Wall and Board Capture

When To Use

The session just ended and everything it produced is on a wall that will be wiped tonight. Use this in a break, at lunch or right after the close, while you can still walk over and check a note. It answers one question: what exactly did the room write, vote and decide, before anyone starts tidying it?

When Not To Use

Not for themes, clusters or reading the result against the brief: that is Session Summary and Decisions, and the capture never interprets. Not while the room is working: nothing here listens to or follows a live session. And no transcript is used unless the Session Agreements and Consent page says participants agreed to recording.

Inputs

• Whatever exists, in any state: upload a photo of the wall or flipchart, paste an export from your online board, typed notes, a voice note written up.
• The blocks of the session and which exercise produced which wall, so each item gets a home.
• What participants agreed to about recording, and who the capture will be sent to.

If you have none of this, I start from a single photo and your one-line description of the exercise, and mark the output as a first draft.

Approach

Capture first, synthesis later: the working standard among facilitators. Transcribe faithfully, label by block, flag and do not interpret. It works from a photo uploaded to the chat, and no reading of handwriting is perfect, so anything doubtful is logged for you to check against the wall. The failure this prevents: two similar notes quietly merged and reworded on the train home, and a sponsor later quoting a sentence nobody in the room wrote.

Workflow

1. Ask up to three questions: which block and exercise each photo or export belongs to, what participants agreed to about recording, and who will receive the capture.
2. Take everything as it comes: bad angles, two languages and half a flipchart are fine. Transcribe in the room's own words: no improved phrasing, no merging of similar notes, no fixing an idea that looks wrong. The wall is a record, not a draft.
3. Log what cannot be read as unreadable, with its location ("top right cluster, orange note"), so you can walk over and check while the wall still exists.
4. Label every item by block, exercise and group. Record votes as counts, exactly as they stand. Keep attribution as the room made it: anonymous notes stay anonymous, with no guessing authors from handwriting or position.
5. Flag ambiguities as questions for you, never as decisions: two clusters that may be one, a note that could sit under two headings. Capture the close in the exact words used: decisions, owners, next steps.
6. List what needs cleaning before anything is sent: names visible in photos, confidential figures on a board, anything the agreements said stays in the room.

Output Format

# Session Capture File
Session: [title] | Date: [date] | Captured: [break or after the close] | Sources: [photos, exports, notes] | Recording agreed by participants: [yes or no; if no, no transcript is used]
## Block [number]: [block name], [exercise]
| Group | Item, in the room's words | Location on the wall | Votes (count) |
|---|---|---|---|
| [group] | [text as written] | [location] | [count or none] |
## Unreadable log and open questions (check while the wall exists)
| Kind | Block | Location | What can be made out, or what is ambiguous | Trainer's answer |
|---|---|---|---|---|
| Unreadable | [block] | [top right cluster, orange note] | [partial text] | [checked against the wall: text] |
| Ambiguous | [block] | [location] | [question, with the options] | [answer] |
## The close, in the exact words used
| Decision or next step | Owner as stated | By when, as stated |
|---|---|---|
| [words used] | [owner] | [date or none stated] |
## Before it travels
1. [Photo or board]: [what needs cleaning: names, confidential figures] | Done: [yes or no]
## Decision
[Trainer name] checks the unreadable log against the wall before [time the room is cleared], answers the open questions, and decides by [date] what is cleaned and who receives the file.

Done When

• Every photo, export and note is transcribed under a block, an exercise and a group.
• Vote counts match the wall, and nothing has been merged or reworded.
• Every unreadable item has a location, and ambiguities sit in the same table as questions, unanswered by Claude.
• The travel list names what must be cleaned before the file is sent.

Quality Bar

• The room's words, not better ones, and no themes, clusters or conclusions: improvement belongs to the summary, where it is visible.
• Unreadable means logged, never guessed.
• Privacy and data points about photos or transcripts are flagged, not ruled on: check with a qualified adviser.
• It captures what the room produced, never who spoke or how anyone behaved, and a transcript is used only where participants agreed to recording.

Next

Run fac-session-summary (Session Summary and Decisions) to write up what it adds up to, within two days, while you can still correct it from memory.

About the makers

This pack is made by 90n (90n.ai): AI-native ways of working and tools for Product, Marketing and HR teams, built by ex-McKinsey founders. We map where time and money leak, redesign the work around AI and stay until your people run it themselves, and we run our own company on Claude. If your team has outgrown the self-serve version, message Pauline (linkedin.com/in/paulinebertry).