Runway till January: Amplify's funding ask to market EA & AI Safety

By gergo, Paul Dodd, Amplify @ 2025-11-18T19:22 (+30)

Summary

• Amplify is an EA-aligned digital marketing agency that has supported 40+ EA and AI Safety fieldbuilding organisations with subsidised or pro bono marketing.
• We secured over 3000 counterfactual expressions of interest to join EA Intro & AIS programmes, helped multiple groups recruit highly engaged participants, and delivered other cost-effective results such as ~$570 per GWWC pledge and ~$1700 per MARS fellow.
• Marketing remains heavily underinvested across the EA ecosystem, and many groups lack the expertise or budget to run effective outreach. For every $1,000 donated, Amplify can support one additional fieldbuilding organisation with end-to-end campaign setup, guidance, and post-campaign analysis.
• Institutional funders have not prioritised marketing, so we are turning to the community to keep Amplify running. Without additional support, we will have to pause or shut down operations in January.

If you think the world would benefit from more people taking EA and AI Safety ideas seriously, we would deeply appreciate your support. You can donate to our Manifund page here, or vote in the upcoming donation election of the EA forum. (But we recommend reviewing other organisations too, so you can make a fully informed decision!)

Who is Amplify?

Amplify is an EA-aligned digital marketing agency supporting Effective Altruist and AI Safety fieldbuilding organisations in attracting talent and building a strong, professional online presence.

What is our track record?

To date, we have provided pro bono or subsidised digital marketing support to more than 40 EA communities, such as EA South Africa, EA Finland, AI Safety Dublin, EA Serbia, EA Sweden, EA Germany, EA Tilburg, EA Madrid, EA France, EA Netherlands, EA Oxford, PISE, Cambridge AIS Hub, Indian Network for Impact, etc.

We have also supported the Centre for Reducing Suffering with a book promotion as well as Mieux Donner (French effective giving organisation, MCF grantee) with a newsletter growth and fundraising campaign..

Highlights

• Over 3000 new expressions of interest to join EA Intro & AIS programmes generated from our campaigns.^[1]
• Amplify has set up a campaign for EA Netherlands’ intro to EA Fellowship. This campaign attracted three participants, two of whom signed the GWWC pledge at the end of the course, resulting in a cost of ~570 USD per GWWC pledge signed.
• Amplify set up a marketing campaign for the MARS (2.0), which is a highly selective research fellowship run by the Cambridge AI Safety Hub. The campaign secured a successful applicant from outside of the group’s existing network for ~1700 USD.
• Amplify ran a campaign for the Centre for Reducing Suffering, promoting the e-book “Avoiding the Worst.” With a total advertising budget of ~$2,036, the campaign generated 464 newsletter sign-ups and book downloads, resulting in a cost per download of ~$4.39.
• Amplify also supported the EA South Africa Summit (October 2024). With a total budget of ~$1,114, the campaign brought in 7 attendees, 3 of whom later volunteered, achieving a cost of ~$159 per counterfactual attendee.

"We have worked together on the promotion of our latest EA Intro programs and saw a significant improvement in the number of professionals we now have attending versus previous cohorts. A great team who have been a positive addition to our marketing work and provide valuable field building services to the EA community." - James Herbert, EA NL

Note that these costs include both direct advertisement spend + the time spent setting up the campaigns, as well as all other organisational overhead^[2].

You can also learn about the cost-effectiveness of 37 past campaigns we ran here^[3] 

What success for a local EA community looks like:

Amplify began as a project within EA and AI Safety Hungary, where each early campaign ran on a budget of under $400.

While time spent on campaign setup wasn’t recorded (therefore, the exact cost-effectiveness can’t be calculated) the following outcomes show the strong results achieved from even small-scale digital efforts:

• An AI Safety hackathon attracted a talented participant who later received a career transition grant from Open Philanthropy and went on to contract with the Centre for Long-Term Resilience and OneDay Sooner.
• That same individual later connected us with an experienced professional who joined the local EA community and eventually became the director of the EA Hotel.
• 15+ attendees reached through paid ads to attend an intro course went on to attend EAG(x) conferences later.
• Two course attendees sourced directly through social media advertising went on to become paid staff at AIS Hungary.

• Two students reached via our campaigns went on to found a new EA university group in Budapest^[4].

What does a 1,000 USD donation to Amplify achieve on the margin?

