Muscology is an overpopulated field. The forced purchase of Twitter by the world’s wealthiest alpha dork has already been taken up from all sorts of angles, with the near-universal conclusion being that he screwed it up.
Waggish Hybris has encountered contractual and financial nemesis, which has led to irritation. Or, to put it in Elon Musk’s preferred form:
However, are there lessons to be learned from less than a decade of public life as Twitter stock is set to be delisted today?
Possibly. Attempts to make money from microblogging predate Musk’s involvement by several years. A summary of what went wrong and when might help determine the reason for the reinstall Corn on the Cob-in-Chief constituted its many prejudices and blind spots.
Ideally, the reflection process also shows a reversed Frankenstein: or how the machine created the monster and how the monster will destroy the machine.
A @Elon Musk However, the Twitter profile first appeared in early 2009 he wasn’t. The US election a year earlier had raised Twitter’s public profile enough to attract prominent CEOs, including jeff bezos, Sundar Pichai and Mark Zuckerberg, as well as speculation about interest from Google and Yahoo. But during the first phase of Twitter’s land grab, Musk remained mostly offline and distant.
He missed some fun. Twitter’s 2009 vintage “About” page offered the following warning:
How do you make money with Twitter?
Twitter has many attractive revenue-generating opportunities, but we’re holding off on implementing them for now because we don’t want to distract ourselves from the more important work of creating a compelling service and great user experience for millions of people around the world World. While our business model is in the research phase, we spend more money than we earn.”
Also in early 2009, Jeffrey Lindsay, celebrity analyst at Bernstein Research, wrote that buying Google or Yahoo Twitter would be an absolutely terrible idea:
Don’t get us wrong – we like Twitter and we think the concept of sending short, standard candid messages detailing your activity is a lot more fun and valuable than it sounds. The problem is that Twitter falls into the broad category of new internet companies that are almost impossible to monetize.
Lindsay’s 2009 guide to internet investing was to avoid social media. Non-niche sites would be poor at guessing their users’ specific desires without siphoning off data at problematic volumes, he argued, so ads could only ever be intrusive. Users would refuse to pay for something that was previously given away for free. The marketing and network effects of celebrities on board could easily reverse:
[Internet businesses that] Making money can be described without referring to “future levels of monetization”, “deeply embedded value stacks”, “preservation as a free tool” or “the 50 percent of users who would definitely pay $5 a month for it”. The ones that make money all seem to involve either shipping stuff to people who paid for it online, or getting endorsement deals from the likes of General Motors.
A year later, while Twitter was still independent thanks to Bonfire funds from VCs like Bezos Expeditions and Spark Capital, Musk took over his eponymous account. He started posting fake quotes and Name dropping but promised to use google+ for anything serious. Halcyon times.
That first viral tweet (in December 2011) was a hoax stolen from Tina Fey. The second (in November 2013, the month of Twitter’s IPO) was lamentable media bias. The third (July 2014) was a nerd memes. The fourth (October 2014) was thick allusion and thus Musk’s personality was fully formed.
In between, Twitter was valued at $14.2 billion when it went public, or 4.4 times its 2012 sales ($317 million). At the end of the first session, it was valued at $31 billion, or 9.7 times sales. That, the commentator agreed, was insane. Twitter was clunky, cold and unfriendly, especially when compared to Facebook, and could only sell its promoted content at flat fees. The estimated 600 advertisers each spent an average of just $30,000 per year; Facebook was making $1 million per full-service customer back then.
Twitter’s float was therefore based on the Field of Dreams principle. Early research found ambitious multiples for its nascent video service (aka Vine) and for MoPub, a mobile ad exchange that Twitter bought just prior to its IPO. (It was recently sold.) Data services were already in the works and subscription fees were never mentioned, as it had been agreed long ago that scaling customer tiers was anathema. There was only advertising.
Here’s a sales forecast from Morgan Stanley’s 2013 initiation note:
Key analyst Brian Wieser estimated that Twitter could capture about 1 percent of global adspend excluding China, warranting a $25 billion valuation once the business reaches maturity in 2018. Perhaps surprisingly, his predictions weren’t that far off.
