.impact's pivot to focus projects

The views in this piece are primarily of Peter Hurford. While the piece was reviewed by several members of the .impact community, the community is large and diverse, and views expressed here should not be taken to be any sort of consensus.

When I started .impact with Ozzie Gooen in the summer of 2013, we were focused on solving a key problem that we saw in EA -- there were easy opportunities to donate your money or your career, but there was no outlet for donating your spare time toward effective projects. We wanted to give new EAs an opportunity to get more involved after they arrived into the movement, which we saw as a neglected problem in the EA community. In short, we were aiming for effective volunteering.

However, in our 2014 review, we found that our case for impact lied almost entirely in our yet-to-be-realized potential as an organization, that the supply of volunteering was more limited than expected, that our projects were more difficult than we originally assumed, and that it was difficult to predict how useful our projects would be.

Since then, we’ve had a change of stride. Initial outcomes of our projects seem much more useful and we’ve developed better working processes to improve the reliability at which these projects get done. In this review, I explain what our current projects are, how we’ve improved our outcomes and processes, and what work remains to be done.

.impact projects

Some of our projects have panned out to be successful, though we still have not yet evaluated them rigorously.

By far our biggest project, The Local Effective Altruism Network (LEAN) has seeded over 100 new EA groups. They’re also supporting new groups to grow and take on new activities, from a first social meetup for new ones to fundraisers. We’re in the middle of evaluating LEAN more rigorously through using the EA Survey and an additional custom study of group leaders to understand the outcomes of EA groups.

The EA Newsletter, which we collaborate on jointly with EA Outreach, now goes out to over 7,000 subscribers and we’re figuring out how to use it to drive more engagement with EA, such as encouraging people to take part in key activities, donate for the first time, take the GWWC pledge, take the EA Survey, run a Christmas fundraiser, etc. Despite having such a large subscriber base, the EA Newsletter has a 37.6% open rate and 8% click-through rate, which is much higher than the MailChimp non-profit average of 22.9% open and 2.4% click-through.

We’ve also run the annual EA Survey for two years so far. This collects and analyzes information on the demographics, beliefs, values, and results of the EA movement. We think this could provide valuable longitudinal data, helping expand our understanding of the growth of the EA movement. The 2014 EA Survey had responses from 813 EAs and the 2015 EA Survey already has over 3000 EA responses. The EA Survey has also proven to be a key community tool for establishing contact with new EAs and tracking our collective donations and actions.

We’ve also gained a lot of signups at the EA Hub, which now has over 3,500 EA Profiles and over 500 recording donation plans on the EA Donation Registry. We also have a map of over 1,700 EAs and over 200 local EA groups.

We also finished collaborating with Mercy for Animals on a large-scale RCT of online ads, which Animal Charity Evaluators has called "the highest quality randomized controlled trial (RCT) so far of an animal advocacy intervention." Unfortunately, we still ran into inevitable problems with sample size and did not produce as definitive of an answer as we had hoped.

Lastly, we’ve maintained and improved the EA Forum, including responding to bugs and issues.

The Focus on Focus Projects

The above was accomplished in 2015 with a very different workflow than we used for 2013 and 2014. Our initial model was to hold weekly (and later bi-weekly or monthly) project meetings where people would come with their own projects and get community support in building and launching them. Unfortunately, many of these projects did not end up coming to fruition and many of those that did end up existing did not end up having as significant an impact as was hoped for, often due to low user uptake.

We then switched to a different model, which we called the workathon, which we hosted 12 in July through October 2015. We aimed to create a regular and large block of time for people to work on their projects rather than merely talk about them. This appeared more successful than before, but we still had trouble retaining people (presumably since it was a large commitment) and we had trouble matching the people who came with tasks they could do.

We also tried our first attempt at an online conference, called the EA Assembly, which was fun to put together and had some good discussion, but it wasn’t as well attended as we would have liked and a few of the speakers bailed on us at the last minute, which was unfortunate.

Our most recent model has been quite different. While remaining open to people running their own projects, we now also have a focus projects team which acts more similarly to most other EA organizations, where the leadership decides which projects they think are most important and proactively aims to create and fill roles for people to help with those projects. We also emphasized looking for paid, full-time and part-time staff rather than volunteers who could only contribute a few hours a week and we saw a dramatic increase in the reliability and speed at which projects were moved forward.

2016 also makes our first year raising money to support our operations. We now have a budget of $67K, supported by 25 different donors.

The .impact of the future

I see the move toward focus projects as a significant step forward for .impact and I think it will remain the large core of .impact for 2016.

However, I don’t want to lose our grassroots roots and I’d like to explore opportunities to continue to engage the EA community in ways that make sense, perhaps with the return of workathons (albeit with no promise of being regularly occurring) and with the continued advertisement of jobs and internships as we find them necessary.

We’ll also continue to support the wider community through our existing infrastructure. We invite anyone to use our Facebook group, our slack channel, and our Hackpad to coordinate. An example of a grassroots project using this is the the AMF Project for Awesome team successfully did to go on to raise $25K for AMF, though we wouldn’t overstate how much our infrastructure helped with this. We’ll also continue to broadcast regular updates about the wider .impact community every two months.

.impact will also continue to be open to incubating any other project and team. We’ve started doing this by hosting our second paid team, SHIC, which will do outreach in high schools, and has totally independent management and funding from .impact’s focus projects.

Lastly, we’re also providing tech services (including free hosting) for valuable projects. In addition to helping SHIC, we recently helped the launch of EA Australia by building their website and online donations system.

Overall, I believe in 2016 we can continue to reach out to the broader EA community and help them out while maintaining good progress on core EA projects.