Supporting online connections: what I learned after trying to find impactful opportunities

By Clifford, Sarah Cheng @ 2022-11-04T17:34 (+65)

I’ve been exploring ideas for how the Forum could help people make a valuable connection over the course of 4+ months. In this doc, I share the different areas I’ve been thinking about, how promising I think they are and what I’ve learnt about them. 

Overall, I don't think any of the ideas meet the bar for focusing on connections via an online platform. My bar for an idea was that it could facilitate roughly an EAG’s worth of quality-adjusted connections each year (~10,000). I’m now more excited about meeting this bar by targeting high-value outcomes (e.g. collaborations, jobs), even if that means a lower number of total connections.

I cover:

1. Overall take
2. Why did we explore connections?
3. What ideas have we explored?
4. FAQ: Why don’t you build a year-round Swapcard?
5. Other ideas we didn’t explore

Overall take

I’m more excited about targeting high-value outcomes that come from connections (e.g. collaborations, jobs) rather than trying to broadly facilitate conversations online.

• I wanted to find a product that could plausibly match a single EAG (Effective Altruism Global) in terms of the number of connections made in a year.
  • A single EAG costs roughly 2 FTEs a year to run (excluding other costs).
  • If Sarah (my colleague) and I are choosing what to work on, you could imagine that the alternative we’re trying to beat or match is running another EAG.
  • Note: this is a relatively ambitious target. I think it would be more achievable to build something that’s still cost-effective but which has less potential for scale. For example, some of the ideas I explore below would be much more cost-effective than an EAG, because it's much cheaper to run online services than in-person conferences but would generate far fewer connections. [Related Forum post.]
• I’ve found it hard to compete with EAG on number of “connections”, where a connection means someone you feel you can ask a favour of (likely after a half-hour conversation).
• At an EAG, each attendee makes an average of 8 connections via lots of 1:1 chats in an environment (in-person) optimised for good conversations.
• To match this, we’d need 10% of our monthly active readership to make 4 connections a year, which isn’t crazy but feels like a stretch as only 8% of that number have ever sent a message on the forum.
• I think it might be easier to compete with EAG on the downstream effects of connections, e.g. a person ended up collaborating on a paper, a person ended up getting hired.
• My best guess is that EAGs get outcomes like this for something like 1-5% of connections. I think it’s much easier for the Forum to get 100-500 additional concrete outcomes (like collaborations or jobs) a year, than 10k connections.

The rest of this contains details on what we explored, why and our assessment of each idea.

Why did we explore connections?

We thought we should work on connections because:

1. Connections are valuable
   1. They are cited as an important reason that people get involved in EA in the Open Phil EA/Longtermist Survey
      1. 35% of people said personal contact with EAs was important for them getting involved
      2. 38% said personal contacts had the largest influence on their personal ability to have a positive impact
2. We had some evidence that people are making valuable connections through the Forum, even though the Forum wasn’t designed well for this
   1. The 2020 EA Survey revealed that a surprising number of people found connections through the EA Forum
      1. 5.9% of people said they found a valuable connection through the Forum, where 11.3% said this of EAG (an event optimised for connections)
         1. Having followed up with a subset of the forum users who said this, I think it’s more like that ~2% of respondents found a valuable connection
         2. That said, it wouldn’t surprise me if we saw a similar reduction in reported valuable connections for EAG if we were to follow up with those people. I just imagine that a bunch of people forget what they were referring to in a survey they took two years ago and fewer connections seem valuable with hindsight 
   2. People in our early user interviews said they found some valuable connections
      1. We reached out to some people who had sent many messages on the Forum
      2. We spoke to 17 people and learned that people had found project collaborators, spotted new talent, and had people reach out to them for advice
3. If we could deliberately seek to enable more connections through the Forum, perhaps we could facilitate a lot more connections

What ideas have we explored?

Caveats:

• Sometimes my reasons for not exploring something are weak and reasonable people disagree (including my teammates)
• I suspect there are interesting angles to take on the ideas I’ve dismissed
• Sometimes I say that I’m less enthusiastic about an idea because I think another group is working on it. In these instances, I don't necessarily think that more work is useless; competition can be really good, and we might be wrong about how much space existing projects are taking up.
• My bar was to find an idea that could turn out to be as valuable as running one EAG a year. This is a high bar and arguably some of these ideas are still worthwhile pursuing though they fall below it, especially considering how cost-effective they are.

I’m sharing a table summary below, and you can see elaborations below. 

FAQ: Why don’t you build a year-round Swapcard?

