If Twitter went another month without an outage, how would you adjust your opinion? How about a year?
The car analogy is amusing, but how much does it really hold up? Have we ever seen another major social media company drop this much of its staff in one go? I certainly can’t think of an example. I think we’re in somewhat uncharted waters here.
A driverless car won’t last long, we know that for a fact. I think it remains to be seen how long a bloatless twitter can last. I’m personally optimistic.
> I think it remains to be seen how long a bloatless twitter can last.
It's also really hard to define 'last' though. Does 'last' mean just for up-time? Does it mean up-time without a major security incident while maintaining the same DAU? Does it mean business as usual on all fronts except number of employees? We know that Twitter already had some security issues with their God Mode admin panel.
I really wonder, for example, whether no angry ex-employees still have access to critical systems or data. This is usually pretty well regulated in large organizations like twitter, but since they've lost the majority of their staff, who knows when the people looking after that left?
> I think it remains to be seen how long a bloatless twitter can last.
I'm not convinced Twitter had a ton of bloat. (Most of the teams actually involved don't seem to think so). Just because Elon can't understand something, doesn't make that thing "bloat".
Twitter definitely had a few weird features that could be cut (the audio podcasting thing, for example). But calling most of Twitter microservices "bloat" is about as dumb as calling a cars Seatbelt and Airbag and Crumple Zones and that spare tire in the trunk "bloat" -- it's only "bloat" if you assume all people will always be perfect and no one will ever make a mistake anywhere, and nothing bad will ever happen.
I think it's well understood that when people refer to Twitter's "bloat" they're talking about its bloat of extra employees, not its extra microservices.
I've told this example before, but my close friend and roommate did a year at Twitter in 2019. He was tasked with implementing versions of simple, relaxing JS games like Tetris which would be used by Twitter's moderation staff when they felt they had accumulated too much stress and needed a break. He was making $300k USD to copy an open-source version of Tetris to be played by paid staff! This was actually one of his more engaging projects, most days he said that he wrote 0 code at all. He got great reviews and was being encouraged by his manager to pursue the senior track; from what I could observe from the outside, there was a complete misalignment of goals at pretty much all levels.
Perhaps we disagree which is totally fine, but I think this type of allocation of eng resources absolutely counts as "bloat". You may see that work as comparable to a seatbelt or crumple zone, but I personally see it as more comparable to all the other expensive, useless nonsense which plagues modern cars.
This is an interesting example but I have to wonder if it may inadvertently highlight a disconnect between the parties involved.
I'm no expert in this area but many stories have surfaced over the years regarding high levels of stress and (in extreme cases) meaningful declines in the mental health of social media content moderators.
I recently encountered a twitter thread from an early LiveJournal employee (1) @rahaeli "about Trust & Safety work, the toll it takes on you, the things you see, and the human misery, suffering, and death that happens when you fuck it up, including murder and child sex abuse." and I am inclined to believe that, while relatively simple and un-engaging, managers and end-users likely would have found it important.
> He was tasked with implementing versions of simple, relaxing JS games like Tetris which would be used by Twitter's moderation staff when they felt they had accumulated too much stress and needed a break.
Oh my god, this is incredible. Thank you for this story.
I have several college friends (developers) that ended up at twitter. Some of them loved it, some of them hated it, one universal thing I pretty much heard from all of them is that they typically did actual work 15-20 hours week.
Now overall I wouldn't say this is unusual in tech. I know plenty of people at other major tech companies, including other FAANGs, that say the exact same thing. But just because its appears to be a industry wide issue does not make it okay. It seems to me like that this would signify bloat. You only have enough work for your developers to be working half time.
Whats the problem with working only half of the time? Work is also a social enviroment and the «not working » part of work a realy important part. If the staff would work 100% of the worktime they would be fucked. Most people I know prefere a good working condition than working ultra hard for some sadistic asholes.
I don't even know where to begin with this comment? So you have no problem being paid for only half the time right? Or are you saying you deserve to steal from the company you work for?
There are a lot of assumptions baked into your position.
1. The employer pays for your time, not your expertise or output.
It'd only be stealing from the company if the company cares about hours worked over output. If we explore this concept in a theoretical sense it's clear that it doesn't hold up.
