SPC Tracking
Home Status Registration Staff Reference Region Video Report

Report

Conference Closeout Report ** FINAL **

Case Study: Implementing a Global Conference Using a Virtual World Venue

I. Executive Summary
II.
Diagrams
    A. Large Group Presentation Design
    B. Small Group Interaction Design
III.
Pictures
    A. Keynote Presentations
    B. Networking Discussions
    C. Expo Area
    D. Multiple Presentation Options

IV. Conference Logistics
    A.
Conference Assignments
    B.
Assessment of Results
V. Appendices
    A.
Cost Savings
    B.
Comments
    C.
Evocative Questions

I. Executive Summary

Bottom Line

The Linden Lab Solution Provider Conference used a virtual world venue to reach a larger global audience at a substantially lower cost while successfully maintaining high value and convenience for participants and staff.

Goal

The goal of the conference was to share Linden Lab company plans and directions with its solution provider community, a large global audience of independent developers. The desired outcome of the conference was that developers have increased awareness and confidence in Linden Lab business directions.

Share Linden Lab plans and directions with SP Community (Outcome: more confidence in LL & business directions)

Provide clear SL marketing & value messages (Outcome: more consistent messaging, better competitive positioning)

Approach

The conference was provided online using the Second Life® virtual world platform, which provided an interactive, three-dimensional, spatial venue available via internet to multiple global participants logged in and interacting in real time.

Intent to make it as highly interactive as possible. People were broken into small discussion groups. Expo area where people walked around booths, had multiple interactive conversations, and interviews. This was doable in one single medium.

Result

The conference was successfully implemented, providing 6 hours of programmed interactive speaker presentation time plus 6 hours of informal participant networking time. Total registration was 142 participants, representing 112 of the 406 listed Solution Providers eligible for the conference.

Value to Participants: A majority of participants reported that the conference was useful and productive (94%), that it met its objective (97%), and that they would attend another such conference (94%) as well as recommend to others that they attend (94%). A majority of participants reported satisfaction with speaker interaction via Q&A (97%) and the ability to network with other participants (94%). All participants reported being able to attend at the scheduled times.

Cost Reduction: Average participant cost to attend via virtual world venue was 30 USD, compared to 750 USD for an equivalent real world venue (96% Cost Reduction). Sponsor cost to reach 140 participants via virtual world venue was 6 KUSD, compared to 45 KUSD for an equivalent real world venue (86% Cost Reduction). An additional 360 participants could have been served (250%) at a moderate (5%) increase in production cost.

Participation: Approximately one-third (28%) of the solution provider organizations eligible for the conference were represented at the conference. More participants were able to attend via the virtual world venue than had the conference been held in a traditional physical location. This is because half the audience were located in the US and half in Europe. Physical location did not matter because speakers and participants were able to interact online. Travel time and costs were eliminated while high content and interaction values were maintained.

Met Conference Goals: Survey and feedback showed that conference goals were fully met.

Conclusions

This conference provides an example of a high-technology company reaching a world-wide audience of high-technology developers (where 'high-technology' includes people interaction methodologies as well as development). This conference format can apply for ANY large company providing a service to a large global audience via a smaller, intermediary, global technical outreach group - where that outreach group may be an employee base, franchise base, independent contractor base, retail base, or other.

The use of a streaming video format within a virtual world venue worked well, providing participants high-quality audio and moderate-quality video sufficient to deliver program content equally to all attendees in all regions, including presentation, interactive Q&A, and breakout discussion sessions. This format allows a scalable methodology able to reach moderate (100-500) and large (1,000) registrant groups equally well, including providing scheduled as well as informal networking activity. Improvements for future conferences are to improve slide legibility and to investigate ability of participants to receive a 4X higher resolution video image.

The use of Second Life gave flexibility to reach 140 people by direct video presentation and facilitate 25 separate breakout discussions and provide a company booth Exposition area, all hosted using one media platform.

Linden Lab required a confidentiality agreement which added some complexity to the logistics of registration. This, plus the delivery methods used, allowed company confidential information to be communicated in an acceptably secure manner.

It should be noted that the conference audience and staff for this conference had prior experience in Second Life. Organizations looking to use a conference virtual world venue with inexperienced staff and participants should allow extra preparation time and cost for incorporating virtual world experience into their implementation plans.

II. Diagrams


(click for larger image)

Large group presentation design used for the SP Conference allowed serving hundreds of participants at a time with equal, high-quality experiences and interactions with presenters.

