Decision tree maker
A decision tree is a flowchart that has stopped pretending there is one path. Ask the question, label every answer, and follow each one to where it actually ends.
No account. Saved in this browser until you export it.
The pieces
What goes on the canvas
- Question
- A branch point with labelled answers
- Choice
- One available option
- Outcome
- Where a path ends up
- Action
- Something to do as a result
Building one
From blank canvas to finished diagram
- 1Start with the question that comes first. Phrase it so the answers are obvious.
- 2Add a branch per answer and label the connector — an unlabelled branch tells the reader nothing.
- 3Follow each branch to an outcome or an action, not to another unanswered question.
- 4The suggestions panel flags any question left with fewer than two labelled answers.
Examples
Start from a real one
Where it helps
Good uses
Support triage
Which questions the agent asks, and what each answer means.
Eligibility rules
Policy written as prose hides its edge cases; a tree exposes them.
Troubleshooting guides
Readers follow one path rather than reading all of them.
Escalation policy
Makes visible the point where a decision stops being yours.
Questions
Asked often
- How is this different from a flowchart?
- A flowchart usually describes a process that loops and rejoins. A decision tree branches outward and does not come back.
- Must every branch be labelled?
- Nothing is enforced, but unlabelled branches are flagged. A reader cannot follow a fork with no signpost.
- Can branches rejoin?
- You can draw that, but if paths rejoin often, a flowchart is probably the better shape.
- Is the tree stored anywhere online?
- No. It stays in this browser until you export it.
Map the decision
Nothing to sign up for. The diagram stays on this device.