2021/Final report on the conference

WikiConference North America was held on Friday-Sunday, October 8-10, 2021.

Objectives

WikiConference North America (WCNA) is the annual conference for Wikimedians in North America, including Canada, the United States, Mexico and the Caribbean. WCNA 2021 was the 8th conference, and the second virtual conference. Building on 2020’s successful virtual event, in 2021 we wanted to expand and improve the virtual experience. Attendees of WCNA include new and experienced Wikimedians editing all Wikimedia sister projects in several languages (especially English, Spanish and French) as well as GLAM and cultural institution professionals, educators, and technologists. The program for the conference was designed to share information, educate and inspire these varied attendees. For many attendees, WCNA is the only large Wikimedia conference they attend.

The conference organizers committed early to certain goals or frames for the conference:

• (a) making the event inclusive, especially to Spanish speakers, by arranging captions or audio interpretation across languages, which would be new to the WCNA conferences;
• (b) a theme of global-and-local frames implemented in several ways;
• (c) supporting local events and being able to sponsor attendees who needed funding to make time or get equipment to attend
• (d) offering workshops and training for both new and experienced editors, not just one-to-many broadcasts of presentations.

Overall outcomes

Over the course of October 8-10, 2021, over 65 sessions were streamed across two presentation tracks and a breakout track, plus edit-a-thons. Two of the presentation tracks were also provided with live translation in English and Spanish, and these tracks were livestreamed on YouTube along with a livestream of their translations into Spanish (or English, for Spanish language sessions). There were three keynotes, two in English and one in Spanish. In addition, there was a special Friday night trivia broadcast with social media star Annie Rauwerda, creator of the popular Instagram and Tiktok feeds @depthsofwikipedia.

186 people logged into the conference site over the course of the weekend. Others watched our videos on YouTube. The attendee experience was positive, as reflected in evaluations and anecdotal comments. The technology for the conference was a combination of a paid platform (HopIn) plus volunteer efforts to run the streams to YouTube and moderate sessions behind the scenes. Learning from and building on the lessons of 2020 WCNA and the 2021 virtual Wikimania conference, the program and technology ran smoothly all weekend, with no major technical problems and a smooth presenter and attendee experience.

Organization team

The core organizers had attended the conference in the past and had compatible visions and goals. In addition to past WCNA organizers, new organizers were recruited and joined the team. Organizers met weekly, and a few volunteers joined for the conference itself to help moderate sessions. The many volunteers for this year's conference are listed here.

Technologies, platforms, suppliers

Many software platforms and services were involved in planning and executing the conference. Learning from other virtual events, we used a combination of technologies that we had used in the past as well as trying new platforms.

Technologies used to produce WCNA 2021

Technology	purpose	notes	free/paid
HopIn	A virtual conference hosting platform that uses Streamyard to stream presentations, with logged in attendees able to chat during talks and go to breakout rooms	Used as the main attendee platform.	paid
Kudo	Software that creates a virtual translation room that the translators could log into to control input/output	Used for translation. Kudo is already connected to HopIn so it was easy to integrate.	paid
YouTube	Used for live streaming talks in English and Spanish	Streams were produced for the two streams plus the translations of those streams	free (WCNA account)
OBS	free/libre software for live streaming to YouTube	Streams from HopIn were recorded in OBS by distributed volunteers on their laptops, and then sent to YouTube, which enabled streaming post-translation	free
Zoom	Video meeting platform	Used for participatory editathons (editona, vaccine workshop, and training workshops)	free (individual accounts)
Tlatolli	live translators	This is the translation collective that was used by Wikimania 2015 in Mexico City; in addition to translation they were very helpful in being involved in organizational meetings	paid
Etherpad	easily editable documents	Used for participatory notes during sessions	free
WorkAdventure	a free/libre platform for live interaction and discussion	Used for after hours social space	free
Menti	trivia platform	Used for live participatory trivia during the Depths of Wikipedia session
Slack	chat platform	Used for organizer and behind the scenes/day of communication	free
Telegram	chat platform	Used for general and attendee chat	free
Pretix	registration / ticketing platform	Used for registration	free
MailChimp	mass mailing system	Used for attendee communication and advertisement to previous attendees	paid

Many videos of talks are on YouTube; final processing of videos and making them available on YouTube and on Commons is still unfinished.

Sessions and spaces

We had about 65 scheduled items on our Schedule.

• Red and Blue stages: These were broadcast presentations, from Streamyard. The audience was not visible and could not speak. Often the presenters showed slides or a browser tab/screen. These were recorded by Hopin and interpreted between English/Spanish, and generally shown live on YouTube. A moderator was present backstage to help speakers share their screens and help with Q&A from teh audience.
• Pre-scheduled sessions: Those in our "Green track", or breakout track, were generally not recorded. They did not have language translation/interpretation. The people attending could choose to be visible generally, unless there was a pre-designated moderator deciding.
• Editathons, lightning talks, and unconference sessions - Open chat, mostly "sessions" on Hopin; a couple were zoom meetings
• Expo: We had a few prepared YouTube videos. Attendance was light but it meant people could see some presentations before or after the scheduled time.

