EAGx Boston 2018 Postmortem
This is a belated postmortem for EAGx Boston 2018, for the benefit of any future organizers interested in our experiences.
We took three months to plan this conference. We began planning in late January 2018, and it took place on Saturday, April 21st, 2018. Videos of the speaker presentations can be found here.
General recommendations for organizational purposes and governance:
- We recommend that you first check out this document, and pay special attention to the timeline.
- Make sure it is clear who is responsible for each task, possibly designating leaders for categories like food and speaker outreach.
- Make sure someone is taking notes whenever your team meets. Possibly assign this role to a specific person.
- Meet at least weekly to keep everyone accountable, in person if possible.
- Substantively, do not be overly sensitive to cost; returns to amenity quality are large.
Contents
- Venue
- Speaker Outreach
- Funding
- Web Presence
- Audio Visual Services
- Food
- Marketing and Design
- Presentations
- Day-of Execution
- Financial Breakdown
Venue
CEA’s advice to secure a venue many months in advance of a conference is well heeded. That was impossible on our time frame, but we did secure some early, though less than ideal, options on Harvard’s campus. These were unattractive spaces with upfront fees (~$1,000) for using them, which we thought we could avoid since we knew at least a few grad schools that offered free, high quality spaces (e.g. Harvard Law and Business Schools, MIT Sloan). Our final venue choice — MIT Sloan’s Tang Center — came through rather late due to difficulty getting a hold of any Sloan students. We ultimately chose it because there was no reservation fee and it was our best aesthetic option. The main drawbacks of the space were its small size and lack of accessibility.
Tickets sold much faster and with much less advertising help that we expected, and we reached our capacity of 144 attendees very quickly. We advise conference planners with CEA backing to be prepared to accommodate 200+ attendees. Do not be pessimistic about your ability to advertise; people are looking for these events.
On the accessibility front, the Tang Center is tucked between two one-way streets and a major highway and surrounded by construction. Many attendees had difficulty finding us, especially since our space was on the third floor. We recommended favoring ground-floor spaces in buildings which are easy to locate by both car and public transportation in relatively well-traveled areas.
Takeaways:
- Choose a venue to accommodate at least 200+.
- Favor central, recognizable buildings easily accessible to both cars and public transportation.
- Favor ground-floor spaces to minimize in-building navigation issues.
- Be more concerned about size, location, and quality than fees.
- Seeking out university-based EAs to help secure spaces will expand options.
- The earlier you search for spaces, the more high-quality spaces will be available.
Speaker Outreach
Securing and accommodating desirable speakers is likely the most important and logistically challenging input to conference success. Knowing approximately what your speaker lineup will look like is also a prerequisite for meaningfully advertising your event and planning its schedule and format. Getting enough highly-sought-after presenters to all come to the same place on the same day was very difficult. Achieving reasonable race and gender diversity among presenters was also a major challenge for us, but one which is incredibly important to CEA and the entire EA community-building effort at the moment.
Our process began by scouring the EA speaker database, available from CEA, and narrowing our options to regional speakers with broad appeal. We limited our focus mostly to locals to avoid cannibalizing upcoming EAG events in Europe and California, and to limit the costs of speaker transportation and housing. We initially hoped to plan our conference around a theme, but were quickly disabused of the notion that we would be able to conform speaker presentations to it. At least on our timeframe, speakers were going to present what they had prepared, and getting speakers to adapt their remarks to some prescribed theme is a hard sell generally. We enlisted the help of all organizing team members and some outside EAs to get a better sense of who was appealing and rated speakers from 1 to 4 based on quality before dividing them by subject matter in order to assure a variety of EA-relevant topics would be represented. We then sent out emails using this template, making sure not to leave more than 10-12 outstanding at a time so as not to overbook or get too lopsided in terms of content. An alternative — and very fruitful — means of securing speakers was working through our own schools and networks. If you have a large planning team, some of whom have been involved in EA for a while, it’s likely you have access to more, and more flexible, presenters than you think!
