Twitter admits it overstated users for years ahead of Elon Musk takeover | Hacker News Reader

Twitter admits it overstated users for years ahead of Elon Musk takeover

axios.com

113 pointsnimbius111 comments

I looked at the actual data and concluded this is a non-story. Twitter overstated users by <1% each quarter the past 2 years. That is not an egregious accounting error. This happened because they double counted users who used linked accounts.

https://finance.yahoo.com/news/twitter-announces-first-quart...

The headline is also misleading, implying that Twitter deliberately revealed this information in advance of the Musk purchase when in fact this was just part of their normal quarterly reporting.

Did they report this error last quarter also? If not.... then what?

To my knowledge they're revealing this for the first time in this quarterly report.

As for "then what", there's always the possibility of a shareholder lawsuit or Musk using this as a pretext to back out of his offer. That said, the OP indicated the impact on MAU is pretty small (I didn't check this myself), so in the end it's probably all sound and fury signifying nothing.

The comments on this link unfortunately show that many people still only just read the headline before jumping in with their 2c.

Thanks for being one of the few readers that does a bit of fact checking on news - keeping the old HN alive.

Somewhere at Twitter HQ, there are two boxes full of notecards:

1. Things that need to be "fixed" in the code, immediately, before Elon gets here. 2. Things we'll need to publicly admit before Elon gets here.

But a public company has higher public transparency obligations than a private one. Why do you think being afraid of Elon will make them more transparent than being afraid of the SEC?

Cynically, the SEC will only issue fines in most cases. Elon will (according to accounts from friends who have worked at his companies) fire leadership without hesitation and possibly publicly shame them, too.

> and possibly publicly shame them, too

You accidentally hit upon what "free speech" means when used by whales. If Musk berates and bullies an employee with endless racist and sexist slurs, he'd get in trouble. But if he points his mook armies at that same employee and subjects them to the same abuse, it's all good. Just freedom of speech.

> But if he points his mook armies at that same employee and subjects them to the same abuse, it's all good. Just freedom of speech.

What does "points his mock armies at the same employee" mean here? Any examples?

https://www.wsj.com/amp/articles/elon-musk-calls-out-twitter...

The WSJ has been busy: https://twitter.com/robbysoave/status/1519407686063988736

Yeah, like when Politico recently pointed out the FDA was told about infected baby formula months before several babies were killed by it.

Obviously Politico is just gathering their mooks to subject the poor gentle souls of Abbott and the FDA to abuse.

We can't disclose wrongdoing - think of the mean things people might say.

This is missing the point so badly it seems almost intentional. Public outcry at the actions of the government is entirely different from incited bullying of individuals with no public presence.

"incited bullying of individuals with no public presence"

Elon Musk never even mentioned her - he simply referred to Twitter's actions as "obviously incredibly inappropriate" and linked to an article that happened to mention her.

> "bullying of individuals with no public presence"

You can't be serious. This person is literally responsible for burying the Hunter Biden laptop story and is personally responsible for banning Trump. That's a public position whether you like it or not. I don't think anyone would think we can stop criticizing Putin the moment he exits office.

Imagine if Jack Dorsey ran Twitter without a Twitter account. That would not negate your ability to criticize his leadership even if he never spoke.

She has a pretty big public presence. She was on the Joe Rogan podcast with Jack. Calling someone out for being terrible never seemed to bother anyone until apparently 5 seconds ago.

Huh? Punching down has never been considered classy.

1. This "employee" is in a very high position and makes $17 million/year. 2. She is in charge of Twitter's censorship team. She can send an email and block and ban every single account that tweeted something abusive at her.

I think she can handle some criticism.

This "employee" is also the very person who chose to A.) Block President Trump and B.) Ban the story about Hunter Bidens's laptop from the New York Post, which was declared real by the NYT two years later.

Furthermore, Elon did not berate her by name or handle, he just said that what Twitter did was "obviously incredibly inappropriate" with a link to a story mentioning her, even though he did not mention her by name.

She's calling it abusive, not because Elon called her names (he never mentioned her, just the story) even though she was behind the biggest bans and censorship. She's upset that her power has been exposed and she might have some accountability for her decisions.

