X

The number arrived like a prank. On April 14, 2022, Elon Musk filed a Schedule 13D with the Securities and Exchange Commission disclosing an offer to purchase every outstanding share of Twitter, Inc. at $54.20 per share — a price that embedded a marijuana joke in a $44 billion corporate acquisition, the "420" a nod to the same cannabis reference that had once earned Musk a securities fraud settlement with the SEC over a Tesla tweet. The board of directors of a publicly traded company serving as the de facto global wire service for heads of state, journalists, and revolutionaries was now contemplating an all-cash tender offer whose per-share price was, at least in part, a meme.

This was fitting. Twitter had always existed in the space between the profound and the absurd, between its role as the communications backbone of the Arab Spring and its origins as a side project built atop SMS text messaging by a group of friends who couldn't agree on who deserved credit for inventing it. The company had cycled through four CEOs before its IPO, survived at least two near-death experiences, never figured out how to grow like Facebook or monetize like Google, and yet became — through some alchemy of brevity, real-time distribution, and the peculiar psychology of public utterance — the most consequential 140-character constraint in the history of media. Now the world's richest man wanted to buy it, with a joke baked into the offer price, and nobody could quite tell whether the bid was the act of a visionary, a troll, or something more troubling.

What followed — the board's poison pill, Musk's attempt to back out, the Delaware Chancery Court lawsuit that forced his hand, the October 27, 2022 closing in which Musk walked into Twitter's San Francisco headquarters carrying a porcelain sink ("Let that sink in!") — is by now the most chronicled acquisition in recent corporate history, dissected in at least three major books, a PBS FRONTLINE documentary, and more journalism than any single media transaction since the AOL-Time Warner merger. But the acquisition is not the story. The acquisition is the hinge. The story is what Twitter was, what it failed to become, and what Musk's ownership revealed about the fragility of a platform whose value was never really in its code or its revenue but in the collective willingness of several hundred million people to treat it as the place where reality was negotiated in real time.

Twitter's creation myth is both over-told and under-understood. The standard version goes something like this: Jack Dorsey came up with the idea, Biz Stone and Evan Williams built it, and nobody remembers Noah Glass. The truth is messier, more interesting, and more predictive of everything that followed.

In 2005, a small San Francisco startup called Odeo — a podcasting company that had the misfortune of launching weeks before Apple added podcasting to iTunes, thereby vaporizing its entire market — was floundering. Evan Williams, the co-founder who had previously sold Blogger to Google and understood the web as a medium for self-expression with an almost spiritual conviction, made an unusual managerial decision: he told his engineers to break into small teams and spend two weeks building whatever inspired them. Williams had been a farmer's son from a small Nebraska town who dropped out of college, drifted through startups in the early web, and stumbled into Blogger almost by accident, creating one of the foundational tools of online publishing before Google paid him for it. He understood, perhaps better than anyone in Silicon Valley circa 2005, that the interesting things happened in the margins.

Jack Dorsey — then a twenty-nine-year-old NYU dropout who had grown up in St. Louis obsessed with maps, dispatch routing, and the invisible choreography of cities in motion — had been nursing a long fascination with status updates. He wanted to build software that let people broadcast what they were doing, a kind of permanent ambient awareness, the urban equivalent of a radio dispatch. Dorsey and Biz Stone, a college dropout from the Boston suburbs who had worked his way from book jacket design to Google's Blogger team and who possessed an almost supernatural gift for making complicated things sound simple, paired up. Noah Glass, another Odeo employee who by several accounts was deeply involved in the earliest brainstorming and even named the service (he suggested "Twitter," derived from "a short burst of inconsequential information" — the chirps of birds), would later be almost entirely written out of the founding narrative. Nick Bilton's account in Hatching Twitter paints this erasure as one of Silicon Valley's original sins: a co-founder who contributed the idea's emotional core quietly excised as the company's myth calcified.

The prototype was built on SMS. You texted a message to a short code — 40404 — and it went out to everyone following you. That constraint, the 160-character limit of a text message minus 20 characters for a username, produced the iconic 140-character limit. The first tweet, sent by Dorsey on March 21, 2006 — "just setting up my twttr" — was not a manifesto. It was a system test. The service launched internally at Odeo, gained traction at South by Southwest in March 2007, and from there the thing just grew and grew and grew, as Biz Stone later put it.