You have two candidates
One candidate has 20+ years of experience doing the exact thing you want them doing. This candidate says they'll work for you for $100k/yr, and they'll work 10 hours a week, complete all the relevant tasks, and very very rarely cause catastrophic errors or user-impacting bugs.
The other candidate is fresh out of school and says they'll work 50 hours a week. They'll complete the same amount of work as the first candidate, but they'll write more bugs, there will be more planning mistakes causing feature delays, and there's a reasonable chance of catastrophic failure due to debugging-in-prod shenanigans. They are also asking for $100k/yr.
Which candidate is better? Under the assumption that employers pay for _time_, the second candidate is better, but I'd argue most companies should prefer the first candidate.
2. More hours worked produces more or higher quality output.
There's a reasonable amount of research and practical anecdotes that disputes this recently (see companies that have gone to 32-hour 4-day workweeks with no reduction in productivity). Enough that at least, this point is seriously in doubt.
3. Twitter maybe pays a senior engineer $400k/yr under the expectation of their output for 40 hours, and if they get less it wasn't a fair deal.
This is a reasonable take, but Twitter (like most for-profit companies) theoretically has a performance evaluation system, managers, deadlines, etc. They're paying an engineer some amount of money for some amount of output. If that engineer produces that amount of output, Twitter is happy, the engineer is happy, there's no issue. If the engineer working 20 hours or less per week caused them to not meet their goals, then Twitter has the right to fire that employee. They don't, so that implies that they're happy with the arrangement.
4. The employee's salary is equal to their expected output/profit
If the Labor Theory of Value is correct, then companies derive their profits almost exclusively from the labor of their employees.
In order for an employee to "steal" from a company by under-producing work, they would have to earn more in salary + benefits than they earn the company from their work.
This is necessarily not the case (on average) in a for-profit company, because if the company makes a profit and uses those profits to grow or to issue dividends to shareholders, they have earned "_surplus value_" from the employees' labor (on average).
---
In any case, it's not necessarily true that you're wrong, but your comment was fairly dismissive and confrontational. There are a _lot_ of cultural and individual assumptions baked into how we exchange salary/wages for labor, and it's worth examining those before firing off moral judgments at one another for not working hard enough or working too much or whatever.
A lot of what you are saying is theoretical but doesn't hold up in the actual work place (at least in the US).
1. In the US at least we know a majority of the workers are hourly. (https://www.forbes.com/sites/johncaplan/2021/03/12/americas-...) So I would absolutely argue those people are paid for their time. I understand the scenario you laid out, but in a big corporation I do not think they would look at it like you are. They would simple look at who is costing the most. Both are costing 100k/yr so compare their output. Anytime their are massive layoffs, lots of important people are let go. This is because often times the people doing the firing, do not know the employees. They are simply lines on a spreadsheet to them. So in your situation above after a mass firing often times the candidate fresh out of school will be the one left.
2. I personally completely agree with you here. From my experience there is absolutely a burn out point. However major corporations do not see it this way at all. They absolutely believe throwing hours at problems brings about solutions. It doesn't matter to them if its making existing employees work 4 extra hours a day or hiring a new employee. It matters which is cheaper in the end. If the position is salaried its cheaper to have existing employees work more. Look at what musk is asking at twitter. If it is hourly, often times its easier to part ways with the burned out workers and hiring new ones. Look at amazons turn over rates.
3. I agree that Twitter gets to decide the expected compensation for work. But musk now owns twitter. He gets to decided what twitter does and doesn't expect from its employees. He made it clear he is very unhappy with how the arrangement was, and what he expects in the future.
4. I would disagree with you here. An employee can steal from a company many ways. For example, an employee could steal the source code from some twitter service that is considered a company secret. In this case, Twitter is paying an employee x amount for y work. Musk has decided what y work is. If an employee decides to not fulfill y work, and still take there whole pay, how is that not a form of stealing? If they're hourly employees they call it timesheet fraud. The idea that an employee can only steal if the are paid more than they bring in is pretty interesting but I think we would be hard pressed to find a single major corporation that views it that way. I think close to 100% of them would go after an employee spending 50% of their time working on non work related tasks.
A vast majority of the assumptions are based on them being the norm in corporate America which twitter is apart of.
> In the US at least we know a majority of the workers are hourly.