For how powerful she is, imagine if we weren't allowed to criticize the President, or a member of Congress, or Jeff Bezos, or even Elon Musk himself because it might hurt their feelings. When you are in a position of that much power, criticism must come with it, and you don't get to complain about that.

Yes, this is totally bizarre. You have enormously powerful people shaping the public discourse not only crying foul at idea they might face the slightest bit of accountability, but also successfully lobbying other people to defend them from any public criticism. Then we have to listen to them sanctimoniously talk about how "disinformation" threatens democracy, or whatever! Faceless Twitter employees shaping political discourse during an election get a free pass, though?

Meanwhile, these same people have no problem with using their power to shape and weaponize discourse against other people! For example, the news media regularly unmasks regular Twitter users simply for having a huge following with non-media approved opinions and then shrugs when they face death threats. But...don't you dare talk about the journalist that did it!

It should be noted that trump had just attempted a coup against the United States when he was blocked. Preventing him access to his bullhorn was and is a good idea.

It should be noted that trump "attempted a coup" without any weapons. It should be noted the only deaths on January 6th were trump supporters. It should be noted that trump's last tweet that got him banned was "let your voices be heard, and go home in peace". I should start watching Netflix and marvel comic movies again so I can exist in your reality.

Attempted murder is a crime. Failing at your attempt to overthrow the government because you tried doing it without weapons does not shield you from consequence. I'm not even talking only legal consequence here; it is completely reasonable to look at what Trump tried to do and refuse to engage with him.

If that were true, someone would have been charged by now, and I'm not talking about the desperate misdemeanor charges they've been pulling out of their asses to save face.

Vijaya committed a federal felony of election interference and should be arrested.

I find it interesting that this comment has flipflopped gray and black. It seems reasonable enough?

It is never a good sign when a leader holds others to standards they themselves aren't held to.

Counterargument: If you hire someone make a painting of your dog, would you hold their work to a higher standard than if you had done the work yourself?

I expect leaders to lead by example and to not punish others for behaving the same way leadership does. Leaders trashing workers isn't them doing the job they are paid to do.

They are worried about the lawsuits that would be filed against the board and the executive if they find fraud after the deal closes.

... which can become things that Elon can use to sue and break the deal

That is a nice thought, however twitter locked down the codebase to avoid disgruntled employees from vandalizing the site / planting backdoors: https://techcrunch.com/2022/04/26/twitter-lock-source-code-e...

This does raise a question - what are all those twitter engineers are doing?

I imagine they are still writing code, just not shipping it. And there’s an exception for fixing things, which this would definitely fall under

When the company I worked for was in the process of being acquired, our codebase was frozen as well. I'm pretty sure this is just standard practice, no?

My previous experience of this was:

1. No user facing feature releases for the time being the deal being agreed and becoming effective (plenty of backend systems, performance optimisations, etc. still got released)

2. Don't touch anything on the day of the deal, in fact, take a free day off.

3. No major feature removals for 3 months after the deal.

Presumably you could still apply some changes. If there was a critical vulnerability Twitter could make the changes necessary. If those changes are allowed, maybe not so critical changes could be made as well.

The only question is raises is how incompetent is this journalist. the source code is locked. It's not everyone is committing stuff...

What I'm learning from this how little anyone posting shit on the internet actually knows. Misinformation from incompetence or just trying to be a know it all everywhere.

So please explain to us how it does work!

i don't really have an incentive to do that, i just wanted to comment on how inaccurate this was written and share a thought about our "news" based off my knowledge. The author shouldn't have written this if they don't understand what they're writing about. Or cant verify their leaked information.

Probably most news is misinformed if you think about.

Code freeze and 'legal documents freeze' are 2 different. There is no way to prove that any traces of biased decisions taken over the past few years were gotten rid of immediately after those decisions were taken.

I am curious what all those twitter engineers are doing as well. Presumably it's secret in some way?

But who is guarding the guards?

Whatever engineers do when their company is acquired in such a fashion: Loitering around the cafeterias and coolers after sending their resumes and scheduling their interviews.

Look at any FAANG/MAANA and ask the same question.

If I had to guess, most of them have so much cashflow that mid and senior level managers are able to do engage in organizational silo-ing by hiring a large amount of expensive engineering talent without much pushback.

