Twitter has around 8,000 employees. I fully understand and respect the high level of technical skills needed to operate and manage such a highly available service, BUT I still can't for the life of me figure what the other 7,950 employees do.
One interesting point here. I noticed that Twitter bought a design agency a few months back [1] who themselves claimed to have dozens of employees. It made me wonder. Why would Twitter need dozens of designers on top of the designers they already had? I would have assumed they likely had a 10-20 person design team. One designer and a few managers for the handful of features they have. But then I read a post online that even back in 2014 they had north of 59 designers. And given their growth since then, they likely have hundreds of designers.
As a founder, engineer, and designer myself. I cannot fathom what 100+ designers do all day for an app like Twitter. They have to be working on so much overlap, and busy work.
I gave a try to Twitter's Ad platform recently. Mother of god, they've built a giant machine behind the curtains. It has terrible UX, but I could, if I wanted to, target a U.S. Military retiree that likes creole food, located in Boston, enjoys watching Golf and uses iPhone 11 Pro with an older iOS version 14. The profiling of users, and the ad-tech machinery is insane. There is a auto-bidding machine, detailed analytics, website tracking scripts, conversion metrics, the works. I had no idea. Imagine you build something amazing, but no one knows about it – how do you get that thing across people that actually want it? It is a hard problem and from my little experiment with their advertisement platform, it is miraculous. Surprisingly, I've not seen any article from mainstream newspapers that exposes what kind of a goliath Big Tech's ad platform is.
You ought to sign up and see it for yourself. Meta & Google's Business Platform is similar if not more powerful.
After this revelation, 8000 people entirely makes sense.
> It has terrible UX, but I could target a U.S. Military retiree that likes creole food, located in Boston, enjoys watching Golf and uses iPhone 11 Pro with an older iOS version 14
So, it's got lots of categories you can target. But...what evidence or guarantees are there of accuracy of the targeting?
There seems to be some misfiring because rate of likes isn't high (~1 like in 100 impressions), but based on the people that like the promoted post it is 100% without exception accurate. So, at the least, it is reaching the right people. I know because there are currently 500+ likes and I've visited every single profile and tried to engage with them. Audience is just developers.
I must say, it has been a very positive ad experience – receive compliments, 10% retweets, and had several long form conversations with strangers on Twitter messages. Really fun.
Surely the legal department alone requires more than 50 employees, considering all the countries they operate in I imagine there are many legal requests.
Generally, things get harder as you scale because something that would have taken a new Postgres table has become a distributed systems problem.
You also hit diminishing returns with new features and user growth, but when you have hundreds of ~~billions~~ millions of users, even marginal engagement gains can multiply out to something big, so you have a lot of teams working on features that might move the needle a bit.
There are also more regulatory hurdles as you grow, so some of the staff just support that.
Yes, things gets harder at scale which is why he said 50 and not 5. Depending on how much infrastructure code you do inhouse you add another 50 for managing that. If you have your own datacenters add another 100 or so if you have a few around the world. If you want world class recommendation add 100 or so data scientists.
That gets us 300 tech people to run the service.
That is for engineering, those numbers are roughly what I saw at Google for these kinds of things. Then Google typically has 1 non tech person per tech person, so add in at least as many people again. Then since Googles customer support and community management is hardly world famous for being good, rather it is infamous for being bad, you probably want even more non tech people than that. But still, do you really need 8000 people for it?
Allow me to ruin your now-traditional joke [1] by telling you what I saw when I worked there in 2016-7.
First, it's a complex product. Just for the user-facing side of things, there are four big codebases: back-end, iPhone, Android, and Web. The back-end stuff is very large, generally for good reason, and sometimes not. A lot of those good reasons relate to performance; with over 100m people using it every day, getting the right tweets to the right people is challenging, especially given that some people have over 100m followers and many users will follow a lot of accounts (I follow over 3300, for example).
But the user-facing side is only part of it. A big way the site makes money is ads. This is in some ways a more complex problem than the user side of things. If you'd like to see, try buying an ad on Twitter. They also have a division that does data products, including a variety of APIs, and another group that has other products for businesses, like tooling for the customer support interactions that people expect to handle on Twitter.
I of course can't forget their SRE folks, who keep all of the machinery humming and make sure all the software is doing what it's supposed to. There a quite a lot of people doing infrastructural work making tools and products that you never hear about and I probably can't list. There's also a good developer tooling group that helps keep the developers above working smoothly. And let's not forget their internal IT group,
Adjacent to engineering are a lot of really sharp product and design people, as well as a design research crew that understands the many ways people use Twitter and examine how it works for them.
We then must turn to people who handle the social side of this. When I worked there, that included a significant staff doing policy work, trust and safety enforcement, and handling darker things like CSAM and terrorism issues. There were also a bunch of people fighting the banal spammers.
And let's not forget the ML people! That alone was a few hundred people doing research, creating and improving models, and applying them to many of the things I mentioned above.
And at last we come to the kind of things that pretty much any large company needs: finance, legal, marketing, government relations, sales, computer security, physical security, admin, management, and the like.
If you ever wonder this in the future about any company, I suggest you look at their jobs site [2], which is always a good way to get an idea of what goes on at a company.
Easiest way to make money would be to fire 90% of them. 800 people is hardly a skeleton crew for running a service like Twitter, even at global scale, and then you would have plenty of profits.
