Is Elon Musk still interested in buying Twitter? Only the person concerned knows the answer to this question, valued at 44 billion dollars. Until Friday, May 13, the agreement between the two parties was on track. But that day, Musk signed a tweet of which he had the secret, referring to the quarterly results of Twitter which informed of the presence of less than 5% of fake accounts and spam on its platform. The entrepreneur-star doubts the accuracy of this figure (yet unchanged for several years) and makes it known to his 94 million subscribers. The sale is suspended. Twitter’s share price drops below $40, down from 54.20 in the original sale deal. The “boss-influencer” has struck again.
Apart from a withdrawal with a loss and a crash – which would cost it at least a billion dollars – everything suggests that this reversal aims to renegotiate the deal on the decline. Musk is currently losing billions of dollars in the fall of his Tesla shares. His new Twitter fad is becoming more and more costly for him. False users, who for a certain number push harmful, even misleading content, can in this context become an interesting means of pressure. Because this scourge, rather well known, remains today difficult to quantify with precision.
“This is an argument that will not be enough on its own to derail the agreement, observes Asma Mhalla, teacher at Sciences Po Paris, specialist in the digital economy. But it can be in order to challenge and bog down the transaction.” A Machiavellian plan that Twitter, currently in difficulty, faced with major departures and a recruitment freeze, will have a hard time thwarting, thinks the American lawyer from Harvard Law School, Alejandra Caraballo, a fine connoisseur of the file. “There will be a lot of estimates and expert reports. A real side show during which Elon Musk will be quiet,” she wrote. on Twitter.
Limited offer. 2 months for 1€ without commitment
battle of numbers
After the announcement of the suspension of the agreement, the boss of SpaceX therefore quickly launched into the battle of figures. On Tuesday, Musk claimed that there would be 20% of these unwanted profiles on Twitter, more than four times the figure given by the social network. The source: a study published a few days earlier by the agency specialized in audience measurement on social networks, SparkToro, which gives precisely 19.42% of fake accounts on Twitter during the last quarter, on the basis of a sample of 44,058 active public accounts over the last three months. This is one of the highest estimates to date. An Israeli company, Cyabra, achieved a rate of 13.7%; others judge it to be less than 10%.
In a long thread messages published on his social network, Parag Agrawal, the current CEO of Twitter, took the trouble to explain on Monday why this information was so difficult to calculate. “The most advanced spam campaigns use combinations of coordinated humans and automation. These accounts also compromise real accounts and then use them to further their campaign. This is all very sophisticated and difficult to detect,” comments- he. An admission of weakness. But Twitter, which says it suspends half a million offending accounts a day, is still best placed to know what’s going on on its platform. The social network “uses both public and private data (for example, IP address, telephone number, geolocation, client/browser signatures, what the account does when it is active…) to take stock of each account”, and determine the real users of the fake ones, continues Parag Agrawal.
Estimates other than those of Twitter would therefore not hold up? “Thanks to their own information, their metadata, Twitter remains the best able to target inauthentic accounts and spam,” supports Anuchika Stanislaus, analyst specializing in digital public policy. If external analyzes are not ineffective, however, they do not have “the same methods as Twitter nor the same definitions of what are called fake accounts, spam and bots”, continues the expert. Not all automated accounts or bots are harmful, for example, like “chatbots”, used by companies for their customer relationship.
But does Musk care? Faced with the arguments put forward by the CEO of Twitter, the entrepreneur responded with an emoji “poop”.
Musk, more than a troll
Reading at this level is simple: anyone who wants to buy Twitter is surely its biggest “troll”. In truth, Elon Musk knows how to put his finger where it hurts. The large number of spam potentially deprives Twitter of significant advertising revenue. “The question of fake accounts is a real political subject: they transform Twitter into a space of informational, psychological warfare”, adds Asma Mhalla. The war in Ukraine recently brought a new illustration of this. Elon Musk had already considered that in his eyes it was the main scourge on the blue bird site.
Be that as it may, Elon Musk’s criticisms are not gratuitous. “This case once again touches on the question of the insoluble balance between the protection of privacy and issues of transparency”, also comments Iris Boyer, secretary general of the Institute for Strategic Dialogue (ISD). “Even if Twitter is one of the platforms that makes the most effort in this area, Elon Musk recalls how these social networks still operate as black boxes, and could obtain more precise results on these fake accounts by opening a little more their data to researchers.”
The multi-billionaire has every interest in disagreeing with the figure on fake accounts presented by Twitter. Even in the event that this would be the closest to reality. Because a very small percentage of fake users ultimately undermines its own vision of the social network, which is deeply degraded. “Twitter claims that over 95% of daily active users are real humans. Does anyone have that impression?” Musk asked on Tuesday. to its subscribers. His personal account is the subject of estimates ranging from 50% to 70% of “fake” subscribers, spam.
The Express app
To track analysis and decryption wherever you are
Download the app
This vision of “his” Twitter is also the one that drives him to want to change the model, gradually abandoning advertising in favor of a subscription system. In Musk’s ideal world, this new “freemium” Twitter and open-source, would thus be freer, and would finally require less intervention from the platform. “While if 95% of users are real, Musk understands that he will surely not be immediately exempt from the heavy, costly and imperfect moderation work that Twitter is doing at the moment,” Asma Mhalla sketches. The issue of fake accounts may be much more than a financial excuse for Musk.