To be honest, I feel like Google and Meta both had far too many employees.
"Too many cooks spoil the meal." "The Mythical Man-Month" etc.
There was a crazy push for hiring and growth, which diluted the quality and increased the communications bottlenecks. A 10% cut won't do it, though. Over-hiring is hard to recover from. Google would need to cut 75-90% of its workforce to be at all efficient.
If it did that, it would be left with a ton of code with no one who understands it, a horrible reputation, and zero morale.
The only way I can think of to manage this is to build out an elite smaller unit, with a smaller codebase, with less technical debt, and then to spin out the deadweight.
Guestimate. You can tell me I pulled it out of my ass, and you'd be right. But it's that type of number.
A better answer is looking at the size of Google when it was doing (approximately) the same things. Google has 20k employees in 2009. It has 155k today. I don't see Google doing much for me in 2022 that it wasn't doing in 2009.
I could probably go a long ways before that too.
A lot of what's happened from 2009 to 2022 is churn. Make-work to keep people employed.
> Also, they have a 27% operating margin, how much more "efficient" do they need to be?
Let me spell this out: In many cases, the same thing can be accomplished in 100 lines of code as 10,000 lines of code. That's largely a function of architecture, discipline, and cleverness. 100 lines of code is orders-of-magnitude easier to maintain and faster to evolve.
The Google codebase is *massive*. If you need a new feature, with good communication channels and focus, you can architect it in. With poor communication channels, it get kludged in.
Communications scales as O(n^2) where n is the number of employees. There's a similar power law for how pieces of code can interact and introduce complex bugs.
By "more efficient," I don't mean Google would be able to do more per dollar. With a smaller, better workforce, and a smaller, better codebase, Google would be better able to maintain old products, add new features, and develop new ones.
"Too many cooks spoil the meal." "The Mythical Man-Month" etc.
There was a crazy push for hiring and growth, which diluted the quality and increased the communications bottlenecks. A 10% cut won't do it, though. Over-hiring is hard to recover from. Google would need to cut 75-90% of its workforce to be at all efficient.
If it did that, it would be left with a ton of code with no one who understands it, a horrible reputation, and zero morale.
The only way I can think of to manage this is to build out an elite smaller unit, with a smaller codebase, with less technical debt, and then to spin out the deadweight.