We're talking about IT employees. Unless someone was a contractor, most IT employees (white collar workers in general) are exempt, and not tracked or paid hourly.
I guess my question is - what are they doing for the rest of the week? Are they straight up just not doing anything, or just not writing code? I honestly don't think I could be writing code 40 hours a week, but I think that the code I do write when I'm writing it is the better for my taking the rest or zone out time. I also don't know if my work would be of a decent quality if I was suddenly asked to be typing all 40 hours a week.
Cannot speak to all of them but I know a couple work on sides personal projects the rest of their time. Doesn't matter if they are in the office or at home.
No one is asking them to write code 40 hours a week. That is not realistic in coding. Most developers' jobs include a lot more than coding. They are saying they spend 15-20 hour a week on all job related work. Whether that is coding, code review, engineering, testing, research, documentation, meetings ect.
maybe the thinking is tear it down to just bare bones and life support and then rebuild? Although, and I say this as a semi-fan, it's really exhausting trying to understand Musk heh.
Company soft killed the product, everybody left, they didn't hire anyone to replace. We went from 20 engs to 3. Worst codebase ever made by ex FANGs hotshots who thought they understood something about system architecture. Very "clever" and complicated. Data consistency issues happening everyday, likely due to misuse of messagging queues. Chargebacks being ignored and mailed in physical letters every month. A couple of millions going through the platform every year.
My task was to run a team of mostly juniors maintaining and adding features to that mess.
I had no clue what that codebase was doing. We just left things as they were, fixing fires as they came. Nothing too bad happened.
Slowly built a leaner replacement for some components. We simplified things over time and we even rebuilt some of the knowledge of the old platform, which helped with the daily outages.
The issues started happening not as often. Eventually. I moved on from that company, removing again a big chunk of knowledge. Over time I've heard tales of other people coming in and rebuilding that knowledge, over and over.
> Worst codebase ever made by ex FANGs hotshots who thought they understood something about system architecture. Very "clever" and complicated.
I call these sort of folks (very) smart juniors. Probably can whip leetcode or whiteboard tests like few others (after some preparation). Then you let them roam wild on your product, because they're of course experts. Then some period passes, and/or they leave, and you can only cry. Complex hard-to-grok approaches to simple problems that have tons of caveats/edge cases, no documentation of work done and why it was done as it was, because they are oh-so-cool and such lowly tasks are for peasants.
Either they have no clue how long term sustainable company looks like or they don't care, in any case not a good fit.
These days I go in opposite direction as probably most here - FAANG type of company (or cargo-culted startup) is a big fat warning sign when hiring. Unless I would be working at some wannabe another FAANG startup, which I am not, I would sure as hell make sure the person can actually deliver long term improvements for everybody and not just upp their CV with another bleeding edge technology and move, leaving more damage than added value and making everything worse.
> If Twitter went another month without an outage, how would you adjust your opinion? How about a year?
It's a tricky one, because on one hand it increase my trust that their system was built robustly, but at the same time the passage of time would increase the chance of unseen/unaddressed "wear an tear" (bot figurative and literal) that might be going unaddressed, or under-addressed. But we have no view into that.
We won't really know until they suffer a major problem whether or not they have enough staff yet to keep sufficient maintenance going that such an event doesn't cascade into something much worse and/or whether or not they will be able to recover from it in a reasonable amount of time.
Horrible systems can survive, but often they survive through sheer luck.
Beyond merely surviving, Twitter has to compete with other companies and enhance their product over time.
What happens if someone like Bytedance decides to launch a Twitter clone that does everything better?
Meta takes a lot of criticism for being Meta but you have to hand it to them, they've copied every other innovative social media feature that has shown up in competitors' products.
Does Twitter have enough staff to launch a major new feature?
The problem is that Twitter is now saddled with debt. It probably can't make investments in new products, and we can speculate whether it has the necessary headcount in the sales staff needed to maintain advertiser relationships that pay 90% of its revenue. [1, 2] Enterprise clients usually expect a dedicated CSM relationship.
That's exactly what happened to Toys R Us, which slid into liquidation due to the debt saddled on it by its private equity firm rather than a fundamental issue with its business.
