Ev Williams gives up
How Medium got stuck in the middle. PLUS: Twitter sues Musk
Few tech CEOs can claim to have steered the course of online conversation more than Ev Williams. In 1999 he co-founded Blogger, which helped to take blogging mainstream with a well designed, free tool that sold to Google four years later. In 2006 Williams and his co-founder followed with Twitter, which remains one of the most influential social networks in the world.
Five years later, after a stint as Twitter CEO and much turmoil, Williams announced his next act: Medium, a publishing platform that sought to split the difference between blogs and tweets: medium-length posts, published occasionally rather than daily, that would seek to answer one question: “Now that we’ve made sharing information virtually effortless,” Williams wrote in 2012, “how do we increase depth of understanding, while also creating a level playing field that encourages ideas that come from anywhere?”
Medium declined to make Mr. Williams available for an interview. He said in his post that he was leaving Medium because “change and renewal are healthy,” noting that August will be his 10th anniversary as chief executive.
“To be clear, Medium’s story is far from over,” Mr. Williams wrote.
Medium said that Mr. Williams would be stepping down as chief executive effective July 20 and that he would be replaced by Tony Stubblebine, the chief executive of the online coaching company Coach.me. Mr. Williams will become chairman of Medium’s board, a new position.
Williams said he would remain chairman of the board but also start a “a new holding company/research lab” to work on other projects.
I last wrote about Medium in March 2021, days after Williams had announced his latest about-face for the company. Two years after launching a subscription-based group of publications dedicated to high-quality original journalism, and just after its workers fell one vote short of forming a union, Williams offered buyouts to all of its roughly 75 editorial employees.
As I wrote at the time, by some measures Medium was succeeding. It had started 2021 with around 700,000 paid subscriptions, and was on track for more than $35 million in revenue from its $5 monthly subscription offering. At the same time, internal data showed that it largely was not high-quality journalism that was leading readers to subscribe: it was random stories posted to the platform by independent writers that happened to get featured by the Google or Facebook algorithms.
Williams frequently spoke of his desire to make Medium a place for high-quality writing, one that elevated the national conversation. His annual appearances in the New York Times proclaiming that the internet was broken and that he intended to fix it, became a running gag among a certain set of media obsessives. (OK, maybe just me.)
And yet by last year Williams was confronted by the fact that his revenues were being largely dictated by what performed well on search engines — the same business logic of the low-quality content farms of the late 2000s. It wasn’t a terrible business, exactly. But it wasn’t the business Williams had set out to build — nor was it the business he had hired for.
And so for the second time in his tenure as CEO — the first had been in 2015, when he laid off 50 people in a pivot away from advertising toward subscriptions — Williams upended the lives of dozens of journalists and flushed them off the platform.
Media is a famously unforgiving business, and Williams is far from the only CEO to struggle to build a sustainable company. And yet over the past decade, few have matched him for the sheer number of changes in direction he inflicted on investors, users, and employees.
Nieman Lab’s Laura Hazard Owen wrote an essential guide to Williams’ whipsawing in 2019. Among the things Medium tried during his tenure, from its launch to the present day:
A Snapchat stories clone (2017).
Hiring exclusive columnists (2018).
Paid newsletters (2020).
“An audio-based learning platform” (2021).
There’s no shame in a startup trying lots of different ideas. But Medium’s ideas were often coupled to to the livelihoods of journalists and the publications they worked for. It’s one thing to have a singular vision and change your tactics along the way; Williams vision for what Medium was transformed almost continuously.
“When I first started at Medium I was really inspired by Ev,” one former employee told me today. “He's a Vision Guy™, and that really shone through for me in the beginning. But vision only goes so far, and over the years it gave way to my increasing perception of him as out of touch and restless for results at any cost (see: the one thousand pivots in strategy over the last 5 years).”
Williams went back and forth on whether Medium should host its own publications or serve as a platform for others to build on. And while he dithered, Medium got caught in the middle.
On the high end, well funded digital publishers from BuzzFeed to Vice to the Atlantic excelled at publishing high-quality journalism. And on the low end, Substack emerged to let solo creators develop thriving, sustainable careers by offering individual subscriptions. (See my ethics disclosure about Substack.) In such a world, Medium had no obvious advantage. With its owned and operated publications gone, it became a general-interest web magazine staffed by freelancers and dependent on Google.
Another former employee noted that, for all the pivots over the years, Williams always seemed a step behind.
“I'd say you could describe the Ev era of Medium as a series of digital publishing experiments that often felt of the zeitgeist without ever defining it,” the employee said. “A lot of the work Medium did over the years genuinely had an impact, but it often felt, for whatever reason, like Ev made it a point not to lean into this work. he meandered and never seemed satisfied. And eventually Twitter evolved to support more of the kind of publishing that had originally been native to Medium, and Substack came along and ate the platform's lunch.”
