he was now returning
to Twitter, a move that instilled a fear in some of them that he might
abandon their company in lieu of his other baby.

Soon after, Twitter employees were told there would be an all-hands
meeting in the cafeteria. And while many of them expected surprising
news, none were quite prepared for the specter of Costolo, Williams, and
Peter Fenton, a board member who resembled a miniature G.I. Joe,
standing before them anointing the return of Dorsey, who had grown a
beard so long that he looked like a cast member of Duck Dynasty.

As Twitter employees listened to the reassurances of their incoming and
outgoing C.E.O.'s, many were in shock. Others were paralytic from the
brusqueness of the regime change. Some even cried. To many in the
audience that day, it was untenable that Dorsey could return to run the
company for essentially the third time. But things were about to grow
even stranger.

Defense Against the Dark Arts

Twitter, which was founded in mid-2006, has always been engulfed in
madness. Its first (and mostly forgotten) leader, Noah Glass, was fired
a few months into the company's life as he sat on a green bench in the
South Park area of San Francisco. When Dorsey took over as C.E.O., he
lasted for a year and a half before he was fired while sitting in front
of an uneaten bowl of yogurt and granola at the Clift hotel, on Geary
Street. Evan Williams lasted 23 months before he was pushed out in a
vicious boardroom coup as he sat, helpless, at a mahogany table at the
company's law offices.

If these expulsions sound like murders, that's likely because many of
them were committed with the same behind-the-scenes planning and
mastery. In every instance, the man who was knocked off had no idea who
was behind the coup that led to his demise. In 2013, after the
publication of my book Hatching Twitter: A True Story of Money, Power,
Friendship, and Betrayal, I was greeted with effusive (or enraged) phone
calls, text messages, and e-mails from co-founders, board members, and
senior employees who were excited to finally learn the true identities
of their tormentors. I still occasionally receive calls from people
inside the company asking me for information about a recent firing.

There are many reasons for Twitter's internal conflicts. Almost
immediately after the concept was formed, in a tiny rodent-filled office
in San Francisco, it was apparent that the strange Web site was going to
grow up to do great things—allow people all across the Internet to
have a voice, to speak up against government action in oppressed
nations, and to engage in a live conversation from any point on the
globe. As a result, everyone in the room wanted their name attached to
Twitter, and each person involved wanted to steer the social network in
a unique direction.

Those desires, coupled with adolescent leadership, led to a company in
constant ferment. For a long time, however, that chaos seemed
inconsequential. It didn't really matter what direction Twitter went
in—was it a media company? a social network? a messaging
platform?—as long as it continued to grow. But a couple of years into
Costolo's rule, Twitter—like virtually every technology
company—found itself at the whims of today's fickle customers. All of
a sudden, at that point, Twitter needed to settle on an identity. And
that's when its schizophrenic nature entered overdrive.

While Costolo had enjoyed the longest tenure of any Twitter C.E.O., he
spent the better part of his time fending off other employees who wanted
to sit on the Iron Throne. During one patch, amid Square's early days,
Dorsey had tried to take control of the site's direction. More recently,
it was Costolo's C.F.O. turned C.O.O., Ali Rowghani, a former staunch
ally, who began guilefully politicking for the expulsion of his boss.
And this was just the top-level drama. Every single product chief at
Twitter—seven or eight people, depending on how you count—has been
fired or forced to resign over the past decade. One former staffer told
me the position is akin to the jinxed Defense Against the Dark Arts
professorship in the Harry Potter saga, where every professor ends up
dead or ousted at the end of the school year. A board member once said
he could use only one word—"Shakespearean"—to describe the
company.

Dorsey's return to Twitter, last July, merely exacerbated the conflicts.
On Costolo's last day, as he walked out of his office, which is named
Kingfisher (a small, brightly colored bird with a long beak), he left
behind a plush L-shaped couch, a sleek desk, a coffee table, and
numerous Twitter accoutrements that made the space feel cozy. Almost as
soon as Costolo had crossed the Golden Gate Bridge, however, Dorsey
asked for a moving crew to come in and completely empty the office and
install a new, large wood-slab standing conference table in the middle
of the room. Dorsey then renamed the Operating Committee by the more
simplified title of "Staff." Subsequently, Dorsey began holding court
with the Staff in Kingfisher.

One of the first meetings Dorsey organized regarded what he, as interim
C.E.O., was going to say to investors at Twitter's upcoming
quarterly-earnings call, which was just a few weeks away. This would
require some delicate choreography. Dorsey couldn't exactly criticize
everything Costolo had done. Since 2010, after all, Dorsey, as a board
member, had technically overseen Costolo's performance.

This conundrum led to a tempestuous discussion among members of the
Staff. "We have zero credibility with Wall Street right now," Gabriel
Stricker, the director of communications, said in a meeting with Dorsey
and top managers. "We have to come clean" about the company's stagnant
growth numbers.

