Layoffs are one thing. People don't enjoy them but they can help a bloated company get back in fighting shape.
Purging half your workforce and then losing half of the half you had left, at this speed, is a different story altogether. I know nothing of Twitter but I am quite sure there is not that much fat to cut. At the very least, Musk's ultimatum needed to be delivered first, to calibrate the layoffs against the number of people willing to stay under the conditions he laid out.
A twitter thread about the issues of >50% workforce losses on a company, from the perspective of a startup (which is admittedly different from a large organization):
TL;DR "the most challenging turnover times were ones when our organization was very specialized, and we ended up losing whole parts of the team, where nobody from a group was left to transfer knowledge." ... "Anyone who thinks a major reboot like this is a "good thing" that's going to result major improvements in the near-term are probably deluding themselves."
The crux of the issue is
how many people are actually necessary to run Twitter?
And, bluntly put, I have a hard time imagining there's enough work for a
thousand people, much less seven and a half.
How many people do you need writing the code?
It's a stripped down forum-board.
One person could write the code for something as simple as Twitter; have a team of three or five to make sure you've got all your bases covered, and you're good. Just to be incredibly over-generous, let's multiply that by ten, so we have thirty to fifty people working on code.
How many people do you need maintaining the hardware?
Twitter is primarily (though no longer exclusively) a
text message platform. The amount of data you need to send and receive per-tweet is less than a kilobyte. Every Terrabyte of HD space you have should be able to hold a billion tweets, let's call it half a billion for safety's sake. Some people tweet prolifically, but the proportion compared to total account-holders is tiny, and modular blade servers are extremely low-maintenance, so even if we assume you need another one every day to store everything.
Assuming Twitter even uses its own hardware for hosting, my rough guesstimate would be six people, let's say two dozen to be generous and make sure you have people on-call at all times.
What about marketing?
What marketing? Twitter doesn't
do marketing, it's the platform that you market
on!
What about handling business with advertisers?
Well, now we're starting to get outside my areas of expertise. I'm something of a techie, but I am
not a marketing or advertising wonk. Still, assuming a decent employee can handle at least one advertising client per day (I expect the real number is at least four times that high), if you have let's say a hundred people working with those who want to advertise on your platform, that hundred is enough to have over two thousand advertising clients. That can maintain you a business model.
Of course, if the system is mostly automated, which basic ads almost certainly are, it's going to be
way faster than that. I'm just deliberately skewing high because it goes against my point.
Let's talk verification process!
Since Elon's new policy makes this such a big deal, let's get
super aggressive and have two hundred people whose dedicated job is verifying that whoever is getting their account verified, is who they claim to be. Obviously this isn't what happened so far, because only an automated system (or a lot of malicious employees deliberately verifying things they shouldn't) could get so many impostor accounts so quickly, especially when they need to have their payment information attached to the back end of the account.
Simple process for verifying accounts; look at the application, see if the credit card data corresponds to the name on the account, if yes, approve, if not, deny. If it's a corporate account, look up the corporation's public contact data, call them up, ask if X person applying for an account is actually them, you're done. Because corporations and the like can be a mess, that might take an hour, or even
hours, but individual accounts should take
minutes. If you're literally just rubber-stamping 'yes this corresponds,' then that should take
less than a minute.
Still, let's roughly assume an average of five minutes per verification based on the above, and each of your 200 employees in that department is averaging 12 verifications an hour, or 96 a day, which comes to 19,200 a day for the lot of them. That'd be 5 million a year. Even if I'm off about the average time by an order of magnitude, that's still 500,000 a year, which is pretty big business.
So, then let's hit legal. How many lawyers do you need?
This, in a more ideal world, should be a handful of people, but let's say you've got a hundred lawyers and less-expensive people working as their aides on full time to advise you, fight in court, and manage 'regulatory compliance' crap. Half of their work is probably going to be blocking specific posts in specific countries that don't have a proper attitude towards Free Speech.
All told, this brings us to 474 people. If we assume you need one manager for every ten people, and we don't count those managers as already productive members of their various departments, that brings our total up to 522 employees at Twitter. Let's add another 28 for 'odd jobs,' and we land at 550 total...
Which is about
eight percent of the 7500 people that Twitter had on staff.
Obviously, I'm not an expert in running a large corporation. Obviously, there's going to be all kinds of factors that play into exactly how many people you need. That's why I was deliberately generous in each estimate, but even then, I could be substantially wrong in any given category.
The point though, is it is very,
very easy to see how Twitter wasn't just bloated, it was
horrifically bloated, and what Elon was talking about months before the deal went through, laying off 75% of people...
Well. Laying off half, and another half quitting, sounds like it's just gotten to what he was talking about in the first place. And it's entirely possible he could still trim more people, and not lose what's important. All of this before we even get into the Pareto Distribution.