[ryandrake]

I thought the company would collapse after firing 80% of staff, and for a long time I kept arguing that Twitter was still in the "Wile E Coyote ran off the edge of the cliff but didn't fall yet" phase where they were running on pure momentum. I must admit, as time goes by, it's harder and harder to argue that. I just can't believe that 80% of the company was really just not needed. Every company I've ever worked at was lean to the point where they almost couldn't get anything done due to lack of people. I can't imagine a company that could lose 80% of their people and keep on trucking.

[ghshephard]

Twitter’s US revenue has dropped -83% from $661M in Q2 of 2022 to $114M in Q2 of 2024 according to: https://twitter.com/RealNeilC/status/1817562915634819464

That's a pretty dramatic collapse.

The thing about Infra - is that if all you want is 99% uptime - that's, with reasonable architectural decisions - relatively straightforward. You can run with a skeleton crew (particularly if you make really smart Infra Decisions like Midjourney, Whatsapp, others have done an outsource 95%+ of your infra to a third party (Discord, Platform Messaging APIs).

As time goes on though, and you go through incident review after incident review, and sharpen things up - and 99% becomes 99.9% you start to get diminishing returns on more Infra Employees - at some point they don't add much reliability value (but boy do they make pager rotation schedules pretty nice).

My sense (from both interviewing and working with them) is that the vast majority of people fired/laid off from Twitter weren't (for the most part - definitely lots of exceptions) core engineers or core infra-people -they were people on the periphery associated with making Twitter a friendly place for advertisers, and just maintaining a healthy work-life balance for the Infra people - a job where you could work your 30-32/hours week without it becoming all encompassing.

When they were fired, Twitter became a very unfriendly place, and the advertisers ran away, and the revenue crashed.

[hcurtiss]

Where would they get revenue numbers on a private company?

[mikeyouse]

Musk took a bunch of outside money during the buyout so they're still reporting results to e.g. Fidelity and other debtholders. Hence why we broadly know that investors have lost >70% of their money: https://x.com/danprimack/status/1774456271871033823

[wonderwonder]

They just announced an anti-trust lawsuit regarding what they say is a coordinated effort to remove advertisers from the site. Not advocating one way or the other as to the merits of the claim but an interesting development regarding the revenue drop you mentioned.

https://x.com/lindayaX/status/1820838625245880634

[ghshephard]

Companies have to be sensitive to the overton window of their customers. You make a mistake - it can be expensive. Let's ignore X/Twitter for a second - look at what happened to one of ABInBev's brands, Bud Light. They stepped outside that window and got smacked down pretty quickly.

I still don't understand why Musk believes he can dictate to his customers who they should do business with. I can kind of understand regulating what vendors do when interacting, particularly with (for the most part) completely powerless customers caught up in monopolies. But I'm looking forward to digging into the theory of law which suggests that vendors can regulate who/what type of business their customers do.

I've heard of Monosopny's - but it just doesn't feel like there is a "single buyer" in this scenario - and, companies are really, really profit seeking - if there was an opportunity for them to make a lot of money by advertising on Twitter/X, and increasing their revenue, and therefore their stock - I challenge you go find me 1 CFO/VP Marketing in 100 who wouldn't jump at the chance. Their political views would be irrelevant.

The problem is - when all these trust and safety and advertising people were let go -their was nobody left to reassure those CFO/VP Marketing types that something horrible wouldn't happen to their brand on Twitter/X. So they just decided to play it safe until things shook out.

[runako]

This lawsuit is ridiculous. If Twitter can sue companies for not advertising, can Tesla sue people for buying vehicles not made by Tesla? Can Google sue companies for advertising on Bing? Can Rivian sue Tesla buyers for not buying Rivians?

The whole notion that Twitter is owed a share of advertising spend (based on what?) is absurd.

[zdragnar]

Don't forget that 2022 was still a good time for tech with people recovering from COVID lockdowns. It's not really much of a surprise that they're back to pre-COVID income.

Aside from that, there's the lawsuit the sibling mentioned, plus the coordinated campaigns from groups like Media Matters and others attempting to scare advertisers away.

[TrapLord_Rhodo]

Those numbers don't mean anything. Revenue does not equal profit.

Tesla makes double what GM makes with a 1/3 of the revenue. Twitter was always a money loser. If they made $661M in revenue, but lost 700m, and now they make $114M a quarter with $80M in profits, i wouldn't call that a dramatic collapse but rather a dramatic revival.

Quote profit not revenue when it comes to Elon Companies.

[runako]

> now they make $114M a quarter with $80M in profits

> Quote profit not revenue when it comes to Elon Companies.

Notably, the company does not release audited financials anymore. If the company were able to go from breakeven to netting 80% margins, great. But nobody should believe such a turnaround without evidence.

