Thanks to Niel Bowerman, Habiba Islam, Lee McClenon, Holly Morgan, Steve Thompson, Kathryn Mecrow-Flynn, Ben Snodin and Yonatan Cale for their thoughtful feedback on a draft of this post (though note that they don’t necessarily endorse our views, and all mistakes are our own)!
Note that Anya currently manages recruiting for Open Philanthropy, but this post represents only her views and not the views of OP.
TL;DR
Over the past 6 months or so, we’ve seen an encouraging uptick in conversations around and initiatives to solve commonly-felt pain points in EA recruitment. In this post, we:
- Give an overview of what we see as currently happening in the space in order to try and create common knowledge
- Share our perceptions of potentially neglected areas of EA recruitment which we would be excited to see more people working on. Where relevant, we outline potential project ideas people could pursue, as well as risks that might come with working on them
If, after reading, you’re excited about pursuing any of the project ideas listed in this post or have ideas in a similar vein you’d like to explore further, feel free to DM Anya and she’ll try to help if she can!
The EA recruitment landscape today
We think one of the best ways the EA movement can add value to the world is by directing talent towards critical, neglected problems. This year, there’s been a notable and exciting increase in energy around outreach to new people who might be interested in EA. And lots of organizations are on the lookout for talented people to fill open roles (e.g. there have been ~2300 roles listed on the 80,000 Hours job board so far in 2022).
Earlier this year, Katie spoke to ~20 people working on various EA projects to get a better sense of the current EA community talent and recruitment landscape. Here’s what we observed across the community:
Individual organizations trying to hire people. This looks like in-house recruiters and hiring managers recruiting for their own orgs (via e.g. asking other organizations for referrals, scanning public lists for potential talent and conducting cold outreach). This also looks like open roles being publicly advertised in the hopes that the right people will proactively apply (see 80K’s Jobs Board, EA Work Club, the EA Newsletter’s ‘applications due soon’ section, the EA Forum’s Who’s hiring? thread, EA group newsletters, Effective Altruism Job Postings on Facebook, Tom Wein’s list of social purpose job boards).
Candidates publicly listing themselves in the hopes of being hired. See the EA Forum’s Who wants to be hired? thread, Pineapple Operations’ PA directory, 80K’s longtermist census).
Some candidate-focused matchmaking efforts. E.g. 80K’s career coaching program might include introductions to orgs that are hiring, and some EA groups have career coaches who mostly advise on making career plans but can nudge specific folk to apply to particular roles.
A number of organizations and affinity groups maintain CRMs and help direct people towards roles as one of their activities. Many EA groups and community-building orgs do some recruitment (e.g. managing their own CRMs, responding to various organizations’ requests for referrals for individual roles and ‘signposting’ people towards relevant opportunities). Beyond local EA groups, we’re aware of the following organizations who do this :
- CEA’s Groups team (identifying promising group organizers) and EA Global’s stewardship program (connecting promising EAG attendees to mentors/opportunities)
- The Global Challenges Project (longtermist-focused outreach to students)
- EA Consulting Network (outreach to consultants)
- High Impact Professionals (support for workplace professional groups)
- Training for Good (diagnosing skill bottlenecks in EA and delivering training programs to fill those gaps)
- EA Cambridge’s cause area seminar programs and AI Safety Talent Pipeline
- EA Software Developers (provides career coaching for developers and actively works to place them in roles)
- EA Pathfinder (provides career coaching for mid-career professionals)
- Charity Entrepreneurship (works to headhunt founders and talent for new high-impact charities and does some matchmaking for those who don’t make it through their application process)
In addition, some hiring-focused organizations are just starting up, and some existing organizations are expanding the degree to which they will focus on recruiting:
- 80K’s one-on-one team are planning to expand their work (though they also want to see more projects in the careers space)
- Non-Linear’s ‘EA hiring agency’ incubation project. In the near term, they are planning to focus on providing supporting smaller organizations without much internal recruiting capacity with operations hiring. Hiring managers and job seekers can reach out to them here.
- Magnify Mentoring recently launched a recruitment form. They plan to do more to encourage women/non-binary/trans people to apply for roles (e.g. checking to ensure they do so before the deadline)
- Tälist is a recently launched organization that provides talent matching and other recruitment services within the alternative protein industry.
There are probably more efforts we’re not aware of — one of the goals of this post was trying to create shared knowledge around existing organizations and efforts (as a resource to both job seekers and hiring organizations) so if we haven’t listed your project, please comment or DM us and we’ll add it!