“He's a little bit of a mystery to me,” the employee aded. “I hope a leadership shakeup is good for the company and the people who work there.”
A third former employee told me my assessment of Williams — essentially, a callous dilettante — was unfair.
“I think he was trying to solve a really hard problem, it kept not working, and he screwed a lot of people over to varying degrees by continually changing his approach,” the employee said. “But he really did try a lot of things, and it wasn't necessarily obvious that they'd fail until somebody with a ton of money tried it.”
Medium declined to comment, as did incoming CEO Stubblebine. (“I need to get my feet under me first,” he told me over direct message. “But after that I expect to be pretty chatty.”)
The challenge for Stubblebine is the same as it was for Williams: after ten long years, there is very little that Medium hasn’t already tried. It was always grandiose to suggest that a humble blogging platform could fix the internet. Now Williams has kicked himself upstairs, and it will be up to someone else to try to fix his company.
Twitter sues Musk
As expected, Twitter sued Elon Musk in an effort to force him to complete his $44 billion acquisition of the company. The lawsuit is a doozy, and if you’ve been following the acquisition closely, I highly recommend giving it a read — it’s surprisingly breezy for a legal filing, and documents a staggering array of bad-faith behavior from Musk and his team beginning almost immediately after he signed the deal.
Some things that stood out to me:
After Musk’s incessant complaints about bots and spam, Twitter tried to set up a meeting with Musk to explain its process for determining which accounts may be fake. In response, Musk said he hadn’t actually read the detailed explanation they had prepared for them, and never met with the team.
Amidst all this chaos, Twitter sought to create a retention program to incentivize employees to remain through the close of the deal. But Musk refused to approve it — and now attrition is going up, Twitter says.
Twitter noted that Musk isn’t allowed to back out of the deal if he himself doesn’t meet his own obligations to it — and then list six ways he has violated it, including by disparaging the company and its employees.
And yes, Musk’s poop-emoji response to Twitter CEO Agrawal did make it into the lawsuit as an exhibit.
Twitter is seeking to go to trial as early as mid-September, and believes the trial can be completed in as little as four days.
Musk responded — if that’s the right term — to the suit by tweeting “Oh the irony lol.”
I’m sure Musk’s team will respond with lots of exasperated spluttering sometime soon. In the meantime, though, the lawsuit is a refreshingly clear-eyed and meticulously evidenced-based repudiation of nearly everything Musk has said about the company and the deal since he signed it.
So how does this resolve? I’d say it makes more likely one of the options we talked about here yesterday: Musk walks away from the deal in a settlement with the company, but it costs him much more than $1 billion to do so.
An anonymous former Twitter employee speaking under oath to the January 6th commission said the company had gone on easy on President Trump while he was in office, declining to take action against his account because the company relished its close connection to Trump. (Drew Harwell and Naomi Nix / Washington Post)
The Federal Trade Commission warned tech companies against misleading consumers by telling them that their personal data has been “anonymized” while still selling it off to data brokers, where it can be easily de-anonymized by combining it with other data sources. (Michael Kan / PC Magazine)
A look at how WeChat became “a powerful tool of surveillance and control” in China — but also grew large that President Xi Jinping came to see it as a threat to his rule. (Lulu Yilun Chen / Bloomberg)
Elon Musk said Donald Trump should “hang up his hat & sail into the sunset,” after the former president called Musk "a bullshit artist.” This is the first known case of me agreeing with both Elon Musk and Donald Trump. (Gareth Vipers / Wall Street Journal)
BLOOM is a new open-access large language model, similar to OpenAI’s GPT-3 or Google’s LaMDA, created by volunteer researchers and available to anyone under a new “responsible AI” license. It’s the first such model to be made available in Spanish, Arabic, and several other languages. (Melissa Heikkiläa / MIT Tech Review)
Meta ordered engineering managers to identify low-performing employees and report them to human resources, a likely first step toward significant layoffs. ““If a direct report is coasting or is a low performer, they are not who we need; they are failing this company,” its head of engineering wrote. “As a manager, you cannot allow someone to be net neutral or negative for Meta.” (Naomi Nix and Elizabeth Dwoskin / Washington Post)
YouTube TV hit 5 million subscribers. That’s up from 3 million in 2020. (David Pierce / The Verge)
Mysten Labs, a startup co-founded by two people who helped build Facebook’s cryptocurrency project Novi, is reportedly raising a $200 million series B round, at a $2 billion valuation. (Aidan Ryan / The Information)
“Clubhouse isn’t dead yet,” this piece argues, based on … interviews with three people who say they’re still using it. (Kaya Yurieff / The Information)
Those good tweets
Talk to me
Send me tips, comments, questions, and Medium posts: email@example.com.