Anthony Noto, the chief financial officer, agreed, but he had another
solution. He wanted to blame the current state of the company on
marketing and messaging, essentially throwing Stricker under the bus.
When Stricker threatened to quit over the verdict, he was fired. Then
the company very publicly made it apparent—or "messaged," in the
lingua franca of the Valley—that Stricker had been pushed out, fearing
that a high-level executive quitting two weeks into Dorsey's return
would be—God forbid—bad P.R. for Twitter.

Behind the scenes, the machinations deepened. Co-founder Evan Williams,
who remained a board member, was trying to persuade the board to buy his
company, Medium, an online publishing platform, for $500 million and
integrate the platform, and possibly himself, into Twitter. (The deal
ultimately didn't happen, for a variety of reasons, the price tag among
them.)

Around this time, Dorsey began holding three-hour-long meetings with
managers in Kingfisher. Notes from the meetings were disseminated to
everyone at Twitter, in keeping with a leadership style that Dorsey had
employed at Square. This, incidentally, is how Twitter employees began
to learn the profundity of their company's growth problem. The situation
was further complicated when the board asked Sard Verbinnen, the outside
communications firm that Twitter had retained, to put out a statement
that the board was going to consider hiring only C.E.O. candidates "who
are in a position to make a full-time commitment to Twitter." It
appeared to be a direct swipe at Dorsey, who had repeatedly declared to
the board that he would sign on as permanent C.E.O. of Twitter only if
he could also remain at Square—and who, a few days earlier, had
believed he was positioning himself to do just that.

Meanwhile, Dorsey was trying to figure out how to stop the user decline.
He was also forced to face the reality that, in the years since he had
last led the service, Twitter had become a vicious, often ruthless
platform. Louis CK had recently thrown away his millions of followers,
saying Twitter "didn't make me feel good." When Stephen Fry
deactivated his account, he likened the site to someone taking a
"turd" in a reservoir. Megyn Kelly repeatedly said that she could no
longer look at Twitter because of the violent tweets from Donald Trump
supporters.

If Dorsey had a magic touch, this would be the time to use it. But in
the early months of Dorsey's turnaround, as summer 2015 faded into
autumn, Twitter's stock began sliding again, with shares falling to an
all-time low of $25, down about 30 percent from Costolo's last day.

"Churn Rate"

Twitter's board may have been enamored with Dorsey's success at Square,
but there was an additional motive in their pursuit of him. In order for
Twitter to grow again, it had to become cool again. And the only person
who could pull that off, it seemed, after an exhaustive C.E.O. search,
was the person who had helped create the magic in the first place. So on
Thursday, October 1, during a private conference call with the board,
Dorsey was told it was official: he was now the full-time C.E.O. of
Twitter (in addition to, vexingly, the C.E.O. of Square). Four days
later, the news was announced publicly.

As Dorsey officially took the reins, it became clear just how difficult
this turnaround was going to be. During his first tenure at Twitter,
when the company was made up of two dozen people, the founders had an
idea to host a weekly meeting called "Tea Time" on Friday afternoons,
during which people would drink tea, sit through a brief presentation,
and hang out. In the early days, the miscreant employees chose to forgo
tea in favor of vodka or beer.

For a while, under Costolo, part of the Tea Time ritual included a
show-and-tell to the employees about the current state of the business.
A projection on a screen would show an animated bird wing, and the words
"We Measure Things" would appear. One notable chart showed the number
of people who were logging in to Twitter each month. On the chart there
were two important lines: a solid line showed the actual number of
people on the platform, and a dotted line depicted the projected number
of new users in the future. That dotted line stretched up past 400
million active users and pointed toward an imaginary half-a-billion
number. But each week, as the slides went up in front of the employees
at Tea Time, the solid line remained almost flat, stagnating around 300
million users. The gap between reality and hope grew so extreme that
this section of Tea Time was quietly phased out.

Twitter has seemingly tried every option to re-start its growth. One
early solution, "Bananaquit"—named after a bird of the passerine
species—involved tracking people online in order to offer up a better
experience when they joined the site. There was a move to try to entice
new users in other countries, but the "churn rate"—a term given to
the number of people who join and then quit—was often too high. (In
places such as India, I have been told, it is especially high.)

Days after Dorsey returned as permanent C.E.O., in October, Twitter
announced what it hoped would be the antidote for its user decline:
"Moments," a new feature that uses humans to corral tweets around a
certain live topic, such as a sporting event or international
demonstration, to stoke engagement. Given that Twitter has always
specialized in what people are doing in real time, Moments was greeted
with great curiosity by the tech press. But while the product helped
prevent some new users from walking out the back door, it didn't do much
to introduce Twitter to a new audience.