Once a given speaker agreed in principle to participate, we began a more personalized correspondence in which we first tried to get a sense of how flexible their schedule was on the day of the conference, then asked about their potential interest in alternative formats to a generic talk. We were hoping to mix up presentation formats to make the conference more stimulating for attendees, but were limited here by what our presenters were willing to do. Still, we were able to successfully run an interview, anti-debate, and two workshops along with with standard panels and solo talks. Negotiation about time slots and format continued until very close to our event date in part because some speakers were difficult to get and keep a hold of. CEA also needed them to e-sign release forms and provide a headshot and bio. We found it easy and stress-relieving to pull headshots and draft bios ourselves and simply ask the speaker to approve, but it seems most had something of their own ready to go anyway.
Takeaways:
- Divide outreach responsibilities over several people. Assign each speaker a point person on your team who is responsible for communicating with them.
- Use your networks to expand your speaker options in both number and flexibility.
- Frontload speaker outreach as much as possible because it will constrain a lot of your next steps. Expect it to take at least 5-6 weeks because of rejections and communication lags.
- Suggest Skyping in to highly valuable speakers who are not otherwise available, if this is technically feasible at your venue.
Funding
The conference was primarily financed by ticket sales and an unsolicited donation from an attendee. CEA’s full backing assured us that we be able to cover any expenses which might arise in our short planning timeframe. We made tickets available for $25 each with $15 early bird pricing for the first two weeks of sales (we only sold tickets in the four weeks leading up to the conference). Most attendees paid the early bird price and processing fees averaged a little under 5% of gross sales.
Our short time-frame and full backing from CEA made seeking outside grants a lower priority with relatively low chances of bearing much fruit. We surveyed the organizing team for knowledge of available school-based grants within the timeframe and applied for nine of them with one of those applications being successful (COOP Gives). A complete list of our applications is provided here. As shown in the final financial breakdown, however, most of what we needed came from ticket sales and private donation from an attendee.
Harvard: IOP, Regan Fund, UC
MIT: ODGE, Sloan Senate, COOP Gives, Sandbox Fund, LEAP Grant, Community Service Funds, UA Fresh Fund
Web presence
The email address boston@eaglobalx.org was a G Suite account managed by one person. This worked okay, although it might have been preferable to have a collaborative inbox instead.
Our Facebook presence consisted of the EAGxBoston page combined with the EAGxBoston 2018 event. We didn’t do much with it, but it worked well enough as something for people to share. A few people messaged the page with questions.
We used ti.to for ticketing. This turned out to be a problem for our capacity-constrained event because its waitlist management features were not very good: waitlist releases had to be done manually, which slowed things down especially near the end, and changing the numbers of seats while there was an active waitlist caused problems. It would probably be a good idea to look into alternative options before doing this again. An additional problem was that CEA gave us a preexisting account which we didn’t realize until later was already connected to the Stripe and PayPal accounts of EA groups in Australia and the Czech Republic, which meant we had to work with those groups to get our money back. This was especially annoying when dealing with refund processing.
Our event website was on CEA’s Contentful site. Initially we were unable to publish updates due to technical problems with the system, and publishing information about speakers took even longer because of permissions problems. It turned out to be impossible to publish the event schedule there, so we had to set up and link to a separate site powered by Netlify. How this should be handled for future events remains an open question.
Takeaways:
- Have a Facebook presence and email for easy attendee access and sharing.
- CEA’s official site and ticketing system will be somewhat unreliable, but can ultimately get the job done.
- You can consider building or using a different system, though this might be more trouble than it’s worth.
Audiovisual Services
For us, AV was largely an appendage of venue choice. CEA wanted all of our sessions recorded, so we needed mics and cameras. We considered searching our networks for amateur photographers who might be able to take charge of recording, but we decided against this because of our short timeline and uncertainty as to quality. MIT gave us a quote for professional videographers that ran about $4,000-$6,000, which was more expensive than we could justify. It was easier and more reliable to use the classroom cameras and have MIT’s formal AV provide mics and technicians, the latter of which they required based on the number of mics we needed. A six-mic rental package was about $180, and the technician time associated with it was about $600. It should be noted, however, that as of January 2019, we still have not been billed the amounts estimated and it is possible MIT covered this expense for us either by error or gratuity. We’ve received no response on our attempts to follow up on this. Your mileage may vary.
