Comment by piva00 | Hacker News Reader

Comment by piva00 | Hacker News Reader

Parent Full thread

After working in large corporations I only partially agree with this, if I take "unnecessary" to its literal meaning. It's impossible to keep everyone at a 80-90% busy rate at any large company just due to the inefficiencies a large system accrues, idling is embedded due to the many layers of communication, coordination requirements, etc.

I agree there's a lot of empire-building in large orgs, and perverse incentives for management promotion where headcount is directly related to their career ladder. At the same time I completely disagree with the "vast majority" take on it, you can't just hire consultants when needed for ramping up a project, institutional knowledge is extremely important in any large org and I haven't seen any yet that has a streamlined and efficient onboarding process for that. Experience matters and that is a reason to keep an overhead of employees that might look unnecessary in accounting/MBA terms but is a pool of workers with enough knowledge to ramp up a project at a moment's time if the organisation needs.

The other side of this would be to keep a ship that's efficient for the current moment but that can't be steered rapidly when needed, that might kill or stagnate an organisation pretty quickly when markets change and the reaction time is too long.

As I said, I don't disagree completely, I'm not fond of upper management in general, and quite tired of the whole career game that is played in large orgs but your point of view is very, very cynical.

> It's impossible to keep everyone at a 80-90% busy rate at any large company just due to the inefficiencies a large system accrues, idling is embedded due to the many layers of communication, coordination requirements, etc.

Also, let's add another charitable explanation to this:

As people acquire more experience in the org, they become more efficient. That means someone who was maybe idling 20% of the time is now idling 50% of the time - they are simply faster and better at doing their job, but for whatever reason they haven't been promoted yet or assigned new responsibilities.

I don't know whether this is "good" or "bad".

> your point of view is very, very cynical

Yes. It is also informed by a 40-year career that has included a pretty even mix of working for both small and large organizations. Cynical != wrong.

Here is another cynical but informed observation: part of the modus operandi of large organizations is specifically to prevent any one person from becoming too important to fire. So as soon as someone starts to significantly outperform their peers, the organization actively moves to squash them because an individual contributor at 3 sigma is an existential threat to good order and discipline.

This is not speculation. This actually come up many years ago when I was interviewing for a CTO position. I was asked: what would you do if you had an engineer who outperformed everyone on their team by a wide margin, but who everyone disliked. The answer it turned out they were looking for (I learned after the fact): fire them. That's not the answer I gave, and I didn't get the job.

I've also tried both strategies (work hard, and slack off) over the years when I've worked at large organizations. I've consistently gotten much better performance reviews when I slacked off.

(This is absolutely not true at small companies BTW, which is why I vastly prefer working there.)

I agree with this. Maybe I've been very fortunate in choice of employers, but every company I've ever worked at, large FAANG to small startup, has had far more work in the queue than staff to do that work. Like MULTIPLES more: 4-10X as many bugs than staff available to fix. 4-10X as many possible features to implement than staff could possible do. So everyone has to prioritize, and the list of bugs and list of potential features always grows faster than it shrinks. People fight for headcount not just to grow their empire, but because there is so much work to do, and they (as managers) are accountable for getting that work done. Of course, not everyone is actually 100% utilized--there's some slack, but as you say, 80-90% utilization seems right. The vast majority of people don't just sit there, twiddling their thumbs for the majority of the time.

EDIT: OK, jeez, I'm willing to admit I'm wrong; but this doesn't even remotely square with any reality I've ever seen at about a dozen companies. People just hired to... do nothing? It doesn't make any sense at all. How many of you have an empty bug tracker? How many have zero tasks waiting for you when you clock out at the end of the day?

> People fight for headcount not just to grow their empire, but because there is so much work to do, and they (as managers) are accountable for getting that work done.

How does this square with a company like Google that is famous for having multiple teams competing with each other on the same product categories, ie the disaster that is their chat platform? Sure, there's work to do - because they're creating frivolous work intentionally. It's nothing but empire building. And the end result is a worse product!

I get that my topic is a bit of a tangent because you're talking mostly about IC productivity and the gist of my comment is on managerial incompetence. But ultimately my point is people are doing meaningless work and the system is designed for it. I've never worked somewhere that valued productivity over politics when it came time to promote people.

Google and Meta over-hired thousands of employees who do 'fake work,' says PayPal Mafia's Keith Rabois

https://www.businessinsider.com/google-meta-staff-do-fake-wo...

Alphabet's Google and Facebook owner Meta had thousands of employees who don't do anything.

"There's nothing for these people to do — they're really — it's all fake work," he said. "Now that's being exposed, what do these people actually do, they go to meetings."

Google, he continued, had intentionally over-hired engineers and tech talent to stop them from moving to other companies, a strategy he described as "pretty coherent." But, he added, that meant engineers had been happy to "be entitled, sit at their desks, and do nothing."

You’re not wrong. But HNers have decided that their employers are terrible tyrants, companies are awful entities that are both efficiently Machiavellian but also incredibly stupid and unproductive, and managers and executives are just foolish people who work for status and have no idea how to do actual real work.

And yet, somehow, these companies have existed for decades and the only time they’ve died out is usually because of technological leaps implemented by a competitor that looks very much like them.

Even beyond ramping up a new project, it can be important for maintaining existing ones. If a service outage costs a company millions of dollars a minute, it can be worth it to keep a team around that can resolve the occasional outage a few minutes faster even if they have almost nothing to do the rest of the year.

It's not even really inefficiencies, it's just straight up queuing theory.

As you approach maximum throughput, average queue size – and therefore average wait time – approaches infinity.

In order to have a quick and nimble organization you need to leave sufficient slack in your utilization of employees.

I’ve worked in big companies where this dynamic was present and appreciate your nuanced reply.

Reply on news.ycombinator.com