What made Twitter different from every other social network was not the technology — it was embarrassingly simple, technically — but the information architecture. Facebook asked "Who are you?" Instagram asked "What does your life look like?" Twitter asked "What is happening?" The asymmetric follow model (you could follow anyone; they didn't have to follow you back) created something unprecedented: a global, real-time, public conversation where a teenager in Cairo could reply to the President of the United States, and both messages occupied the same feed with the same formatting. The 140-character constraint, initially a technical artifact of SMS limitations, turned out to be a design masterpiece — it lowered the barrier to posting, rewarded wit and compression, and made every tweet feel like a dispatch from the front lines of someone's consciousness.

The company that built this extraordinary communications tool was, almost from inception, a catastrophe of governance. Twitter cycled through leadership the way a failed state cycles through juntas: Jack Dorsey was CEO, then forced out in favor of Evan Williams in 2008; Williams was CEO, then pushed aside for Dick Costolo in 2010; Costolo ran the company through its IPO and its period of maximum cultural relevance, then was eased out in 2015, at which point Dorsey returned as CEO — while simultaneously running Square (now Block), a $40 billion payments company, splitting his time between two of Silicon Valley's most complex organizations.

Each transition carried the same pattern: the sitting CEO proved unable to solve the growth problem, the board lost patience, and the musical chairs resumed. Dorsey's first tenure ended because the board believed he was more interested in fashion and yoga than in running a company. Williams's ended because, despite being the strategic visionary who understood Twitter's potential as a media platform, he couldn't execute fast enough. Costolo — a former improv comedian from Detroit who had been CEO of FeedBurner before Google acquired it, and who brought genuine operational discipline to a company that had almost none — managed to take Twitter public in November 2013 at a $31 billion valuation but couldn't solve the core existential problem: user growth was decelerating, and the product was too confusing for mainstream adoption.

Dick Costolo later reflected on the structural challenge with characteristic candor: the product that power users loved — a raw, chronological, real-time firehose of information — was precisely the product that confused and alienated new users. The timeline was noisy, the onboarding was terrible, the abuse problem was metastasizing, and the company seemed constitutionally incapable of making the hard product decisions that would have sacrificed the experience of its most passionate users to grow the base. Facebook, meanwhile, was crossing a billion users by algorithmically curating the feed — doing exactly what Twitter's culture refused to do.

This mattered enormously. By the time Twitter filed its S-1 in 2013, it had roughly 230 million monthly active users — impressive, but a fraction of Facebook's scale, and the growth curve was already bending. Revenue was $665 million in 2013, almost entirely advertising, which meant Twitter was selling attention it couldn't grow fast enough to meet Wall Street's expectations. The stock peaked near $70 in early 2014, then began a long, grinding decline that would take it below $15 by 2016. The market was telling a story: Twitter was one of the most culturally important products ever created, and one of the most disappointing businesses.

Twitter's business model was always a paradox. The platform's power derived from its openness — anyone could read tweets without logging in, journalists embedded tweets in articles, politicians used it as their primary communication channel, and the real-time firehose powered an entire ecosystem of financial data providers, news aggregators, and social listening tools. But openness is the enemy of monetization. Facebook could show you ads because it knew your age, your location, your relationship status, and which brands you'd liked. Twitter knew you were interested in... whatever appeared in your chronological timeline, which might be Syrian civil war footage followed by a joke about a cat followed by a CEO's earnings commentary.

The advertising product improved over the years. Twitter developed promoted tweets, promoted accounts, and promoted trends. It built increasingly sophisticated targeting based on interest graphs, keywords, and follower lookalikes. Revenue grew from $665 million in 2013 to $3.46 billion in 2019, and then to $5.08 billion in 2021 — the company's pre-acquisition peak. Gross margins were healthy. The platform finally turned its first GAAP net profit in 2018 and remained profitable (narrowly) through 2019.

But the comparison to peers was always painful. In 2021, while Twitter generated $5 billion in revenue, Facebook's parent Meta generated $118 billion. Google generated $258 billion. Even Snapchat, which had fewer daily users, was generating comparable revenue on a per-user basis. Twitter's average revenue per user hovered around $25 in the U.S. — respectable, but roughly half of Facebook's and a third of Google's. The company's total addressable market seemed, paradoxically, to be smaller than its cultural footprint suggested.

The underlying issue was engagement depth. Twitter's most engaged users — journalists, politicians, activists, finance professionals — were extraordinarily valuable per capita, but they represented a relatively thin slice of the global population. The vast majority of accounts were lurkers, reading but never posting. And the product's learning curve — understanding @mentions, hashtags, quote tweets, threads, the unwritten norms of each sub-community — acted as a natural ceiling on growth. Twitter reached approximately 238 million monetizable daily active users by Q1 2022, a number that had taken fifteen years to achieve and still represented less than a tenth of Facebook's user base.

