I really like the specific numbers people are posting. I’ll add my own (rough estimates) from the ~5 months I spent applying to roles in 2018.
Context: In spring 2018, I attended an event CEA ran for people with an interest in operations, because Open Phil referred me to them; this is how I wound up deciding to apply to most of the roles below. Before attending the operations event, I’d started two EA groups, one of which still existed, and spent ~1 year working 5-10 hours/week as a private consultant for a small family foundation, doing a combination of research and operations work. All of the below experiences were specific to me; others may have gone through different processes based on timing, available positions, prior experience with organizations, etc.
CEA (applied to many positions, interviewed for all of them at once, didn’t spend much additional time vs. what I’d have done if I just applied to one)
~4 hours of interview time before the work trial, including several semi-casual conversations with CEA staff at different events about roles they had open.
~2-hour work trial task, not very intense compared to Open Phil’s tasks
1.5-week work trial at CEA; there were approximately as many open positions as there were work trial candidates, and I’m not sure anyone went through a trial of this length and wasn’t hired (though this might have happened). I was paid at a standard hourly rate for this, so it came out to ~$1500.
Open Phil (research)
They reused my conversation notes and charity evaluation test from a previous GiveWell application (those took me ~8 hours total, so perhaps I should count it as ~4 hours per application)
The first interview took ~30 minutes (and was more of a Q&A for my benefit, not something that required too much preparation).
The next work trial was ~12 hours (I worked until almost the maximum time permitted; we were instructed not to spend more than this, and to submit an incomplete application if we ran out of time).
The second interview was ~75 minutes, and pretty intense, but not something I was asked to study for in a particular way.
When you include the resume + initial submission, this adds up to 18-22 hours (depending on how you count the reused conversation notes), for which I was paid $1100 ($300 for notes, $800 for work test). That was better than my freelance writing rate at the time, so while it was time-consuming, it wasn’t totally unsustainable.
These hours were spread out over months of waiting time, which wasn’t ideal, but given the many hundreds of people who applied, I’m not surprised the process took a while (I’d guess that staff spent something like 500 hours grading research tests and conducting follow-up interviews with the last round of candidates, which is a full month of work for three people).
Open Phil (operations)
Started with a ~45-minute informational interview (mostly for me to ask questions, didn’t require much prep)
Work test in the 2-8-hour recommended range, paid at $24/hour for up to 8 hours (which read to me as a strong signal of “don’t spend more time than this”, though I understand the pressure to keep going). It took me 2.5 hours for the four-page assignment; it was an email rather than a research report, so a bit less stressful to finish.
I joined that hiring process fairly late, and someone else was hired before I got any further. When a new position opened a few months later, Open Phil asked me to come in for a one-day visit, and they were flexible enough that I was able to combine this with another trip to the Bay for interviews (the price of flexibility, of course, is that everything takes longer for each applicant—it’s a tough tradeoff).
The visit was a full day, but didn’t involve much “work”, per se; there were ~3 hours of interviews, with the rest of the time spent on between-interview breaks, casual lunch with other operations staff (no interviews), and a visit to the daily morning meeting for ops staff.
Total: Counting travel as “half time”, 10-12 hours.
One-hour interview to learn more about the position; I was also asked some questions, but this was more of a screening for “do you understand what MIRI does, and why”.
Initial one-hour test as part of their standard recruiting process for all staff (30 minutes of quantitative reasoning, 30 minutes of logic puzzles). I don’t think this was a stage in and of itself, but I could be wrong (I think I was always going to do the work test).
~4 hours of work tests in the MIRI office, plus a ~one-hour interview with a MIRI staffer I’d likely have worked with as part of the role (very casual, mostly me asking questions).
Total: 8 hours with travel as “half time” (this was done as part of a trip I made to the Bay to work through several interviews). I was paid $120 for my time on the work tests.
Ought (COO role)
Three interviews of ~4 hours total, which were fairly “work-like” (I was answering more questions than I asked, or discussing trial tasks)
~3 hours spent on two trial tasks; no time recommendations were given, but the work was light (“think about this brief technical article and be ready to explain it”, “think about whether we should hire someone with this resume”—I wasn’t submitting any writing, just discussing the assignments in my interviews)
I didn’t get to the “work trial” stage for this position (though I don’t know whether there was one—they may have just trialed their one favorite candidate).
Total: ~7 hours of work, all remote, and I was paid $250. Ought gets bonus points for giving me very good feedback on the ways in which my last trial task wasn’t up to par.