In 2021, Twitter had a net loss of $221.4 million, and in 2020 they lost over $1 billion. Revenue was around $5 billion in 2021, about 90% from advertisers. [1]
The company (not Elon) is now loaded with $13 billion in debt. [2]
> Last year, Twitter’s interest expense was about $50 million. With the new debt taken on in the deal, that will now balloon to about $1 billion a year. Yet the company’s operations last year generated about $630 million in cash flow to meet its financial obligations.
>
> That means that Twitter is generating less money per year than what it owes its lenders. The company also does not appear to have a lot of extra cash on hand. While it had about $6 billion in cash before Mr. Musk’s buyout, a large portion of that probably went into the cost of closing the acquisition.
>
> That gives Mr. Musk little wiggle room, Mr. Pascarella said. “They are essentially going to take all the financial resources of the company and just pour it into servicing the debt,” he said.
I personally don't see how Twitter gets out of this without bankruptcy. I think Twitter can go the next 5 years without an infrastructure outage and still likely ends up bankrupt.
ByteDance I could see, but Meta and Google are hamstrung by the innovator’s dilemma to a degree: Both companies have the money and the technical smarts to pull it off, but their management structures would wreck it immediately because it poses a threat to other internal empires.
My loose guesstimate is that Elon has cut $2-3bn in expenses so far. I don't know the revenue hit yet, but $1bn in interest expense with $2.6bn in cash flows seems entirely sustainable
I think the income statement shows that it would be difficult for Twitter to find $2-3 billion of expenses to cut and still remain a competitive, functional company. You're talking about cutting total expenses by over 50%, including expenses that are not salary.
2021 total expenses of $4.8 billion on a revenue of $5 billion. $1.8 billion cost of revenue, $3 billion operating expense. Out of the operating expense, $1.175 billion is in Selling & Marketing Expense. $1.25 billion in R&D.
Let's say Twitter cuts R&D to $0, that's a $1.25 billion savings. How long can a social media company remain competitive putting $0 into R&D? Are there any examples of any software-adjacent company surviving that spends $0 on R&D?
If they cut SG&A, that will impact revenue negatively. Activities like marketing and sales have an ROI. You spend money to make money. Twitter was spending $600 million a year on SG&A in 2014 when they had less than half their current level of daily active users (DAUs). Twitter has a product, its daily active users. If they can't sell that product to its customer (advertisers) because it doesn't have enough sales staff to physically make the required phone calls (yes, they do that sort of thing with large advertisers), they risk entering the death spiral.
Cost of Revenue: I don't think this line item can be cut beyond a certain level. This is the direct cost of delivering the product. Anything that isn't employee salary can't be cut very easily. Twitter can't turn off servers and sell off data centers without impacting the product.
Twitter had more than 7,500 employees in 2021.
If they cut 80% of staff, that's 5,000 employees gone (and don't forget that those cut employees will still count as 1/4 of an employee for the next year due to the severance payment).
If each employee represents $300,000 in total expense (a very generous estimate), that's only $1.5 billion in savings, and for the first year the savings is only $825 million due to the severance payments.
Okay, maybe Twitter can raise revenue with Twitter Blue. They'll need to pick up a smidge over 15 million paid Verified users in order to cover the $1 billion interest expense. Twitter currently has 400,000 Verified users, ~240 million DAUs, so they need 6.25% of their global user base to purchase a subscription at $8/month. Out of Twitter's DAUs, you'll need to omit most users from countries that won't generally pay $8/month, like India, Brazil, and Indonesia. [3]
For reference, Spotify costs $1.46/month in India.
We could make a rough guess at this by assuming that the 155 million people in the US, Japan, and UK will pay $8/month for Twitter Blue, so you'd need close to 10% of the affluent user base paying for Twitter Blue.
For a benchmark, Discord makes almost all of its revenue from Nitro subscribers. In 2020 it had 14 million DAUs, with $130 million in annual revenue, which means with the $99/annual fee Discord had 1.3 million paid subscribers. In other words, about 10% of Discord users are paid subscribers. [1]
If Twitter can get the same subscriber rate with Blue, it'll reach $1 billion they need to pay off their loan interest, assuming Twitter Blue incurs no additional cost.
