Decision trees turn complex choices into clear Yes/No paths. Learn structure, labeling rules, examples for support and product, and how to build them free online.
A decision tree flowchart is a flowchart optimized for branching choices. Instead of documenting a mostly linear procedure, you document questions that route someone to an outcome: approve/deny, troubleshoot, classify, or recommend.
Decision tree vs regular process flowchart
Process flowchartDecision tree
Primary shape mixMany process rectanglesMany diamonds
Question“What happens next?”“Which answer path?”
EndingsOften one success endMany terminal outcomes
AudienceOperators executing workPeople classifying or diagnosing
Most real diagrams are hybrids — and that is fine. Just know which mode you are in.
Rules for readable decision trees
Phrase every diamond as a closed question — “Is amount > $500?” not “Amount check”.Label every outbound arrow — Yes/No or concrete categories.Prefer binary splits when teaching beginners; multi-way is OK if labels stay short.Order questions by information gain — ask the question that eliminates the most paths first (severity before cosmetic checks).Terminal nodes state outcomes — “Refund full”, “Escalate fraud”, “Close — cannot repro”.Avoid cycles unless you intentionally model “ask again” loops (and label them).Example: pricing plan recommender (simplified)
StartDiamond: “Team size > 10?”Yes → Diamond: “Need SSO?”Yes → End: EnterpriseNo → End: BusinessNo → Diamond: “Need cloud sync?”Yes → End: ProNo → End: FreeExample: basic IT troubleshooting
Start: “App won’t load”Diamond: “Internet works in browser?”No → End: Fix network firstYes → Diamond: “Error code shown?”Yes → Process: Lookup code → …No → Diamond: “Worked yesterday?”No → Check recent installsYes → Clear cache → Retry → …How to build one in FreeFlowCharts
Open /editor.Place the root question near the top.Expand Yes/No branches evenly to reduce crossings.Use consistent horizontal spacing for peer outcomes.Color terminal outcomes: green success, red hard stop, amber needs human.Export PNG for help center; export Mermaid for developer docs.Testing your tree
Hand the chart to someone outside the team with three scenarios. If they reach the wrong ending, either a question is ambiguous or an arrow is mislabeled. Decision trees fail in language more often than in structure.
When not to use a decision tree
Pure hierarchies → org chartOpen brainstorming → mind mapProportional mix → pie chartSet overlap → Venn diagramRelated resources
Flowchart symbols guideHow to make a flowchartDiagram type comparisonPractice exercise
Pick a policy your company explains in a long wiki paragraph (refund eligibility, data retention requests, discount approval). Convert it into a decision tree with at most twelve diamonds. If you need more, split into a subprocess chart linked from a single “See refund policy tree” node.