If Twitter was a mediocre business by Silicon Valley's standards, it was something close to irreplaceable as infrastructure. By the late 2010s, Twitter had become the de facto nervous system of the global information ecosystem — not because it had the most users, but because it had the right users. Presidents and prime ministers announced policy on Twitter. Journalists broke stories on Twitter. Financial markets moved on tweets from executives, central bankers, and the account @DeItaone. When a plane landed on the Hudson River in 2009, the first image came from Twitter. When protests erupted in Egypt in 2011, the organizing happened on Twitter. When the pandemic hit in 2020, the epidemiological debates, the policy fights, the raw data — all of it played out in real time on Twitter's infrastructure.

This made Twitter the world's most peculiar media company: one whose content was produced entirely by its users, whose editorial function was performed by an algorithm and a trust-and-safety team that together constituted a kind of accidental newsroom, and whose cultural influence was wildly disproportionate to its financial returns. A single tweet from Donald Trump could move equity markets. A viral thread could reshape a political narrative in hours. The platform was, in every meaningful sense, a public utility that happened to be owned by private shareholders and funded by advertising — a combination that created irreconcilable tensions.

The content moderation problem was particularly acute. Twitter's commitment to being an open, real-time communication platform meant it was the preferred venue for harassment campaigns, coordinated disinformation, and the sort of pile-on dynamics that could destroy a person's life in an afternoon. The trust-and-safety teams grew steadily — by 2022, hundreds of employees were dedicated to content policy, enforcement, and the geopolitically sensitive work of identifying state-sponsored information operations from Russia, China, Iran, and elsewhere. But every moderation decision was a lose-lose proposition: remove a tweet and you were accused of censorship; leave it up and you were accused of enabling abuse. This was the structural trap that would define Twitter's fate.

Elon Musk had been one of Twitter's most prolific and influential users for years before he decided to buy it. His account, @elonmusk, was one of the most-followed on the platform, and he used it the way a jazz musician uses an instrument — riffing, provoking, joking, occasionally announcing things that moved billions of dollars in market capitalization. He tweeted about Tesla production numbers, SpaceX launches, cryptocurrency, and, increasingly, his grievances with what he perceived as Twitter's political bias and excessive content moderation.

Musk was a figure of almost absurd biographical compression: born in Pretoria, South Africa, brutalized by childhood bullies so severely his nose required corrective surgery, estranged from a father he has described as evil, emigrated through Canada to the United States, co-founded Zip2 (sold for $307 million), co-founded X.com (merged with PayPal, sold to eBay for $1.5 billion), founded SpaceX (now valued north of $350 billion), became CEO of Tesla (market cap peaking above $1 trillion), and somehow found time to run Neuralink, The Boring Company, and eventually xAI — all while tweeting constantly. Walter Isaacson's biography, reviewed by the Guardian as an "insight-free doorstop," nonetheless made one thing vivid: Musk's "demon moods" and the way his management style oscillated between visionary intensity and destructive chaos. The bullied child had internalized the bully's logic.

The decision to buy Twitter appears to have crystallized around Trump's suspension. When Twitter permanently banned Donald Trump on January 8, 2021 — two days after the Capitol insurrection — Musk saw it as a "turning point," as Zoë Schiffer reports in Extremely Hardcore: Inside Elon Musk's Twitter. For Musk, the ban was evidence of a platform captured by what he called the "woke mind virus," a company whose trust-and-safety apparatus had become a political actor rather than a neutral enforcer. Whether this diagnosis was correct, self-serving, or both, it provided the ideological framework for what came next.

Musk began quietly accumulating Twitter shares in January 2022. By March, he held 9.2% of the company — enough to become the largest shareholder. On April 4, Twitter disclosed Musk's stake. On April 14, he made his $54.20 offer. And then, in one of the most dramatic reversals in corporate history, he tried to walk it away.

The months between Musk's April offer and the October closing constitute a legal thriller that would strain credulity in fiction. Having signed a binding merger agreement with no financing condition and an unusually limited ability to terminate, Musk spent the summer trying to escape a deal he no longer wanted — or claimed he no longer wanted. He cited bot accounts. He demanded data. He accused Twitter's management of misrepresenting the platform's user metrics. On July 8, 2022, his lawyers sent a letter purporting to terminate the merger agreement.