This has been suggested a number of times by different people, because:

1. People use Swapcard during EAGs and this is seen as valuable
2. Offering the same service year-round might therefore be even more valuable
3. EA Hub was like a year-round Swapcard and was used a bit (~1 message sent/day)
4. Perhaps the EA Forum could do an even better job than EA Hub did (because it has more users and traffic)

My view on this:

1. The biggest issue with building a directory is getting people to actually use it.
   1. I think this is the biggest problem with many products.
   2. See the “community map”, which only has a couple of messages sent a month.
2. A timebound conference creates an artificially highly liquid market for conversations.
   1. Demand - people have set aside two days to do nothing but talk to people. They request conversations they would never normally have the impulse to book.
   2. Supply - people have set aside time to respond to requests and are more receptive to speak to people.
   3. The fact that people don’t use Swapcard much outside of the conference is evidence for this.
3. Outside of a conference, I think it’s much rarer that people think “I should schedule a chat with X to get advice on my career” (although it does happen).
4. So I think it’s important to design a directory around a burning user need that people have outside of conferences (e.g. “we need to find candidates for a hiring round”).
   1. I also think the features will look different depending on the user, and we should build around that.

I think there is also an argument that goes:

1. Don’t build a directory for a specific use case (e.g. getting specific career advice).
2. Instead build a general-purpose directory that will be used for multiple purposes e.g. making friends, reaching out about project ideas, getting general advice and recruitment.
3. This will benefit from network effects (you’ll have more people using this).
4. You may also observe people using the directory in ways you don’t expect.

I feel pretty tempted by this argument, and it’s nice to build stuff, but I think without a specific use case you’re kind of just hoping people will use your new service.

It might be worth saying that it is possible to search for Forum users by keywords in their profile and any tags they’ve added to their profile here — and you can message people! If this gets a lot of use and I learn of valuable outcomes from this search, it could change my mind about whether we should spend more time on something like a general-purpose directory.

Ideas explored in depth

We explored these ideas in some depth, building prototypes that people used. All these ideas showed some promise, so I could imagine changing my mind about them.

Get cause-specific career advice

Why? 

When people are looking for advice they often want to speak to people who know the cause area. One of the ways that 80,000 hours provide value is referring coachees to cause area experts who can bring them up to speed and make further introductions. It seemed plausible that it would be useful for people to reach out to these experts directly.

This was one of the use cases I tested most thoroughly. We ran a pilot to connect people interested in biosecurity with more senior people in biosecurity (currently doing PhDs or working in the field). If you’re thinking of doing something like this, I’d be happy to share more info and discuss.

Take

I think this was a pretty useful thing to do, but the results after running for a month weren’t amazing and I couldn’t see a way to scale it up to be EAG-level good.

Others at CEA disagreed with my overall assessment of the pilot and thought the service could be worth continuing to run, even if at a small scale.

It’s possible someone else should take this on as a side project, or individual interest groups (e.g. consulting, psychology) should have something like this.

Evidence

• I asked someone senior to review the talent attracted and outcomes in terms of plan changes and they thought it was approximately break-even on advisors’ time.
• On the other hand, there was demand from advisees, advisors enjoyed the calls and there were some signs of minor plan changes.
• It's possible that the time the advisors spent mentoring didn't trade off against working hours, in which case there was less downside.

Get advice from a professional

Why?

We’d heard of people reaching out to forum authors to ask for advice. I think getting career advice from a professional is one of the common conversations people have at EAG. I had a guess that people would reach out if they felt they had permission to. We were very close to launching this, but then decided the numbers wouldn’t be large enough for it to be worth it.

Take

There was a real demand for a professional advice program, but I don’t think enough to warrant scaling one up. I think this was worse than the biosecurity advice pilot we ran, because we had no filter on who could apply and so people took it less seriously.

Evidence

We did a cheap test by listing people in different careers on a forum post: “Book a chat” with an EA professional. The results looked quite a bit worse than the biosecurity advice pilot:

1. There were fewer exciting people who weren’t already involved in the community.
2. There were fewer reported plan changes.

That said, it was promising that some subset of advisors were happy to do many calls and felt they were a decent use of time. I think there’s something to be said for giving people permission to reach out to the subset of senior people who enjoy doing these calls. A counterargument is that it’s possible these senior people aren’t valuing their time highly enough and we’re distracting them. It should be noted that most of them did not want to spend more time having EAG-like conversations (they saw EAG as their give-back budget).

I also looked into how the EA London directory was used, because it seemed like the best existing example of the get advice service. It seemed like it was most often used when someone was going through a career transition (e.g. finishing a PhD or switching from Finance to Software) and wanted to speak to someone who worked in that area to get practical advice and leads for jobs. This got a surprisingly low uptake though, given it was heavily promoted by the EA London community leaders. We estimated that 40 messages were sent a year and that 2.5% of visitors to the page sent a message. If 2.5% of all monthly Forum visitors (25k) sent a message, this would be 500 messages a year, which seems optimistic and wouldn’t be close to our target. We’d need about 8x as many messages.

