Hacker News Reader

> To put a very fine point on it, the reason they had to fight (like everyone else there) is because if they didn't come up with something to do that was quantifiable and impactful in line with org goals, they'd be terminated within the year. It was their job to find a way to be useful. The system culls those who do not.

You're saying the approach "one of the most efficient companies in the history of the world" takes to producing value is: let's just hire people and let them figure out the value they can provide, otherwise we'll fire them.

Am I getting that right?

Before I've even hired someone I've already scoped and quantified the impact I need them to have. I then go find someone that is qualified to do that, and then I motivate them to do more. It is _extremely_ inefficient from an _everything_ perspective to do so in reverse. Not harsh. Not impersonal. Not even ruthless. It's objectively n. e. fish. ent.

I mean .. you are correct to do so when adding 1 or 3 employees, and basically completely wrong at the scale of hires we're talking about.

For example this is how every tech "body shop" hires. The value is just out there in the ether of modern enterprise.

Bring in 20% more people year over year. Keep attrition rate for them and existing hires around 10%. Go knock hat in hand at every Fortune 1000 company and wrangle whatever buzzwords Gartner has thrown at them into a project or managed service you can slip as many billable hours as possible into.

It's the same with Meta, they can try literally about anything so when GPT or crypto or Web3 or AR or whatever comes along they just want to be able to throw a bunch of people at it to figure out how to socialize, advertise, and monetize it at massive scale.

>They hired as though the COVID trajectory would continue

Some would argue that management's sole goal is to effectively measure which direction the wind is blowing and predict course accordingly. The fact that so many did so poorly may make some question why management gets reimbursed as well as they do.

Just because Meta is pretty revenue efficient already doesn't mean it couldn't be more revenue efficient. Why is $2M per employee good? It's actually quite bad if Meta should be generating $10M per employee.

I would say an 80% margin is objectively good, and that moving it really doesn't make much difference even if you can make the number feel bigger.

Again, their system is designed to be ruthlessly efficient. They probably over-hired like 10-15%. I'd strongly suggest listening to the Odd Lots episode with Patrick McKenzie titled roughly 'what happens to tech workers in the age of layoffs.'

I get it's fun to speculate from the outside but there's no reason to look at it and go 'yep should be 5X higher' simply because they have a lot of staff. They have a lot of staff because they have billions of monthly active users and optimizing every aspect of their business is something they can do, and something that yields more profit than cost.

A lot of this armchair quarterbacking comes from folks who never worked there. Find someone who works at Meta and ask them about PSC.

I work at Meta :) The growth I watched in headcount over the last 4 years has been pretty amazing. Before Meta I worked at Reddit. More people were in my bootcamp class at Meta than the entirety of Reddit at the time. And they did bootcamps like that every week for years.

I don’t totally disagree with you. People at Meta work hard and they optimize everything and psc for better or worse definitely encourages performance. But I don’t know where numbers like “overhired by 10-15%” come from. I don’t think things were falling through the cracks and not being optimized before the company doubled in size.

I worked at Meta for five years, and fighting for work was never, ever my experience. When I started there was far too much to do and you were basically forced to ruthlessly prioritize in order to be successful.

Towards the end this started to change for me, but it happened a bunch earlier in engineering/product. Fundamentally, the standard way to grow headcount was revenue per head, and while that works for sales and ops it's a terrible, terrible idea for engineering because you end up fighting for work.

In fact, I'm part of an ex Facebook group, and all of the old timers are incredibly shocked that this was allowed to happen, and a bunch of newer people are confirming it did.

Just like companies can grow forever and stocks only go up. This is what MBA's are taught to do and makes work toxic by chasing numbers and metrics.

Their monopoly position generates the revenue. You could divide it by whatever number of employees they have and get whatever number you want.

Then why hasn't Zuckerberg fired everyone?

Because you need some people to keep the lights on and shareholders happy (see twitter).

>Meta is one of the most revenue efficient companies in the history of the world. Each employee generates $1.929 million in revenue, and they have an 79.63% gross profit margin vs a sector median of 49%. That's each employee including the recruiters, EMs, etc.

Do you have a source for this? Does this include contractors, as well?