Categories
News

Conference Stats & Reflection

The following stats were collected at the end of the conference day, April 25, 2020 at 11:00pm.

Registration

  • 462 registered users (479 form submissions)
    • 80 registered today
  • 233 login success today
    • 229 login failures (54.7% success rate)
    • But this doesn’t take into account the users who may have already been logged in previously and saved their login information (i.e. no need to login again that morning)
By Conference End
  • May 3, 2020:
    • 501 form submissions
    • 492 registered users

Notes:

Registration / Login is one of the most difficult things to “get right” but it’s also one of the most important.

Although I spent a few days working on the best way to get users registered and logged in with entirely FREE methods, I didn’t have a lot of time to test and double-check the registration / login flow for errors and bugs. Common issues / complaints included:

  1. No “activation email” sent / received
    • Some users had to check Spam messages
    • Some users appear to have never gotten a mail (problem with the outgoing mail perhaps?)
    • I personally helped 5-6 users login by giving them a new password when they were having trouble
  2. Facebook login was set to “Developer Mode” until the day of the conference, so visitors using that login method couldn’t login before that
    • Easy fix, I changed it to “Live Mode” that morning. Didn’t realize I needed to do that until someone brought it to my attention.
  3. After registration / login some users were not redirected well
    • Mostly fixed by redirecting all logins to the main conference schedule on this site – but that wasn’t implemented until the day before the event, so some people got lost when trying to login early.
  4. There was no clear message to users that they were logged in at all (no “Welcome back” message, no user icon in the corner, etc)
    • Given the time constraints, this couldn’t really be avoided. With a little more time, I would have liked to show logged in status in the top bar.
  5. There was confusion about WHICH website people were actually logged into. It wasn’t immediately obvious the differences between KOTESOL’s various websites:
    • https://kotesolconf.com = Promotional site, built very quickly, used only for promotional purposes. Impossible to login here, though its similarity with the Live site was great enough to make many people confuse the two.
    • https://live.kotesolconf.com = LIVE site, where all the “protected” content was stored. Users were required to register / login to THIS site to access conference content.
    • https://koreatesol.org = Main KOTESOL site. Some users thought that user logins were automatically linked between this site and the LIVE site, but that was not the case. Given more time, that would’ve been the ideal situation (single sign-on), but the main KOTESOL site may not yet be capable of this (using Drupal 7).
  6. The plugin I chose only allowed login with a username, not the registered email.
    1. Some people (and sometimes me) forgot their username, although they would easily remember their email. This (I believe) led to the higher level of failed login attempts. In the future, I should choose a different plugin, or modify the code to also accept user emails.

Website Usage

  • 444 unique site visitors (from different IP addresses)
  • 569 unique sessions (some IP addresses were the same, meaning the viewer left the site, then returned later)
  • 2,235 unique page views (different visitors visiting different pages – returning to the same page is not counted)
    • Meaning each visitor viewed between 4-5 pages each on average
    • Average session duration (total site viewing time): 4:03
By Conference End

Possibly due to timezone differences, we actually had MORE visitors the day AFTER the conference:

  • April 25:
    • 718 site visitors
    • 1,160 sessions
    • 5,077 page views (4-5 pages / user, 6:20 average duration)
  • Cumulative totals by May 3, 2020:
    • 1,234 site visitors
    • 2,146 sessions
    • 8,635 page views (4 pages / user, 5:00 average duration)

Notes:

I originally estimated between 200-300 online attendees. Last year, I believe the National Conference in Jeonju drew over 300 in-person attendees. But this being KOTESOL’s first ever online conference, I estimated that less than that number would attend:

