Cost-Effectiveness Of Outreach Channels That Brought 100+ Applicants to EA Turkey Fellowships

TL;DR

EA Turkey’s first mass outreach for our 5-week intro fellowship brought ~100 applications, of which ~40 were accepted into 6 cohorts. Personal referrals, debating community outreach, and influencer marketing drove ~70% of applications. Moderate-impact channels (cold outreach, uni WhatsApp groups) brought ~25%, while Meta ads and org networks had almost no effect. Biggest lesson: friend referrals and trusted community endorsements are the most effective channels. We’re excited to explore LinkedIn paid ads, stronger referral incentives, and targeted Google ads for future outreach.

Introduction

EA Turkey had run different outreach strategies for our remote nationwide fellowship. We ended up getting ~100 applications for the 5-week intro fellowship, accepted ~40 of them, which are currently in our 6 cohorts. Unfortunately, our data is quite messy, so we’ll be giving ballpark figures on the impact of different actions. I’m limiting the amount of detail to keep the post short so send me a quick message if you'd like to use any of the channels below!

This was EA Turkey’s first mass outreach. Here is what we had done, and which of them worked best. You can find our posters and marketing messages through this link.

Hight Impact Channels

The channels below brought about 70% of our applicants.

Personal Referrals

Personally messaged all of EA Turkey’s most engaged members (36 people) to share our intro fellowship application with their friends, on their Instagram, and in the WhatsApp groups they are in. Most of them ended up sharing the announcement with their friends and on their Instagram (thanks again!!).

• Impact: Referrals were the biggest channel where people heard about and applied to our fellowship.
• Lesson: Friend endorsements and referrals would be your number one place to bring more people into your community. Generate ways to optimize this channel, like monetary or donation incentives for referring (see “ideas we are excited about” section for more).

Debating Community Outreach

We shared the intro fellowship marketing messages in debating community WhatsApp groups with links to the application form and EA Turkey announcements WhatsApp chat. Community members posted Instagram stories about the program.

• Context: The debating community is connected through many different WhatsApp channels from previous tournaments and debating clubs in universities and high schools. Many senior and mid-career members of our community have high recognition and connections in the Turkish debating community due to their organizational and educational efforts.
• Impact: Our WhatsApp group grew significantly, and we believe we got around 20 applicants due to this.
• Lesson: Utilizing the influence and connections of high-status members in small intellectual communities worked pretty well. Our senior and mid-level members in the debating community are a worked example of this approach. I believe the same could be applied to different intellectual groups like STEM Olympiads, chess, robotics/rocket groups, competitive coding groups etc. Just engaging (or finding someone engaged) with these communities could be a good investment for future outreach.

Influencer Marketing

Reached out to seven influencers who create career/intellectual content which had a potential EA audience. We tried to find the most personal way to contact these people using our network, but if we didn’t have any, we emailed them. Only two of them agreed to share on their main channel (both were reached personally), and we didn’t get responses from the others.

• Impact: One influencer sharing our poster and application form brought us 35 applications, 82% of which were women! However, the quality of the applications was lower than other methods (probably due to being more mainstream). We believe the sharing of the other influencer didn’t bring any applications, since they shared many different things and also shared about global politics, which Instagram’s algorithm makes fewer people see.
• Lesson: Influencer marketing could be a very powerful outreach tool. However, try to find more personal channels to reach these people.

Paid Newsletter Ads

We cold-reached a digital media company that owns and curates content for many different newsletters and news apps. We ended up publishing paid ads in a newsletter focusing on world issues and their solutions with 100,000 readers and 30,000 ad viewers. The paid ads were heavily discounted thanks to the courtesy of the company.

• Impact: Due to a communication error, the ad was sent after the deadline for applications, so only 5 people applied, but even though it was late, they were admitted. Three of these applicants were mid- to senior-level professionals in high positions. One of them had a fat-tailed significant impact potential due to their background.
• Lesson: Newsletter ads with a proto-EA audience could be very impactful. Try to get a discount since EA is a do-good cause where people are sympathetic.