I have been told by people close to the company that, in the face of
mounting pressure from Wall Street, Twitter occasionally resorted to
what most start-ups do when they need to goose the numbers: they kind of
faked it. This happens at virtually all social networks; the company
sends an e-mail to inactive users who haven't been on the service in a
few months, informing them there is a problem with their username or
account, which leads people to log in to fix the situation. Magically,
those people become monthly active users even if they were not.

And while Dorsey wasn't employing that trick, his magic was not yet
apparent to investors on Wall Street. Months into his turnaround
campaign, user growth was relatively flat and Twitter's stock was now
down nearly 60 percent from where it had stood when Costolo was
convening his staff in Waterthrush. Twitter, which once had a market
valuation of nearly $40 billion, was now worth about half that.

"Jack Dorsey's Reputation Is Crushed . . ."

This is the point where I become part of the Twitter story. For several
years, Dorsey and I were friends. We went to dinner together and
endlessly explored the sanctums of San Francisco and New York City with
several mutual acquaintances. But in 2012, when I told him of my plans
to write a book about the founding of Twitter, a very different side of
Dorsey emerged. He immediately tried to kill the project. He told
everyone at Twitter, and anyone associated with the company, not to talk
to me.

As I started reporting, I realized why. Dorsey, who was so charming in
person, had been a bully behind the scenes. Countless former employees
came out of the woodwork to recall his role in their ousting. Or, in a
fate that is even worse in Silicon Valley, how he had seemingly erased
their contribution from the company record.

The book left a lot of people wondering if Dorsey was really a talented
innovator, or just lucky. In my reporting, indeed, I learned that Dorsey
was working as a programmer for an Alcatraz-tour company when he
serendipitously spotted Evan Williams in a San Francisco coffee shop.
Williams, by then, had already sold a company to Google and was becoming
minor tech royalty. Dorsey, on the other hand, was applying for a job
selling shoes at Camper. Taking a chance, Dorsey e-mailed Williams the
résumé that he was going to use for the Camper job. (He erased any
reference to shoes before hitting Send.) That e-mail would lead to a
company that eventually became Twitter. The amount of drama that ensued
afterward, however, surprised me, and became the basis for my book. When
it was published, one headline noted: "Jack Dorsey's Reputation Is
Crushed in Nick Bilton's Book on the Early Days of Twitter."

I was pretty sure that Dorsey would never speak to me again. But in
early April, when I reached out to see if he would be open to meeting
for this article, I was surprised by his response. "Let's do it!" he
replied in an e-mail. We met at the offices of Square, down the block
from Twitter. In the same manner that everything at Twitter is named
after a bird, everything at Square is shaped like a square: the little
cubbyhole desks, the conference-room tables, the bricks on the exterior
of the building, are all quadrilaterals. We met in one of those square
cubbyholes on the sixth floor and took the rear stairs down to the
street level, where Dorsey said he wanted to go for tacos at a nearby
food truck.

It had been a strange week for Dorsey. Shares of Twitter's stock had
fallen another 16 percent in a single day after the most recent
quarterly earnings statement detailed slowing ad growth and only a
minuscule user increase. And yet, shares of Square had rallied, up 16
percent. Dorsey was, as one investor noted on Twitter, the C.E.O. of
both the best- and worst-performing tech stocks in the United States
that week.

"You should try the crispy beef tacos," he told me as he ordered from
a bright-red food truck that was parked on the sidewalk. "Sure, I'll
have two," I replied and then jumped into a question I had for him:
"Do you ever look at the Twitter stock chart on your phone and turn it
upside down and dream of a day?"

After brief laughter, he said he doesn't look at the stock chart, ever.
"I know there are people in the company that do, but I don't because I
can't control that."

"Never?"

"Nope," he said. "Never."

I then asked him the question that had been on my mind for months. Why
was he doing all this? He was already worth hundreds of millions of
dollars, and more than a billion on paper. He was only 39, with a whole
life ahead of him. Most people are more than content to run one public
company, but he wanted to oversee two—including a significant
turnaround effort.

Dorsey replied that his work wasn't done on the product he had helped
start. Indeed, he now spends a large portion of his day trying to pitch
people—investors, new recruits, current employees who might be on the
verge of quitting, the board—a narrative of what Twitter can still be.
"I want people to wake up every day and the first thing they check is
Twitter in order to see what's happening in the world," he said between
bites of his first crispy beef taco. "It's a metaphor for checking the
weather. Twitter has a similar potential."

If there's one thing that hasn't changed about Dorsey since I first met
him, a decade ago, it's his ability to think big. He isn't exaggerating
when he says that he wants people to check Twitter each morning as if
they were wondering whether they needed an umbrella. When I asked him
how he's going to get there, Dorsey said that he plans to double down on
what the company is best at, which is being the platform where people
weigh in on live events. "If you were to describe what Twitter is," he
said, now moving on to his second beef taco, "it's live news,
entertainment, sports, and chat."

