Subscribe to discover Om’s fresh perspectives on the present and future.
The Land of Imperfect Data.
The technology press is once again in an uproar over half-truths and incendiary comments by Elon Musk, the owner of Twitter. He indicated that he’s going to remove the “block” feature from Twitter (formerly known as X) and instead will force users to opt for “mute.”
The bigger question is: why would Twitter undertake such a move? I suspect the service’s traffic and engagement are declining. In theory, the end of blocking should lead to more engagement in timelines, but I’m not so sure. Most of the accounts I’ve blocked are spambots, known trolls, and other unwanted online entities. I don’t really have any desire to engage with them. So, ending the “blocking” feature is kind of counterproductive and reduces my Twitter usage.
Still, it will help create a false impression of growth and usage on Twitter. The platform has been on a decline since Musk took over in the first half of 2022. This isn’t just a problem for Twitter; it’s a widespread issue across our entire technology ecosystem.
But before I continue, take a moment and watch this video. It left quite an impression on me when I was much younger, and I have never looked at data or numbers with naiveté since.
Scottish poet Andrew Lang once said, “Politicians use statistics in the same way that a drunk uses lamp-posts—for support rather than illumination.” The same argument can be made for technology companies.
The half-truths in the data trotted out by technology companies show inflated growth, perceived engagement, and give a sense of overall well-being. Whether you’re an upstart or a search giant, a social king or a baron of the clouds, you can manipulate the numbers to show what you want. Given the speed of our news cycle, everyone knows that no one will take the time to dig deeper.
I have doubts about the data and metrics shared by cloud companies — none of them are very transparent, and the data isn’t consistent across the industry. It gives a skewed idea of what growth is, who is growing, and why. Start digging in, and you’ll find that not all clouds are equal.
For instance, Google faces legal challenges due to its advertising practices, with some numbers raising doubts about its overall advertising revenues. They are not the only ones: a report from Business of Apps estimates that total online ad fraud will approach $100 billion in 2023. According to Interceptd, 31% of iOS apps and 25% of Android app installs are fraudulent — but that doesn’t stop either Apple or Google from touting the overall growth of their app stores and their app economy.
And no one is more egregious than Facebook, which buries the facts about fake accounts and duplicates at the end of its filings with the regulators. It’s hard to discern what’s true and what’s a mirage with that company. In 2018, when I first reported on Facebook’s fake accounts, I pointed out that back in 2016, the company had 6 percent of its total monthly active users as duplicates, fake, or bot accounts. By 2017, this number had risen to about 13 percent.
By 2020, 16 percent of all Facebook accounts were false or duplicates — about 432 million. In 2021, that number stood at 465 million. At the end of 2022, it had climbed to 469.28 million. By the end of Q1 2023, the number had decreased to 426 million. One thing is certain: nearly 16 percent of Facebook’s monthly active users are “false or duplicate”. However, it’s challenging to determine if the activity from these accounts influenced daily usage or skewed other data in favor of Meta.
I don’t mean to single out Facebook; Twitter isn’t immune to these problems either, and nothing illustrates this better than Elon Musk’s own data.
Last year, when Elon was considering not purchasing Twitter, he sent out a tweet lamenting the number of bots and spam accounts on the platform. (Twitter’s management at the time claimed that less than 5 percent of Twitter’s user base consisted of spam, fake accounts, or bots.)
“20% fake/spam accounts, while 4 times what Twitter claims, could be *much* higher.” Elon Musk, May 17, 2022.
That tweet is ironic; it’s either true, coming from a place of personal experience, or both.
Since taking over Twitter, Elon’s influence has expanded, with his follower count now standing at 153 million. However, not all of these accounts are real or meaningful, as data researcher Travis Brown (via Mashable) discovered.
“the 153,209,283 X accounts following Musk at the time the data was collected, around 42 percent of Musk’s followers, or more than 65.3 million users, have zero followers on their own account. Just over 72 percent, or nearly 112 million, of these users, following Musk have less than 10 followers on their account.” Over 62.5 million Musk followers have zero tweets.
….if an account uses the default avatar or the auto-generated username that usually includes a string of numbers, there’s a good chance you haven’t come across a real, active user. Just around 25 percent of all of Musk’s followers, or more than 38 million, use the default profile image that X provides for newly registered accounts. More than 40 percent of Musk’s followers, or just under 50 million, have 4 or more numbers in their @ handle on X.
One line, in particular, stood out to me: “median number of followers for all 153 million accounts following Musk is just one (1) single follower.”
Having such a multitude of fanbots can prove useful and profitable. In April 2022, preliminary research from David A. Kirsch, a professor at the University of Maryland, showed that Twitter bots played a significant role in creating the cult of Musk and, in the process, boosted Tesla. He highlighted that “about 1.4 million tweets from the top 400 accounts posting to the “cashtag” $TSLA, 10% were produced by bots. Of 157,000 tweets posted to the hashtag #TSLA, 23% came from bots.”
In his research, “Fanbois and Fanbots: Tesla’s Entrepreneurial Narratives and Corporate Computational Propaganda on Social Media,” Kirsch concluded that “This computational content might have protected the Tesla narrative from an emerging group of critics, reduced downward pressure on the Tesla stock price, and heightened pro-Tesla sentiment from the firm’s IPO in June 2010 to the end of 2020.”
Musk’s adeptness at using Twitter to influence sentiment around Tesla (and other topics) has enabled him to use the platform as his bully pulpit. (I have discussed this topic on several occasions.)
I know you believe what you believe, but given the data analysis, there is little doubt that bots and spam are a big problem on Twitter — and only going to get bigger. The company has stopped spending money to fight that problem, and by ending “block” and other such features, it is only going to use false data to paint a picture of growth and prosperity.
August 20, 2023. San Francisco
Related Recommended Reads
One thought on this post
Comments are closed.
Good piece. Now I get it.
Btw, I think you mean: Twitter (formally known as X)