Moderate Impact Channels

These channels combined brought about 25% of our applicants. To note, all these channels required high resources, making them low in cost-effectiveness.

Cold Messages to People with EA Interest

24 people with an interest in EA were personally messaged and invited to the intro fellowship. 15 of 24 were reached via EA LinkedIn groups (see example). Individuals from Turkey were found and messaged. 9 of 24 were reached via School of Moral Ambition’s online community platform, which is very similar to the EA Forum’s people directory.

Impact: 3 out of 24 people ended up enrolling in the fellowship.

Lesson: Cold outreach does not work well in this setting. Maybe sending emails could have been better.

WhatsApp Groups in Target Universities

Found people in our network who were enrolled in our target universities and asked them to share messages in potentially EA-related WhatsApp groups (economics undergrads, AI society, etc.). We also had messaged Instagram pages of EA-related university societies individually

• Context: These WhatsApp groups are great outreach channels since they can bring applicants from good universities who have an interest in EA-related fields. It was quite hard to find people in these universities who had access to internal WhatsApp groups. It was also hard to make them send messages into the relevant channels, since if they weren’t already in those groups, they had to use their own networks to find others, which people were unmotivated to do.
• Impact: We believe around 5 people applied after seeing messages in these group chats.
• Lesson: This approach could have worked better if we knew more motivated volunteer contacts who’d take ownership and use their networks in university. Or we could have given higher incentives like Amazon gift cards or money to people we asked to share our message.

Low Impact Channels

These channels combined brought about 5% of our applicants. Again, all these channels required high resources, making them very low in cost-effectiveness.

Meta Ads

Created Meta ads, ran them for 2 weeks, and got 0 applicants. We didn’t have much previous experience with Meta ads, but we spent time learning the basics and targeted locations and people with EA-related interests. Still, running Meta ads on a national level with no experience wasn’t a good idea.

Internal Networks of Different Orgs

Used our network to send messages to internal WhatsApp groups of relevant organizations. Most of the contacts were entrepreneurial or leadership education fellowships with an emphasis on creating an impact or sustainability in general. We reached out to 8 organizations, and 2 of them ended up sharing the fellowship on their internal channels. We didn’t get any applications from them.

Channels We Are Excited to Explore More

These are other ways we are excited to try but haven’t yet tested.

LinkedIn Content

Creating LinkedIn content and potentially paying for ads has a lot of potential. A post of mine celebrating a year of giving and introducing the 10% pledge had 8k interactions, and I know at least 2 people who had taken the pledge after seeing that post. Creating high-quality LinkedIn posts, making our network repost them, and potentially paying for LinkedIn ads could be a cost-effective way to find good applicants for the intro fellowship.

Personal Referrals with Higher Incentives

Personal referrals could have even more impactful if people were incentivized to refer others. These incentives could be free Amazon gift cards, or a $5 donation to a charity of their choice if their referral gets accepted. Advertising referrals with these incentives could be quite useful.

End-of-Form Referrals

The same referral system as above, but included at the end of the application form, so applicants can also refer their friends.

Luma Front Page

Luma is an event platform where you can create event pages quickly. Some cities have a special page showcasing certain events happening in that city. Only a few events are featured, but if yours gets listed, you can attract many people. An “Intro to EA” social event could bring in many people and encourage them to apply to the fellowship.

Google Ads

Haven’t tried this yet, but if you can filter for people who have already engaged with EA websites, this could be very useful.

EA Forum Map

EA Forum also has a people directory page. By checking it regularly, you may find new users and invite them to the local fellowship.

Fresher's WhatsApp Paid Ads

In Turkey, universities have WhatsApp group chats for new freshmen. These groups are usually opened in the summer and can only be accessed by incoming students. They are often found through Instagram, with names like “Koç 2025 Fresher’s.” These channels typically include more than half of the newcomers, and reaching out to the WhatsApp admins—whether by paying or not—to post a message in their groups could work very well.

This is a great channel because it allows direct access to freshmen, the group most interested in student organizations. Moreover, since it’s a WhatsApp chat, it provides a high-trust environment.