Module 6 — Knowledge and Thinking Tools · Lesson 6.2
Whiteboards and Mind Maps
Two canvases, two different jobs, and when to convert what is on them into work
~8 min
What you'll learn
- Choose between a whiteboard and a mind map for a given problem
- Run a collaborative session on either surface
- Recognize when a canvas has stopped generating and started being maintained
- Convert the output into tasks and documents
Some thinking does not fit in sentences. A dependency graph, a customer journey, the shape of a system, the branches of an idea you have not pruned yet — all of these are faster to draw than to write, and much faster to discuss when drawn. Kavanah has two canvases for this and they are not interchangeable.
Whiteboards: unstructured space
A whiteboard is a freeform canvas: draw, add shapes and text, collaborate.
Use it when the structure is the thing you are trying to find. Architecture sketches, customer journeys, process maps, anything where the relationship between the pieces is spatial and you do not yet know what it is.
The strength is that it imposes nothing. The corresponding weakness is that it imposes nothing — a whiteboard six months old is usually unreadable to anyone who was not in the room, because the meaning lived in the conversation rather than the artefact.
So the discipline is to close a whiteboard session by writing down what it concluded, somewhere that is not the whiteboard.
Mind maps: structured branching
A mind map structures ideas into node trees: a centre, branches, sub-branches.
Use it when the structure is hierarchical and you are trying to be exhaustive. Breaking down a large piece of work, exploring the options under a decision, mapping everything that touches a system.
The tree constraint is the feature. It forces every idea to attach somewhere, which surfaces the ones that do not fit — and an idea that does not fit anywhere on your tree is usually either the most important thing in the session or evidence the centre is wrong.
Mind maps age better than whiteboards because the structure carries meaning on its own. A tree is readable without the conversation that produced it.
Knowing when to stop
Both surfaces have the same failure mode: they become artefacts that get maintained rather than tools that generate.
The signal is that a session ends with 'we should tidy this up' rather than 'so we should do X'. A canvas that is being tidied is a canvas that has finished its job and is now consuming attention.
The honest end of a canvas session is a short list of things to do and, if it produced a durable insight, a document. Both of those live better elsewhere: tasks on the board, the insight in Knowledge.
That does not mean deleting the canvas. It means not treating it as the record.
Converting output into work
After a session, do the conversion immediately rather than later. Later means someone re-reads a diagram in a fortnight and reconstructs what it meant, badly.
Tasks go on the board with an owner. Decisions go into a decision record. The diagram itself, if it is worth keeping, gets linked from the relevant document so someone finding the document finds the picture.
This is also the point at which the AI agent becomes useful: describe what the session concluded and ask it to draft the tasks. It will produce something close, you edit, and the twenty minutes of transcription that usually kills the follow-through does not happen.
Use the canvases
- 1
Sketch a system on a whiteboard
Use it when you are trying to FIND the structure. Close the session by writing down what it concluded.
- 2
Break down a large piece of work as a mind map
The tree constraint forces every idea to attach somewhere, and the ones that do not fit are the interesting ones.
- 3
Convert the output into tasks the same day
Or ask the agent to draft them from your summary and edit what it produces.
- 4
Link the diagram from the document that references it
A canvas nobody can find is a canvas nobody reads, however good it was.
What to watch
- Session-to-task conversion
- Share of canvas sessions that produced tasks within a day.
- Healthy signal: Most of them. A session that generated nothing actionable was a conversation, which is fine — but two in a row is a pattern.
- Canvas re-read rate
- Whether anyone opens a canvas again after the session that made it.
- Healthy signal: Low for whiteboards, higher for mind maps. That asymmetry is expected and is why the conclusions go elsewhere.
Key takeaways
- ·Whiteboards are for finding structure; mind maps are for exhausting a structure you already have.
- ·The tree constraint is the mind map's feature — ideas that do not fit are the interesting ones.
- ·A canvas being tidied has finished its job; the record belongs elsewhere.
- ·Convert to tasks the same day, or someone reconstructs the meaning badly in a fortnight.
- ·Ask the agent to draft the tasks from your summary rather than transcribing them yourself.
Next: the single most consequential piece of writing in the product — the workspace KVN charter.