  • Theoretically MORE than that number could attend (especially with it being FREE), however
  • I also guessed that fewer people would choose to join an online conference than an in-person conference, especially due to the limitations of the online format.
Limitations of the online format and how we overcame them:
  1. Presentations: how can we do “live” presentations, and include speaker / attendee interaction, like a Q&A after the event?
    • YouTube LIVE = can easily accommodate a large number of viewers at once (Zoom and other options are limited)
    • YouTube Premiere = similar to LIVE but allows for a pre-recorded video to be played LIVE
    • Zoom = limited in size with rooms up to 100, and rooms with 50 or more start getting quite crowded and more difficult to manage
    • YouTube Pre-recorded webinars = By setting the videos as “Unlisted”, only those with the link can find them – and we didn’t release the links until the day / hour of the conference
  2. Networking: how can we enable conference attendees to interact with one another, ask questions, socialize in an online format?
    • Zoom lunch table = Open up a Zoom chat room during the lunch hour
    • Slack “hallway” = Utilize Slack (see below) as a virtual “hallway” where people would be able to mingle and ask questions

Slack

(i.e. The virtual “Hallway” of the conference)
Graph from Google Analytics, but shows where the site’s visitors came from that day.
  • 186 participants in 16 time zones
    • Almost 100 new users that morning
    • This means almost half the website visitors also joined Slack
  • 335 Slack messages
By Conference End
  • May 3, 2020:
    • 190 members
    • 2,085 messages sent

Notes:

I’ve used Slack for a number of purposes in the past, and it’s a great technology with numerous advantages over similar chatting apps:

  1. Join link = people can opt-in to join, and there’s no need to invite them individually (although you can). i.e. you don’t have to already be “friends” with them to add them to a chat, they can add themselves
  2. Multiple channels = in a single Slack space, you can open multiple public channels (rooms) that people can opt-in to join, or private channels that can be reserved for specific groups of people (team members, presenters, etc)
  3. Privacy = there is no need to “friend” someone, have their contact info (email, SNS ID, etc), or even know them before communicating. Once in the Slack space, anyone can DM anyone else. So, in some ways it’s a little like Twitter – but a closed-off Twitter space where only certain people with the join link can interact with each other.

For the reasons listed above, I felt that Slack was the ideal solution to address the problem of quick and instant communication, conference-wide announcements, questions & answers, and personal interactions and networking.

However, Slack was a new technology to many of these users, and I don’t think many of them really grasped from the beginning how useful the tool would prove to be. The following were a few things I observed when trying to get Slack up and running:

  1. Pre-conference communication: I wanted to open Slack up early and use it immediately for instant, mass communication with the presenters. Few presenters joined early, and fewer still used it as a communication tool before the conference.
    • In fact, most presenters still channeled their questions through a third-party that they already had contact with (via email or Facebook) before they got to me. So, rather than being in quick communication with presenters, I found I was still relaying information through other people.
  2. During the Conference: Nearly half the people who joined Slack did so ON the day of the conference, and we had rolling registrations throughout the day. This means that many people entered the chat without much experience, and so I had multiple repeated questions throughout the day. Therefore, I found it important to:
    1. Create announcements with the @everyone flag in the #general channel
    2. Pin posts to the channel with the most useful / frequently asked questions (even though most users probably didn’t even notice the pinned posts, it allowed me to quickly access the same information over and over again)
    3. Create Posts in Slack with longer information (like ALL the Zoom sessions links and passwords) that could be easily shared with a link
  3. Post-conference: I was a little surprised that the end of the conference also brought almost the end of the Slack communication. After we sent the “Thank you” messages at 4:00pm, there was a brief suggestion for a virtual “After Party”, as well as the creation of a group to share Pre-recorded webinar links and feedback, but since then, not much additional communication has really occurred. I’ve gotten only 2 DMs.
    • Still, keeping the Slack chat open seems like a good idea for now in case there are still conversations going on behind the scenes (DMs between attendees). And it allows users to still have access to all the links, announcements, files, conference discussion, or whatever else was posted.

