4 ms·

There is a theory about large complex systems which seems to be true in biology and maybe applies here. Intentional downsizing during times of stress works whe

There is a theory about large complex systems which seems to be true in biology and maybe applies here. Intentional downsizing during times of stress works when it preferentially targets defective or dysfunctional components of the larger system. The system improves because the worst parts were removed. Layoffs don't help companies unless they can reliably remove the worst parts. At most large public companies, the cancerous bureaucracy protects itself and the parts removed are closer to median performers, or even above-median performers. The system gets smaller and less efficient. Layoffs can be necessary to get the company to fit through a certain sized hole (in the form of cash flow constraints), but it won't be better at what it does on the other side of the hole, it will just continue to exist. Layoffs work when there is an accurate discriminating mechanism for who stays and goes. The best example of this (outside of private equity turn-arounds that are not widely known) is Twitter. Outside engineering talent was brought in as an oracle, immune to Twitter's bureaucracy. It reliably discriminated between value-adding and not. As a result, the company became incredibly lean and even consistently profitable.

sumeno 2y ago
Twitter was profitable before the takeover and now it is not
- alphazard 2y ago
  Maybe we disagree about what "profitable" means. Posting an authoritative source, so others can decide for themselves. https://stockanalysis.com/stocks/twtr/financials/ https://stockanalysis.com/stocks/twtr/financials/
  - sumeno 2y ago
    So over the 4 years before the takeover they had a net profit of $1.3 billion. What's their profit been in the 4 years since?
  - hkpack 2y ago
    I don't have that much of an experience analysing similar reports, but it seems to me to be a healthy and growing company? If you mean a negative cash flow and operating income - it doesn't mean much, as we need to know why. Taking into account the jump in "Revenue", positive "EBIT" and so on, I would imagine that it was because of some sort of "strategic investment", like some acquisitions probably? What am I missing? I would definitely be happy with such reports if that was my company.
    - monocasa 2y ago
      The parent only provided financials from before the acquisition.
  - monocasa 2y ago
    Unless I'm missing something, your data you've voted seems to be pre-acquisition. Also, by all accounts their revenue has dropped so much they're having issues even covering the loan payments, much less having enough for any semblance of profit.
bdangubic 2y ago
twitter is just about the worst example you could have used - by wide margin…
- dehrmann 2y ago
  It depends if you mean keeping Twitter the service operational or Twitter the product relevant.
  - bdangubic 2y ago
    neither can be comparable to what twitter was - not service not product
Ologn 2y ago
By the late 1960s, software projects had finally reached a certain size, and the particularities of managing these large projects were discussed at NATO software engineering conferences, and codified by Fred Brooks in 1975 in the Mythical Man Month. The upshot being, software projects at corporations can't be run in the same manner as non-software projects. Yet at Fortune 500 companies today, and in the past two years that includes at least some of the FAANGs, you see them trying to run software projects and treating programmers in the same manner they treat non-programmers, and it doesn't work. I just happened to watch some old talk or interview with Gabe Newell recently, and he said industry thinking at the time Valve was founded was of a certain kind, and Valve took the opposite approach from the industry - the industry was looking to get cheaper programmers, and Valve went the other way and looked for the most expensive programmers, and so forth. Valve has probably less than 400 people working there (don't know the 2025 headcount), but makes billions a year in profit.
logsr 2y ago
Large tech companies are inefficient and have very poor engineering productivity but this is largely for systemic reasons. They definitely do not have any mechanism for identifying high performers. The basic problem they face is that they pay a fixed wage for complete IP ownership (horrible deal) and so they intentionally do not measure and reward actual performance because they do not want to compensate high performers based on the value they create. There isn't really a solution because these companies are making rational profit maximizing decisions, it just happens that mediocrity and managed decline of locked in revenue streams is profit maximizing for them.
- mattgreenrocks 2y ago
  > they intentionally do not measure and reward actual performance because they do not want to compensate high performers based on the value they create. Ten years ago, an oft-repeated quote: “you’re not paid what you’re worth, but for the value you create.” In light of that, is the argument that performance is institutionally obscured so that devs don’t realize how much leverage they really have? I can see traces of that in seeing shipped products as less the result of one/few devs working hard and more as multiple teams, as you now have two levels of personnel abstraction to diffuse potential leverage (individual/their team, and the teams among themselves).
  - rightbyte 2y ago
    If you work in an org the share of the success of something that you could claim is really abstract. In some sense the cleaning lady can put claims on "my" success. Most stuff I have done would probably be worthless without the sale people and customer relations in place. Like 0$ value.
- intelVISA 2y ago
  Well, it goes both ways: a fixed wage for complete IP ownership is great if your main dev skill is "bringing key stakeholders into conversations at the appropriate time".
CPLX 2y ago
There's absolutely no reason whatsoever to think Twitter's financial performance improved after going private.
hobs 2y ago
Outside engineering is not ever an oracle, and it has its own interior motives and bureaucracy, firing people is generally a one way street. Twitter has literally issued authoritarian threats to advertisers to come on its platform or else because of how much its valuation has dropped and how much money has been lost - this is the worst example of "engineering oracle fixing things" I could possibly imagine.
cjaackie 2y ago
Could I be missing something, but are you seriously promoting layoffs? You must be trolling by using twitter as functional example of why they work? I think sometimes I miss sarcastic tones and I really hope i did with this one
- rightbyte 2y ago
  I think Twitter is a bad example to use, since so much other "direction changes" could arguably shadow the reduction in workforce.
srpablo 2y ago
lmaoooooo buddy "oracle" implies they know something; the people brought in to fire people at Twitter spent almost no time in determining who was good or even how the company worked, completely undermining your thesis. It was madness: people were instructed to print paper copies of their code to bring into an office, like it was 1995. Remember geohot in a Spaces saying "the main problem with Twitter is that you can't run it locally?", as if any company of that size has run that way at any point in the last decade? Additionally, the horrible communication and chaos made it even harder for performers to perform. As others have pointed out, its stock lost a ton of value, it performs worse financially, and as a product by virtually every other metric has gone to hell (outages, CSAM safety, spam, bots...). You tell a decent story at the start, but your choice of example couldn't be worse.