Connecting with someone who lives nearby

Why?

This seems like one of the key features of the EA Hub. I think having a friend who is interested in EA can be very motivating and matching people can lead to new EA groups. So we created a map for Forum users who wanted to share their approximate location.

Take

The feature hasn’t been used much. It’s possible we should delete this feature, as it’s distracting to have.

You could argue that we should have done more to promote it and worked harder to get more people on the map - the usefulness of this is of course only as valuable as the number of people on it. But I think we did a decent job promoting, and demand is just genuinely low.

Evidence

We created the map in May, and a couple messages a month are sent through the map.

Ideas explored in medium depth

We spoke to a decent number of users and/or ran tests, so we feel somewhat confident about our positions on these ideas.

Advertising jobs (still exploring)

Why?

We know posts on the Forum have played a role in people applying for jobs, and we think this might be one of the main sources of the Forum’s impact.

Take

I think we could play a role in generating more leads for recruiters. I particularly think there might be an opportunity to do this with EAG attendees as well as the Forum. It seems like there are a decent number of people who aren’t aware of jobs that are a match for them or can be nudged to consider jobs.

Evidence

• Top-level posts on the Forum and a couple of posts in the “Who’s hiring” thread led to people being hired.
  • We’re still gathering data on this - if anyone is interested in the details, feel free to message me.
• It’s too early to say whether anyone has actually been hired from our recent experiments, but we know people have applied.
  • We sent a survey asking software developers if they were interested in 4 jobs — 3/12 applied for a job we listed.
  • We sent a similar survey asking EAGxRotterdam attendees interested in entrepreneurship to consider applying to jobs and 13/100 emailed applied to a job.
  • We posted job ads in the comments on the Forum - these mostly didn’t get many clicks and we think it's unlikely they resulted in new applicants, though we are waiting to hear back from the hiring managers.

Recruiting candidates (still exploring)

Why?

Helping hiring managers find counterfactual hires seems particularly valuable. We’d heard of people spotting talent through well-written Forum posts. We also know that the EA Hub was used by hiring managers to source candidates.

Take

I think there are a subset of hiring managers who are very excited about new sources of leads. Others have a relatively easy time hiring and don’t need additional help. There are a number of recruitment projects starting and I suspect they will fill the gap, but I think we may be in a position to help provide lists of candidates for recruiters.

Evidence

When we spoke to 10+ hiring managers about recruiting people, it didn’t seem like we found a hair-on-fire problem. About half were uninterested in receiving a long list of candidates. A couple were very enthusiastic. This mixed reaction, paired with it looking like other people were going to enter the space, made us back away.

That said, we still think CEA’s data (from people who have agreed to share it with other organisations) would be useful for recruiters (though perhaps not as its own product) and we are working on making data available to hiring organisations where users have explicitly opted-in to sharing.

Help online group organisers get their members to engage

Why?

I spoke to a couple of people who made valuable connections through Slack groups. We wondered if we could help online group organisers better connect people online. The theory being that perhaps individual members of an online group (e.g. EA Creatives & Communicators) don’t have a burning need to better connect, but the group organisers might have a stronger desire to help people connect.

Take

There was some interest from online group organisers in having an alternative to the many Slack groups. It seems plausible that the EA Forum could host this, and this is being experimented with currently (see Subforums).

Evidence

Some online group organisers we spoke to were interested, and subforums are getting some use.

Help community builders connect members to the broader EA community

Why?

We explored how we could help group organisers connect their members to the broader EA community. I was excited about this based on chatting with David Nash, who connects new EA London members to relevant people in the community on a daily basis. David uses his personal LinkedIn to do this so it seemed like replicating this could be valuable.

Take

Unfortunately, after chatting to 20+ group organisers, it turns out:

1. They don’t connect as many new people as David.
2. They prefer to connect people within their existing network.
3. Some prefer to connect people within the group.

However, I think smaller university groups that have leaders without an established network could benefit from some kind of directory. It's possible that the CEA groups team will take something like this on in future, depending on other priorities.
 

Evidence

1. We spoke to larger local groups and national groups, and mostly they prefer to connect people to each other locally
   1. We spoke to UCL EA, EA Cambridge, EA Netherlands, EA DC, EA NY, EA London
   2. Prefer to use their own network because it’s easier to make the ask
   3. Prefer to introduce someone people can reconnect with easily/bump into locally
2. A genuine opportunity is providing a network of experts for smaller local groups to refer members to
   1. We spoke to two smaller local groups (EA Durham, EA Warwick) who lacked their own personal network and lacked experts in the local area, and so were very enthusiastic about being able to connect students to knowledgeable EAs who they knew were open to chat
   2. A spreadsheet like this used to exist among local group leaders a few years ago, and it was used

Ideas explored in shallow depth

We only had a couple of calls about each of these ideas, and I could imagine changing my mind about them.