Vox (journalist and engagement manager positions, Future Perfect)
~7 hours on an initial work assignment, plus a 30-minute phone screening for me to ask questions. The journalist assignment took roughly the same amount of time as the engagement manager assignment.
I didn’t move beyond that in the process. Amusingly, the only non-EA organization I applied to led to my doing the greatest amount of unpaid work.
AI Impacts (operations/research role, it’s a tiny organization and I’d have been a jack-of-all-trades)
~2 hours of initial interviews before being offered a work trial
Because the role was nebulous, I wound up planning my own work trial together with AI Impacts staff. I estimated that the work would take ~20 hours total (paid at $30/hour), but wound up accepting a CEA position before starting in on the tasks.
CHAI (communications/executive assistant role)
Two interviews of ~2.5 hours total (of this, 1.5 hours was talking to Stuart Russell, which was much more exciting for me than for him).
...and that’s it. I received an offer (I think they had very few candidates) without a further work trial.
BERI (project manager)
Two interviews of ~1.5 hours total, nothing beyond that (no offer)
I also had some exploratory conversations with people at a couple of other organizations, but accepted the CEA position before getting to a formal interview.
All told, if I throw in ~5 hours for updating my resume and writing a few brief “cover letter” notes (huge props to the orgs I applied to for not requiring formal cover letters), I spent ~70 hours interviewing (with travel at half time) and was paid $1530 (outside the CEA work trial, which was another 60 hours and $1500). I’m not sure how to think about time costs from travel, but I got to meet a lot of interesting people and eat some free meals along the way.
I didn’t find any process especially aggravating, though there were small adjustments I’d suggest for some organizations (mostly the small ones that hadn’t done much interviewing). I think I was compensated fairly, and most of my interviews were genuinely useful to me, both for learning about the particular organization and for getting a better sense of my own strengths, weaknesses, and goals.
I agree with some of the criticism on this page, but I also want to point out some really good things EA orgs do with hiring:
Open Phil passed me along to CEA as a possible operations candidate when they hired someone before I finished making it through their pipeline, and I wouldn’t have applied for most of these positions if they hadn’t done so.
Open Phil also reused my GiveWell tests so I didn’t have to write new conversation notes.
CHAI passed notes from one of my interviews to BERI when I was still applying for the latter role.
After I volunteered at a CEA event after the operations retreat, they passed my name to MIRI, who hired me to work on operations for some of their retreats, which helped me learn about their open position. These organizations talk to each other, and in my experience, that’s been a good thing.
No cover letters! (I said this once before, but it’s worth saying again.)
Compensation for time spent on work trials! It’s possible that orgs should also compensate for interviews, but the work-trial payments put EA leagues ahead of some other industries. I certainly never got paid for any of the cover letters I wrote in college, or the hours of math tests and debate prep I had to do while applying for jobs at investment firms.
Not having everything be interview-based! I don’t interview well, and spent a lot of time in college wondering whether I’d just screwed something up in an interview without noticing. My work trials, on the other hand, are concrete and visible, and if I don’t get accepted to a position, there’s at least a chance that I can learn something by reviewing my work.
I’d say that very careful, non-unilateral EA outreach could be leveraged.
I read almost every Tweet where anyone mentions effective altruism. One frequent type of message which makes me wince is when someone @s a very wealthy/famous person something like “have you heard of effective altruism?” when that person mentions charity in any context.
(I’ll refrain from linking to any specific examples, but I tend to see it happen at least once a week, and the messages come from many different sources.)
I doubt that any of these Tweets have caused specific problems yet, but if we want very wealthy people to become interested in EA, we should distinguish ourselves from movements/groups that do a lot of random solicitation.
Fortunately, we are at this point well-known enough that a lot of high-net-worth individuals naturally hear about us in the course of looking up giving opportunities, and we have groups like Effective Giving that practice careful outreach to promising prospects. But a sufficiently pushy/annoying message from an individual talking about EA could still create a bad first impression. I’d hope that our first reaction as a movement to news like this would be “what are this person’s interests, and is there a way EA can help?” rather than “I wonder how we can get donations from this person?”
(I don’t mean to say that your post implies you think in this way—I’m just taking this chance to lay out something I’ve seen a lot, and which has been bothering me for a while.)
Note: I work for CEA, but these views are my own.
People interested in this topic should read this post about how EAF ran their hiring round. They seem to have a good setup: only 25% of applicants were asked to take a work test, and only 25% of those tested were invited to a work trial (the other 64 of 68 were estimated to have invested <10 hours).
I’d love to see other organizations within the EA community publish similar numbers, though I understand that it can be tricky (e.g. someone might learn they were the only person interviewed who didn’t make it to the next stage).