The hand-held mics we rented generally achieved their purpose, but were obviously inconvenient and unattractive for speakers, who would sometimes fail to hold the mic close enough to be picked up. For this reason we strongly recommend securing lapel mics for all presenters. Handheld mics work best for audience questions.
Rob Mather’s Skype call-in was our biggest AV challenge, but with the help of MIT’s technicians and the in-room conference call software and projector, all his slides and the man himself appeared as desired. We were happy with the clarity, lag, and ability to get live questions back to Mather. Our dry run with Mather two days ahead of the conference was essential to making sure everything went well.
Ultimately, our choice to use the classroom cameras had costs and benefits. It was much cheaper than our relevant alternative, and the recording equipment didn’t get in the way of attendees watching presentations and moving in and out of the classrooms. The quality was lower, however — consider this sample. Also, the recording windows had to be preset remotely, so they were not very flexible in terms of changing stop-start times. Fortunately we managed to build in buffers large enough to prevent missing any content. After the conference, the recordings themselves were sent directly to MIT’s STS office. They sent us the raw files, which we sent to CEA for editing. Be advised that CEA took at least several months to edit our files, so if quick publication is important to you, consider recruiting your own video editor.
Takeaways:
- Start exploring AV options early to give yourself more options on cost and quality.
- Make sure your venue has simple, accessible PC connections to project slides and video that organizers can use without needing login credentials they don’t have.
- Always pretest AV elements more complicated than simple recording, such as speaker Skype calls.
- Lapel mics are strongly preferable to handheld mics.
- Consider recruiting your own video editor for faster publication.
Food
Choice, delivery, and variety of food options was probably our second largest logistical item, and it prompted more intense attendee feedback than any other item on our agenda. This is the area where our monetary cost aversion hurt us most, and was probably our biggest mistake.
Our full day of programming called for three meals: a welcome breakfast, lunch, and afternoon snack. Lunch was fully catered and breakfast was ordered from Rosenfeld’s Bagels, but we put together the snack ourselves from grocery store supplies and supplemented the bagels with Trader Joe’s vegan cream cheese and Dunkin Donuts coffee. We recommend catering as much as possible to all future organizers.
Breakfast: 320 Rosenfeld’s Bagels, 36 Trader Joe’s Vegan Cream Cheese, 18 Dunkin Donuts Boxes o’ Joe. Two bagels per guest was a considerable overestimation, as were 1.5 cups of coffee per guest. We received complaints about the lack of gluten-free options, tea, fruit, and options other than carbs. Picking up bagels and coffee was not all that difficult, and given the sort of hacking one might have to do to get a good vegan breakfast catered, we think it wouldn’t have been too difficult to focus our personal efforts on pulling off breakfast and leaving the rest to professionals.
Lunch: Rhythm ‘n Wraps, a vegan food truck and caterer. We very severely under-ordered lunch, which was otherwise tasty, if delivered significantly too early. The under-order can be put up to having one person managing food solo. Our food lead miscalculated based on the EAGx 2016 order and we failed to follow through on concerns about the low price point. Rhythm ‘n Wraps was also very confusing to communicate with and there may have been a misunderstanding by one party or the other at some point.
Snack: Bread, chips, pretzels, baby carrots, vegan butter, hummus. This was an area where catering would have been a vast improvement. Our presentation and variety were lacking here, and again we had some complaints about overloading on carbs. A school or adjacent caterer probably could have made a nicer spread for everyone.
We bought 320+ bottled waters and 150+ sodas, which we kept available throughout. Attendees had difficulty keeping track of their drinks while moving between events, so we recommend future organizers buy stickers or markers for attendees to mark their drinks with.