Potential gaps in the space
Looking at the current landscape, we believe there may be insufficient capacity being allocated to:
- Creating strategic clarity around the biggest talent needs/opportunities across problem areas + communicating these insights to relevant people
- Centralized coordination to help make headhunting easier (e.g. a ‘one stop’ CRM of talent that both organizations and jobseekers can opt into)
- Building a talent pipeline of especially strong candidates (including targeted outreach to mid-career ‘proto-EA’ professionals)
- Developing the career capital and skills necessary to become either in-house recruiters or external headhunters
Strategic clarity
We think one issue with the existing landscape is that no one is responsible for conducting meta-level research or creating strategic clarity around talent needs. Instead, talent bottlenecks within problem areas tend to be identified in a kind of organic, ad-hoc way (see ‘EA needs more x!’ forum posts which create a burst of interest in a particular career path). By the time attention gets paid to the talent bottleneck, it’s a live concern that is holding organizations back from making progress on their goals, and people are scrambling to address them.
Developing relevant skills takes time; possibly years or decades. When it comes to responding to evolving talent needs in a small community of young people, more of a heads up would be helpful.
We think it might be helpful to have a person or group of people with a primary focus and responsibility to conduct research on EA talent gaps. i.e.:
- Talk to EA leaders to forecast how problem areas might develop over the next 2-5 years (i.e. volume of new organizations that might be launched, funding supply, areas of growth, etc.)
- Run a survey with existing orgs to assess bottlenecks and forecast demand
- Figure out how to model expected supply of talent (e.g. by surveying community members about their needs and current + future plans, conducting ‘talent mapping’ type activities to calibrate on the quality of talent etc.)
- Communicate these insights to the community so that other EA groups and organizations working on community-building, recruitment and training can benefit.
In previous years, CEA and 80K have surveyed EA orgs on their talent needs, but (according to representatives of these organizations we spoke to) it’s not currently a priority for either. This year Training for Good explored taking up this mantle and are currently looking for someone to drive it forward.
We’re uncertain how tractable it is to conduct this kind of high-level research for the EA community as a whole versus prioritizing and focusing on a limited number of specific problem areas (or conducting such research on the level of each cause area), but we’d be curious to see further investigation.
Coordination between hiring organizations
We’re not sure what the optimal level of coordination between hiring organizations in the EA community is, but we suspect it’s more than exists today. Increased coordination (e.g. an organization focused on headhunting across cause areas, or a cluster of cause-specific headhunters who are set up to coordinate with one another) could reduce duplication of effort and save candidates/recruiters time and energy. Below we outline some pain points that stem from today’s low levels of coordination and how they might be improved by having more centralized hiring services in the community (and then we talk about some potential projects and risks).
- 1. Candidate time and frustration with applying and reapplying to many EA organizations. Some people in the community have reported frustration with repeatedly applying to many related organizations (many of which have unusually time-intensive application processes, involving multiple rounds of work tests), not feeling calibrated about what their odds are of being hired at any one of them and sometimes not receiving much feedback (see discussion here and here). Hiring organizations, on the other hand, are often capacity strapped and may struggle to provide candidates services such as in-depth feedback or matchmaking to other roles.
- How more coordination could help:
- An EA “hiring agency” (similar to traditional staffing agencies) that worked with organizations across the community could ease some of these pain points by more proactively working to match candidates with roles. By working with the same set of organizations over time, they would learn a lot about their hiring standards and preferences and help job-seekers narrow their list and identify the roles and organizations that are most likely to be a good fit for them. Since they would be working on behalf of the candidate as well as the organization, they could do more to provide them with some feedback and coaching throughout the process (and take the labor and legal risk of conveying feedback on performance back to candidates off the hiring organizations).
- An organization could run a centralized “common app” and provide at least basic data with a bunch of organizations, who could then decide who to reach out to.
- An even more maximally scoped organization could take some of the work of candidate vetting (e.g. interviews or work test administration) off hiring organizations for roles that tend to vary less across organizations (e.g. operations roles), thereby dramatically reducing e.g. the number of applications and work tests individual candidates needed to participate in.
- 2. Information sharing about strong candidates between organizations tends to be ad hoc and inefficient. Currently organizations mainly contact each other one by one to request referrals from past open rounds (in Anya’s experience the OP recruitment team gets ~5+ requests for referrals from other organizations per month, and often sends just as many). This creates many nodes and points of contact to keep track of; each request is time-consuming to respond to, and many fall by the wayside in favor of more urgent work. This ad hoc approach also creates privacy concerns around candidate data and makes it harder to establish a principled set of norms when sharing candidate information between organizations.
- How more coordination could help: It would streamline the process considerably if recruiters could feed information about strong candidates back to a single organization or working group (and then ask that same organization/group for recommendations when they have an open role, trusting that that organizations’ referrals have already incorporated suggestions from a variety of other organizations in the space). Reducing how much ad hoc information sharing about candidates is taking place could also have benefits for candidate privacy — organizations would be asking candidates for permission to share their information with just one entity, making it easier to set clear expectations about what is being shared and how it will be used.