Twitter sued in Delaware Chancery Court. Chancellor Kathaleen McCormick scheduled a trial for October. The legal consensus was overwhelming: Musk had signed a contract, the specific performance provisions were ironclad, and he was going to be forced to close. Rather than face a trial he was almost certain to lose, Musk reversed course in early October and agreed to complete the acquisition at the original price.

The merger proxy statement filed with the SEC laid out the terms: $54.20 per share, all cash, representing an approximately 38% premium to Twitter's closing price before Musk's stake was disclosed. The total deal value was approximately $44 billion. Musk funded it with roughly $13 billion in debt financing from a syndicate of banks (Morgan Stanley, Bank of America, and others), approximately $7.1 billion in equity from outside investors (including Larry Ellison, Sequoia Capital, Andreessen Horowitz, Binance, Qatar Investment Authority, and Saudi Prince Alwaleed bin Talal, who rolled over his existing Twitter stake), and the rest — more than $20 billion — from Musk's personal wealth, largely through the sale of Tesla shares.

The $13 billion in acquisition debt would prove to be a millstone. At the interest rates locked in during the deal, Twitter-now-X owed approximately $1.2 billion per year in interest payments alone — against a revenue base that was about to collapse.

On October 27, 2022, Musk closed the deal. He walked into Twitter's San Francisco headquarters carrying a bathroom sink. He posted a video. He fired CEO Parag Agrawal, CFO Ned Segal, and Chief Legal Officer Vijaya Gadde immediately — reportedly learning some were still in the building and having them escorted out. Within days, he brought in what amounted to an occupation force: Steve Davis, president of The Boring Company, who reportedly slept at Twitter's offices with his partner and newborn child while overseeing cost-cutting; his cousins James and Andrew Musk; Tesla and Neuralink engineers who had no experience in social media; and a group of loyalists tasked with conducting what amounted to a corporate audit at gunpoint.

The layoffs began almost immediately. On November 4, approximately half the company — roughly 3,700 people — received emails informing them their access had been revoked. The cuts were indiscriminate, hitting engineering, trust and safety, communications, legal, sales, and infrastructure teams. Former head of trust and safety Yoel Roth later described learning about major decisions by watching them unfold on the platform itself. Employees discovered their termination when their laptops locked. In some cases, teams were eliminated entirely, then partially rehired days later when Musk's team realized they had fired people responsible for keeping the site operational.

The scale of the workforce reduction was staggering: from approximately 7,500 employees pre-acquisition to roughly 1,500 within months — a 75% cut that went deeper than any comparable corporate restructuring in recent tech history. Entire functions — content moderation at scale, advertising sales operations, communications, policy research, the teams that managed relationships with government regulators in markets from the EU to India — were either gutted or eliminated.

NPR's Ryan Mac and Kate Conger documented in their book Character Limit what this looked like at the human level: employees in the New York office running out of toilet paper because janitorial services had been cut, weekend conference calls where Musk went through the company budget line by line demanding justifications, an atmosphere of fear and ideological loyalty tests that one former HR employee compared to a gestapo.

If the layoffs were the initial shock, the advertising collapse was the slow bleed. Twitter had always derived the vast majority of its revenue — approximately 90% — from advertising. And advertising, uniquely among revenue models, is a confidence game: brands pay to be associated with environments they trust, and trust is fragile.

The sequence of events in Musk's first months was an almost perfect algorithm for destroying advertiser confidence. He reinstated banned accounts, including Donald Trump's. He disbanded the trust-and-safety advisory council. He removed labels identifying government and state media accounts for Russia and China. He launched a paid verification system — Twitter Blue, then X Premium — that replaced the old verified badge (which signaled identity confirmation) with a purchasable checkmark (which signaled willingness to pay $8 per month), instantly degrading the information-integrity infrastructure that advertisers relied on. Hate speech surged. Research from the Institute for Strategic Dialogue found anti-Semitic tweets in English more than doubled after the acquisition. A European Commission report found engagement with pro-Kremlin accounts grew 36% in the first half of 2023.

Musk himself acknowledged the damage. In the summer of 2023, he stated publicly that advertising revenue had fallen approximately 50%. But rather than reconcile with advertisers, he escalated. At the New York Times DealBook Summit in November 2023, when asked about the advertising boycott, Musk delivered what may be the most extraordinary statement ever made by a CEO to his own customers:

He singled out Disney CEO Bob Iger by name. The moment was crystallizing. Major brands — Disney, Apple, IBM, Comcast, Warner Bros. Discovery — suspended their campaigns. The revenue decline accelerated. By January 2024, Fidelity, which had been instrumental in financing the acquisition, marked down the value of its X investment for the fourth time, valuing the company at roughly 71.5% below the $44 billion purchase price — implying a valuation of approximately $12.5 billion.

