Why Your Workshop Produced a List Nobody Acted On

The structure of a session determines what it can produce. Most default formats are not designed to help.

AI generated image of abstract multicoloured waveforms.

Signal Boost: "The Sound of Silence" by Simon & Garfunkel
People talking without speaking. People hearing without listening. It is a poetic description of every conventional meeting, with the "neon God they made" mapping uncomfortably to the presentation screen!


There is a familiar pattern in workshops.

The room fills. Someone presents. A few questions are invited. Two or three people contribute. The session closes with a list of actions that look reasonable on the day and quietly disappear by the following week.

Nobody is being difficult. The people in the room are often thoughtful and capable. The problem is the format.

Most meetings and workshops default to one of five structures: presentations, managed discussions, open discussions, status reports, or brainstorming. These are so established that we rarely examine what they actually produce. A few voices carry the session. The rest wait or disengage. The conclusions reached are usually close to the ones assumed before anyone arrived.

This is not a failure of intent. It is a structural outcome.

The structure selects who participates.

That is the insight behind Liberating Structures, a set of 33 facilitation methods developed by Henri Lipmanowicz and Keith McCandless over more than two decades. Each one is defined by precise mechanics: a specific invitation, a defined group configuration, a time boundary, and a participation model built to distribute contribution rather than concentrate it.

Conventional formats produce predictable outputs because they cannot produce anything else. If you want different thinking to emerge, you have to design the conditions for it.


The idea of a string

Individual structures are useful. The concept that changes how you design a session is the string: a deliberate sequence of structures, each building on what the previous step produced.

A practical example. You are running a change session with a leadership team. Open with Impromptu Networking. In twenty minutes of structured paired conversation, you find out what is genuinely in the room before any framing takes hold. The concerns that were present but not yet in the conversation.

Move to Wicked Questions to name the paradoxes explicitly. Most change initiatives fail because the inherent tensions are never surfaced. Wicked Questions forces the group to state them rather than work around them.

Close with 15% Solutions. No collective action plans or commitments that need sign-off. Just what each person will do in the next two weeks within their existing authority. Small, owned, immediate.

Each step produces something the next step requires. That is the difference between a session that generates energy and one that produces something still useful by Monday morning.


How I use this with Claude

Earlier this year I started working with a structured reference document covering all 33 Liberating Structures, formatted specifically as a knowledge file for Claude. The original concept comes from Ruben Hassid, published on his Substack, How to AI. I have adapted and extended it for my own facilitation and advisory work.

Load the file into a Claude project. Describe your situation: the context, group size, time available, and what you need to achieve. What comes back is a sequenced session design. The structures are named, the steps explained, and the rationale for the sequence is set out clearly. That last part matters most. A room never moves exactly as planned. Understanding why a string is ordered the way it is tells you what you can change and what you should not.


The document

The reference file covers all 33 structures with timing, group configurations, and virtual guidance. It includes seven pre-built strings for common scenarios including change sessions, strategic planning, new service design, and large group sense-making.

It is available to download. Load it into Claude as a project knowledge file and describe your challenge.

If you already work with Liberating Structures and have a view on what works in practice, I would be interested to hear it.


What It All Comes Down To

Most facilitation problems are design problems. The room has the thinking in it. The question is whether the session is built to find it. A well-designed string gives the group a genuine chance at that. Most default formats do not.


Credit: The original prompt engineering reference work was developed by Ruben Hassid and published on his Substack, How to AI. Well worth following.