(click for larger image)

Small group interaction design allowed participants to form many small discussion groups after each keynote presentation. In addition an Expo area was provided where people walked around to visit booths, give interviews, and engage in multiple individual conversations. This was all done within one single medium, the Second Life® platform.

III. Pictures


(click for larger image)

Keynote Presentations: M Linden, CEO, and Glenn Linden, Director of Developer Relations, answer questions about Business Directions during Live Q&A. Moderators on each conference region forwarded questions from 120 Solution Providers live to M and Glenn.

(click for larger image)

Networking Discussions: Following each presentation, Solution Providers assembled in small groups to discuss the material and its implications for their work. Additional, informal discussions were held between keynote days facilitated by Glenn and Zero Linden.

(click for larger image)

(click for larger image)

(click for larger image)
Expo Area allowed Solution Providers to set up booths describing themselves and their companies and to see what others are doing. Seventy (70) company booths were set up in one day! M Linden toured the Expo following his keynote and spoke with several participants in front of their displays. The Expo remains available through Summer 2009.

(click for larger image)

(click for larger image)
Multiple Presentation Options: The use of video streaming allowed multiple options for feeds including recorded, live, and SL. T Linden's presentations were transmitted from recordings made in San Francisco prior to the conference but his Q&A's were transmitted live via SL from Brighton UK.

IV. Conference Logistics

A total of 30 staff were required to reach 142 conference participants. An additional 360 additional participants could have been served (250%) at a moderate (5%) increase in production cost (6 more event staff).

A. Conference Assignments

Task Conference and Event Staff
(21 people)
Linden Lab Staff
(9 people)
Project Management Conference Manager Linden Lab Contact
Speaker Preparation and Management - Linden Lab Contact
Presenters (5)
Audience Management Conference Manager
Registration Manager
Event Staff Manager
Hosts (6)
Moderators (6)
Event Transcriber #1
Linden Lab Contact
Event Transcriber #2
Audio Video Conference Manager
Video Stream Provider
Camera (Recording)
Camera (Live)
Linden Lab Contact
Presenters (5)
Video Web Casting Manager
Event Information Conference Manager Linden Lab Contact
Regions Conference Manager
Signage and Graphics Manager
Builder
Linden Lab Contact
Concierge

B. Assessment of Results

Item Assessment Discussion
Overall
Conference
A WORKED WELL: Live video streaming of speaker presentations and Q&A; Speakers audio quality, Speakers visual image quality
DIDN'T WORK WELL: Slides visual legibility
WORKED WELL: Event information, Region setup, Registration, Event staff, Speaker content and presentation, Expo and Discussion

Final approved conference management specification is documented here.

A key concept of the specification was to use streaming video to serve all participants on all sims an equal interactive experience with speakers. A benefit of this streaming video presentation is easy scalability for medium to large to extremely large numbers of participants.

The video resolution used was 320x240 to assure that all participants were able to handle the stream via their local internet connectors. Participants reported audio quality was satisfactory but that image quality was moderate, especially when information slides were shown during the pre-recorded presentations.

A recommended modification for next time would be to re-test the feasibility to use a 4X higher video resolution of 640x480. (640x480 was tested prior to this conference, however, the test audience reported that half COULD see the image OK but half could NOT. However, ALL could see it at 320x240.)

A recommended modification for next time would be to use simpler slides as part of the video presentation; and provide more detailed images, as needed, separately. For example, the "roadmap" slide was posted in world as a stand-alone image visible to all during the presentation.

At issue here is the use of detailed slides for medium to large conferences delivered in-world. High resolution imagery is always bandwidth-intensive, whether delivered as part of a dynamic video stream or as a static image slide show. For this conference, high-resolution slides were provided as part of the conference archive.

One participant suggested using their own guideline for slides: "Put no more on a slide than can be seen and read on a tee-shirt."

Event Information A WORKED WELL: Event emails sent weekly; Registration status updated daily

Conference information web site is documented here.

Event communications consisted of emails from the Director of Developer Relations, describing conference opportunities and timing, combined with a Conference Website, providing detailed program information including registration form, registration status, and email addresses for additional program and registration inquiries.

Region A WORKED WELL: High carrying capacity of overall region; Ease of re-tasking individual sims;

Final region settings are documented here. Final region key map is documented here.

A region consisting of nine (9) sims was established to serve keynote presentations to a specified potential capacity of 500 participants and staff (~55 per sim).

Registration projection based on trajectories (number of registrations per first week, second week, etc.) indicated 135 actual participants were likely. Therefore, six (6) regions were kept for keynote presentations, one (1) was re-tasked for an Expo and Discussion area, one (1) was re-tasked for Conference Archive; the remaining region was held as spare.