- 3. Individual hiring organizations are incentivized to do the bare minimum, especially when it comes to sourcing and vetting unknown candidates. It doesn’t make sense for individual organizations’ recruiting teams to attempt to sort through and take calls with all the possibly stellar and interested people in the world just to fill just the roles they have open. Sourcing and vetting unknown candidates is extremely labor-intensive work and in-house recruiting teams tend to be busy and under time pressure to fill roles as quickly as possible. This creates a tendency to focus efforts on the sourcing that seems “safest” from an ROI perspective (which might mean e.g. saying no to lots of calls with people who look interesting on paper but who we don’t have much signal about).
- How more coordination could help: There are lots of other things that would be an unreasonable strain on an in-house team’s recruiting capacity that would make sense to do on behalf of a long list of client organizations as the odds of finding a match will be much higher. A community-wide headhunting agency could undertake activities that are currently being duplicated across many organizations and do a much more thorough job, potentially helping the community build a larger pipeline.
Brainstorm of projects
Here’s a list of projects we think could be promising; these could sit under one organizational umbrella, but don’t necessarily have to:
- Building an up-to-date CRM: a list of potential candidates and their hiring-related information (Linkedin data, skills, roles they’re interested in etc). Whilst this would only be as useful as it is good, we’re interested in the claim that there’s a potential gap for someone to create, manage and use a community-wide CRM built for headhunting
- Re-routing candidates who made it to the final stages of hiring rounds (but don’t get the job) to other relevant orgs
- Developing talent pipelines for specific problem areas, including conducting proactive outbound outreach to potential talent and having screening calls (especially with ‘new/unknown’ but potentially stellar candidates) with the intention of placing them in open roles
- Systematically routing potential candidates to relevant orgs (including figuring out a way to collect, store and share references which could help hiring managers uncover cruxes about candidates earlier in the process)
- Figuring out how to gather feedback from hiring managers and deliver it to candidates to improve the candidate experience
- ‘Talent mapping’ of existing staff at EA orgs and developing relationships with them so that if they’re seeking a role change, they can receive support
- Contributing to the headhunting of top candidates from diverse backgrounds to increase the diversity of orgs’ hiring pipelines
Potential risks
Creating a more coordinated hiring or vetting organization dramatically raises the stakes of that one organization’s work and raises the cost of mistakes. Some ways this could go wrong:
- A centralized vetting organization could consolidate power. If the process inadvertently optimized for the wrong things but hiring organizations took the assessments seriously, this could lead to worse hiring decisions by many organizations. (Though it seems more likely in the short run that such an org would just struggle to create a process that a large number of organizations felt they could trust and buy into.)
- With any kind of centralized information collection, communication around candidate data would need to be handled very sensitively in order to avoid creating a scenario in which candidates are disincentivized from using the service at all (out of real or imagined fears that it would damage their reputation or get them rejected from all EA organizations at once vs. having a bunch of different chances to impress specific organizations).
- It will be hard for any centralized CRM to keep up with community growth and stay organized and exhaustive for very long; it may just inevitably devolve into a bunch of more niche CRMs maintained by different organizations.
- In general, this is difficult work with large overheads; it requires balancing lots of people’s interests and building trust networks among stakeholders whose incentives are not aligned (e.g. hiring organizations are competing with each other for the best talent; candidates are competing with each other for the best roles; candidates would probably prefer less information sharing about their performance if it could hurt their chances while hiring organizations want to avoid mistakes). We could imagine it generally just being hard to get something sufficiently high quality and trusted off the ground.
(For more discussion on the risks and benefits of a Common App specifically, see discussion here and here. For more on the benefits of more coordination with EA generally, see this post.)
Disclaimer: These are just some rough ideas. Before moving forward with any kind of initiative to coordinate hiring activities across organizations, people would need to ensure they were in compliance with data privacy, antitrust and employment laws in all applicable jurisdictions.
Finding and cultivating especially talented candidates
Beyond more effectively allocating existing talent, we believe that there’s value in experimenting with trying to proactively identify the candidates currently most highly valued by relevant labor markets, who are not familiar with EA ideas. These candidates can be classified as generically “hard to hire:” they have already established career capital outside of EA and have access to numerous exciting opportunities and are unlikely to be looking to leave their current job. While it’s true such people might also be less open to EA ideas or pivoting their careers, the upside of successfully moving them into impactful roles may be high enough to justify some extra effort to convince them.
Anecdotally, we’ve seen several examples of people in this category who successfully transitioned out of the business world and into EA careers (despite having little EA context to start with) and very quickly began adding large amounts of value which younger, highly engaged EAs would be unable to replicate for years. We would love to see such success stories replicated more often.
