Colay / Guides
Choose a facilitation format by examining the blockage from several angles
A team says the problem is communication, but the actual blockage might be unclear decision rights, an overloaded handoff or incompatible expectations. Before scheduling another open discussion, use Colay to compare possible explanations and the evidence each would require. The result is a facilitation brief with a purpose, a suitable format and a testable next step—not a model’s verdict about who is at fault.
Describe observable events without assigning motives
Write a short situation note: what happened, when it happened, which roles were involved, what consequence was observed and what is still unknown. Replace “the team resists ownership” with something you can check, such as “two decisions were returned without a named approver.” Remove names and unnecessary personal details, and use only material you are permitted to share.
Keep reports from different people separate. A manager’s summary is one perspective, not the whole situation. Ask the models to propose questions that would distinguish explanations rather than infer personality, intent or emotional state from a few messages. This makes the eventual conversation easier to correct when participants add missing context.
Use eight lenses to find different questions
Try these lenses: shared goal, decision rights, handoffs, workload, incentives, information quality, expectations and meeting participation. Eight is a way to structure the request, not a claim that Colay supplies eight independent facilitators. Select available agents, compare separate answers, and keep useful differences rather than merging everything into a diagnosis.
A repeated explanation is still a hypothesis. If every answer says “unclear ownership,” ask what observation would show that ownership is actually clear and another bottleneck is responsible. The strongest preparation often comes from discovering which missing fact would change your choice of meeting.
Help prepare facilitation, not judge people. Situation note with observable events: [note]. Review it through 8 lenses: goal, decision rights, handoffs, workload, incentives, information quality, expectations and participation. For each give one tentative explanation, one missing fact, one neutral question and one possible facilitation format. Label assumptions. Do not infer motives or diagnose personalities. Finish with the smallest information-gathering step needed before selecting a meeting format.
Match the meeting to an output
In a fictional example, product and design both believe the other owns final acceptance, and a release waits while comments circulate. This does not establish a relationship conflict. The immediate deliverable might be a decision map rather than a general discussion about trust.
| Question to resolve | Possible format | Concrete output |
|---|---|---|
| Who can make the acceptance decision? | Map roles around one recent decision | Named decision owner and consultation path |
| Where does work wait? | Reconstruct one handoff timeline | One changed handoff with an owner |
| Whose information is missing? | Individual reflection, pairs, then group sharing | Distinct observations to investigate |
| Which trade-off should the team accept? | Compare explicit options against agreed criteria | Recorded choice and review trigger |
Let people supply the perspectives that AI cannot
Liberating Structures describes 1-2-4-All as a sequence from individual reflection to pairs, small groups and collective discussion. It offers a concrete format for gathering participant contributions. Use the original instructions when that purpose fits; do not assume a format solves every conflict.
Bring the hypotheses into the room as questions. Tell participants what the session can decide and what is outside its scope. Where there is a concern requiring an established people or conduct process, use that process instead of turning a group workshop into an improvised investigation.
Close with one observable experiment
For the fictional release blockage, the experiment could be to name the acceptance owner before the next handoff and inspect whether the same return loop occurs. Colay can help compare facilitation plans and summarize permitted notes, but the team supplies its experience and makes the decision. That division keeps the tool useful without mistaking a fluent analysis for knowledge of the people involved.
- Write the agreed problem in language participants recognize.
- Choose one change, a responsible owner and a review date.
- Specify what observable result would support keeping or changing the experiment.
- Record remaining disagreements instead of calling silence consensus.
Questions, answered
Should I ask the models who is causing the conflict?
Ask for competing explanations and questions instead. A short account cannot establish motives, and assigning blame prematurely can distort the facilitation brief.
Do I need eight models?
No. Eight lenses are an optional structure. Start with available agents and a few contrasting questions; add more only when they reveal useful differences.
What is the final artifact?
A brief containing observed events, unresolved questions, the chosen session purpose and format, and the output participants are expected to produce.
Sources and methodology
- Liberating Structures: 1-2-4-All
The method’s own description of its participation sequence. The example format-selection matrix is an original planning aid, not evidence of a guaranteed conflict outcome.
Bring your next question to Colay
Choose a model, use Auto, or bring several perspectives together with Consensus.