If this is the case, there's likely a large pool of talented and highly compensated employees doing work that doesn't touch on either a profit center or high-impact research & development. And what work is being done is likely influenced by Parkinson's Law.

I wonder how many people are going to be arrested at Twitter for all the federal crimes they've committed over the years?

Twitter needs to get smaller to get better, specifically they need only real people (no bots). While this usage misstatement is minor it points to a longstanding directional problem whereby they focused too much on vanity metrics. I'd argue this a reason why they didn't do a better job challenged all the fake/bot accounts because those are great at pumping up vanity metrics.

I'm still waiting for an explanation on why liberal accounts have lost thousands followers but conservative accounts have gained thousands.

I'm fairly liberal myself, but seeing what is clear shadow banning/suppression especially of politicians is borderline voting interference and should be punished heavily at both the corporate and personnel level.

Do you have any evidence or examples of this?

I don't know one way or another but Twitter and Twitter users strike me far more as liberal heavy.

On the conservative side:

https://twitter.com/SethDillon/status/1519688581261967366

Not evidence but people claiming this happened:

https://twitter.com/cbouzy/status/1518953731961802752

https://twitter.com/DonaldJTrumpJr/status/151900201165147751...

https://twitter.com/MZHemingway/status/1518946715407106048

I already got downvoted for posting this yesterday, lol.

I agree with the other reply. Unless you have evidence to support this, it's just a conspiracy theory.

Can someone explain the conspiracy theory? I don't get what it implies...that twitter was overstating liberal followers but understating conservative followers in a way to juice liberal tweets? That thumb on the scale was then removed before Musk took over? Seems pointless and much easier to achieve through other means.

Ultimately this nonsense just confirms to me that Musk is right and twitter's algorithm needs to be open source and the rules surrounding banning people need to be clear to ensure trust.

>Conservative Twitter accounts got boost in followers after Musk acquisition, data shows

https://www.theverge.com/2022/4/27/23045005/conservative-twi...

>This repository contains data for The Verge's article

https://github.com/corintxt/post-elon-twitter-follows

Why is this such a mystery to people like you?

The biggest news item of the past week has been Elon buying Twitter.

Prominent conservative accounts had followers go up by ~1%

Prominent liberal accounts had followers go down by ~1%

People who don't like the idea of Musk owning Twitter left (though probably temporarily) and people who previously left due to Twitter's policies returned because they read the news and like the idea of Musk owning it.

>People who don't like the idea of Musk owning Twitter left

really? thousands of people just left twitter at the idea of Musk owning it, and thousands joined out of love for Musk? within a week?

Maybe rhetoric like this is why it's a mystery.

Thousands are a drop in the bucket when you talk about a site with user counts in the 200 million range. I can totally belive 0.01% of Twitter's userbase (20000 users) made kneejerk decisions based on the ownership.

Yes, that is exactly what happened. It's clear and you can see by the likes of people such as Jameela Jamil leaving explicitly because of Elon buying it.

If you are on twitter, and follow people on the left, a big topic right now is whether people should leave as a direct result of the Elon purchase.

> It's clear and you can see by the likes of people such as Jameela Jamil leaving

Its incredibly obviously clear, given the thing you say has happened, hasnt happened. [0]

[0] https://twitter.com/jameelajamil

Did you read the tweets on that twitter? Here, I'll help you out:

https://twitter.com/jameelajamil/status/1518641966661177344

If someone makes a twitter post, saying that they "will FINALLY leave", I think it is fair to describe that person as being in the process of "leaving".

Thus, that is fair evidence that yes, at least people are talking about leaving twitter over the elon purchase, and that this might effect user stats.

It is also proof that many people say they will do something on their twitter account and their twitter account remains.

I have presented conclusive and definative proof that they have not left twitter at hte time of this posting. Do you have proof that they are "in the process of leaving twitter"?

The thing is that it's very hard to prove that someone left twitter if you only define leaving twitter as "accounts being deleted"... as their account is deleted. That makes it something hard to prove, and if that is the only way that you will believe that anyone has left twitter there's no way I'm going to spend that amount of time digging up previously active accounts that have since been deleted

The topic in question is why some accounts have lost followers - the explanation is people have left twitter - the "proof" is jameelajamil left twitter.