Brainstorm of projects
- Creating more custom ‘entry points’ for mid-career professionals with targeted messaging for this demographic. Mid-career professionals face particular challenges entering the EA community; they don’t have ‘EA capital,’ may be poorly networked and might feel alienated by current messaging, much of which is targeted towards students or people who haven’t already deeply invested in a career path.
- Headhunting proto-EA mid-career talent with relevant skills and experience for suitable roles. We expect most successes to be more ‘high-touch’ and networks and relationships-based than usual; this would likely mean working outwards from who is already in the community.
- A service could be developed to advise EA organizations on how to do effective “values-vetting” interviews with mid-career professionals in the later stages of job applications without relying on superficial “community involvement” indicators (and without creeping the candidate out). The goal would be to assess whether a candidate has the “proto-EA” orientation required to be easily assimilated into EA orgs.
- We think more ideas could be generated by studying best practices from other industries in which competition for top talent is the norm (this would be another project someone could take on).
Beyond especially capable mid- and late-career professionals, we suspect this demographic-targeted headhunting approach could also be replicated to source and recruit excellent candidates from other underrepresented backgrounds.
Potential risks
- It may be that getting such people into EA roles is so difficult that it’s not a good use of anyone’s time to focus exclusively on this, and it should just be done opportunistically (via connections and relationships). (When we checked in with 80,000 Hours, they suggested that this consideration was part of what caused them to back off this type of work, but they also said they’d be interested in seeing more people experiment with it.)
- EA organizations may be justified in being wary of candidates who didn’t self-select into the community; bringing in mid-career or senior people who are e.g. not primarily motivated by altruistic considerations, or have been steeped in an environment where values like truth-seeking are not encouraged, could dilute positive things about the movement’s culture we’d rather preserve.
- That said, we think EA organizations today might have an unreasonable degree of risk aversion when it comes to hiring highly talented individuals who are not culturally embedded in the community. Some candidates really do have incompatible goals and values; some will just make a lot of insider-baseball faux pas when interviewing with an EA organization because they’ve never spent time with EAs before. There are definitely risks and trade-offs to consider here (and for some roles not just alignment but EA cultural context is important on day one). But there are also tradeoffs to passing on people with hard-to-find skills due to risk aversion about “culture fit.”
Skillbuilding to do this work
We believe that, as a community, we need more professional recruiters and headhunters to figure out how to place (or facilitate the placing of) suitable candidates in these open roles. This is a smaller point, but being included for completeness: in our experience it currently seems quite hard for EA organizations to hire aligned talent who already have recruiting skills and a solid understanding of best practices, and we don’t think we’ve ever read this advice written down. (This is based on Anya’s experience hiring at Open Phil and anecdotal evidence from talking to other organizations who were trying to hire recruiters.)
In our experience, being a good fit for generalist operations work doesn’t perfectly map onto being a good fit for recruiting/headhunting/hiring. We think people who have already done this work in a professional context could bring unique value to EA organizations. (E.g. The following orgs are currently seeking recruitment talent: Recruiting Manager at Open Philanthropy, Recruiter at Ought, multiple roles at Schmidt Futures, multiple roles at Anthropic.)
This isn’t a path that is likely to make sense for tons and tons of people, but we believe it would, on the margin, be quite helpful for some people drawn to this work to get jobs at high-end talent acquisition firms, skill build for a couple of years, and then come back to the community significantly more able to help. (See also: this recent Recruiter job profile, and this post about the value of recruiters).
Strongly upvoted. This kind of post is extremely valuable and I would encourage anyone to write more of these kinds of posts: It helps us as a community to coordinate much better. Thanks Anya and Katie for all your efforts and research putting this together and sharing it.
Reg. "There are probably more efforts we’re not aware of" There are EA-aligned newsletters including job openings relevant to that newsletter's scope, e.g. like this.
Reg. "We’re not sure what the optimal level of coordination between hiring organizations in the EA community is": I guess the different points of the spectrum is something like
1. Fully uncoordinated hiring agents (uncoordinated & decentralized)
2. Well coordinated network of hiring agents (coordinated & decentralized)
3. Having one big "EA hiring agent" (centralized)
Ideally we would be at 2. (I'm uncertain whether it's better to tend towards 1. or tend towards 3.)
3. Has many downsides (incl. the ones you've mentioned) whereas the prototype for 2. could look something like this:
*I'm less certain about what the hiring scope of each agent ought to be, but it seems best for different agents to optimize for different hiring scopes (cause area, specific org, anything that the EA community might need in the future...)
Please see this comment as unstructured, half-baked thoughts I had after reading this post and wrote in a couple of minutes vs. a well-thought idea.
Thanks for the kind feedback, Christina and for sharing your thoughts!
And I think I generally agree that the optimal place is somewhere around 2, with some centralized provision of certain goods and services that are valuable to lots of decentralized hiring agents.