Actual registration was 142 served by 14 staff on five (5) sims (~35 per sim).

Keynote sims used a varied terraform but identical presentation areas consisting of open kiosks, video screens, video test areas for participants, flags, and signage. A benefit of this sim build approach is easy scalability for medium to large to extremely large numbers of participants.

Registration A WORKED WELL: Standard communication notes from Registrar to participants at each stage of registration. Standard reminder notes sent weekly then daily. Listing of registrant status on the conference website, updated daily.
WORKED WELL: Communication between Registrar and Linden Lab on questions of eligibility, NDA, fee adjustments.
WORKED WELL: Communication between Registrar and Conference Manager to monitor and maximize confirmed registrations.

Registration communications to participants are documented here. Final registration status is documented here.

Registration required several complexities: Applicants had to be listed Solution Providers, requiring communication with a Primary Registrant on file with Linden Lab. The primary registrant could then add two additional staff from their organizations. ALL registrants had to complete and return a signed NDA by fax or scan attached to email. SOME registrants qualified for a waiver of registration fee based on records with Linden Lab. The remainder were required to pay a registration fee that was discounted for registrations received before a time limit; a full fee was required after the time limit.

A key learning from this conference was that a significant fraction (~10%) of participants and staff were hesitant about providing signed confidentiality agreements. With additional clarification provided where needed, many of these concerns were alleviated and registrations were confirmed. However, 14 organizations originally submitting registration forms never supplied requisite NDA's, despite repeated follow ups.

Event Staff A WORKED WELL: Immediate audience assistance with program questions, audio video setting questions, positive energy
WORKED WELL: Communication between Event Staff Manager and Event Transcribers (1 in world, 1 in San Francisco presentation room)

Event staff recruiting, training, and scheduling are documented here.

Each presentation sim had one (1) Host and one (1) Moderator. Both assisted participants to answer program questions and assure they could see and hear the video. Both assisted participants to forward questions for the speakers during live Q&A.

Event staff were in communication at all times during the presentations with the Event Staff Manager and Conference Manager. The Conference Manager was in communication with the conference presentation room in San Francisco. Thus a two-way feedback loop allowed almost immediate adjustments in stream volume, individual reassurances to participants, etc.

Positive staff morale was positively remarked by several participants. In fact the informal staff motto to take care of participants was, "Keep 'em chirpy".

Audio Video Streaming A- WORKED WELL: Professional recording and live camera, gentle movement (panning and zooming) during live portion to keep audience eyes refreshed

Audio video service specification is documented here.

See also Conference Management Specification for discussion of 320x240 vs. 640x480 video resolution.

A key learning of this conference is that real world video production companies, even ones that advertise expertise in "webcasting", are NOT fluent in Second Life needs. In particular, a high-quality high-capacity video stream may be had from Second Life companies that meet the need for, literally, less than one percent (1%) of what some real world companies are charging. That said, a quality video setup and camera for recording and live transmission are well-done by these companies and should continue to be used.

Audio Video Logistics B WORKED WELL: Do a full technical rehearsal 24 hours in advance
DIDN'T WORK WELL: Matching audio-video camera formats to video-web-casting took more time than predicted.

The stream inside of Second Life was easy. Getting the video feed TO the stream was a bigger deal than expected. For example, the tape used for the recording camera did not work in the camera used for live broadcast. The feed from the camera did not match the computer used for video-casting. All of these were resolved successfully but required several extra hours during setup.

Speakers A WORKED WELL: Pre-recording presentations allowed for increased review, rehearsal, and content coordination
WORKED WELL: Live Q&A with each speaker immediately following their pre-recorded presentation.

Speaker content and presentation were positively remarked by participants.

A key learning of this conference is that pre-recording keynote speakers serves multiple purposes well. First, it requires presentations to be prepared and rehearsed early. This improves focus and consistency across speakers because the speaker coordinator is aware of, "Who is covering what" and because speakers know, "This is serious, I'm being recorded." Second, pre-recording provides a backup in case a key speaker is suddenly unavailable during the conference.

A second key learning is that live Q&A with speakers can be incorporated easily into the time slot immediately following the pre-recorded presentation. An delivery rate of one question delivered-and-answered-per-minute was achieved for all six keynote speakers. Breaking up presentation with Q&A also kept participants refreshed in what was an intense program schedule. Finally, live Q&A could even be achieved with the speaker calling in from a remote location and being present as their SL avatar but still video-streamed to the 142 participants. Something NOT possible using traditional one-avatar-one-region approaches.