We have established the account is still there - we agree on that. They tweeted "they were leaving" - we agree on that.

However, they have not deleted their account and have not unfollowed everyone they previously followed [0]. So if this is the example of "what is happening" how does that action (saying you are leaving, not deleting your account and not unfollowing anyone) lead to accounts have lost thousands followers

[0] https://socialblade.com/twitter/user/jameelajamil

Like I have said, I am not going to invest time into figuring out a way to find previously active twitter accounts that were deleted in the past ~4 days to have definite proof. I don't care enough, and it seems like it would take a time investment to figure out how to do that that a silly internet debate does not merit.

I am asking you to take the step from "Many people are saying they will quit twitter and have stopped posting" -> "It is reasonable to assume that there are some amount of people who have quit twitter and deleted their account". If you are unwilling to make that conclusion, they by all means go ahead; but I think you are being unreasonable to not make that relatively safe assumption

> Do you have proof that they are "in the process of leaving twitter"?

The point being here, that you are being pedantic, and borderline bad faith.

It is perfectly reasonable for someone to say "Hey, there is this person who is leaving twitter", in reference to a post of them saying that they are doing that.

And you are coming in, saying "Well, actually! Technically speaking, I don't see a deleted twitter account yet, therefore by the laws of debate land, you are wrong!" just comes off as you being a pedantic idiot.

The original interpretation, is a reasonable one, for regular human beings to reasonable agree is mostly true. And pedantic "well actually" people, like yourself, are contributing nothing to the conversation, and just trying to get a 1 off, bad faith gotcha on a technicality.

Stop. Nobody likes you when you do that, and nobody is fooled by your behavior.

Well I for one am glad Musk is taking over and putting horrible little people like you, who want to censor others in their place. Finally speech will be free again.

Please feel free to take your leave twitter, move to canada or whatever else you snowflakes like to do.

https://twitter.com/jameelajamil/status/1518676806261477386

I'm using her declaration as a high-profile example of someone deciding to leave twitter simply due to Elon's purchase.

She hasn't deleted her account, most likely because she doesn't want to let someone else get her account name and impersonate her or something similar, but for non-celebrities it's really safe to assume that many people have deleted their accounts as a result of elon; and that many have created accounts as a result of elon; and their choices on who to follow will closely correlate with their politics.

Can you provide any evidence that it's a grander conspiracy like you alleged?

> Can you provide any evidence that it's a grander conspiracy like you alleged?

I alleged absolutely nothing of the kind. I simply pointed out that X person leaving twitter had not actually occured.

I could, use your cited example to point out that, while many people say they would leave twitter, like those that say they will leave the country if the other guy becomes president, they havent. As proven, obviously, by whoever this jameelajamil person is.

Ok, fair enough I thought you were on the side of the commenter who made this comment:

> I'm still waiting for an explanation on why liberal accounts have lost thousands followers but conservative accounts have gained thousands.

> I'm fairly liberal myself, but seeing what is clear shadow banning/suppression especially of politicians is borderline voting interference and should be punished heavily at both the corporate and personnel level.

> really? thousands of people just left twitter at the idea of Musk owning it

I know a fair few people said they would, for a few different reasons. Particularly as, given the size of the user-base to start with, “thousands” is a pretty small number.

> and thousands joined out of love for Musk?

More likely because of the “freedom of speech” promises that were being made. I'd wager that many right leaning people who had previously left (or not joined in the first place) over certain other people being kicked, off are quite enamoured by the idea that their brand of rhetoric might be less unwelcome there in future. The converse (wanting that sort of talk to continue to be frowned upon) is one of the reasons some others talked about leaving. Certain messaging is key, not Musk specifically (although it was him giving off those messages).