Verifying someone for a grant

We thought the Forum could be helpful for verifying grantees and seeing their work, but this didn’t seem very important after chatting with a few people. At best it saved grantors some time, but didn’t make the difference between granting or not.

Finding projects/grantees for regrantors

We thought it might be helpful to source potential projects for regrantors to fund. We were excited about regrantors because there are a lot of them. After speaking to a few, it seemed like most were not interested in spending additional time seeking out opportunities, and instead funded projects that they came across organically as part of their work.

Help people find jobs

We spoke to 6 people sourced from the “Who wants to be hired” thread and people who ticked “seeking work” in their Forum profile. Mostly people seemed to feel ok about their job hunt and didn’t seem to need additional help. I’m a bit surprised by this as I would have guessed that finding an EA-related job is hard. That said, the sample size was very small so I wouldn’t be surprised if we changed our mind on this. 

Connecting peers with mutual interest

There are a number of existing initiatives to help people connect around interests e.g. online Slack and Facebook groups, someone in the EAG DC Slack created a channel for software engineers.

One user we spoke to wanted to meet more peers (students) interested in biosecurity because there were a limited number in his university. We were pretty excited about this and contacted 16 students who were interested in biosecurity to ask if they wanted to connect with him. He had calls with 3 of them and got value out of the conversations. I got the impression that this provided a different kind of value to talking to someone more senior as

1. You have more space to ask dumb questions and express doubts
2. You’re more likely to build a mutual relationship that’s valuable for your network

In response to this, we added a list of “peers” in biosecurity on the biosecurity advice page. Unfortunately, no messages were sent on the back of this section. This doesn’t feel like a super fair test because we didn’t advertise the page very well (it has had ~500 unique page views since we launched and 38 profiles were viewed), but I think getting traffic to this page would have been hard and even if we did well, we could probably expect to get a similar number of messages sent as the community map gets (a couple a month).

I could imagine the subforums helping with this. Perhaps also EA chats would be useful.

Help people figure out which jobs are the highest impact

Yonatan has highlighted problems with 80k’s job board here. I thought the Forum might be able to help by creating a space for discussion around which jobs were the highest impact. I spoke to Yonatan more about this and to Kush, who manages the 80k job board.

Overall I think it would be good for 80k’s job board to be clearer about which jobs are good for direct impact vs career capital and to highlight which are controversial, but I’m not convinced it would make a big difference to what jobs people take.

Other ideas, not explored

Here’s a list of ideas we didn’t explore because I wasn’t excited, but possibly we or someone else should explore them in the future.

Build a network with algorithmically assigned chats

• “EA Chats” was done in the past and sounds like it went quite well. Vaidehi was happy for us to survey participants to find out what the outcomes are 3 years on. It’s possible she’d be happy for others who are thinking about doing this to do the same.
• I feel less optimistic about this than other people seem to. I’m sceptical that algorithmically assigned chats (even if based on interests) would turn out very well. I also think that people tend to drop off these services over time.
• I’m happy that EAGxVirtual and EA Anywhere facilitate this.

Help people find co-founders for projects

• This doesn’t seem to work in the for-profit world and I don’t expect the EA world to be much different.
• I don’t think matching is the bottleneck for founders.

Make friends with people with similar interests

• Seems potentially very valuable to link people without a local group.
• That said, I’m sceptical of how well online friendships will do.
• I think EA Anywhere and EA Gathertown partially address this need.
• Maybe reading groups are the best approach? It would be interesting to follow up with Virtual Programmes participants to ask.
• Some people are experimenting with a slack channel called “Study Buddy” which connects people who are interested in studying something together. I think this seems like a decent approach.

Get advice on a thesis topic

• I would guess Effective Thesis has this covered.

Group organisers want to know which speakers they can invite to their group

• This seems helpful to me, but it’s not a way to connect lots of people.

Professionals want to build their network to find jobs

• I like the two ways we have to do this:
  • Networking events, where attendees have to clear a bar to enter.
  • Forum posts where you establish credibility through the quality of your writing.
• There may be others.

EA related orgs want to know recommended professionals for outsourced jobs

• Orgs might want to find someone to do a specific task e.g. build a simple website or give legal advice.
• I don’t think this is great for connecting lots of people.
• A nice feature would be that professionals who are building career capital can have a way to stay connected with EA.

Following up with contacts made in person

• Maybe making it easier to follow up with people post-EAG would make EAG 10% more valuable?

Reaching out for a favour

• One Forum user said he preferred to reach out to people on the Forum because it felt more personal. Maybe we could facilitate more interactions?
• Maybe there are missed opportunities, where people don’t know each other’s emails, and so don’t get in touch.

Other

• Ben West had a bunch more ideas in this post: Creating Individual Connections via the Forum.