- Home
- Research skills
- Research Debrief and Workshop Design
Research Debrief and Workshop Design
Design the session where stakeholders actually engage with the evidence and leave having made decisions, with owners, dates and disagreements on the record.
Free. Works on Claude, ChatGPT, Gemini or any assistant that accepts a skill file.
What this skill does
The method, encoded.
The debrief is where most research projects are really decided, and it is usually designed as a presentation. Slides are narrated, a few questions are asked at the end, everyone says it was useful, and nothing is decided.
This skill designs the session backwards from the decision it must produce. It starts by choosing honestly between a presentation, a working session and neither, then works back through what must be true for the decision to be made, what evidence the group must encounter, and which exercises make them encounter it rather than receive it.
The core structural rule is that evidence must be understood before implications are debated. A room that jumps to solutions has skipped the finding, so the design includes an explicit understanding gate, a parking mechanism for the ideas that will arrive early, and exercises that cannot be completed without looking at the evidence.
It also covers the parts nobody plans for: the senior person who disagrees, and how to tell a legitimate challenge to the evidence from an unwelcome conclusion being resisted; the room that wants to redesign the product; capturing decisions live; hybrid sessions; and the tension of facilitating a session about your own findings.
Best used for
- Turning a findings readout into a session that produces decisions
- Getting several functions to a shared reading of one evidence base
- Handling a senior stakeholder who disagrees with a central finding
- Designing exercises that make stakeholders engage with evidence, not a summary
- Running working sessions across virtual and hybrid settings
- Capturing decisions and owners live rather than in a later write-up
- Regular research readouts that have become a ritual nobody acts on
Typical inputs
What you give it.
The decision or decisions the session must produce, stated as choices, Final quality-checked findings with references, bases and confidence levels, Attendee list with roles and decision authority, Time available and the physical or virtual setting, Optional record of what each attendee currently believes, Optional known positions, commitments and sensitivities, Optional raw evidence in usable form (verbatim cards, clips, data tables), Optional previous session records and their follow-through, Optional independent facilitator
Typical outputs
What you get back.
Session design document with purpose stated as a decision, Run sheet by segment, mode, precondition served, materials and facilitator, Evidence pack with base, population and confidence on every item, Exercise specifications with purpose, materials, grouping and output, Risk plan covering anticipated challenges and private briefings held, Decision record completed in the session, with owners, dates and revisit triggers, Session record issued within one working day in the organisation's decision format, Register of disagreements left open, with the holder named
Method coverage
What the skill works through.
- Presentation, working session, or neither
- Designing backwards from the decision
- Who must be in the room, and who should not be
- Pre-reading and its realistic uptake
- Sequencing evidence before implications
- The understanding gate
- Exercises that force engagement with evidence
- The senior stakeholder who disagrees
- Challenge to the evidence versus challenge to the conclusion
- The room that wants to redesign the product
- Capturing decisions and owners live
- Facilitating a session about your own findings
- Virtual and hybrid session design
- Follow-up that turns a session into action
Download
Free skill. One file.
Enter your email once. Every skill you download after that takes a single click.
How to install
Add the skill file and the five kernel protocols to a Claude Project, a ChatGPT Project, a Gemini Gem, or paste them at the top of any assistant conversation. Then give it your real research material, not a description of it.
Download skillQuestions
Common questions.
What is the difference between a research presentation and a working session?
A presentation transmits findings to an audience that does not need to act together, and is right when the evidence is uncontested and one person will decide later. A working session produces a shared reading and a decision from a group, and is right when people must act separately on a common understanding, when trade-offs need the group present, or when the barrier is belief rather than knowledge. A useful test: if the session did not happen, would anything be different tomorrow?
How do I stop stakeholders jumping straight to solutions?
Sequence the session in phases and put an explicit gate between evidence and implications: before any solution is discussed, ask participants to state in their own words what the evidence shows and where it is weak. Then handle the solving instinct structurally rather than by suppressing it. A visible parking board that captures ideas with the proposer's name, plus a segment at the end that genuinely returns to them, works. Suppressing early ideas just makes the proposer stop participating.
How do I handle a senior stakeholder who disagrees with the findings?
First distinguish two things. A challenge to the evidence names a specific sample, base, wording, period or method concern; it is legitimate, often right, and should be answered directly and conceded where it holds. A challenge to the conclusion because it is unwelcome looks the same but does not resolve when the evidence question is answered, and reappears in a different form. That one is not won in the room: acknowledge the position, state what the evidence does and does not establish, and move to what would change their mind. Above all, brief them privately before the session rather than surfacing it in public.
What makes a good workshop exercise for research findings?
The test is whether a participant could complete it without looking at the evidence. If they could, it is a discussion prompt. Exercises that work include asking the group to predict a result before revealing it, having them sort verbatim or finding cards and compare their grouping to the analysis, immersion in raw clips or transcripts before any aggregate is shown, and asking small groups to build the strongest case that a finding is wrong.
Can the researcher facilitate a session about their own findings?
Usually yes, provided the dual role is declared rather than hidden. Say at the start that you are neutral on what the organisation decides and not neutral on what the evidence says, and signal the transition when you switch between answering as the analyst and chairing as the facilitator. For contested or high-stakes sessions, an independent facilitator with the researcher present as evidence witness is the stronger configuration.
How much pre-reading will actually get read?
Assume a minority. Design so the session works if nobody read it, and so those who did gain something rather than sit through a recap. A short question set ("which of these findings would most change your team's plan, and which do you find hardest to believe") produces far more useful preparation than a summary, because it asks for a position rather than for recognition.
How do I make sure a workshop leads to action?
Capture decisions live, on a visible surface, in a fixed format that includes the owner, the date, the evidence it rests on and what would cause it to be revisited, and read each one back to the room. Protect the last ten minutes for this. Then issue the record within one working day, in the format the organisation's own decision machinery uses, and route each decision to its owner individually.
How do you run a hybrid findings session without losing the remote participants?
Hybrid fails in one predictable way: the people in the room have a conversation the remote participants cannot join. Two fixes work, and both cost something. Either make everyone remote, including those in the building, so the medium is equal, or give every remote participant an in-room partner responsible for bringing them in. Choose deliberately rather than putting a laptop at the end of the table.
The skill chain
Works well with.
Research where people already are.
Analyse it where you already work.
Yazi helps researchers conduct surveys, AI interviews and longitudinal research directly through WhatsApp.
%202.png)