Coordinating shopping times, car acquisition and fridge space proved to be time- and labor-consuming pains that weren’t worth only moderate cost savings, hence the advice to cater more — especially for food served later in the day, since cold storage then becomes an issue. Food quality and quantity seem likely to have a significant impact on people’s enjoyment of, and energy levels during, the conference, so we recommend that future organizers be less afraid to spend for better quality.
Takeaways:
- Here is a breakdown of what we provided.
- DIY food is more acceptable at breakfast.
- Cater as much as possible.
- Have a wide variety of both food and drink options.
- Do not skimp on food, this appears to greatly affect attendees’ experience.
- Have markers or stickers for keeping track of drink ownership.
Marketing and Design
Our constraining factor on attendee number turned out to be facility space, not interest. Marketing for the event was relatively minimal because of how fast ticket sales accumulated once we opened them. The best marketing is getting CEA’s sponsorship, a place on their website, and an email address from them. Pragmatically, build a Facebook page for your event and get in touch with school- or area-associated EA groups for mailing lists.
We originally expected to focus on university students, because we had our most reliable EA network connections with them. Our marketing lead searched out and contacted some college groups more than a month in advance of the conference to see about organizing group trips to Boston. A few of these partially materialized, but out-of-school professionals who learned about us through CEA dominated our attendee population. It may have been relevant that we held the conference just before finals season began for many schools, and we advise future organizers to keep academic calendars in mind when choosing a date.
Our design lead here created her own visual theme, which you can see in the Facebook page and event, and which we tried to be consistent with in our on-site imagery. We printed out some nicer 17x11 posters to mark the rooms where talks would be held and placed similar ones elsewhere in the building to direct guests to the main conference space.
Lanyards were our other major design piece. Customlanyard.net allowed us to upload our design and got printed lanyards with card holders to us in about a week for $1.11 per lanyard. We created, printed, cut, and inserted the paper nametags ourselves. This was somewhat time-consuming, especially given the fact we left registration open until the last minute and had to account for changes.
Takeaways:
- CEA’s endorsement will make marketing easy.
- On your end, have a Facebook presence and access to EA mailing lists.
- Do not assign someone exclusively to marketing too early in your planning process, or they may have little to do.
Day-of Execution
Our major tasks on the day of the conference were delivering and distributing food, directing and registering guests, managing transitions between rooms and events, and cleaning up.
We had a reliable team of 10-12 volunteers to help move food from cars to fridges and our event space. For a conference of our size, managing food distribution could easily be a 4-5 person job all day. Our fridge space in the building was several floors away and the coffee shop was a few blocks down the street, so food transport and management was a major project. Having lunch catered reduced this burden, so again we recommend catering as much as possible to minimize complication, excessive labor, and increase food quality. We also suggest that you put some thought into cleanup before your event starts: we did not have a plan for what to do with excess trash and had to simply stack it out of the way.
We had additional volunteers in the building’s main lobby to direct guests up to our third floor space, and some signs to help with this. Given our event size and hour-long arrival window (“breakfast and registration” 9-10am), one person at a time was able to manage the registration table. This task consisted of helping people find their premade badges, making some new badges for last-minute ticket purchasers, and fielding attendees’ logistical questions.
Once the presentations got started, some volunteers were responsible for short introductions of our presenters and announcements before and after events. Our opening remarks provided a link to the day’s schedule for guests to access electronically, along with a description of the rooms and the general plan and goals for the day. We also pointed out who the volunteers were so that guests knew where to direct concerns. We tried sharing a hastily-made and untested Google Sheet with nearby restaurants so guests could sign up for dinners together at local veg-friendly restaurants without overwhelming any few destinations. Unfortunately, this ended up failing to come together. We also used the opening remarks to announce our next-day meetup. It would probably have been better to announce both the dinner plans and the meetups earlier via email to get better turnout and stimulate more post-conference engagement.
From then on, our main tasks were introducing speakers and announcing breaks. At the end of breaks, guests also required a considerable amount of goading to take their seats for the next events. We had one fifteen-minute break and two 25-minute breaks apart from our 50-minute lunch. These were shorter than guests wanted. Making people go to talks during each block may have been a mistake, since networking is such an important component of EAGx and conversations with other attendees may have been higher-value than speaker presentations. We ended up sticking quite close to our formal schedule and never got more than 10 minutes behind, but we had to be the bad guys (and reduce guest value!) to do it.
