[dx034]

That is only possible because of outsourcing. If you only employ the highest paid earners and use contractors for all support staff, you artificially inflate your median salary. I'd bet that FB has much lower salaries if you count anyone working full time for Facebook.

[camillomiller]

All the "Facebook" content screeners in Berlin, Germany, are actually employees of a contractor (Arvato-Bertelsmann) who is mostly hiring them through job agencies. They're in charge of deciding what has to be banned according to Facebook Community Guidelines. They get to see gore, suicide videos, racial slurs etc. all the time, with no proper psychological training. They earn minimum wage (around 10$ per hour) and they're completely disposable.

Nonetheless, when Facebook had to face a small content-related crisis in Germany in 2016, that didn't stop Zuckerberg from saying the the company hires teams of hundreds in Germany to stop the spreading of hate content. They don't.

Source: Inside Facebook - Im Netz des Bösen, Süddeutsche Zeitung, 2016. (sorry, paywall). http://www.sueddeutsche.de/digital/inside-facebook-im-netz-d...

[dx034]

Thanks for the example. I'd be surprised if content screeners are employed with Facebook anywhere in the world. Most don't make that job for a long time and they probably want to avoid being sued for compensation, I could imagine viewing those videos/images without proper training will lead to long-term psychological issues for many. It's easy if FB can just hire a subcontractor that files for bankruptcy once claims come in.

[kossae]

I’ve always been curious as to what “psychological training” would consist of to prepare for seeing such explicit content day in and day out. Perhaps something similar to military/war psychological practices?

[whb07]

Even in the military there is no such thing as this type of training.

source: was infantry.

[jaclaz]

Not exactly what you asked for, but similar enough:

https://www.forensicfocus.com/Forums/viewtopic/t=2329/

More than "training" it is "assistance" and "psychological support".

[fasj82]

Wow, racial slurs. I'm sure that can trigger some kind of trauma on them

[speedplane]

What percentage of Facebook workers are contractors?

[dx034]

I don't know and they don't disclose it. It's the vast majority of what they pay for physical security, building maintenance, cleaning, hospitality, etc. This happened across all industries. A few decades ago, janitors and cleaners worked for the company, now they're outsourced to keep them out of scope for unions and to be more "flexible" (i.e. hire and fire at will without negative press).

The NYT has a great article on that, not related to Facebook but it shows the general trend: https://www.nytimes.com/2017/09/03/upshot/to-understand-risi...

[badpun]

He probably means cleaning staff, cooks and other auxiliary functions.

[Buge]

Cleaning staff, cooks, etc, scale with the size of buildings. Customer support staff, content flag reviewers, etc, scale with the number of users. Since Facebook has so many users, I would expect the second category to be much larger.

[dx034]

Facebook runs a low-support model. At the same time, they offer a lot of perks to their developers. I haven't been to their HQ but from my experience with other companies I would guess that a significant part of the people working there (20-30%+) are working in maintenance, security, hospitality, cleaning, restaurants, etc.

Keeping large offices running smoothly and in nice conditions takes an enormous amount of manual work, a lot of which other employees never see (much of cleaning/maintenance is done over weekends and at night).

[lucideer]

> Facebook runs a low-support model.

Not for large advertisers they don't. Here in Ireland for example their outsourced advertising support & soft-sales, a large part handled by Accenture, are huge. Possibly more numerous than the Facebook EU HQ here. I'm sure the case is similar elsewhere.

[aje403]

I'm sure most people reading HN scanning the title have SDE's come to mind immediately, without taking support staff into consideration, so the median is probably a good estimator for this audience, given your point.

Maybe it doesn't belong on the same table as a company like Wells Fargo since it's a different basis of comparison