Cost-effectiveness analysis of ~1260 USD worth of social media ads for fellowship marketing
By gergo @ 2024-01-25T15:18 (+61)
TLDR.: I spent ~1260 USD on social media ads (Facebook/Instagram) over ~1,5 years.
We got an additional 53-57 applicants this way, resulting in a cost-effectiveness of 22,1-23,8 USD per applicant.
Disclaimer: I wanted to capture 80% of the value of what I have to say without putting a lot of time into writing. This means that the post is somewhat rough around the edges, but hopefully, it will be still useful.
I have been very excited about experimenting with paid ads to reach out to people who would otherwise not hear about our Into to EA and AGISF programs. This post is a summary of how I spent ~1260 USD on paid ads for social media, and a botec of what it bought us. Please take all results with a grain of salt, as the data is limited and one thing that works for us might not apply in other contexts. That being said, I’m quite confident that groups that want to increase the number of talented and diverse applicants to their programs should at least experiment with using paid ads.
Cost-effectiveness
I overall spent ~1260 USD, which resulted in 53 additional applicants to our fellowships over 1,5 years (23,8 USD per applicant). At least 4 of these 53 applicants also invited a friend along with them to our program, and if we count them as well, we got overall 57 additional applicants, which slightly improves the cost-effectiveness to 22,1 USD per applicant.
You can take a look at the raw-ish data here as well as see the breakdown by campaign and course type (EA vs. AIS).
Impact
I think most of the expected impact of this will come from the ~30% of the overall applicants who engaged with the courses very seriously and took a lot of value from them.
Unfortunately, I didn’t do an amazing job keeping track of what % of the original 53-57 applicants never started the course. I would estimate this to be around 20-35%. Given that many people never start the course, I think it’s really valuable to encourage people to sign up for your newsletter[1] as part of the application process - or if they have a good application, reach out to them in the next round.
As for the rest of the applications, I think it’s pretty similar to the usual fellowship experience, some people drop out after a couple of sessions, some finish it but end up disappearing after the course, etc.
It goes without saying, but of course, this is not a judgment on people’s intrinsic value!
Additional points and caveats:
- Note that the courses I was advertising were 4 sessions only, and sometimes it was an intensive 1-week course - which I think partly improved the cost-effectiveness but can have other drawbacks, see the discussion here and here.
- With paid ads, we got to reach out to many talented international students from 3rd world countries, which is awesome - and otherwise, we would have likely not reached them.[2]
- If you have data on the cost-effectiveness of your social media ads (or want to start gathering such data) make sure to reach out!
Conclusion
Based on this, I will increase our marketing budget, as well as probably expand it to cities where we don’t have an EA presence yet in the country. I think it’s possible that once I have more data, these ads won’t seem as good as now, but even if I’m currently overestimating the cost-effectiveness by 10x - they would still look pretty good.
If you would like to use social media ads for your national/city/university group, feel free to shoot us an email at info[at]eahungary.com
- ^
- ^
In Hungary, there are a lot of international students from 3rd world countries who are here on a scholarship. This means that they have already had to go through a filter, and they are more ambitious than the average student. Given that they are in a new country, they are looking for opportunities to take and communities to join. They also already have more of a “big picture - global mindset”, which is needed for effective altruism. Unfortunately, there is no real “social infrastructure” provided by the universities to welcome these students into Hungary, so many of them can feel isolated, and understandably so. For a lot of them, our EA and AIS community was able to provide this community, which I’m very happy about. Given this isolation issue, at the moment we have also no other good means of outreach to them, although going to English-speaking university classes and pitching our programs is something I want to do more of. We also found that national students who found us through paid ads are also great, and are at least as motivated (if not more) than hungarians finding out about us through other forms of outreach.
Bella @ 2024-01-25T17:40 (+14)
Hey, thanks for sharing! Are you up for sharing any more detail on the ads, for example:
- What targeting settings did you use on the ads?
- What creative (images/videos) did you use?
- What copy (text that goes in the ad) did you use?
- Did any of those do especially well?
gergo @ 2024-01-26T16:13 (+11)
Hey Bella, thanks for your questions!
What targeting settings did you use on the ads?
We targeted university-age people (so around 18-26) in Budapest (+~40km radius). In the future, I also plan to include people up to 35 years old (or more).
Here are the tags I used. Many of them come from EA CZ's @Jakub Sechter who did structured interviews with their members about what stuff they follow on Facebook - so I think those must be pretty good. I also added a bunch on my own which may or may not be useful. (Unfortunately, I can't find Jakub's original list, but what I have in the docs includes all of his suggestions)
As a side note, I think you can also set up ads specifically to one university if you set the radius very small - but I haven't tested this yet. If you do that though, you shouldn't use any tags as that would filter out too many people from an already small sample.
What creative (images/videos) did you use?
I used either general posters that are available on CEA's Canva (such as this and this) - or pictures of our group photo. My general impression was that a friendly group photo gathers more clicks /$ spent, but I don't know if that also translates to getting more motivated applicants, unfortunately. I haven't used videos before, but I was told that they tend to perform better (ofc the cons being is that they are more time to make)
One mistake I made though was that I always used just one poster/picture for a campaign in the past. Since then I learned that you should upload several posters, group photos in different angles/proximity etc. - and then after the first hours/days, Meta's algorithm will automatically prioritize those that perform best. (this is true for different descriptions, headlines etc. too)
What copy (text that goes in the ad) did you use?
I added them to the second half of this doc, both for the EA and AIS group :)
Did any of those do especially well?
Unfortunately, I didn't get to compare different messages yet, apart from what I wrote about the pictures vs posters stuff! I recommend using several descriptions etc!
Edit: After writing the initial reply I realized most of the additional "tips" I wrote will be pretty obvious to you, as I just saw that you work in marketing haha x) - I guess I will leave the comment as it is, in case it's helpful for others!
Bella @ 2024-01-29T09:35 (+2)
Hey, thanks so much for replying in such detail — really appreciated! :)
(haha, yes, I had heard of some of the extra things you mentioned, but I think for someone just starting out they're really valuable to learn!!)
JSWinchell @ 2024-01-25T22:08 (+1)
I'd be curious to see this as well