The problem is, it can't do that without that pesky R&D that we were talking about earlier. Discord is built around the concept that Nitro offers tangible benefits to paid membership. What features can Twitter build with a skeleton crew that will convince 10% of their users to pay for the service? Twitter Blue currently represents very basic functionality. [4]
I haven't even talked about the fact that I'm just talking interest payments for the debt, and the fact that Twitter wasn't profitable to begin with!
> Okay, maybe Twitter can raise revenue with Twitter Blue.
This made me laugh. Twitter Blue will have to be great for people to pay. Out of all your numbers, I wonder how many of those 155M users in the US actually post anything that would find value in Blue? My guess is that we're seeing Pareto in action where 80% of the Twitter content comes from 20% of the users.
I heard an interesting take today from someone in the social space. Basically as soon as they heard the price paid for Twitter, they knew it was over. It's just too much debt to both get out from under and move the company in the right direction.
I've started to come around that maybe it is that simple. Musk saw the debt payment and just cut the most easily cuttable expense in the short term - people. Bankruptcy seems inevitable.
I 100% agree with you. I don't even know how seriously Musk can try to avoid bankruptcy.
The leaked company communications are damning. Here's your brand new owner, richest guy in the world, implying the company might not survive another year or two, for a company that's been public, mature, and stable for years. It's completely insane.
Like I alluded to my comment, Discord's revenue story is essentially a best-case scenario type of benchmark. It's very unlikely that Twitter or anyone else in the social media space can replicate Discord's paid user share. Discord has been designing their product around a specific niche and optimizing for non-advertising revenue since inception. Twitter is designed for advertisers. It was never intended to be a paid product. They'll be lucky if 1% of their users bite.
I told you what I think he cut. I also said I don't know what that does to their revenues, but you don't either.
If you buy the narrative that 80% of employees were bloat and you can cut them without impacting top line, then it will work. If you think he's gutted essentially activities that will sink the ship, then it won't
> Twitter can't turn off servers and sell off data centers without impacting the product.
A friend of mine tried to tell me that Elon had some type of new computing technology that would let them turn off most of their servers because they could fit all of the tweets in less space.
He blocked me after I told him he had confused Elon Musk with gzip.
I think your friend gets at why people think Twitters IT costs can be cut though (ignoring his idea about compression): They think Twitters product is mostly the delivery of tweets.
But's not. Twitters product is ad inventory and their ad placement platform, and the engine driving engagement that boosts their ad inventory by reordering the timeline to keep people scrolling. Those are the hard parts.
"Just" delivering and storing tweets is easy. If you ignore the nasty business of the moderation. And most people have never even visited analytics.twitter.com and seen how much data is available to them about their own tweets, much less looked at ads.twitter.com and seen how precisely they can be targeted, and the precision Twitter offers in what kind of things you can pay for (engagement, follows, media views, clicks). And they've certainly not tried running ad campaigns in those categories, and seen how good Twitter are at showing your ads mostly to people doing what you're paying for.
Musk's actions makes me wonder how well he understood the complexity of Twitter too. Surely he must have looked at those other aspects of Twitter before he made his bid.
> If Twitter went another month without an outage, how would you adjust your opinion? How about a year?
I think it's hard to just use time as a measure. If there are no security issues and they add no features, then things should run fine. Which, ironically points to how solid the team was that was fired.
Of course if Twitter not only limps along, but thrives in this new setup then I'll definitely change my opinion. Being in the US, this might end up the case while Twitter for the rest of the world falls apart.
Keep in mind, that I think Twitter was bloated and needed a big shakeup. Randomly dumping people and those who tried to correct me is not the heuristic I would have used.
With that said, Twitter still has 2 huge problems. No vision and saddled with an enormous amount of debt. Right now, Musk is taking the PE approach to cut and milk what's there. The problem is, there isn't much to milk.
yes, maybe a car without mechanics rather than without a driver. drivers can be really bad, too. so bad that they forget to steer or steer badly, etc. but the steering wheel is categorically the interface made for non-engineer usage.
The car analogy is amusing, but how much does it really hold up? Have we ever seen another major social media company drop this much of its staff in one go? I certainly can’t think of an example. I think we’re in somewhat uncharted waters here.
A driverless car won’t last long, we know that for a fact. I think it remains to be seen how long a bloatless twitter can last. I’m personally optimistic.