Breakout Discussions A WORKED WELL: Participant breakout discussions; formed easily and focused on questions easily
DIDN'T WORK WELL: Confusion in forwarding a summary of discussion results rather than a full transcript

Following speaker presentations, participants were split into small groups of about seven (7) to address a focus question.

A key learning of this conference was that focus groups CAN be done using text only, but that it must of necessity go slower than using voice. That said, the groups formed and began their work quickly and intently.

At issue for another time is whether to use SL Voice. For this conference it was rejected as a potentially disruptive detail (using streaming video with voice disabled in the region, then suddenly re-enabling voice and asking participants to turn it on). Another factor is the a substantial number of participants (~20%) do not like to use voice or have equipment issues with voice. Even on the Discussion day (where no video was streamed), the majority of participants stayed with text only.

The first day of breakout sessions enjoyed full participation. The second day about half left early. However, the participants remaining were as focused and intent as the first day.

Expo and Discussion A WORKED WELL: Expo and Discussion use of spare sim

Participation in the informal discussion times was much higher than expected. Two (2) two-hour sessions were scheduled and well attended. Some participants stayed a total of six (6) hours.

Half of registrants also set up information booths (70) in the Expo area.

Conference Archive A WORKED WELL: All presentations, Q&A, and slides available the week following the conference.

A conference archive area was set up where participants could revisit ALL six video presentations and live Q&A. High-resolution copies of slide images were also made available. A benefit of having a conference archive is reaching participants in extreme time zones, and participants unable to make the scheduled times.

V. Appendices

A. Cost Savings

Use of virtual-world-enabled conference technology reduced cost to participants AND staff by over 90%.

Item Virtual World Venue Traditional (Physical) Venue Cost Reduction
Management Fee
- Conference and Event Staff
- Audio Video    
Subtotal

Grand Subtotal


4.4 KUSD
1.8 KUSD
6.2 KUSD

6.2 KUSD


17.6 KUSD
7.2 KUSD      
24.8 KUSD

24.8 KUSD




 

18.6 KUSD (75%)

Venue Fee*

Grand Subtotal

-

-

2.0 KUSD/Day

4.0 KUSD (2 Days)

 

4.0 KUSD (100%)

Per-Staffer-Expense
- Air Fair
- Lodging        
Subtotal

Grand Subtotal


-
-                
-

-


270 USD Domestic Round Trip
300 USD 2 Nights
~750 USD/Staffer

22.5 KUSD (30 Staffers)




 

22.5 KUSD (100%)

Per-Participant-Expense
- Registration
- Air Fare
- Lodging        
Total

29 USD
-
-                
~30 USD/Registrant

4.2 KUSD (140 Registrants)


165 USD
270 USD Domestic Round Trip
300 USD 2 Nights
~750 USD/Registrant

105.0 KUSD (140 Registrants)





 

100.8 KUSD (96%)

Grand Total 10.4 KUSD 156.3 KUSD 149.9 KUSD (93%)

*Venue Fee: Linden Lab used its own Second Life® region for the conference.

B. Comments

Participant comments while in session substantiate assessments and flesh out statistical results.

For quote by name, ask permission. Else use anonymous. Else invite an interview.

Source Conference Pluses Conference Minuses Conference Neutrals
During Conference (Chat) Overall Conference
Gus Plisskin: great conference. (yes really)
Kim Anubis: It turned out great
Kim Anubis: So glad you chose to do this in world.
Gambino Amsterdam: what stood out was to meet all this cool people
Ham Rambler: the whole team did an amazing job

Use of Video Streaming
zFire Xue: Very interesting to see some Lindens on video for the first time

Ability to Network
Ham Rambler: the SP group is so diverse... and just sitting talking to [other SP's ]makes me feel we can really learn from each other
iLike Loon: I think it's interesting to exchange opinions and suggestions now
Patrice Cournoyer: great sessions earlier
Patrice Cournoyer: making some good connections for future I think
Fim Fischer: Guys, it was great to be here at the conference with you today. I hope to see all of you again tomorrow
Patrice Cournoyer: great day
Kim Anubis: Have a great day, everyone. It was good chatting with you.