Yes, it's hard to believe that twitters users can be emotional and impulsive drama kings

FWIW, media organizations seem absolutely giddy to be able to run articles about how to delete a Twitter account. All of these are from the past week:

https://www.cnet.com/tech/how-to-delete-your-twitter-account...

https://www.washingtonpost.com/technology/2022/04/26/twitter...

https://www.mentalfloss.com/posts/how-to-delete-twitter

https://www.independent.co.uk/tech/how-to-delete-twitter-acc...

https://www.newsweek.com/how-delete-your-twitter-account-170...

https://www.express.co.uk/life-style/science-technology/1601...

https://www.pcmag.com/how-to/delete-twitter-and-your-terribl...

https://9to5google.com/2022/04/27/how-to-delete-twitter-andr...

https://www.marketwatch.com/story/how-to-delete-your-twitter...

https://finance.yahoo.com/video/delete-twitter-account-15483...

https://blog.mozilla.org/en/internet-culture/how-to-delete-t...

https://www.thestreet.com/investing/how-to-delete-twitter

https://www.pcworld.com/article/695148/how-to-delete-your-tw...

This reminds me of that massive and coordinated push to get people to delete Facebook many months ago. I have no idea if it actually worked but most people I know wound up getting hooked on TikTok around the same time, which is arguably even worse.

Twitter has ~200m daily active users. I can't find recent churn numbers, but older reports suggest it is quite high[1].

Generally, churn would not have much of an ideological bias. On average people would leave for personal reasons and so you get an even-ish spread. However, if the major events of the day do have an ideological character - you would expect those events to have an outsized impact in one direction or another on people along those ideological lines.

So on a normal week you'd see both liberal and conservative accounts get +500 and -500 followers, but on this week the liberal gets -1000 and the conservative gets +1000. Same churn, but driven by what is on peoples' minds.

[1] https://www.vox.com/2017/10/26/16552368/twitter-new-user-gro...

I've seen conservative acquaintances (IRL, not just people I "know" online) posting about how excited they are to use Twitter again/more after Musk takes over, so yeah, seems plausible.

[EDIT] Meanwhile, I've also seen people on the left (though not any I personally know—not many people I know, tweet) announcing that they'll be reducing their Twitter use, in anticipation of what they think Musk will do to it.

If you go on reddit there are hordes of people announcing their departure.

I'm not sure what the alternative hypothesis is. That angry twitter employees are implementing changes in advance of the Musk purchase?

The (conspiracy?) theory I've seen is that Twitter was secretly modifying the displayed counts (inflating for some, deflating for others), and now are getting rid of that manipulation so that Musk doesn't see it when he comes in.

What's with the "people like you"?

In context, it reads as "people who peddle this conspiracy theory".

The parent comment had "seeing what is clear shadow banning/suppression especially of politicians ..."

The response, I think, is trying to group the other people who remain belligerently insistent that this must be some shadowy conspiracy, not just random shifts in usage, or a reaction to news, or something else we don't know.

I started following a bunch of right-wingish accounts just because it was aired that they had been banned so I wanted to see what they were all about for a while. OTOH the Musk hysteria became annoying so I unfollowed a lot of blue checks and leftish people.

For this to be true I think you have to make the argument that people went in and deleted their accounts, rather than just deleting the app on their phone. The former involves a couple more steps and the latter is pretty costless and easy. Same logic applies to the follower gains, but I think that one is easier to make since the people who like the idea of Musk owning twitter may actually be new users.

I don't know, I don't work at twitter.

I understand conservatives gaining followers. It was in the news, conservatives who would otherwise not install Twitter now installed the Twitter app and created the account.

But you're saying that the same amount of liberal users didn't just delete the app, like you expect every normal non-techie user to do. They actively went through the trouble of deleting the account hidden behind dark pattern?

Not buying that, sorry.

> I'm still waiting for an explanation on why liberal accounts have lost thousands followers but conservative accounts have gained thousands.

Because one side has had lots of people pushing a “quit twitter” movement (to the extent that voices on that side, or marketing to them, who depend on Twitter have been posting desperate pleas for people not to leave) in reaction to the Musk takeover, and one has been rushing in because they perceive a more favorable environment?

> why liberal accounts have lost thousands followers but conservative accounts have gained thousands

Since what reference time? Any explanation for since Musk took over will differ from one about, for example, the start of this year, or the last, etc.

Why do you conclude shadow banning and suppression?maybe libs are more likely to leave the flamebait social media platforms? I did.

Well, personally I'm still waiting for an explanation for why George W. Bush Sr. and Bill Clinton were both on the grassy knoll with rifles on that fateful day, but none is forthcoming.