I asked him if he worried that Mark Zuckerberg has lately been using
that same word—"live"—on his earnings calls, noting that this is
also Facebook's new focus.

"Yes," he said candidly. He did very much.

Dorsey then mentioned something more revealing. The stagnant user
growth, he conceded, was due in large part to the constant turmoil at
Twitter. "There has been an ever shifting leadership, platform, and
strategy, and it's hard to see any momentum in that," he said, digging
into his third beef taco. I agreed with him. One of the main issues with
Twitter, both inside and outside the company, has always been that
Twitter doesn't know what Twitter is. Ten years on, that existential
question persists: Is it, in fact, a media company? A social network? A
messaging platform? Maybe it's all of the above. But in order to
persuade people to go through the somewhat laborious process of signing
up for an account, learning the vernacular of the site, and sticking
with it, this needs to be articulated to the public, whose behavior is
the only thing that can convince Wall Street.

Plan B

There are few things about Twitter's future that anyone can say for
certain, but I'll offer one prediction with absolute assuredness: there
will not be a fourth Jack Dorsey era. Recently, when I met with
executives at the company—including the executive chairman of the
board, the chief financial officer, and the director of
communications—there was one query that seemed to catch everyone off
guard. What was Plan B, I asked, if Dorsey couldn't turn the company
around? "There is no Plan B," I was told. "This is it."

The solution to Twitter's problems, they all reiterated, along with
Dorsey, is that word "live." "We now know what inhibits usage, and
what doesn't," Dorsey explained to me. He said he has a slew of new
features—including hosting live video from the N.F.L., where people
can talk about the game as they watch it—that will grow the audience
and focus on that single, live strategy.

Twitter is betting a lot on this relatively simple notion. And if it
doesn't work, there actually is a logical Plan B, even if it's one that
few at Twitter want to ponder: a sale of the company. But I've spoken to
dozens of outsiders about this scenario, and it's unclear who the likely
suitor would be. The late Bill Campbell, a confidant of Larry Page's,
told me not long before his death that he had tried to press Google to
buy Twitter many times, but Page had "zero" interest in the social
network. There are people close to Facebook who have told me that, while
Zuckerberg could still be interested in buying the company, he wouldn't
want to get into a bidding war. Apple is an option, perhaps, but a lot
of people in the Valley believe it's got bigger challenges ahead, and a
social network isn't going to help sell millions of iPhones. Then there
are the not-so-sexy possibilities, such as Microsoft, Alibaba, or
Verizon.

But it's unlikely that Twitter will relent easily. While Dorsey and his
co-founder Williams don't always agree on everything, the two have been
resolute in their determination not to sell, and remain so. (When one
board member suggested a sale, around the time of Costolo's departure,
Dorsey and Williams refused.)

During our conversation, Dorsey tried to convince me that there was a
better future for Twitter. He noted that Apple, at its low, was worth
$271 million. Then Steve Jobs returned and set it on a path to a market
capitalization of $774 billion. Disney, he also noted, was in the sink
before Bob Iger retooled the company and led it to a valuation of more
than $200 billion.

Indeed, just as there are two different scenarios for Twitter, there are
two different versions of Dorsey. There's the eccentric artist who used
to walk around San Francisco wearing a T-shirt with his phone number on
it to see if anyone would call him; the same man who once pitched the
idea of a start-up that was a "massage parlor for programmers," where
one person would write code while the other would give him or her a
shiatsu backrub. And this is the same Dorsey who thought it would be
neat if people could share what they were doing at any given moment, no
matter how mundane—an idea that eventually became Twitter.

Then there is the guy who is able to manage thousands of employees, and
occasionally dabble in the dark arts of the boardroom, seemingly without
breaking a sweat. The first Dorsey was the one who ran Twitter in its
heyday, only to be pushed out. The second built Square. The question
now, it seems, is whether Dorsey can embody both versions of himself.

Toward the end of our taco dinner, dusk was beginning to settle over San
Francisco, and I said something to Dorsey that had been on my mind for a
long time. I explained that, while it's my job to tell these stories, it
didn't bring me pleasure to report some of the harsher details in my
book. I then asked if he had any regrets about all that chaos at Twitter
over the past decade. He paused for a moment. "There is really nothing
I regret," he said.

But when I pressed further, he wistfully talked about the group of
people, mostly friends, who helped hatch Twitter in that rodent-infested
basement. Some of them became billionaires, others ended up with
nothing, but most no longer speak to one another. "It was such a good
team. It just became so screwy, and confusing. I don't know what
happened. I don't regret it. I feel sad about it," he said, his voice
trailing off into the night.