Separately, if the company were netting $320mm annually (using your hypothetical), AND we assign the P/E of best-in-class Meta (which is growing, not shrinking), the company would be worth $8B[1]. Under these generous assumptions, Musk has presided over a $36B (82%) destruction of value in under 2 years.

1 - That's not accounting for the outstanding debt used to finance the deal. Including that makes the value of the enterprise negative.

[Aperocky]

So what's the profit?

Assuming all $114M is profit, Elon dropped what, $43B on this?

Which ever way you look at it, it looks pretty bad.

[Lutger]

Not able to comment on the technical side, but for sure a lot people who got sacked were on moderation. Some countries like mine eventually lost all moderators specifically working for that country. And it does show in the amount of spam an unchecked racism etc, which is a big reason why many advertisers left.

From a moderation point of view, Twitter arguably did collapse. The technical side is not all there is to it when running social media.

[valval]

[flagged]

[psd1]

Bollocks. The word "censorship" has connotations of "suppressing truth". There's no uptick of truth. It's not censorship to flush a turd.

[muddi900]

Not all censorship is bad. For example, I doubt you would like to be exposed to CSAM or ISIS videos when scrolling Twitter.

[dang]

They're paired like this: https://en.wikipedia.org/wiki/Emotive_conjugation.

[orblivion]

Depends on your standards I guess. You could cut it down to one person and have a website running. As it stands if you look at most people's profiles when logged out you see tweets starting a couple years ago other than the pinned tweet. I haven't used it much since a bit before Elon took over but generally as I understand it's more buggy and spammy than before. If that's good enough, I guess you could say it's still trucking.

Funny thing is that I took the opposite side of that bet. I figured Elon would slash things that people thought were important but actually weren't, and make it more efficient. It's mostly played out, except that the site is still rickety. But maybe from a business perspective that's not important, which is a shame for us.

[ljsprague]

Definitely more spammy but I have not found it to be more buggy.

[padthai]

I think Twitter is down 60% in advertisers and 30% in users? Elon personality is part of it, but losing so many people handling community, clients, institutions… I am sure the company has lost most of its institutional knowledge.

[Andrew_nenakhov]

They have recently reported an all time high usage record, that unsurprisingly happened after the assassination attempt on Trump.

[irksomehails]

I wonder if that record takes bots into account because the site is now teeming with bots, even more so than before.

[gopher_space]

> I can't imagine a company that could lose 80% of their people and keep on trucking.

Musk didn't buy Twitter to make money or learn how to run a successful business. "Keep on trucking" isn't what Twitter is supposed to be doing right now. 20% workforce is more than enough to run the operation in maintenance mode, which is exactly what's being asked for.

How many dev-ops roles would it take to just keep the lights on at your org? A dozen? Three? You certainly wouldn't have a need for decision-makers or heavy lifters.

[durandal1]

By "maintenance mode" you mean shipping more features per year than the pre-Elon Twitter? Remember, the pre-Elon Twitter was completely stagnant and seemingly unable to ship even the most desired features.

[WheatMillington]

Musk has openly said many times he wants X to be an "everything app" and sees huge growth in its future. What makes you think he ever intended to put it in "maintenance mode"?

[jokethrowaway]

I thought the same when they fired a bunch of people at my old employer (and 98% of the developers left).

Well, they just outsourced everything to cheap devs in India and things kept rolling. No new features and some new bugs, but most things work.

Turns out you don't really need that much to keep lights on.

[Lutger]

That is true. But you'll:

a) slowly lose to competitors as you can't keep up with increased demands in the space. b) take on more and more existential risks

For a lot of companies, that is exactly what they want to do. Its called the exploit phase, I forgot what business lingo this came from. Do a practical feature freeze, cut costs to the max, and squeeze all the value out the product for as long as it lives. Informally known as enshittification. Its all about cost-cutting rather than market capture.

You can last a while though, especially because there aren't many changes so there's also less operational risk.

[lokar]

This is the Broadcom business model

[Zigurd]

It's hilarious that Broadcom bought what was left of Computer Associates. It's like they needed to absorb that PE-with-a-tech-fig-leaf vibe.

[recursive]

People always find a way to justify their existence. In sufficiently large companies you can find people whose job is to satisfy policies invented by other people in other departments. In one sense, they were all "needed" because of the domino effects of some policy created long ago.

But if you get rid of all of it at the same time, it might be tough to see the difference from the outside.

[albedoa]

Hey man, checking in on this: https://x.com/elonmusk/status/1823152153445404990

[onion2k]

80% was needed to grow the business. If Twitter chooses not to grow them it doesn't need those people. It only needs enough to keep the lights on.

If Twitter ever gets a competitor with some traction it'll be dead in months because it won't be able to react. It seems like new social networks aren't a thing any more though so it's probably quite safe.