The drama of Twitter spam accounts

@paraga So how do advertisers know what they’re getting for their money? This is fundamental to the financial health of Twitter.

The news on Elon Musk the acquisition of Twitter caused a huge stir in the company, prompting numerous meetings and announcements, employers fearing the possibility of layoffs, and, of course, more Musk tweets. Then one Tweeter from the CEO of SpaceX and Tesla made a louder boom: “The Twitter agreement temporarily suspended pending details supporting the calculation that spam/fake accounts indeed represent less than 5% of users. Then Twitter CEO Parag Agrawal gave us a thread of tweets addressing the issue and how the company is handling it. Musk’s first response? A poo emoji.

Twitter and bots: what makes fake accounts tricky for the Blue Bird app?

Parag Agrawal made a detailed explanation some of the challenges Twitter is really facing with bots on Twitter. These bots are not just what we think of them. They can be complex in terms of types and their behaviors. According to him, they can be a combination of automation and human coordination, and they even compromise real accounts, making things more complex than they already are. With this, the Agrawal clearly emphasizes that the solutions to combat them are “dynamic”, and the bots do the same: they also evolve. “You can’t build a set of rules to detect spam today and expect them to work tomorrow. They won’t,” the tweet read.

Agrawal reports that the company suspends more than half a million accounts a day and locks millions of suspected spam accounts every week. Apparently, challenges remain despite the sophisticated system Twitter has built to stop these fake accounts.

“The difficult challenge is that many of the accounts that appear superficially fake – are actually real people. Our team is constantly updating our systems and policies to remove as much spam as possible, without inadvertently suspending real people or adding unnecessary friction for real people when they use Twitter: none of us wants to solve a captcha every time we use Twitter. Now we know we’re not great at catching spam. And that’s why after all the spam removal I talked about above, we know that some are still slipping in. We measure this internally. And each quarter, we’ve estimated that

How do I know the proper rating for Twitter bots?

Independent researchers claim that more than 20% of accounts on Twitter are fake. That’s a huge number compared to the 5% that Twitter claims for its 226 million monetizable daily active users in the first quarter. If true, it means a big loss for Musk, who is offering to buy the company for $44 billion, which is more than Twitter’s real value.

With that, Parag Agrawal and his team are badly in need of proving the numbers. According to him, the company follows a set of specific standards to produce the estimates and determine the actual accounts of those that are not.

“Our estimate is based on multiple (repeating) human reviews of thousands of accounts, which are randomly sampled, consistently over time, from *accounts that we count as mDAUs*. We do this quarterly , and we’ve been doing it for many years,” Agrawal says in the tweet. “Each human review is based on Twitter’s rules that define spam and platform manipulation, and uses both public and public data. private (e.g. IP address, phone number, geolocation, client/browser signatures, what the account does when active…) to make a determination on each account… Our actual internal estimates for the last four quarters were all well below 5% – based on the methodology described above. The margins of error on our estimates give us confidence in our public statements each quarter.

The problems

Twitter uses private data to correctly determine actual spam accounts on its platform. The problem is that this type of data cannot be shared by Twitter with external researchers. With this, claims that the platform has a higher number of bots cannot be validated externally.

Here’s another problem related to this: Twitter is the only one who can tell the exact number of bots it has. If he wishes, he can increase the number for his benefit to attract investors, attract more attention and much more. Enter Musk with his poo emoji.

“So how do advertisers know what they’re getting for their money? asks Musk. “It’s fundamental to the financial health of Twitter.”

Musk has every right to ask the question. After all, he is the one who will buy Twitter. But, somehow, it makes you question the action of the man. For someone who is a billionaire (and the richest man in the world), a businessman and a man with all the power to hire a finance team, isn’t it weird for him to ask that ? In short, why would you buy something you don’t have a clear idea about? Or maybe he really knows how things are. So is it just Musk trying to back out of the original deal or is he hoping to renegotiate the offer at a lower price? That’s what people think, especially with the drop in value of Twitter shares over the past week.

So, what is the real reason behind this drama? We can only tell once the offer is fully finalized or withdrawn.