The table below shows Pivotal’s pre-float expectations versus actual performance; The graphic below, via Bloomberg, provides the big picture.
So Twitter increased revenue as expected, but couldn’t cover costs and therefore continued to burn money. That’s the story of the charts, but it’s not one that investors will recognize. Quarter after quarter, the company would disappoint the Street on sales, not expenses. Management had to constantly downplay ad expectations. User numbers have always lagged behind forecasts, partly due to over-optimism, but also because the definition of a user was constantly changing.
JPMorgan had estimated in its initiation research that the active user base would more than double to 500 million within five years; For the fourth quarter of 2018, Twitter reported 126 million daily active users. Recent discussions about counting bots have kind of overlooked all of this, and finding out the size of the audience has turned out to be almost entirely pointless.
Twitter’s share price slipped below the float price level in mid-2015 and stayed under the water for another two years. There was a bit of a rebound in 2017 as GAAP profitability took a backseat and video and live streams looked like they could be useful new products. But results remained weak and CEO Jack Dorsey’s eccentricities caused concern.
It wasn’t until March 2020, shortly after Twitter released guidance on pandemic uncertainties, that the stock started to climb. The appeal of Doom scrolling meant that user numbers soared and advertisers had a literal captive audience. Price chart from Bloomberg:
It’s worth comparing the above with a logarithmic perspective of Musk’s personal Twitter experience:
Here are some possible takeaways. The first, and perhaps the most obvious, is that Musk uses Twitter as a transmission medium. His subsequent accounts took five years to double and another five years to double again. Musk only wants to listen when spoken to, which is fair enough. He’s probably very busy and very rich.
The second takeaway is that unlike most Twitter users, Musk’s audience is growing exponentially with his output. Total tweet count has doubled roughly every two years, while follower count has doubled annually. For any attention vampire, that alone is a strong positive ROCE, making Twitter a wise use of time and energy.
Tweets per day show that Vogel app addiction had subsumed Musk’s personality by 2019, having first become problematic around 2014.
Musk’s personal golden age of tweeting was 2014-18, which coincided with Twitter’s worst years in terms of share price. The delay meant Musk missed the hope and hype. Instead, he only caught business failures, largely driven by user goals, and the powerful impact of social networks, largely political in nature.
In that context, why wouldn’t Musk think he’s better on Twitter than Twitter? Rapid audience growth despite nerdyness, excitement, and mediocre shitposting proved he could do what Dorsey couldn’t.
But Twitter’s problems as a public company have been misdiagnosed. Revenue growth wasn’t far from original targets, despite quarterly agonies over ad metrics and active user counts. Where the IPO models went wrong was in the cost. They went up, not down, because maintaining a comfort zone for advertisers turned out to be an ongoing expense that grew in proportion to the user base.
To be fair, Musk attacked the cost issue with zeal. But not much foresight. Whatever the motivations for bidding, his approach to running the business — the indiscriminate layoffs, audience goading, firebombing of advertisers’ comfort zones — stemmed from his atypical experience with the site.
The story could have offered a clearer perspective on why, for example, charging $8 a month to watch a rich man humiliate himself is not the gamble imagined. But it seems that during the months he’s been desperate not to buy Twitter, Musk has only been thinking about what to tweet next.
Honestly, where to start? For The Verge, Nilay Patel has an unparalleled overview. For the FT, Hannah Murphy and Ian Johnston provide the inside track. Marietje Schaake also summarizes the key regulatory points for the FT. For platformers, Casey Newton and Zoë Schiffer take a close look at job cuts. For Tocqueville21, Megan Jones dissects Elon’s personality. On Medium, Dave Troy has a secret conspiracy theory. Matt Levines works piecework for Bloomberg. On Twitter, Andrew Gawthorpe has this Geopolitical conspiracy theory. And finally, there’s this tweet:
It just struck me that the beauty of this week’s timeline is that this is a show about the costliest midlife crisis of all time, but told from the point of view of the sports car itself
— zatapatique (@zatapatique) November 6, 2022