The U.K. financial filings, among the few public windows into X's financial performance since it went private, tell the story in granular detail. Revenue from X's U.K. arm — historically a reliable proxy for global trends, accounting for roughly 5.3% of total revenue — plummeted from $282.9 million in 2022 to $95.2 million in 2023 (a 66% decline) and then to $39.8 million in 2024 (a further 58% decline). If the U.K. trajectory holds as a global proxy, X's worldwide revenue may have fallen from roughly $5 billion pre-acquisition to somewhere in the range of $1.5–2.5 billion — a financial catastrophe made existential by the $1.2 billion annual debt service.

X's own U.K. filings acknowledged the cause with unusual candor: "The significant decrease in the performance of the company is a result of the decline of advertising revenue primarily driven by a reduction in spend from large brand advertisers due to concerns about brand safety, reputation and/or content moderation."

In July 2023, Musk renamed Twitter to X. The bird logo — one of the most recognized symbols in technology, a design so embedded in the culture that "tweet" had become a verb used by heads of state — was replaced by a stark, monochromatic X. The rebrand was executed with the same velocity and lack of deliberation that characterized every other post-acquisition decision. Larry the Bird, gone overnight.

Linda Yaccarino, the former head of global advertising at NBCUniversal whom Musk had named CEO in June 2023, framed the rebrand as liberation. "The rebrand represented really a liberation from Twitter," she told CNBC. "A liberation that allowed us to evolve past a legacy mindset and thinking." She described a vision of X as an "everything app" — integrating social media, payments, commerce, video, and messaging into a single super-platform, a Western WeChat.

The comparison to WeChat was revealing, and not in the way Yaccarino intended. WeChat's success as a super-app was inseparable from the specific conditions of the Chinese market: a massive population coming online for the first time through mobile, limited incumbent competition in digital payments, and a regulatory environment that favored domestic platforms. None of these conditions existed in the United States, where consumers already had deeply entrenched habits across dozens of specialized apps (Venmo for payments, YouTube for video, iMessage for chat, Instagram for photos) and where regulatory scrutiny of tech platforms was intensifying, not relaxing.

The rebrand destroyed what was arguably Twitter's most valuable non-financial asset: its linguistic moat. "Tweet," "retweet," "tweetstorm" — these words had entered the global vocabulary, appearing in dictionaries, news broadcasts, and everyday speech. They were free branding. No one says "I x'd about it" or "Let me re-x that." The killing of this linguistic infrastructure was, from a branding perspective, one of the most mystifying decisions in corporate history — equivalent, as one CNBC interviewer suggested, to Johnson & Johnson renaming Band-Aid.

Musk's ownership of Twitter created the competitive opening that Twitter's rivals had spent a decade failing to create on their own. For years, the social media landscape had a Twitter-shaped hole that nobody could fill — not App.net in 2012, not Mastodon in 2017, not Parler or Gab or Truth Social, all of which were either too niche, too technical, or too ideologically coded to attract mainstream adoption. Twitter's network effects were formidable: the journalists were there, so the politicians were there, so the newsmakers were there, so the journalists were there. Breaking the loop required an exogenous shock.

Musk provided it. Meta launched Threads on July 5, 2023 — a product that had been in development since January, built as an Instagram-adjacent text-based platform that leveraged Instagram's existing 2 billion+ user base for frictionless onboarding. Threads hit 100 million sign-ups in its first five days, the fastest-growing app in history. Bluesky, a decentralized social network that Jack Dorsey had originally incubated as a Twitter research project, gained traction among the journalists and power users who formed Twitter's cultural backbone. By late 2024, Bluesky had surpassed 20 million users — small by platform standards, but disproportionately populated by exactly the high-influence users whose presence had made Twitter valuable.

The fragmentation was real, even if no single competitor had yet delivered the killing blow. As Fortune observed: "I don't foresee any rival outright devouring X's lunch, but it's evident that people are growing weary of the unpredictable billionaire's theatrics." The platform's web traffic, measured by Similarweb, fell roughly 14% year-over-year by September 2023. The decline was not uniform — X retained enormous reach, still commanding billions of monthly site visits — but the trajectory was unmistakable, and the users leaving were disproportionately the ones advertisers most wanted to reach.