YouTube

(i.e. The presentation “Rooms” and “Live” aspect of the conference)
  • 1,129 YouTube views
    • From our 444 unique visitors that day, this equates to an average of 2.5 YouTube views from each visitor
  • 152.4 hours viewed
  • Individual videos:
    • Opening ceremony:
      • 70ish Live viewers
      • 308 by the end of the day
    • Plenary:
      • 110 or so Live viewers
      • 219 by the end of the day
    • Meet the SIGS:
      • 30-50 Live viewers
      • 95ish by the end of the day
    • Hwang (Legal):
      • 50ish Live viewers
      • 95ish by the end of the day
    • TNKR:
      • 35 or so Live viewers
      • 125 by the end of the day
    • Pre-recorded:
      • 30-40 views by the end of the day on average
  • Devices (also applicable to the main website):
    • 45% Computers
    • 55% Mobile devices
By Conference End

Notes:

First of all, it’s important to note that the number of mobile device users exceeded computer users.

This means that any future KOTESOL conference or even workshop that seeks to do something online like this needs to take into consideration the prevalence of mobile devices that will be joining in. Therefore, the website and all utilized technologies need to be either optimized for mobile screens, or allow for the downloading of mobile apps to access the content.

Another thing to note is that many people were LATE to view some part of the conference.

Even though we had scheduled times posted, and some of the videos included 2 minute countdowns, many people “arrived late.” In one instance, this is because the end of the Plenary session overlapped by about 1-2 minutes with the Meet the SIGs video Premiere.

Next time, rather than trying to schedule multiple videos to Premiere quickly in a row, it would be better to combine those videos into one, OR give more time between the videos – because people need to go and click on another link to get over to the next video.

Note about YouTube Premieres: When a user comes to a Premiere late, the screen seems to automatically load the current video location – i.e. mid-talk. Most users would probably not consider trying to restart the video from the beginning, but would just continue watching from that place in the video. They could later go back and re-watch the whole thing again, but most would not consider trying to click to the start of the video if it’s already mid-video in the Premiere.

All-in-all, it seems that the success of this conference comes down more on:

  • The presenters, and how well they can manage their own technology
  • The attendees, and how well they can access the links to each video / Zoom session

Therefore, it is really important to try to make it VERY CLEAR:

  1. Where to find the links,
  2. Where the links lead, and
  3. The times that each link will be active (perhaps we should delay everything additional 5 minutes in order to get more people time to get over there).

Zoom

(i.e. The speaker / attendee “Interactive” element of the conference)
  • 13:00 sessions: between 20-30 participants
  • 14:00 Lisa Hunsberger: 60+ participants
    • Note: Lisa (among others) has agreed to record her presentation to make it more widely available on YouTube

Notes:

  • Passwords: Setting a password on the sessions, though intended to increase security (i.e. prevent trolls), actually hindered some participants from entering the sessions on time or at all.
    • In the future, perhaps passwords don’t need to be assigned. The join link alone will be enough, especially when that is only available for a short time behind a login.
  • Access: People had the most trouble accessing the Zoom sessions. I received numerous (almost a dozen or more) questions in Slack about it. To the point that I collected ALL Zoom info into a single document to be shared on Slack for the conference.
  • Analytics: I wish there was a more effective way to collect some participant data (like Google Analytics) on the Zoom attendance other than just asking the moderators to take a head count.

On the bright side, I noticed a few really creative uses of Zoom in the conference apart from a “regular” Zoom meeting:

  1. Individual recording: Reece Randall had trouble booking a room at his university to record his video, but was able to record himself, alone in a Zoom meeting while doing a screen share.
  1. YouTube Live: Vanessa Virgiel was in charge of helping TNKR with their session, and was able to hook up OBS Studio to record a private Zoom session between herself and the TNKR presenters, including a screen share, and live Q&A time.

Conclusion

By the end of the day, we’d also received 25 responses to the Conference Feedback survey with 100% positive impressions of KOTESOL (the only unanimous answer).

All in all, the majority of attendees seemed to enjoy participating and left very positive feedback about how smoothly everything seemed to run.

