Why Twitter never made money

The scoreboard is four CEOs and $5 billion in ten years, with declining users, executive turnover, infighting and no vision. The board repeatedly put the company up for sale and Google, Facebook, Disney and Salesforce all walked away over concerns deeper than price.

The power struggles began at the founding. Of four founders in 2006, Noah Glass — who came up with the name and the idea — was fired and written out of the company's history entirely, never once publicly credited while the other three became billionaires.

Jack Dorsey's first turn as CEO was a mess. He ran the prototype build in production with no backups, tracked company expenses on his own laptop and got them wrong, ran up six-figure monthly SMS bills dogfooding, and left early for drawing classes, hot yoga and fashion design. Ev Williams told him plainly: *"You can either be a dress maker or the CEO of Twitter."*

Ev removed him in an ambush over breakfast — severance, stock, all voting rights stripped, and a silent chairman title where nobody was obliged to listen. Ev then shut off Jack's company email, a petty move that guaranteed the grudge.

Ev was the best product CEO the company ever had, and the numbers say so. Rather than forcing innovation, he watched what power users invented and made it official: hashtags (which Jack had refused as too complicated), retweets, lists, @mentions, verification, location and the first iOS app. Unique visitors went from 2.9 million under Jack to 44.5 million in a year, and users from 500,000 in 2008 to 175 million by 2010 at 370,000 signups a day.

Jack's exile was a seven-year public relations campaign. He told Vanity Fair the idea came to him at age 8, told 60 Minutes it came from a fascination with trains and maps, told the LA Times he had pitched it to classmates in 1984, and demanded corrections from any article that omitted him — all while running Square and holding no role at Twitter.

Ev knew and did nothing, which cost him the company. He judged that removing Jack risked a public relations disaster that would damage growth, and swallowed it. Meanwhile Jack was in employees' ears at coffee shops and at his own house, collecting grievances to build a case to the board.

The genius of the coup was not asking for the job. Jack argued Twitter needed a mature operator rather than another product-minded engineer — and proposed Dick Costolo, Ev's own friend and hire, with himself returning to run product. Dick knew it was a coup and took it anyway. He had joked on his first day as COO: *"Task #1: undermine CEO, consolidate power."* He has since deleted the tweet.

Costolo's worst decision was killing the API. To count every impression before the IPO, Twitter banned third-party clients — the developer ecosystem that had carried the product through Jack's neglect and supplied its feature ideas. It removed the one source of outside innovation, then he lost his job three years later for failing to innovate.

The IPO was the high point and the exposure. Revenue had grown from $30M in 2010 to $664M in 2013, headcount from 200 to nearly 3,000, and the stock ran from $40 to nearly $70. Then the full year landed: over $500M on R&D, a quarter of a billion on sales, and a net loss near $700M on revenue short of his own billion-dollar projection.

The comparison that ended the Facebook narrative: in the same year Facebook spent twice as much on R&D and four times as much on sales and marketing — and made a profit on $8 billion. Per user, Facebook was worth $8 and Twitter $2.75. Advertisers had priced the difference in user quality precisely.

The structural problem is that most of Twitter is free to read. Over 500 million people visit monthly without logging in, and Twitter only earns from those who do — while Facebook collects real name, birthday, gender, school, employer and location at signup. It cannot put a registration wall up, because reach creates virality and virality is the reason anyone posts at all.

Which limits what can be sold. Twitter works for brand awareness and fails at direct response — if you want app installs or purchases, Instagram, YouTube, AdWords or Facebook are simply better. Even Jack eventually questioned publicly whether Twitter could ever compete there.

Social TV was an expensive fantasy and an NBC executive killed it with data. At the 2014 Sochi Olympics, NBC broadcast 1,500 hours to 21 million viewers against 10.6 million tweets from 3 million users — his conclusion being that the show drives the social activity, not the reverse. *"The emperor wears no clothes."* The Nielsen Twitter rating has not been mentioned since.

Then everyone threw bodies at it. Costolo doubled sales and marketing from $300M to over $600M in a year and pushed R&D past $800M, burning nearly $2 billion annually, while MAU crawled 288M to 320M. Dorsey returned with a blank cheque and MAU went 320M to 319M, revenue growth hit 14% — the lowest in company history — and losses stayed near half a billion.

Profitability, when it finally came in 2018, was luck rather than product. Cord-cutting pushed television budgets into online video, Facebook and Instagram ran out of ad inventory and got expensive, and Twitter was attractive by existing — cheaper video ads with adequate reach. Net income of $1.2 billion, and the company stopped reporting MAU. Wall Street read it correctly as macro tailwind rather than innovation, which is why the stock never recovered and Elliott Management arrived in 2020.