Where most companies that lose half their advertising revenue pursue reconciliation, Musk pursued litigation. In August 2024, X filed an antitrust lawsuit against the Global Alliance for Responsible Media (GARM), an advertising industry initiative focused on brand safety, along with member companies including CVS Health, Unilever, Mars, and Ørsted. The suit alleged an illegal conspiracy to boycott X and withhold "billions of dollars in advertising revenue." GARM, a small organization with limited resources, ceased operations rather than fight the lawsuit. Unilever settled on undisclosed terms.

Musk expanded the legal campaign in February 2025, adding Nestlé, Colgate-Palmolive, Lego, Shell, and Tyson Foods to the suit. "We tried peace for two years, now it is war," he posted on X. Bruce Daisley, Twitter's former VP of Europe and U.K. managing director, offered what may be the most succinct assessment of the strategy's novelty: "I can't remember an example in the history of marketing where someone from a platform has threatened to go to law and sue people who don't spend money with him."

The litigation served a dual purpose. It signaled to advertisers that leaving X would carry legal risk — an unusual deterrent, to say the least. And it reframed the narrative from "advertisers are leaving because the platform is toxic" to "advertisers are conspiring to suppress free speech" — a framing that resonated with X's increasingly politicized user base and with Musk's broader ideological project.

That ideological project — the thing that made X's valuation collapse simultaneously irrelevant and central to Musk's broader ambitions — became fully legible during the 2024 U.S. presidential campaign. Musk endorsed Donald Trump, hosted him for a live conversation on X Spaces, donated over $100 million to pro-Trump political action committees, and used the platform as an amplification engine for right-wing political content. When Trump won, Musk was rewarded with proximity to power: he was appointed to lead the Department of Government Efficiency (DOGE), a cost-cutting initiative within the federal government.

The political calculus reframed the entire acquisition. If X's value was measured purely as a media business — revenue, users, engagement metrics — it was an unmitigated financial disaster, a $44 billion bet that had lost perhaps 70% of its value in two years. But if X's value was measured as a political instrument — a platform that gave one individual the ability to shape public discourse, amplify allies, attack critics, and maintain a direct channel to hundreds of millions of people — the economics looked different. Not good, exactly. But strategic in a way that transcended the income statement.

Fortune noted the tension explicitly: X "has seen its valuation plummet to about $9.4 billion in October 2024, though it also has brought Musk major strategic benefits by providing him with a massive communications platform to push forward the controversial political agenda of his close ally President Trump." The platform that once styled itself as the world's "common digital town square" had become something closer to a privately owned propaganda channel — one that also happened to host millions of ordinary conversations about sports, weather, and vegan peanut butter cookies.

In early 2026, X is a strange object. Still enormous — hundreds of millions of monthly users, billions of monthly site visits, a presence in the daily lives of world leaders, journalists, and ordinary people in virtually every country. Still the default venue for breaking news, political combat, and the peculiar form of collective sense-making that no other platform has quite replicated. Still carrying approximately $13 billion in acquisition debt that generates over $1 billion in annual interest payments against a revenue base that may be less than half what it was three years ago.

The "everything app" vision remains aspirational. X has added payments functionality, video features, an AI chatbot called Grok (built by Musk's separate company xAI), and subscription tiers including X Premium. Linda Yaccarino, still nominally CEO, has brought in new executives — a CFO from Tubi, a marketing chief from Hyundai, an advertising head from Verizon — to professionalize the commercial operation. But the fundamental challenge remains: X's value proposition to advertisers depends on brand safety and premium audiences, and both are diminished. Its value proposition to users depends on network density — the sense that the important conversations are happening here — and that density is eroding as alternatives fragment the landscape.

Wikipedia founder Jimmy Wales, speaking at the Web Summit tech conference in 2023, captured the structural concern: "Twitter was, and now I guess X sort of is, in a way, the default public square for the world. And if it's being overrun by trolls and lunatics, it's not good for any of us."

The story of Twitter — from a side project built on SMS to the world's real-time information backbone to a billionaire's $44 billion political instrument — is not a story about social media, exactly. It is a story about what happens when something that functions as public infrastructure is owned as private property, when the gap between cultural importance and business performance becomes a canyon, and when the person who buys the thing values it for reasons the income statement cannot capture.

In the U.K. Companies House filings for fiscal year 2024, on a line labeled "revenue," there is a number: $39.8 million. Three years earlier, it had been $282.9 million. The math is simple. The meaning is not.