The schedule incorporated breaks and five-minute pauses which helped with switching presentations and making sure that the behind-the-scenes technology ran smoothly. Moderators used the HopIn chat to remind attendees to add their questions and comments, as well as reminders of the safe space policy.

The keynotes for the conference included "Breaking the gender gap on Wikipedia / Rompiendo la brecha de género en Wikipedia" (Carmen Alcázar, from Wikimedia Mexico); "Trust and Knowledge on the Global | Local | Glocal Level" (Connie Moon Sehat, from the Credibility Coalition); and "New Maps for an Inclusive Wikipedia: Strategies to Counter Systemic Bias" (Carwil Bjork-James, English Wikipedia and Vanderbilt).

Attendance

See Figures 1-3. (We may not be able to replicate these on meta, depending on copyright stuff, but we can cite them.) Overall over 300 registered, but this includes a number of duplicates and people who could not or did not actually attend. 186 logged in to our Hopin event at some point. The peak attendance at one time was 92 people, early on Friday afternoon, possibly when Carmen was giving her invited talk. In the first chart, attendance never drops near zero even at night because there was no need to log out. On Thursday evening we invited attendees for a social/test time on the platform.

• (Three?) editathon events linked from the conference program were held on Zoom. These gathered perhaps 20 participants overall, most of whom were also conference attendees.

• English/Spanish interpretation: In (about half) of the sessions, those on Hopin's "stages" (our red and blue tracks), attendees could click for interpretation and select a language, English or Spanish. It went well overall, with some glitches. The systems are complicated. Show diagram. Interpretation services were offered in the large presentation rooms, which we called Stage Red and Stage Blue, but not in the "breakout rooms" for workshops, editathons, unconference discussions, or lightning talks.

• Safe space matters: There weren't safe space issues during the conference. We had prepared. A team went over training materials in advance and wrote up a document of what our practices would be at game time. We edited the safe space policy. Volunteers signed up for shifts and sessions. We think the document is worth sharing.

• Local events: We had in-person events on Sunday afternoon in various locations. NYC picnic (photos) and Mexico City, funded by the grant. A parallel picnic was held in San Diego. Our grant anticipated having more local events. There was less interest than anticipated, and we did not push the point; there was interest in the online event, and caution about the ongoing covid pandemic.

Evaluations by attendees

We distributed survey questions of attendees, and got 35 responses between Oct 11th and Oct 19th. We invited ratings from 0-10 on their overall conference experience and 28 of the responses were high, in the 8-10 range. The main platform, Hopin, got 25 responses that high, but a number of attendees had difficulty and expressed frustration with it. We asked about their experience with the live audio interpretation, and 14 said they used it, and of these 12 said it worked well. We invited specific comments about difficulties or matters of special interest, and there were a variety of interesting answers. They are hard to summarize, but generally positive and informative, and worth reflecting on for next time.

Local events

We funded in-person Wiknics in NYC and Mexico City, and a small one was also held in San Diego. Add detail and pix please

Plans not achieved

We meant to give scholarships for attendees to cover child care or other needs to enable them to attend, but we could not manage an application process. It was not clear that there would be much interest; people seemed to take the time to attend some sessions and just skip others. We were willing to hold more in-person wiknics, but partly because the main event was online, and partly because of covid concerns, there were only two major proposals. We made major platform decisions late which was stressful and meant we could not fully plan and train ourselves. We have not uploaded most of our videos yet, but we can link to the youtube streams for several sessions.

Budget

Our original budget was set by a grant from WMF, totaling $US 24,508. Our actual expenses exceeded this by 12%. Savings from previous years and donations made up the difference.

The effort to scale up for over 200 registrants and English/Spanish interpretation raised our costs over other conferences. We made decisions late in the process to scale up, getting Hopin's "business plan" (which lasts a year and includes some support), and the high-end Kudo software service to integrate interpreters. These raised costs but it was not clear we could succeed with smaller-scale, lower-end components. It left us with a year-long subscription to Hopin which our component organizations can use again, and plan to use for Meta-day.

Expenses

Expense	Cost in USD	Notes
Hopin	$10,000	Main conference software platform
Kudo	$9,047	Live interpretation software platform
Tlatolli Ollin	$6,635	Interpreters at conference (en<->es)
Amtrad	$931.75	Interpreters at WikiCari (en->es)
Zoom	$121.51	Experimental conference platform
Mailchimp	$21.30	Publicity communication
WMMX picnic	$294.73	Picnic in Mexico City
WMNY picnic	$371.73	Picnic in New York City
Total	$27,423.02	11.9% ($2915.02) greater than grant

For the future

• It would help us to hold future events if WMF Events could scale up the services it offers, so as to offer some prearranged subscriptions to platforms and services, more of which would have open-source commitments. We were short of time and skills to evaluate these platforms and did not have bargaining power to negotiate for our one event -- but WMF affiliates hold events every week around the world.

Reports from previous years, for reference

• 2016, 2017, 2018, 2019