In short, for every 1,000 USD Amplify receives, it can support one EA/AIS fieldbuilding organisation.

Marketing is a rare skill and is often misunderstood in the EA space, which means it is consistently underinvested in. We have been having to turn away many potential clients due to their and our lack of funding.

For every client, Amplify can provide end-to-end campaign setup, using field-tested messaging and promotional materials, email marketing and post-campaign cost-effectiveness analysis. This is our baseline product.

Beyond direct campaign support, engaging with us helps organisations understand the long-term value of marketing and how to budget for it in future fundraising rounds. Our goal is to make the ecosystem self-sustaining, reducing reliance on external subsidies over time. This has already happened with some organisations we supported. For example, our first campaign for EA Netherlands was run on a pro bono basis, but after seeing the results, they hired us to promote EAGxAmsterdam.

We’re also in the process of onboarding a dozen volunteers to support our work and help them gain experience doing marketing in the EA space. This initiative is a fieldbuilding intervention itself, as it provides career capital and an opportunity to build a network for people to pursue a high-impact role in the field. Our recent call for volunteers received over 160 responses, reflecting a strong supply of talent looking for opportunities in the space.

Note that the $1,000 above figure does not include direct advertising spend, which we ask organisations to cover separately. We generally recommend they allocate an additional $500–$1,000 for ads. We also normally ask organisations to cover around 50% of our operational costs. As a result, a $1,000 donation to Amplify effectively enables us to support two organisations. When including the direct ad spend paid by clients, the total investment per organisation returns to roughly $1,000 each.

What is your budget?^[5]

What is Amplify’s theory of change?

Click here to zoom in.

How do you measure impact and cost-effectiveness?

Imagine an organisation comes to us for support on promoting their introduction to an EA course. After setting up their campaign, these are the outcomes we track:

Short-term (immediate results, quantitative)

• Impressions and reach (campaign visibility).
• Clicks and conversions (initial interest).
• Number of Short-form signups to event/course

Mid-term (Program engagement, both quantitative and qualitative)

• Event or course attendance generated through improved outreach.
• Share of highly engaged participants at the end of a course

Long-term (Milestones and career transitions)

• Share of participants meeting career-milestones, such as volunteering for their group or another EA org, attending EAG(x) conferences, or seeking 80,000 Hours advice).
• Fraction of participants who transition their careers into EA- or AI Safety-aligned roles as a result.

Who is on your team?

Paul Dodd - Senior Marketer / Co-director

• Responsibilities: Advising and setting up campaigns for clients. Sources new clients.
• Track record: 15+ years in marketing, with 7 years leading digital-growth teams in tech. Strong record of scaling paid media across Google, Meta, LinkedIn, and TikTok. Previously led marketing teams at Nova and Nova Growth Capital, driving record growth in applicants, investment, and returns.
• Strengths: Highly seasoned digital marketer with deep expertise in growth, analytics, and advertising across multiple platforms.
• Weaknesses: Has been involved in EA for only ~1 year; completed an intro fellowship and short AIS course but still developing familiarity with the ecosystem.
   

Dennis Howell — Head of Operations

• Responsibilities: Ensures smooth operations, builds and optimises organisational workflows, and supports cross-team coordination.
• Track record: Also works with the European Network for AI Safety; founder of AI Safety Tulsa and co-founder of The Dutch Network for AI Safety. Extensive experience scaling remote teams, operations, and digital strategy across startups, nonprofits, and fintech. Built operational frameworks for high-growth organisations, and served on the boards of several non-profits. Recently conducted psychology research at TU/e and is completing an MBA in AI, Data & Analytics (University of Amsterdam).
• Strengths: Strong operations background, experience with process design, workflow optimisation, and organisational scaling.
• Weaknesses / Risks: Has been active in the space for ~1,5 years, which means he is still developing context about the field.

Danielle Reynolds — Junior Marketer

• Responsibilities: Supports campaign execution, content distribution, and digital marketing tasks.
• Track record: Several years of marketing experience despite her junior title. Selected from a pool of 150+ applicants after a four-round hiring process.
• Strengths: High potential for rapid progression; strong foundational marketing experience.
• Weaknesses: Has engaged with EA for only 6–9 months; completed the intro fellowship and self-study, but still building context.
• Risks: Hired recently on a 3-month trial contract, so the team has limited experience working with her. However, her onboarding has been really smooth so far and we are really excited about her work.

Gergő Gáspár — Co-founder, Co-director (volunteer capacity)

• Responsibilities: Oversees projects, develops big-picture strategy, advises staff, and leads fundraising efforts.
• Track record: Director at the European Network for AI Safety; founder of Amplify; author of The Field Building Blog. Active in fieldbuilding since 2019; founded EA Hungary and AIS Hungary (now with >2 FTEs). Former volunteer charity analyst and analysis coordinator at SoGive. Incoming director of EA UK.
• Strengths: ~5 years of experience in fieldbuilding and a strong understanding of the EA ecosystem.
• Weaknesses: Limited project-management experience, which meant some initial Amplify processes were not optimal; professionalisation (e.g., data pipeline) only recently completed.
  Risks: Transitioning to become Director of EA UK and leaving Amplify at the end of 2025. This may reduce our ability to bring in clients or navigate the EA landscape.
• Mitigations: Gergő will continue attending strategy calls and will have weekly 1-1s with Paul.
   

Milán Alexy — Co-founder (volunteer capacity)

• Responsibilities: Acts as a safety net in case of large demand in a short period.
• Track record: Co-founder of Amplify; Director of EA Hungary and AIS Hungary. Active in EA fieldbuilding for the past 3 years.
• Strengths: Strong familiarity with local EA ecosystems and group operations; well-connected within community-building circles. Offers additional stability and continuity within the founding team.
• Weaknesses: Limited experience in long-term organisational leadership and professionalised project management.
• Risks: Operating in a volunteer capacity may limit availability during periods that require sustained or intensive involvement.

What other organisations are there in the space?

Good Impressions

Good Impressions is the most established marketing agency in the Effective Altruism space. However, they have not supported regional field-building organizations and do not have plans to.

While they are working to scale, there are still many projects they do not have the capacity to support. Therefore we think there are many opportunities in the space that are worth supporting but Good Impressions are not able to take on.

User-friendly

Userfriendly is focused on branding, design, and logo development, and therefore has little overlap with our work.

FAQ:

1. Why doesn’t CEA fund this work?

• CEA is a large organisation, and while we’ve had ongoing conversations with them about increasing marketing support for groups, substantial changes may take some time.
• Their current guidelines cap marketing spending at $100 per program, and they’ve indicated that updating this is not an immediate priority.
• CEA is still building its in-house marketing and communications capacity, which understandably affects how quickly it can adjust its support.
• We’ve seen some early progress and are currently working with them to support 6 of their priority universities, though overall change has been gradual.

2. Why don’t groups or clients fund this themselves?

• Marketing is generally underappreciated in field-building, so many organisations don’t realise how much impact effective marketing could have.
• The highest counterfactual impact comes from working with organisations that could benefit but haven’t budgeted for marketing due to a lack of understanding.
• As JS from Good Impressions told us:
  
  “Willingness to pay is not as strong a predictor of commitment or perceived value as I thought it would be.”
• This creates a chicken-and-egg problem: funders expect clients to pay, but clients lack the means. This grant would bridge that gap, helping them see marketing’s value and plan for it in future budgets.
• Therefore, we are asking the community to look at our results and decide whether Amplify is a project worth supporting. It’s safe to assume that none of the outcomes we achieved would have happened without external funding.

3. Why not hire a regular (non-EA) marketing agency?

• Most EA orgs are not able to pay for the fees of for-profit marketing agencies; we charge a significantly smaller rate to clients who can pay for our work (50 USD per hour). For-profit agencies usually charge between 150-250 USD per hour.
• We are a value-aligned organisation with deep context on EA priorities. Investing in us is investing in the careers of marketing professionals who will contribute to the field in more and more ambitious ways in the years to come.
• Our experience across many EA/AIS groups gives us unique insight and transferable lessons that make our work more effective. Organisations hiring an external agency need to start their onboarding from scratch.
• Amplify also conducts field-building work directly, including training new volunteers, which further strengthens this pipeline of skilled, mission-driven contributors.

4. Why not fund the fieldbuilding orgs themselves?

• Organisations in the space have been systematically underinvesting in marketing for several years — often lacking the expertise to build sustainable digital strategies.
• Our support acts as a trajectory change: once equipped with strong communication systems and professional branding, these organisations can achieve greater impact per dollar for years beyond the initial grant.