Program Content
Hiro Pendragon: It's VERY encouraging to see Linden Lab tailoring its future efforts to specialize for different uses
Maccus McCullough: maps looks interesting
Vasilisa Shilova: oh yes
Arian Writer: I like it
Roland Francis: applause on that
Goldie Goodman: Nice shadowing...
Maccus McCullough: outstanding - content reuse is a significant concern for our organization
Debs Regent: this is great
JoJa Dhara: great answers!!
Aki Shichiroji: ok. YES, we NEED Flash support. and interactivity with web in SL
Patrice Cournoyer: and the hits keep coming
Aki Shichiroji: yaay. document editing in world.
Elsa Dickins: oh cooooooool
Count Bayliss: OK - now they showed me Mesh and shadow - I wish they had not - I Want it now

Expo
Vryl Valkyrie: I have enjoyed seeing all the booths
Vryl Valkyrie: Seeing what everyone has to offer
Vryl Valkyrie: It really expands the horizons and is very inspirational to say the least.

Participation
Patrice Cournoyer: I am impressed at the number of SP
Vryl Valkyrie: Doesn't this almost feel like a kind of fair.. so love it It's been great chatting with you all.. cya later.. I'm off to my red [breakout discussion] area. :-)

Live Q&A
Vryl Valkyrie: I'm so honored they answered my question and SHOCKED!
Vryl Valkyrie: Thanks Glenn!
Vryl Valkyrie: and Moderator Ware :-)
Motorato Ware: just doing our job, m'am :-)
Ariella Languish: GOOD question
Vryl Valkyrie: Of course this will put LL in the market again as far as corporations are concerned who need the secure isolated server. This product will have great potential.
Shy Robbiani: .-'`'-. APPLAUSE APPLAUSE .-'`'-.

Future Conferences
Vryl Valkyrie: Yes I agree, I have enjoyed the entire conference and would love to see them held more often. It's a great learning experience.

Audio Volume
Dom Masala: audio very quiet
Zeus Zetkin: very low audio
iLike Loon: yes

Slide Legibility
Maccus McCullough: The video on the end is too blurry to get anything meaningful out of the picture.
Arian Writer: I agree
Goldie Goodman: The first rule of making graphics for video is, don't put more on the screen than you would on a T-shirt.
Larry Pixel: the slides are useless
Larry Pixel: can't read a word

Program Content
Roland Francis: that's old info
Larry Pixel: the last three custom last names I requested were all turned down -- they are not anything you like

Live Q&A
Anders Wildcat: Why are you fielding questions via IM and giving them to the speaker without attributing them to the person who ask them, I get the impression that these are pre-scripted questions. Using chat for Q&A is so much more dynamic...
Vryl Valkyrie: if any of them even get there.. seems like these are prescreened questions

Conference Archive
Anders Wildcat: The entire point of a Second Life meeting is to have a dynamic conversation in REAL TIME, your comments about "you get the PPT on Thursday" we'll answer the questions later in email, defeats the idea of a Second Life meeting!

Overall Conference
Anders Wildcat: Why didn't we just use WebEx?
Roland Francis: It shows what a conference can and can't do today, and that in itself is interesting when you read between the lines
Anders Wildcat: So if video conferencing is bad, why are we running this as a video conference? :-)

Program Content
Gavrael Hawkeye: this is the same as with new technologies in the web. I was a web designer for years and I hated to dumb down my wonderful software all the time for the noobs in the web
Gavrael Hawkeye: but on the other hand we need to please them too

Live Q&A
JoJa Dhara: How do we explain this to our clients this goes beyond usual ... a laptop with a frog is answering great questions?

C. Evocative Questions

Question Discussion
1) Does SL work for a medium sized group? We had 150 people and were prepared for 500. The only complexity in reaching larger groups is being sure to add extra staff on each additional region to facilitate small group breakout discussions.
2) How can you structure a conference in SL so the participants are participants not recipients? Use one large presentation parcel and five small voice parcels per region.
Large parcel is Media Yes, Voice No;
Small parcels are Voice Yes, Media No.
Always have one of the small discussion areas be Text Only.
3) How do you best deliver content to a group in SL? We chose to use streaming video as optimal for this purpose but there are other options.
4) What is the participant experience of a conference in SL? Participants in this conference had low expense, high convenience, and survey showed that their goals for attending were fully met.
5) How do you deliver content in multiple languages? Prerecorded presentations can use multilingual subtitles or audio streams. Live Q&A can use live interpreters. Increased preparation time and expense will be required.

As an interim practice: Holding discussions in Text allows multilingual participants to follow better than in Voice.

Executive Summary
Diagrams
Pictures
Conference Assignments
Assessment of Results
Cost Savings
Comments
Evocative Questions

 
Staff Confidential: This area is for managing conference preparation and implementation.