> George W. Bush Sr.

George Bush Sr. is George H.W. Bush. There is no George W. Bush Sr. (at least, not a famous one that was president of the US.)

That’s what “they” want you to think …

Even better for the point being made! ;D

Alternatively, "I'm still waiting for an explanation for why he goes by George H W Bush when his birth certificate reads George W Bush Sr"

They probably shouldn't show "follower counts" publicly.

> Twitter also said it accidentally over-counted the number of monetizable daily active users because of a feature that allowed people to link multiple separate accounts together in order to conveniently switch between them. It counted those two separate accounts as two users for more than three years.

Feels like they can be technically two users. Is there a strict regulation how they should be counted for tech giants like Twitter? definition for

There is no legal definition of MAU. However the industry standard is that it represents total number of customers, so it wouldn't make sense to count one person twice if they have two accounts, because they are still just one customer.

often they are agents that manage social media for different businesses. they should be counted as separate

They are still one set of eyeballs and one brain to advertise to. Since the metric is mainly for the purpose of "how many eyeballs might see these ads" it still makes more sense to count them as one.

Feels like a quirk comes from the fact that a lot of prominent pseudonymous Twitterers, that they make accounts per communities and go extra miles to dis-link each ones, especially from real life identities.

Tangential, but twitter's customers aren't their users, it's their advertisers. The users are the product.

As a guy who follows sports it's really baffling to me how with politics there is not one iota of self-criticism among the political party voters and especially among single issue voters.

Like, sure maybe Twitter and all social media are part of a giant conspiracy, but had Republicans presented Ted Cruz or Marco Rubio as the nominee in 2016, there are 70% chances are they'd still be in the WH.

In sports you first focus on the low hanging fruit which is the stuff that you can control, this is why Trump's friend and legendary coach Bill Bellichick has 8 SuperBowl rings.

If you are at the SuperBowl and field a QB which lasts only one half then having referees which penalize you becomes much less relevant.

> it's really baffling to me how with politics there is not one iota of self-criticism among the political party voters and especially among single issue voters.

Uh, there's a whole lot of internal, including self, criticism within political parties.

There are strategic reasons much (but very much not all) of that is deliberately done in a way to be outside of the public eye.

When HRC lost I maybe heard 5% of Dems voters saying that she was a bad choice.

When Trump lost 0% of Republicans voters admitted that they fielded a QB which lasted one half in the SuperBowl.

It's all about referee blaming and accusing the other team of deflating footballs, not one iota of self-criticism.

I can't help but be impressed by the remarkable hoops people will jump through to make everything about US politics.

Here we have a simple story about Twitter's financials, but nope, apparently we gotta debate US political party structure...

>had Republicans presented Ted Cruz or Marco Rubio as the nominee in 2016, there are 70% chances are they'd still be in the WH

Cruz or Rubio might have been like every other Republican candidate since 1988, in being unable to win the midwestern swing states PA, MI, and WI that Trump won for the first time in three decades.

When due diligence is done then these lies about user count would have been exposed so they got to come clean about this "accident" now.

Curious if Musk can sue them for a better price. This could get Twitter in some regulatory trouble...

He can still back out of the deal. This is part of the due diligence. I suspect it's coming out now because they know that Musk will reveal it if he finds out through due diligence.

I don't think so. I've read that he doesn't have a due diligence provision in the merger agreement.

And, I don't see any such provision in the agreement, which I _think_ is here: https://www.sec.gov/ix?doc=/Archives/edgar/data/0001418091/0...

He could back out paying the $1 billion breakup fee, and I suppose he might have some cause of action if there were serious problems with the reps and warranties. But IANAL.

>He can still back out of the deal.

That would cost him (or Twitter) $1B.

I assume Mr. Musk waived his right to any due diligence as part of his agreement with the Twitter board.

It's a public company so all the information should be there, due diligence isn't really a thing in this case.

It's more the aspect that Twitter likely committed a financial crime, defrauding shareholders and the acquiring party.

Accidentally misreporting linked users - less than 1% ?

If they were financial accounting errors, maybe. But for an audience measure that’s just a rounding error.

Just like reddit did before their bait & switch.

"accidentally"

Reply on news.ycombinator.com