In Fall 2023, EA Purdue barely existed. In fall 2024, EA Purdue had 35 people attend its intro meeting. In Fall 2025, EA Purdue had 33 new members attend its intro meeting despite doing dramatically more outreach.
Below, outreach method and scale is specified. It is followed by a discussion of the implications for university group organizers, community builders, and a request for help. If you’re not an organizer, feel free to skip to the “Why I think this happened” section of this post, as it should be more relevant.
Outreach Method
Scale in 2024
Scale in 2025
Club Fairs: Purdue has an all-club tabling fair and a computer science tabling fair. EA Purdue attended both of these in both years. These fairs yielded 100 and 32 mailing list signups respectively in fall 25. In fall 24, the fairs yielded 64 and 16 signups.
Side note: At Purdue, the freshmen tabling fair is incrediblyimportant. It’s 3 hours long and ~40% of the current club was first contacted there. We don’t expect this to change. Try really fucking hard at this fair. Eat caffeine beforehand. If you were being scope-sensitive about how hard you focused on these conversations we estimate you would focus 214x harder during the club fair than your average conversation :).
2 fairs
2 fairs
Lecture Outreach: After emailing 53 professors, I was able to give this 3-minute pitch in the beginning of 16 intro classes. With an average size of 135 students in attendance, I think 2,000 students genuinely listened to the pitch. I held ~80% of the class’s attention for the full duration of the pitch.
Each pitch yielded an average of 3.5 mailing list signups, for a total of 57 signups.
0
16 lectures
Departmental Emails: We emailed and followed up with 70 departments asking if they could include our blurb in their mailing list or newsletter. Close to half of the departments responded saying they had included us in some way. This yielded 11 mailing list sign-ups.
18
~70 Depart-ments
Professor Emails: We emailed roughly 70 professors with large intro classes who we did not email for lecture pitching. We asked them to advertise our club in an announcement to their class on Brightspace, Purdue’s LMS. Only 13 of them responded favorably to our email, and Brightspace announcements led to 8 email signups.
0
~70
Profes-sors
Tabling: Before classes started, I tabled during freshman orientation handing out candy. Our tabling display had info about the club, and 3 buckets where visitors could place a club-sponsored $2 bill. This quantity of money was eventually donated.
We tabled for a total of 7 hours. ~200 people stopped by the table, 72 people interacted with this game, and 11 people signed up for the mailing list.
0
7 hours
Posters: We were able to get posters up in nearly all of the dorms on campus, often with 2 on each floor. Our poster was usually posted next to about 10 other posters that the mostly-freshmen on-campus students would have walked by every day. We received 6 mailing list signups from this. Posters are a really bad method of outreach.
6 posters
300 posters
Word of Mouth: It’s hard to judge this. At least 4 of the new attendees at the intro meeting were confirmed to have been dragged there by friends who had previously been involved with the club. EA Purdue also had ~15 active members in 2024 that certainly mentioned EA to their friends.
0
More
Background on Purdue
Purdue has ~45,000 undergrads with about a 50% acceptance rate. About 60% of Purdue students have a STEM major, and 30% of Purdue students are engineers. From conversations with other organizers, I feel strongly that EA has more appeal to more ambitious students in more competitive schools. I think this effect is very significant.
Most schools are smaller than Purdue. Only 22% of degrees are STEM nationally. Your school might have a higher or lower acceptance rate. In a small school, diminishing returns to outreach are probably even higher.
The idiosyncratic factors at Purdue might affect how much you expect these results to generalize. I would caution you not to over-rely on “selection bias” when considering if this might apply to your school. While this might account for some of the magnitude of our results, a different line of reasoning is necessary to dismiss the phenomenon entirely.
Why I think this happened
Strong diminishing returns to outreach: People are really good at getting the information they need to know. Charlie Kirk came to Purdue and thousands of people attended in spite of an negligibly small amount of advertising from Turning Point USA. If a student has ever died on your campus, the entire campus will hear about this within a week. The information often spreads before official announcements come out.
People almost always know about the opportunities that they might be interested in. After arriving on campus, students work extremely hard to surround themselves with friends and information environments that correspond to their interests. Figuring out which clubs on campus exist that you might be interested in is not a difficult task. It’s clearly worth the time, and most students take the time to do this. This isn’t just true for clubs. There is a reason companies reach an efficient quantity of customers and usually spend less than 10% of revenue on marketing.
You might object that it’s hard to communicate what EA clubs do and that this merits more outreach. I would counter with the fact that our lecture outreach included an in-depth 3-minute pitch with slides in front of a curious audience. About 3 of the 2,000 people that listened to it came to our callout.
Implications
For organizers, perhaps we should stress less about doing extensive outreach. I think you should have a strong prior that outreach is important and are not sure how much this should update you. However, it’s important to note that our “study” was quite high-powered, with no obvious confounders. Depending on how you measure it, our message got out to 5-10,000 people in 2025. Each of them then decided whether to apply for the seminar or attend the callout.
For university group community building, this should update us in the direction of thinking that a strong majority of our counterfactual impact is in starting/helping nonexistent/very small EA groups as well as improving the quality of programming for more established groups.
Request to you
If you have any insight whatsoever about why dramatically more outreach might have yielded the same results, I would appreciate you sharing it in the comments. Please disagree with me. University group organizing is really important and getting it right is crucial.
An Order-of-Magnitude Increase in Outreach Resulted in a Smaller Club Intro Meeting
In Fall 2023, EA Purdue barely existed. In fall 2024, EA Purdue had 35 people attend its intro meeting. In Fall 2025, EA Purdue had 33 new members attend its intro meeting despite doing dramatically more outreach.
Below, outreach method and scale is specified. It is followed by a discussion of the implications for university group organizers, community builders, and a request for help. If you’re not an organizer, feel free to skip to the “Why I think this happened” section of this post, as it should be more relevant.
Club Fairs: Purdue has an all-club tabling fair and a computer science tabling fair. EA Purdue attended both of these in both years. These fairs yielded 100 and 32 mailing list signups respectively in fall 25. In fall 24, the fairs yielded 64 and 16 signups.
Side note: At Purdue, the freshmen tabling fair is incredibly important. It’s 3 hours long and ~40% of the current club was first contacted there. We don’t expect this to change. Try really fucking hard at this fair. Eat caffeine beforehand. If you were being scope-sensitive about how hard you focused on these conversations we estimate you would focus 214x harder during the club fair than your average conversation :).
Lecture Outreach: After emailing 53 professors, I was able to give this 3-minute pitch in the beginning of 16 intro classes. With an average size of 135 students in attendance, I think 2,000 students genuinely listened to the pitch. I held ~80% of the class’s attention for the full duration of the pitch.
Each pitch yielded an average of 3.5 mailing list signups, for a total of 57 signups.
16 lectures
~70
Profes-sors
Tabling: Before classes started, I tabled during freshman orientation handing out candy. Our tabling display had info about the club, and 3 buckets where visitors could place a club-sponsored $2 bill. This quantity of money was eventually donated.
We tabled for a total of 7 hours. ~200 people stopped by the table, 72 people interacted with this game, and 11 people signed up for the mailing list.
Background on Purdue
Purdue has ~45,000 undergrads with about a 50% acceptance rate. About 60% of Purdue students have a STEM major, and 30% of Purdue students are engineers. From conversations with other organizers, I feel strongly that EA has more appeal to more ambitious students in more competitive schools. I think this effect is very significant.
Most schools are smaller than Purdue. Only 22% of degrees are STEM nationally. Your school might have a higher or lower acceptance rate. In a small school, diminishing returns to outreach are probably even higher.
The idiosyncratic factors at Purdue might affect how much you expect these results to generalize. I would caution you not to over-rely on “selection bias” when considering if this might apply to your school. While this might account for some of the magnitude of our results, a different line of reasoning is necessary to dismiss the phenomenon entirely.
Why I think this happened
Strong diminishing returns to outreach: People are really good at getting the information they need to know. Charlie Kirk came to Purdue and thousands of people attended in spite of an negligibly small amount of advertising from Turning Point USA. If a student has ever died on your campus, the entire campus will hear about this within a week. The information often spreads before official announcements come out.
People almost always know about the opportunities that they might be interested in. After arriving on campus, students work extremely hard to surround themselves with friends and information environments that correspond to their interests. Figuring out which clubs on campus exist that you might be interested in is not a difficult task. It’s clearly worth the time, and most students take the time to do this. This isn’t just true for clubs. There is a reason companies reach an efficient quantity of customers and usually spend less than 10% of revenue on marketing.
You might object that it’s hard to communicate what EA clubs do and that this merits more outreach. I would counter with the fact that our lecture outreach included an in-depth 3-minute pitch with slides in front of a curious audience. About 3 of the 2,000 people that listened to it came to our callout.
Implications
For organizers, perhaps we should stress less about doing extensive outreach. I think you should have a strong prior that outreach is important and are not sure how much this should update you. However, it’s important to note that our “study” was quite high-powered, with no obvious confounders. Depending on how you measure it, our message got out to 5-10,000 people in 2025. Each of them then decided whether to apply for the seminar or attend the callout.
For university group community building, this should update us in the direction of thinking that a strong majority of our counterfactual impact is in starting/helping nonexistent/very small EA groups as well as improving the quality of programming for more established groups.
Request to you
If you have any insight whatsoever about why dramatically more outreach might have yielded the same results, I would appreciate you sharing it in the comments. Please disagree with me. University group organizing is really important and getting it right is crucial.
Thank you!