EA Summit: Wellington 2026 - Retrospective

Published on October 20, 2026 3:06 AM GMT This retrospective highlights details and learnings from EA Summit: Wellington 2026 It is written by Tom Hutton - Director of Community at EANZ, and lead organiser for the Summit. While I use “we”/”our” to refer to the Summit as an event/the people who made it happen, all opinions are mine alone/not explicitly endorsed by anyone else. I’ve tried to note currency throughout. If currency is not noted assume figures are in NZD unless stated otherwise. You can read this as a google doc here. The content is the same, the tables are formatted slightly differently It is written in a governing thoughts structure; the contents serve as a high level summary  EA Summit: Wellington was a one-day Summit, on Saturday 19th September 2026 at Tiakiwai Conference Centre This was the second EA Summit in New Zealand, and aimed to act as an introduction to EA ideas and the EA community Our core target audience was proto-EAs or early stage EAs who had not yet engaged with EA as a community. Our programme was designed with newcomers in mind. Experienced EAs were in attendance mostly to act as mentors/experts, though them being able to network was a positive externality. The inaugural EA Summit was run in Christchurch at a similar time last year, and had different organisers. You can read their retrospective here. We performed well on all our feedback metrics Numbers are averages except where obviously not averages   2026 2025[1] Attendance# of Attendees6970      Feedback Survey# of Respondents4727  Likelihood To Recommend - x/108.858.33  NPS59.633.3     New Connections  Average per attendee4.045.18   Total (New Connections x # of Attendees)[2] 279362  Impactful Connections  Average per attendee1.892.11   Total (Impactful Connections x  # of Attendees)[3] 130   148       Felt Welcome - x/54.724.74  Felt Respected - x/109.409.33  Women and Gender Minorities Respected - x/109.499.42  Unwanted sexual/romantic - x/100.000.73  Food - x/109.339.00     Use of time (vs counterfactual)  Better91.49%96%  “Notably” or “Much” better76.6%54%  Worse0%0%      Understanding of EA (self rated 0-10)[4]  Mean before6.37.0  Mean after7.98.3  Difference - all attendees1.61.3  Difference - attendees with initial understanding 0-5 (n=15)4.1  We came in under budget   20262025 BudgetEvent costs only - NZD$24,560   All expenses - NZD$48,110  ExpensesEvent costs only - NZD$15,742        All expenses - NZD Event costs plus organiser time[5] and Travel Support[6] $35,310   $30,547    Cost per attendee - NZD$511.74$436.40   Cost per attendee - USD[7] $288$245 Ticket Sales- NZD[8] Benefactor - $3002   Supporter $1506   Standard - $8030   Student - $408   Free (volunteer / on request)32  Our total cost, including organiser time and travel support was $35,310 (USD 19,860) Our biggest costs in order were: organiser time, venue, and catering These “big three” make up $30,921 or 88% of the total expenditure. Including the fourth biggest expense of travel support accounts for 98% of total expenses. On the advice of CEA we opted for a professional and better located venue over a free but difficult to access student space. Ticket sales generated $4,220.00 NZD in income The event was ticketed with free and volunteer tickets were available on request. We also offered a student discount. We decided to have tickets by default primarily as a commitment device, and ended up with 8/77 registrations no-showing[9]. Total registrations were 85, with 8 cancellations ahead of time. We had fewer attendees than we hoped We had 69 attendees on the day, with 8 no-shows, and 8 cancellations. Of 45 feedback survey responses[10], 24 (53%) had not attended an in person Summit, EAGx, or EAG previously 12 (27%) had attended 1-2 of these events, 9 (20%) had attended 3+ We performed some paid marketing leading to counterfactual attendees We ran boosted posts and a series of ads on Instagram and Facebook. We found the boosted posts had better engagement and click through, though that could be due to timing and the rushed nature of getting the ads over the line[11]. LinkedIn marketing had too high of a base cost, and we’d heard from others that they’d had more success with Meta. In total we spent ~$200 on boosted posts, and ~$130 on ads. We believe this resulted in at least 10 counterfactual registrations, all from the boosted posts. The counterfactual registrations were approximately evenly spread across Facebook and Instagram as their source. In addition to actually leading to more (10 vs 0) registrations, our boosted posts did a better job than ads of driving traffic to Luma. Most of our attendees found out about the conference through word of mouth This chart shows all 85 people who registered for the summit. Data is primarily self reporting during registration, supplemented by utm_source, and my insider knowledge/judgment[12]. “In the EA ecosystem” includes group organisers, speakers, local highly-engaged EAs, and other people