Talks with investors have cooled in recent weeks as Musk’s camp has raised doubts about recent “firehose” data — a set of data sold to corporate customers — that they obtained from Twitter. Musk’s team’s doubts about the spam numbers signal that they don’t think they have enough information to assess Twitter’s prospects as a business, the people said.
Now that Musk’s team has concluded it can’t check Twitter’s data for spam accounts, one of the people said, it is expected to take potentially drastic action. The person said a change in direction from Musk’s team is likely to come soon, though he didn’t say exactly what they think that change will be.
Spam accounts aren’t the only reason Musk tried to get out of the deal. Twitter’s share price has fallen sharply since its takeover bid in April, giving the impression that it is overpaying. Musk also runs two other major companies, Tesla and SpaceX, along with some startups.
But the terms of the deal mean it won’t be easy for Musk to step down. Musk has agreed to complete the deal unless something major happens to Twitter’s business, and legal experts doubt the bot issue will qualify. Twitter, which initially fought Musk’s takeover bid, will be a weaker company if the deal falls apart than it was when Musk first bought a stake, and experts expect Twitter to struggle to do so. Twitter itself has said it intends to complete the deal. Even if Musk convinces a judge to let him go, he could still be on the hook for the $1 billion breakup fee.
Twitter spokesman Brian Poliakoff declined to comment, but referred to a statement the company made in June.
“Twitter has and will continue to jointly share information with Mr. Musk to complete the transaction in accordance with the terms of the merger agreement,” Twitter said in the June statement. “We believe this agreement is in the best interests of all shareholders.” We intend to close the transaction and enforce the merger agreement at the agreed upon price and terms.”
Musk rocked the social media world earlier this year with his unprecedented proposal to take the company private, claiming he would be able to grow Twitter and make it more open and, he said, politically neutral. He said he would let former President Donald Trump back on the service and argued that content moderation practices infringe on free speech. Musk waived his right to take a deeper look at the company’s finances when he signed off on the deal.
But soon after, questions arose as to whether he would actually continue. A global sell-off in tech stocks has severely dented his personal net worth, which he had used to get the debt commitments he needed to buy Twitter.
Musk’s enthusiasm to go through with the deal has been in question since at least May, when he said the deal was “on hold” until he could determine whether Twitter’s claim that fewer than 5 percent of accounts were bots or spam. is true. He accused Twitter of withholding information, while the company said it acted in good faith and provided everything required by the terms of the deal.
“Twitter is not cooperating,” said a person familiar with the discussions, speaking on condition of anonymity because of the sensitive nature of the conversations.
The focus on bots plays into Musk’s favor by driving down Twitter’s stock price and potentially helping him force Twitter to renegotiate the deal at a lower price.
The debate over Twitter bots is not new and has been a large part of the public conversation surrounding the company for years. Musk himself faces a lot of spambots in replying to his tweets. Twitter has long said about 5 percent of its accounts are bots or spam, while outside researchers sometimes say the number could be much higher. Because of how quickly the tactics for creating and concealing the nature of fake accounts change, it’s difficult even for experts to make definitive statements about who is right.
Twitter defended its process for measuring unwanted accounts, including in a news briefing Thursday morning.
Twitter said that every three months it samples its “monetizable average daily users,” the base of users the company plans to charge advertisers to reach. He analyzes this sample by hand to determine if they are fake or not. He said it was always comfortable to keep the total below the 5 percent threshold.
Twitter doesn’t ban all bots that include accounts that post photos of otters about the time or temperature in a particular location. Instead, it looks for indicators that suggest fake or coordinated spam activity, such as mass account creation or coordination between people to artificially boost a particular tweet, set of tweets, or topic.
The terms of the briefing precluded direct quotation or naming of any of the experts.