It was unclear what Sloan expected from us in terms of cleanliness, so we erred on the side of caution and spent about an hour tidying up. Again, our excess trash was a problem, but we apparently kept it tidy enough that we received no complaints. The problem of excess food was eased by MIT’s free food mailing list, which produced a few hungry students to carry away some of our leftovers.
Takeaways:
- Have a team of several uncommitted volunteers to handle food management and odds and ends that cannot be covered by organizers with other responsibilities.
- Cater as much as much as possible to minimize logistical strain.
- Have people with food and supply lists to make sure everything you want at the conference is being delivered.
- Allow longer breaks, and don’t coop guests up in talks for too long.
- Have a plan for on-site food storage if needed.
- Have a plan for how to move waste quickly and neatly away from your event space throughout the day.
- Announce post-conference events in advance of the conference and make sure your plans are functional.
Financial breakdown
Income
Ticket Sales: $1,915.00 Private Donation: $2,000.00 COOP Gives: $500.00
Direct CEA reimbursements: $370.91
Total: $4,785.91
Expenditures
Food: $1,551.91
Speaker Travel: $669.91
Processing Fees (tickets and donation): $192.44
AV: $780.00
Lanyards: $177.63
Posters: $9.81
Total: $3,381.70
Net Income: $1,404.21
Thanks for the excellent write-up! I hope that anyone who comes across it in the future takes special note about your points on food (as someone who's handled catering for a few EA events, I winced at the thought of a bagel-only breakfast, but it's an understandable error to make). I'm really glad that the team behind EAGx Boston was so thoughtful in recording their experience; it sounds like 2019's conference will be even better.
Questions:
1. What was the story behind the $2000 "unsolicited donation"? Did anyone on the team know the donor personally? Was the money for defraying conference costs, or for the next conference?
2. Who actually ran the conference? I may be missing something, but I don't see any organizers or groups named in this post or on the EAGx website.
3. How did attendees review the event in general? Did they like the speakers? Did you notice any particular feature of the content (not the logistics) that was especially well-received, or that didn't work out as well?
No one on the team knew the donor, though he had donated to EA causes in the past and was acquainted with relevant people at CEA. We offered him VIP tickets and then he put $2,000 in the pay-what-you-want box in our online ticketing system. I think it was primarily thought of as defraying conference costs, and indeed we came in less than $2,000 under budget.
The organizers included Matt Reardon (OP and lead organizer) from Harvard Law School, Jen Eason and Vanessa Ruales from Harvard College, Juan Gil from MIT, Rebecca Baron from Tufts, and myself (no institutional affiliation).
When writing this postmortem, we actually did devote a section of it to a discussion of how the content was received, including individual presentations. Because most of the speakers were invited guests, this section will not be made public. I can share a few overall conclusions.
Overall, reception of the content in aggregate was positive. Some attendees were surprised by, and in a few cases critical of, the proportion of it devoted to animal welfare. This was not by design; most of the conference organizers are interested in animal welfare, but not moreso than other EA focus areas. Rather, it was determined primarily by the availability of speakers (most notably keynote speaker Bruce Friedrich). A few talks were also criticized by some attendees for being overly technical or of narrow interest.
Most of the panels were moderated by members of the organizing team; I think it would have been better to have these be moderated by people with deeper knowledge of the respective topics.
The anti-debate was an interesting idea whose specific workings we kind of just made up ad-hoc. I'd like to see it tried again, but only after further refinement of the format and clarity on how exactly it is supposed to work.
What was the issue with the anti-debate?
I don't think there was necessarily anything wrong with it, I'd just encourage future organizers to consider more explicitly what the goal is and how to achieve it.
This post was awarded an EA Forum Prize; see the prize announcement for more details.
My notes on what I liked about the post, from the announcement: