For years I’ve been successfully facilitating in-person group discussions at meetings, using the simple fishbowl and fishbowl sandwich processes. These techniques work because at any moment, only a small, clearly defined, (but constantly changing) group of people are involved in the discussion. As a result you can moderate an interesting, orderly discussion with hundreds of people, any of who have an equal opportunity to speak.
Online group discussions bring a new set of challenges.
We have all experienced poorly facilitated online meetings, where people unilaterally turn on their microphones and speak away, colliding aurally with others and monopolizing the conversation. An experienced moderator can minimize this behavior with a starting set of clear agreements that participants will follow during the discussion.
But however good the facilitation, there is far less environmental and body language information available online than in-person. The subtle cues we’ve all learned for moving between listening and speaking in a conversation are largely absent. (Stephen Mugford and Pamela Kinnear go into more detail here.) This makes creating a useful, flowing discussion harder.
Existing solutions and their limitations
Some of the fancier online meeting platforms provide functionality that can support simple fishbowl process quite well. Typically they use the “panel on a stage” model. A moderator moves audience members who raise their hand in some fashion into a panel (speaking) seat. When people have finished speaking, they leave the stage and the moderator can fill their seat with someone else.
Currently, though, such platforms don’t make it easy to move people in and out of pair or trio share groups: a requirement for the “bread” portions of the fishbowl sandwich.
One of the reasons I like to use Zoom for online meetings is its reliable and easy ability to quickly move people into breakout room groups for sharing. Zoom is a great tool for a discussion session’s opening and closing small-group fishbowl sandwich sharing. But how can we moderate discussion amongst a sea of faces during the fishbowl itself?
How to effectively hold a discussion online in Zoom
I’m indebted to Stephen Mugford for suggesting a simple and effective way of moderating fishbowl in Zoom.
In-person fishbowls use “people sitting in the 3 – 5 chairs up front, facing the group” to indicate who can speak at any moment. For a Zoom discussion, the same delineation can be made. Simply ask everyone except the facilitator/moderator to turn off their camera and microphone at the start.
When someone wants to speak, they turn on their camera and microphone. A nice feature of Zoom is that their picture will then jump to the top of Zoom’s speaker or gallery view. This makes them easy to spot.
The moderator guides the order of speaking and discussion with those who are “live” in the usual way. When people have finished sharing for the moment, they turn off their video/microphone and return to listening.
Sometimes when I run a fishbowl in person there are many who want to speak. I have them queue up in a short line at the side of the chairs. That provides feedback to the folks in the chairs that maybe it’s time to give someone else a turn!
Similarly, you can use Zoom’s text chat to queue up people who wish to enter the fishbowl. This allows:
those who are calling in by phone to signal they want to speak; and
the facilitator and group to see how many people are waiting to speak.
In practice, I’ve found the suggested live/listening camera/microphone protocol works very well. I only add using text chat as a signaling channel when there are participants who are calling in by phone.
This is a simple and successful way to implement fishbowl and fishbowl sandwich discussion process in Zoom. I recommend you try it! And if you have used other platforms to run these processes successfully, please share in the comments below!
What’s the best way to facilitate a community discussion? Recently, I had to answer that question at short notice. My task: design and facilitate a two-hour community discussion in response to a bombshell announcement made by the largest employer in my tiny rural hometown of Marlboro, Vermont.
[Update: Want to know how to do this online? See this post!]
The community was in shock. Consequently, I felt it was important to use a discussion format that:
Supported respectful dialog from a variety of constituencies;
Created an environment that was as safe as possible for people to share;
Minimized the likelihood that people would monopolize the meeting;
Allowed both short statements and controlled impromptu conversations; and
I’m in San Antonio, Texas, having just run two 90-minute “panels” at a national association leadership conference. I say “panels” because at both sessions, the three “panelists” presented for less than five minutes. Yet after both sessions, participants stayed in the room talking in small groups for a long time. That’s one of my favorite signs that a session has successfully built and supported learning, connection, and engagement.
You may be wondering how to effectively structure a panel where the panelists don’t necessarily dominate the proceedings. How can we let attendees contribute and steer content and discussion in the ways they want and need? There’s no one “best” way to do this of course, but here’s the format I used for these two particular sessions.
Each session was designed to discover and meet wants and needs of the executive officers and volunteers of the association’s regional chapters’ members in an area of special interest. The first session focused on a key fund-raising event used by all of the participants. The second covered the more general topic of chapter fundraising and sponsorship.
Room set has a huge effect on the dynamics of a session. Previous sessions in our room had used head tables with table mikes and straight row theater seating (ugh; well, at least it was set to the long edge of the thin room.) I had the tables removed, the mikes replaced with hand mikes, and the chairs set to curved rows. We included plenty of aisles so that anyone could easily get to the front to speak (see below).
Welcome and a fishbowl sandwich
After a brief welcome and overview, I began a four-chair fishbowl sandwich format, which turns every attendee into a participant right at the start, and ensures that they end participating too. This format is my favorite way to create amazing panels. Fishbowl allows control over who is speaking by having them move to a chair at the front of the room. Check the link for a description of this simple but effective way to bring participation into a “panel”.
Next, I used body voting, to give participants relevant information about who else was in the room. For example, I had everyone line up in order of chapter size, so people could:
discover where they fit in the range of chapters present (from 80 to 2,600 members);
meet participants whose chapters were similar in size to their own; and
give everyone a sense of the distribution of chapter sizes represented.
Additional body votes uncovered information about:
revenue contributions from dues, events, and sponsorship;
promotional modalities used;
member fees; and
other issues related to the session topic.
I also gave participants the opportunity to ask for additional information about their peers in the room. This is another powerful way for participants to discover early on that they can determine what happens during their time together. I used appropriate participatory voting techniques (see also here, and here) to get answers to the multiple requests that were made.
Several weeks before the conference, I scheduled separate 30-minute interviews with the six panelists to educate myself about the issues surrounding the session topics and to discover what they could bring to the sessions that would likely be interesting and useful for their audience. After the interviews were complete, I reviewed our conversations and determined that each panelist could share the core of their contributions in five minutes. So I asked each panelist to prepare a five-minute (maximum!) talk that covered the main points they wanted to make.
During the first session I brought up the panelists to the front of the room individually. As each panelist gave their talk, I allowed questions from the audience, and, as I should have expected, each panelist’s five minutes expanded (by a few minutes) as they responded to the questions. So for the second session, I tried something different. All three panelists sat together with me, and I asked the audience to hold questions until all three had finished. Each panelist gave a five-minute presentation, and then I facilitated the questions that followed.
In my opinion, having only one panelist at the front of the room at a time creates a more dynamic experience. But on balance, I think the second approach worked better as there was some overlap between what the panelists shared, and when questions ended there was a more natural segue to the next segment of the session.
At this point we switched to a fishbowl format. I had the panelists return to front row audience chairs, from where they could easily return to the “speaking” chairs. (They were frequent contributors to the discussions that followed.) I identified some hot issues, and listed them for participants. I then invited anyone to sit in one of the three empty front-of-the-room chairs next to me to share their innovations, solutions, thoughts, questions, and concerns. Anyone wishing to respond or discuss joined our set of chairs and I facilitated the resulting flow of conversation. Some of the themes I suggested were discussed. But a significant portion of the discussion in both sessions concentrated on areas that none of my panelists had predicted.
The capability of fishbowl process to adapt to whatever participants actually want to talk about is one of its most attractive and powerful features. If I had used a conventional panel for either session, much more time would have been spent on topics that were not what the audience most wanted to learn about, and unexpected interests would have been relegated to closing Q&A.
During my opening overview of the format, I explained that we might have time for some consulting on a participant’s problem towards the end of the session. We didn’t have time for this during the first session — given a break, we could have probably taken another hour exploring issues that had been raised — but we had a nice opportunity during the second session to consult on an issue for a relatively new executive officer.
Another option that I offered, which we didn’t end up exploring in either session, was to share lessons learned (aka “don’t do this!”) — a useful way to help peers avoid common mistakes.
With a few minutes remaining, I closed the fishbowl and asked participants to once again form pairs and share their takeaways from the session. The resulting hubbub continued long after the sessions were formally over, and I had to raise my voice to thank everyone for their contributions and declare the sessions complete.
When an audience collectively has significantly more experience and expertise than a few panelists — as was the case for these sessions (and a majority of the sessions I’ve attended during forty years of conferences) — well-facilitated formats like the one I’ve just described are far more valuable to participants than the conventional presentations and panels we’ve all suffered through over the years. Use them to create amazing panels, and your attendees will thank you!
Ten minutes after I’d finished facilitating a large national association meeting hour-long fishbowl sandwich discussion on solutions for a persistent industry problem, the conference education director walked in. His jaw dropped. “The attendees are still here talking to each other! That never happens!” he exclaimed.
Well it happened this time. Many small groups had formed and people were chatting energetically. Business cards were being swapped. When I left to catch my flight home twenty minutes later, conversations were still going on all around the room.
How did I build and support this level of interaction and engagement?
I used a fishbowl sandwich. What’s that? Read on!
The components of a fishbowl sandwich
A fishbowl sandwich, like any good sandwich, has a filling surrounded by bread and spread (or accompaniment). The filling is the fishbowl technique, the surrounding bread is comprised of pair-shares at the start and end, and the accompaniment is the facilitative language that segues between the bread and the filling.
How I began the fishbowl sandwich
As people trickled into the room I asked them to pair up by sitting next to someone, preferably someone they didn’t know. I lightly repeated the request several times before the session started.
For the first piece of sandwich “bread” I asked everyone to think of something they had done, small or large, which was a (probably partial) solution to the challenges the industry faced. After about 30 seconds I asked one of each pair’s members to spend 30 seconds sharing what they had done with their partner. A final 30-second share from the second partner to the first wrapped up the opening pair-share.
As usual, it was hard to get everyone to turn back to the front of the room for the next bite of the sandwich!
At this point, everyone had switched, at least for a while from “listening” to “participation” brain mode—they were ready to engage.
Time for the fishbowl
I was sitting on a low stage with three empty chairs besides me, wearing a headset mike, with a couple of wireless stick mikes at hand, and took a minute to share the rules of fishbowl:
You can only talk if you’re sitting in one of these chairs.
If you have something to say, come and sit in an empty chair. You don’t have to wait for someone else to finish talking.
When you’ve finished what you have to say (for the moment, you can always return) vacate your chair.
If all chairs are full, when someone new walks up, the person who’s been talking longest should leave.
And we were off. For the next fifty minutes a constant stream of people came up and shared their ideas and experiences. Sometimes they shared with the audience; sometimes they spoke with each other while the audience listened. No one “hogged the mike”.
A woman wearing a large backpack shared a novel approach that could be implemented regionally. I ran a hand poll to see how many people had done something similar—only about 20% of the audience. I asked those who hadn’t how many would be willing to do the same. Most hands went up, and people looked thoughtful. An industry leader told the woman he wanted to interview her for the association’s national magazine.
After about 40 minutes I said that we had heard an incredible amount of good ideas and advice and it was clear that there was a tremendous amount of expertise and experience in the room. I asked if anyone wanted help with specific problems. Two brave souls came up and shared their individual frustrations. Sure enough, several folks came up and supplied helpful suggestions.
Finishing the fishbowl sandwich
It was time for the final pair-share slice of bread. To conclude, I asked each pair member to share with their partner their single best takeaway from the session. Once again, a buzz of conversation arose, and after a couple of minutes I announced that the session was over.
Dipesh Mody, writing from Mumbai, India, asks five great questions about event process design. I’ve interspersed my answers after each question.
I have now read both your books and have truly enjoyed reading them. Your work has been very inspiring to many; and I am certainly one of them.
While your book is very well written and structured, I had a few questions for you and I am hoping that you will find the time to respond.
After the peer group session sign-up and once the time and space is allocated, who decides which technique to use? Is it the volunteer facilitator of the peer group? If yes, what if the volunteer is not familiar with these techniques? Will he invariably choose a roundtable technique?
Yes, the volunteer facilitator(s) of a peer session is/are responsible for determining the format used in the session, and, as covered in The Power of Participation, there are a number of basic formats you can use. For many years, I’ve given every attendee a one-page peer session facilitation handout (free download) at the start of the event. This short document explains session facilitation, offers a suggested step-by-step process, and includes some tips for effective facilitation.
Analyzing thousands of evaluations of Conferences That Work format events, it’s very rare to see a complaint about the quality of peer session facilitation. So I believe this simple handout is an effective tool for volunteer facilitators to do a decent-to-good job at facilitating a peer session. While I could include some additional opening techniques such as Post It, described in The Power of Participation, it’s possible that making the handout longer might reduce its overall effectiveness.
In India, and other regions where organizational cultures tend to be more hierarchical than those in North America and Europe, participants may be less comfortable taking responsibility for leading a session. Under such circumstances, taking twenty minutes at the opening of a peer conference to explain basic peer session leadership techniques can be helpful.
From what I understand that certain sessions only a trained facilitator can run them such as world café, fishbowl or a human spectrogram? Is my understanding correct? If yes, then such techniques can only be used involving the entire group. For e,g, if the conference size is 50 people then all 50 people need to be in that one session when a human spectrogram technique is being used? Is my understanding correct?
I think it depends on what “trained” means. I have not received any “formal” facilitation training, but I experienced World Café, fishbowl, and human spectrogram process run by others before I attempted to facilitate them myself. I think many people who have experienced a human spectrogram once could successfully facilitate it under similar circumstances, and there are plenty of good resources (including The Power of Participation😄) for other group work techniques.
As participative techniques become more frequently used at conferences, attendees are increasingly likely to be capable of facilitating them, and I expect the requirement for a “trained” facilitator will decrease over time.
About the beginning and the end sessions, I am quite clear but for the middle sessions is there a particular sequence (s) that works best based on your experience? For e.g. use fishbowl to gain a deeper understanding of top six issues and then follow it up with world café to discuss solutions to these issues (assuming we have 6 tables with five people on each table: Conference size 30 people). Then use a human spectrogram to vote on the proposed solutions and to select the most plausible ones.
Again, the answer to your event process design question depends on the circumstances—in this case a session’s desired outcomes. It sounds like you are asking about process to explore and choose solutions to problems. Because we hold meetings for many different reasons, there’s no single process sequence that’s appropriate for every situation.
The Conferences That Work format, for example, works very well for a group of peers who are meeting to learn and connect for individual reasons, determine common ground, and discover and act on opportunities available to the group.
If, as per your example, the meeting is to learn and discuss six pre-determined important issues, you might well use techniques like fishbowl and World Café as opening and mid-course process. If attendees don’t know each other well, an opening roundtable would be useful. Or if the important issues were unknown or unclear at the start of a meeting, introductory educational sessions plus affinity grouping might be appropriate.
As far as discussing solutions is concerned, while human spectrograms are a useful tool to gauge sentiment, outcomes are more typically determined by process prescribed by the norms of the group, organization, association, or corporation stakeholders.
About world café or human spectrogram or voting, while a volunteer team can assist in framing the right questions as pre-work but my experience shows that getting them to contribute on the questions is difficult as they don’t have time to devote on such pre-work activities due to work related and other commitments. Further, on page 222 of Power of Participation, you have identified questions for collective attention, for finding deeper insights, for forward movement etc. In light of this, would it be a good idea for the attendees to frame the questions during the conference beginning? In your experience would this work?
In my experience, if you are going to use World Café at an event, pre-work defining good table questions is essential. While there are frameworks that can be helpful in devising Café question rounds (e.g. those for sense-making by Chris Corrigan and strategic planning by John Inman), I think it’s very hard to build consensually-good questions on the fly at the event unless participants are patient and willing enough to spend a significant amount of time. It’s akin to bringing a large group of people to a building site and asking them to collectively design and erect a building from scratch. Not impossible, but difficult!
While your book does provide model conference schedule but it falls a bit short of getting a real sense of what a real schedule looks like. It would be really great if you could add a few real examples of conferences you facilitated. It would indeed be useful to get a sense of how you mixed and matched various techniques (fishbowl, world café, spectrograms etc.) during a lets say three day conference around a particular theme. It would be a great addition to what a truly amazing book it already is.
Dipesh, I think that’s a good idea in principle. However, I’m wary supplying such examples unless they include extensive background on why the specific types and flow of process techniques were used in the event process design. The danger of providing condensed examples is that some readers will be tempted to copy them verbatim for events that involve participants, logistical constraints, and desired outcomes that are significantly different from those that generated the example design. End result—a design that doesn’t satisfy stakeholder needs, leading to poor evaluations and, perhaps, the conclusion that these new-fangled event designs “don’t work.”
There are so many factors involved in creating a good event design that I estimate a useful case study of a single event design that comprehensively covers the reasons for the design choices made might require 10,000+ words and many days of work! A worthy project, but one that may have to wait a while…
Dipesh Mody, India
Thanks for your thoughtful questions about event process design, Dipesh. I hope these answers help a little in your quest to produce fine events in India!
Another issue of an occasional series—Dear Adrian—in which I answer questions sent to me about event design, elementary particle physics, solar hot water systems, and anything else I might conceivably know something about. If you have a question you’d like me to answer, please write to me (don’t worry, I won’t publish anything without your permission).
In a recent post I described using a standard fishbowl to focus group discussion. There’s a less common fishbowl variant, which I call the two sides fishbowl, that can be used to explore differing or opposing viewpoints in a group. A two sides fishbowl allows representatives of a point of view to listen to and question representatives of an opposing viewpoint for a period of time, after which the two sides switch roles.
You can use a two sides fishbowl to explore introverts’ experience of extraverts and vice versa, to examine two alternative proposals for solving a business problem, or to go deeper into divergent views on a social issue, etc.
A two sides fishbowl uses a chair layout of two concentric circles as shown below. Since, in general, the number of people on each side will not be known in advance, this layout will need to be set up on the fly once the sizes of the two groups are known. If the groups are not approximately the same size, participants will need to reposition chairs appropriately when the two sides swap places.
Use either a single facilitator for both sides, or chose a facilitator from each group to lead the inner circle discussion.
Running a two sides fishbowl Once the groups for the two sides fishbowl have been established (a one-dimensional human spectrogram can be used for this), decide which group will go first and have them sit in the inner circle of chairs. The other group sits in the outer circle.
The rules for a two sides fishbowl are simple. The inner circle does most of the talking. Inner circle members, guided by a facilitator, discuss, explain, or champion their viewpoint for the benefit of the outer circle group. Outer circle members are not allowed to respond to what they hear, except to ask questions that clarify the inner circle discussion.
After a useful discussion has been held, the groups change places and an approximately equal amount of time is given to the new inner group to repeat the above process.
Photo of Jan Steen’s “Argument Over a Card Game”: Flickr user johnmcnab
Most of us have had to suffer through a “discussion” occurring in the presence of a large number of people, most of whom never get an opportunity to speak. Here are three tips for creating an effective focused discussion with a group.
Use a fishbowl A fishbowl provides a simple, ingenious process for focused discussion.
The advantage of a focused discussion over informal discussion is that it greatly reduces the cross-conversations that frequently occur when many people want to respond or comment on something that’s been said. And it manages this feat without limiting discussion to a few voluble people, as it provides all attendees an equal opportunity to contribute.
The term “fishbowl” can refer to a couple of different techniques for focused group discussion. In this post I’ll describe the standard fishbowl design, which assures that the conversation at any moment is restricted to a few clearly defined people while still allowing others to join the discussion in a controlled manner.
A standard fishbowl requires a chair for each participant, with chairs set in a horseshoe or circle, as shown in the diagrams below. See the second tip to decide which layout to use.
The number of chairs in the mouth of the horseshoe or the center of the circle is typically four or five. The fishbowl facilitator sits in one of these chairs for the duration of the fishbowl.
At the start of the fishbowl, the facilitator sits alone in the small group of chairs. They explain how the fishbowl works by saying something like this:
“We’re about to start a focused discussion using a fishbowl. If you want to talk, you must come and sit in one of these chairs next to me. If all these chairs are full and no one has yet spoken, wait a little. Otherwise, when you come up, someone sitting here must go back to a chair in the [horseshoe/outer circle]. Also, if you’re sitting up here and have finished what you have to say, go back to a [horseshoe/outer circle] chair. When you’re up here, you can talk to someone else in these chairs or the whole group—the choice is yours.
[Pause for questions.]
The discussion is now open. Who would like to start?”
You’ll probably find that some attendees will want to talk from their chair in the horseshoe or outer circle. When this happens, gently interrupt and gesture for them to come up and sit next to you. If they’ve interrupted someone in the conversation chairs, steer the conversation back to the folks up front.
Once people get the hang of the fishbowl, everyone’s likely to be surprised by how well it works. Those who tend to monopolize unstructured discussions invariably become aware how much they’re doing so in a fishbowl. If they run on it’s easier to gently ask them to leave the speaking chairs for a while. Participants appreciate how:
the format focuses the discussion;
contributors change as needed;
the front row or inner circle shows who may talk; and
it’s clear when the conversation on a topic has run its course.
Use the best layout for the fishbowl I prefer the circle version for square rooms, and the horseshoe layout for rooms that are significantly longer than wide. If both versions can be accommodated, I like the circle version for general discussions, and the horseshoe version when decisions may be made or if you are scribing discussion points onto flip charts, which can then be placed next to the small row of chairs for all in the horseshoe to see. If you have less than twenty people in the fishbowl, use the horseshoe layout to avoid some participants having to stare at the facilitator’s back the whole time.
Have a set of topics to review Don’t run a fishbowl at the end of a session or conference without first creating a preliminary list of topics. This can prove frustrating when people with different topics in mind occupy the speaking chairs, leading to a conversation that jumps around from topic to topic as each person speaks. To avoid this, I’ve found that it’s best to precede a fishbowl with techniques like plus/delta or affinity grouping (see Conferences That Work: Creating Events That People Love). Both techniques create a list of opening topics for the fishbowl to address. The facilitator then uses the list as a roadmap for discussion.
What tips do you have for improving the quality and effectiveness of group discussion? Share them below!
What approach should we use to teach participation techniques for meeting sessions?
With the rise of social learning and the decline in importance of formal learning, perhaps we should use experiential learning. On the other hand, in the same time needed to experience a limited set of participation techniques we can comprehensively describe many more. There again, perhaps experiencing a participation technique directly is a more effective way to cement both learning it and truly understanding its relevance. So, if we are teaching participation techniques, which of these two approaches is a better path for learning?
J’s light-bulb moment Earlier this week I led a workshop at Meeting Professionals International’s World Education Congress (WEC). The 150-minute session covered a variety of techniques that foster and support meaningful participation during meetings. Participants spent most of their time using these techniques to learn about and connect with each other and explore questions about their experience at WEC and in the session itself.
As the workshop progressed, and I heard from the forty-six participants, it became clear that one of them, whom I’ll call J, had considerable prior experience with the techniques I was facilitating.
Near the end of the workshop I ran Plus/Delta (described in Conferences That Work: Creating Events That People Love): a method that provides a fast, public evaluation of a session or entire meeting. As an advocate for transparency and feedback, I chose the subject of our Plus/Delta to be a group evaluation of the workshop itself. During the evaluation, J commented that he had hoped that I would cover more techniques by talking about them rather than having attendees experience them directly. He then contributed a simple and ingenious way to extend Plus/Delta that was new to me.
My heart sank, just a little. Here was J, an experienced facilitator of participation techniques, proposing that I should spend the workshop talking about techniques rather than facilitating experiences of them. Could I be going about this wrong?
I moved immediately into the last technique of the workshop, running fishbowl: a simple way to facilitate focused discussion with a large group. All participants sit in a large circle of chairs, but only people in the “fishbowl”, a small circle of chairs at the center, can speak. After a few minutes of comments, J entered the fishbowl.
J said that he had read about fishbowls many times before and he understood how they worked, but he had never tried one.
And then, to my surprise and delight, he told us that experiencing the fishbowl had been a revelation to him, because he had directly experienced the power of the technique in a way that significantly enhanced his understanding of it, which he had previously believed to be sufficient. It was poignant for me to hear J express a new point of view that contradicted what he had said only a few minutes earlier, and I admired his courage in sharing his learning with us all.
I too have struggled over the years to define the best balance between understanding techniques through description and understanding them through direct experience. J’s light-bulb moment fits for me; these days I am content to let attendees learn participation techniques, first through direct experience and then, if necessary, via reflection and discussion.
Postscript At the end of the workshop, J hung around and we talked while I was packing up my equipment for a flight home.
He told me that his fishbowl sharing had unexpectedly reminded him of a session he had once attended, entitled “One hundred icebreakers in one hundred minutes”, consisting of rapid descriptions of a hundred ways to introduce attendees to each other.