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.
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.
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.
From a moderation point of view, Twitter arguably did collapse. The technical side is not all there is to it when running social media.
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.
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.
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.
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.
But if you get rid of all of it at the same time, it might be tough to see the difference from the outside.
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.
2014 1.4
2015 2.2
2016 2.5
2017 2.4
2018 3
2019 3.4
2020 3.7
2021 5
2022 4.4
2023 3.4
2015 304
2016 313
2017 310
2018 298
2019 312
2020 347
2021 362
2022 401
2023 421
That is obviously in decline but does alone not capture how much worse off the company is today than before the acquisition. In 2022, the company did $4.4B in revenue and had negligible debt. In 2023, the company possibly had debt service payments in the range of $1.2B for the year. If you back out the mandatory debt service, 2023 looks more like $2B in topline.
Put another way: Twitter 2022 was significantly more capital-efficient than Twitter 2024.
Similarly, one can run a simple model of the value of the company now. Once you include the $13B in outstanding debt, you will end up with a negative number.
Yes, this business is currently in "collapse."
So... you mean the bot detection is now more broken?
Further, the brand has been tainted and Threads was allowed to pop up. Now threads is around 1/4 to 1/3 the size of Twitter MAU. It may not have replaced Twitter, but the door has been opened and that seems like an unforced error.
This election season looks poised to further drive long term disengagement as the platform is going to be very toxic and very unmoderated.
Otoh, profit might actually be up - if revenues are down 50% but costs down 80%, it may make more money. I suspect, like other private equity investments, this will not work for too long. With how much Musk has put his personal brand onto the site, it may also be difficult to unload the pieces at a profit as per the normal PE playbook.
Main thing I consume on X nowadays is space journalism. Plenty of big name space journalists seriously posting on X (e.g. Jeff Foust, Eric Berger). I had a look at Threads, I can't find any of that content, mostly just people posting silly memes and mind-numbingly misinformed takes on the topic, at best people just reposting stuff that you'll already find better coverage of on X.
> This election season looks poised to further drive long term disengagement as the platform is going to be very toxic and very unmoderated.
I don't find much "toxic" on X. If you only read stuff posted by people you follow, and are selective in who you follow, you can have a pretty curated experience.
Check the actual graphs! It went down by 50% briefly, but then increased again. It's still lower than before, but not by that much.
The brand was never not tainted. Twitter has long been known as one of the cesspools of the Internet, actively contributing to the degradation of the social fabric. It would be a great blessing if Elon did actually kill it the way his detractors predicted. Twitter delenda est.
https://x.com/elonmusk/status/1820849880283107725?s=46&t=alZ...
There's no mystery here. The company cut Trust & Safety so that people could post whatever they want under the guise of "free speech." This means more objectionable (to advertisers) content on the site. Advertisers do not want to run ads next to objectionable content.
Worse, Twitter was never big enough to really matter to big ad spenders. So it's small, doesn't have the ability to move the needle for any big advertiser, and now CMOs/ad agencies have to worry about their ads being run next to 4chan-quality posts? Easy decision to stop spending there. There's really no conspiracy involved here.
At base, this is a simple conflict of ideas. Twitter management wants to create a space where people can post ideas that may be broadly seen as objectionable, even if they are legal. Big advertisers are well-known for restricting their ads to more sanitized spaces (this practice predates the Internet). The goals of Twitter management are directly in conflict with the goals of big advertisers. This was the obvious way for this situation to unfold.