Conference by the numbers:

  • 462 website registrations
    • 233 successful logins (on the day, not counting those who’d previously logged in)
  • 444 unique website visitors
    • 569 unique sessions (125 users returning in the same day)
    • 2,235 page views (4-5 pages per visitor @ 4:03 per visit)
    • 45% computer / 55% mobile devices
  • 186 Slack users in 16 timezones
    • 335 messages exchanged
  • 1,129 YouTube views
    • 152.4 hours watched
  • 20-30 attendees in most Zoom sessions
    • 60+ in at least one Zoom session

Thanks for reading!~

By Conference End
Lessons learned:
  1. Making the conference FREE and open to all registrants, as well as hosting all the presentations on YouTube allowed KOTESOL to gather a larger crowd than originally anticipated.
  2. By leaving the website completely open (including registration) for a week after the date of the actual event, the actual size of the attendee crowd was more than DOUBLED even only ONE DAY after the “live” presentations.
  3. Hosting the conference “live” in the Korea Standard Timezone (UTC +9) also likely affected the number of people who were able to view the videos at their “live” times.
  4. Day One and Day Two of the conference drew the most active participants and engagement. And although Slack was quite vibrant in the public channels on Day One with 335 messages, there was actually a lot going on behind the scenes with Direct Messages and in private channels by the end of the conference (2,085 total messages sent).
  5. Although we had nearly 500 registered users on this site, there were more than 1,200 site visitors and nearly 2,200 YouTube video views. We can assume this means two (or three) things:
    1. About 50% of the site visitors either did not register for the site, or were unsuccessful (i.e. many may have just stumbled upon the site through Google Search or a Facebook link out of curiosity and saw no need to register)
    2. Each of the nearly 500 registered users likely watched an average of 4-5 YouTube videos over the course of the week (about the same number of presentations they may have attended at an in-person event)
    3. It is also possible that as “Unlisted” videos, the YouTube links could have been shared separately outside this website’s login page (i.e. if someone registered for Slack and NOT this site, they would still have access to the videos, or if any attendees chose to share a particular video, they would just need to share the link. Once the videos became “Private” on May 3, 2020, accessing videos with a link was no longer possible – now, invitation only, until links are made “Unlisted” or “Public” again).
By the numbers:
StatDay One
(April 25)
Day Two
(April 26)
End
(May 3)
Registered462492
Slack186190
Slack msgs3352,085
YouTube views1,1292,191
YouTube hrs152.4258.6
Website visitors4447181,234
Sessions5691,1602,146
Page views2,2355,0778,635

Sample Feedback

From Joanne McCuaig at the University of Birmingham (Promotional session)

Hello, I wanted to share my feedback to the conference team from the perspective of a presenter.

• The conference book and website were really well organised.

• The presentation options were helpful

• Zoom works better for our particular needs.

• We’re happy with the turnout we got – some attendees were late which was fine as it was live so their particular questions could be answered individually.

• Overall, we’re really pleased on our end and want to thank KOTESOL for the planning, organisation, and time that the volunteers have put into the conference.

Joanne McCuaig
PhD. Candidate, University of Birmingham
English Language and Applied Linguistics
https://www.birmingham.ac.uk/joannemccuaig

By Aaron

Aaron Snowberger is the Technical Chair for the KOTESOL 2020 National Conference. As such, he's building the conference website(s), as well as providing consultation and recommendations for all technical aspects of the online format. In 2019, he was the Conference Chair for the National Conference held in Jeonju, where he also created the conference website, app, posters, and book for the event.

Aaron is a 14-year Jeonju resident, a Google Certified Educator, Trainer, and GSuite Admin, and a professional programmer, graphic designer, and online educator. He has taught ESL for 14 years, and computer-related design and programming courses for 8 years. He is currently employed by Jeonju University.

1 reply on “Conference Stats & Reflection”