5. Has Amplify tried fundraising from institutional funders?

Amplify got its seed grant from OP and EAIF; however, this was not renewed. Our funding needs won’t be fulfilled by institutional funders, which is why - like all other organisations taking part in Marginal Funding week - we turn to the EA community.

EAIF

We have gotten some valuable feedback from EAIF that you can read below. We did our best to address all of their points throughout the post.

"I'm sorry that we decided against funding Amplify. Briefly, the main reasons that we decided against funding were:

- It wasn't obvious to us that different organisations working with Amplify would be more valuable than them working with for-profit digital marketing agencies. - See FAQ #3.

- We think that it's likely that there are some high value opportunities for this kind of work, though we didn't think there would obviously be a lot of these opportunities (we thought that there would likely be significant demand for support from Amplify, but that much of these would come from lower value opportunities). - See FAQ #8 and #9.

- We didn't think that the cost-effectiveness in terms of $/attendee made this competitive.  - See our highlights section and our link to or cost-effectiveness analysis. We think these results are extremely promising, and we ask our donors to make up their own judgment! See our additional thoughts on the value of a career transition at FAQ #7.

More recently, we sent another application to EAIF to produce a marketing guide for fieldbuilding organisations, but it got rejected, and we didn’t get feedback.

Open Philanthropy

We tried fundraising from OP, but weren’t successful, and we didn’t receive feedback.

MCF

We have tried fundraising from Meta Charity Funders earlier this year before, but weren’t successful, and we didn’t receive feedback.

SFF

Amplify has gotten a 30k USD speculation grant in SFF’s last funding round, but didn’t receive its full funding ask. These are the funds that we are currently using to operate.

6. Does Amplify have other planned or pending funding applications?

• We are planning to submit a grant application to EA Funds about supporting broader communications efforts in the EA/AIS space with a similar business model.
• Amplify has another pending application with MCF about creating a guide for EA/AIS communities and subsidised support for 10 organisations (15k USD).
• Amplify has 2 pending grant applications with ACX, one supporting the fieldbuilding ecosystem (15k USD) and another tongue-in-cheek application to promote the ACX substack itself (15k USD).
  • We haven’t fundraised from ACX before and we are not sure of their taste for marketing work.

We think this shows that funders currently don’t prioritise marketing to a sufficient extent. We expect this to gradually change, but currently, it’s safe to assume that Amplify won’t be able to continue supporting fieldbuilding organisations and risk having to shut down.

7. What’s the value of having another person join the EA movement? Aren’t there very few available roles anyway?

The base rate of people landing full-time roles in effective altruism is low. Field-building interventions are heavy-tailed: most of the value comes from a small number of highly engaged individuals. The same is true for the organisations we support.

However, these people will then go on to create an outsized impact. See these highlights from last year's Meta Coordination Forum Talent Need Survey:

• The average value to an organization of their most preferred over their second most preferred candidate, in a typical hiring round, was estimated to be $88,737 (junior hire) and $455,278 (senior hire).
• The average value to an organization of making a hire compared to not filling the role at all was $272,222 (junior) and $1,450,000 (senior).
• The average value to the community of a person with equivalent expected lifetime value to an organization’s typical hire joining the community was estimated to be $2,037,500 (junior) and $7,308,333 (senior). This suggests that the value of recruiting ‘hire-level’ EAs to the community is estimated to be extremely high.

Historically, most EA field-building has focused on universities, which naturally generates a large pipeline of junior candidates. On the margin, however, the community would benefit more from recruiting senior professionals.

This has not happened largely due to limited marketing expertise. Universities have built-in channels—newsletters, career fairs, student networks—that make it easy to promote EA ideas, but no equivalent infrastructure exists for reaching experienced professionals who could fill senior roles. Digital marketing is the most promising way to fill this gap.

It’s also worth noting that there are many high-absorbency career paths where even junior talent can make a meaningful difference.

8. What are some ways in which Amplify is trying to improve?

• Tracking our impact sounds simple, but it’s one of the most challenging parts of our work. We’ve always known what data we want to collect, but for a long time, we lacked a complete, automated pipeline to gather impact stories from the organisations we support. Due to our capacity constraints, we often collected this data post-hoc by following up with clients individually instead of using an automated system.
  • Long-term, qualitative data about career transitions is what we value most, but it’s also the hardest to obtain. Organisations sometimes struggle to share this information, so we often have to rely on shorter-term proxies. Smaller organisations, in particular, are less likely to get back to being able to track their impact in a rigorous manner.
• We have also been lacking an automated mechanism for collecting feedback from clients systematically. We have only made a form for this now, and plan to work it into our workflow by the end of the year.

9. What are some mistakes you made?

• We could have been more selective in choosing which organisations received subsidised services. Many applicants were volunteer-run, making their track records harder to assess and limiting our confidence in their impact. Data collection from these groups was also more difficult.
• Smaller organisations often provide less consistent, lower-quality support to their course graduates. This increases uncertainty about longer-term outcomes.
  • However, we do not think that recruiting people into EA in these communities is not valuable. Historically, many people have landed EA roles with little external support.
• Going forward, we plan to prioritise organisations backed by funders to increase our confidence in their impact. This is necessary for us to do, as one reason our grant application got rejected was that many of the groups we supported didn’t meet their bar.
  • This will, however, exclude more small groups, especially those in low- and middle-income countries.
    • We would be excited to receive restricted grants from donors who want to specifically support smaller or volunteer-run organisations, allowing us to continue working with them despite their higher uncertainty.
• We should have invested more time in proactively sourcing additional (paying) clients instead of relying on inbound demand.
  • This would have required rejecting even more small clients, but it would have made our funding situation more stable by increasing the share of paying clients.
• We also found it challenging at times to maintain clear boundaries around the scope of our support. Clients often asked for small additional favours, not realising that these requests could require significantly more work on our end.
  • While it’s natural for them to seek help and for us to want to provide it, this sometimes led to us taking on more than we initially agreed. Going forward, we plan to be more proactive in explaining the workload behind these requests and setting clearer expectations, so we can support organisations effectively without overextending ourselves.

10. How does Amplify decide who to work with and offer subsidised support to?

As mentioned above, we plan to focus on small- to mid-size organisations that have some existing funding but still can’t afford full-price marketing services. These groups represent a sweet spot because:

• We can be more confident in their impact
• Supporting them signals institutional funders that there is real demand for marketing support
• Our contribution makes them more cost-effective and helps shift their trajectory so that they budget for marketing in their future funding rounds

In rare cases, we will make exceptions to this, and support volunteer-run initiatives if we think they are particularly promising.

11. Even if recruiting highly engaged participants is cost-effective, isn’t the total investment per career transition still much higher? After all, the organisations Amplify supports also need to invest in supporting their program participants.

• That’s true, but the organisations we support have already built the infrastructure to help these people (e.g., mentorship, programs, career advising etc.).
• Their main challenge is not having enough qualified participants, not the cost of serving them.
• This means that, on the margin, adding additional engaged people is much cheaper than if they had to build the support systems from scratch.
• While many career transitions take years, we also see more immediate outcomes such as the signing of GWWC pledges. This means there are returns on shorter timescales as well, with minimal support needed from other organisations.

Additional questions?

Please feel free to share any questions in the comments, and we will get back to you as soon as we can!

1. ^{^}

   Data taken from our CE study, plus subsequent campaigns to date. An expression of interest is someone responding to our advertisement and sharing contact details (name & email) and requesting further information about the course / programme.

2. ^{^}

    It costs us ~1k USD to support a client+overhead, so we have used this number on calculating the cost-effectiveness of the different outcomes. With this information in mind, it’s,easy to calculate the direct ads spent of the campaigns

    With a very conservative estimate, if just 10% of these EOIs go on to participate in the course, we will have secured more than 300 participants.

   We haven't yet gathered enough data to feel confident about sharing a representative number for the cost per course graduate or per course participant and these vary greatly based on region.

   However, we do have an older cost-effectiveness analysis from Hungary. This older analysis suggests approximately $25–30 per course participant and roughly $100 per highly engaged graduate (someone who went on to further interact with or contribute to the community).

   Please note that this was conducted before the war and before Covid, and in a country where Meta advertising is relatively inexpensive. We therefore do not consider this analysis representative of current conditions, but it is the only systematic data point we have at the moment.

3. ^{^}

    Note that these figures only include the direct ad costs and only include shorter-term quantitative results.

4. ^{^}

    Unfortunately this university group is now dormant, as it often happens groups in the space.

5. ^{^}

   The number of projects supported should be interpreted as baseline product-equivalent (as desrtibed above). For example, if we run a campaign that is twice as elaborate as our baseline product, we would count that as two baseline-equivalent projects. This allows us to normalise across different campaign sizes and maintain a consistent measure of output.