sumeno 2y ago

Twitter was profitable before the takeover and now it is not

alphazard 2y ago
Maybe we disagree about what "profitable" means. Posting an authoritative source, so others can decide for themselves. https://stockanalysis.com/stocks/twtr/financials/ https://stockanalysis.com/stocks/twtr/financials/
- sumeno 2y ago
  So over the 4 years before the takeover they had a net profit of $1.3 billion. What's their profit been in the 4 years since?
- hkpack 2y ago
  I don't have that much of an experience analysing similar reports, but it seems to me to be a healthy and growing company? If you mean a negative cash flow and operating income - it doesn't mean much, as we need to know why. Taking into account the jump in "Revenue", positive "EBIT" and so on, I would imagine that it was because of some sort of "strategic investment", like some acquisitions probably? What am I missing? I would definitely be happy with such reports if that was my company.
  - monocasa 2y ago
    The parent only provided financials from before the acquisition.
- monocasa 2y ago
  Unless I'm missing something, your data you've voted seems to be pre-acquisition. Also, by all accounts their revenue has dropped so much they're having issues even covering the loan payments, much less having enough for any semblance of profit.

bdangubic 2y ago

twitter is just about the worst example you could have used - by wide margin…

dehrmann 2y ago
It depends if you mean keeping Twitter the service operational or Twitter the product relevant.
- bdangubic 2y ago
  neither can be comparable to what twitter was - not service not product

Ologn 2y ago

By the late 1960s, software projects had finally reached a certain size, and the particularities of managing these large projects were discussed at NATO software engineering conferences, and codified by Fred Brooks in 1975 in the Mythical Man Month. The upshot being, software projects at corporations can't be run in the same manner as non-software projects. Yet at Fortune 500 companies today, and in the past two years that includes at least some of the FAANGs, you see them trying to run software projects and treating programmers in the same manner they treat non-programmers, and it doesn't work. I just happened to watch some old talk or interview with Gabe Newell recently, and he said industry thinking at the time Valve was founded was of a certain kind, and Valve took the opposite approach from the industry - the industry was looking to get cheaper programmers, and Valve went the other way and looked for the most expensive programmers, and so forth. Valve has probably less than 400 people working there (don't know the 2025 headcount), but makes billions a year in profit.

logsr 2y ago

Large tech companies are inefficient and have very poor engineering productivity but this is largely for systemic reasons. They definitely do not have any mechanism for identifying high performers. The basic problem they face is that they pay a fixed wage for complete IP ownership (horrible deal) and so they intentionally do not measure and reward actual performance because they do not want to compensate high performers based on the value they create. There isn't really a solution because these companies are making rational profit maximizing decisions, it just happens that mediocrity and managed decline of locked in revenue streams is profit maximizing for them.

mattgreenrocks 2y ago
> they intentionally do not measure and reward actual performance because they do not want to compensate high performers based on the value they create. Ten years ago, an oft-repeated quote: “you’re not paid what you’re worth, but for the value you create.” In light of that, is the argument that performance is institutionally obscured so that devs don’t realize how much leverage they really have? I can see traces of that in seeing shipped products as less the result of one/few devs working hard and more as multiple teams, as you now have two levels of personnel abstraction to diffuse potential leverage (individual/their team, and the teams among themselves).
- rightbyte 2y ago
  If you work in an org the share of the success of something that you could claim is really abstract. In some sense the cleaning lady can put claims on "my" success. Most stuff I have done would probably be worthless without the sale people and customer relations in place. Like 0$ value.
intelVISA 2y ago
Well, it goes both ways: a fixed wage for complete IP ownership is great if your main dev skill is "bringing key stakeholders into conversations at the appropriate time".

CPLX 2y ago

There's absolutely no reason whatsoever to think Twitter's financial performance improved after going private.

hobs 2y ago

Outside engineering is not ever an oracle, and it has its own interior motives and bureaucracy, firing people is generally a one way street. Twitter has literally issued authoritarian threats to advertisers to come on its platform or else because of how much its valuation has dropped and how much money has been lost - this is the worst example of "engineering oracle fixing things" I could possibly imagine.

cjaackie 2y ago

Could I be missing something, but are you seriously promoting layoffs? You must be trolling by using twitter as functional example of why they work? I think sometimes I miss sarcastic tones and I really hope i did with this one

rightbyte 2y ago
I think Twitter is a bad example to use, since so much other "direction changes" could arguably shadow the reduction in workforce.

srpablo 2y ago

lmaoooooo buddy "oracle" implies they know something; the people brought in to fire people at Twitter spent almost no time in determining who was good or even how the company worked, completely undermining your thesis. It was madness: people were instructed to print paper copies of their code to bring into an office, like it was 1995. Remember geohot in a Spaces saying "the main problem with Twitter is that you can't run it locally?", as if any company of that size has run that way at any point in the last decade? Additionally, the horrible communication and chaos made it even harder for performers to perform. As others have pointed out, its stock lost a ton of value, it performs worse financially, and as a product by virtually every other metric has gone to hell (outages, CSAM safety, spam, bots...). You tell a decent story at the start, but your choice of example couldn't be worse.