How is a decision tree different from a flowchart?
A flowchart can include steps that always happen. A decision tree is mostly questions. If you have a long chain of rectangles and one diamond, you have a flowchart. If every split is a yes or no that changes the next action, you have a tree.
The decision-tree template is a support example: billing or not, then charged twice or not. Open it and replace the questions. Do not add a third arrow to a diamond. A third answer is a different question or a list, not a diamond.
How deep should it be?
Two questions is a tree a person can hold in their head. Five questions on one page usually means two procedures. Split them. The reader should be able to point at their answer without tracing every other branch.
If both answers lead to the same next step, it was not a decision. Delete the diamond and keep the step.
Ready to draw?
No signup. Opens instantly in your browser.
What do you write on the leaves?
An action, with a verb. Refund the duplicate. Explain the invoice. Route to product. Not Outcome A. The person following the tree at a desk does not know what Outcome A means.
If the leaf is really another team's process, name the handoff and stop. Their process is their chart.
What are the steps?
- Write the first question. It should split the work. Billing issue? is a question. Handle the ticket is not.
- Draw both answers. Each answer leaves the diamond and lands on a different shape. Yes and No are enough when they are opposites.
- Add the next question only on the path that needs it. Charged twice? belongs on the billing path, not on every ticket.
- End on an action. The leaves are rectangles: refund the duplicate, explain the invoice, route to product. A tree that ends on another question is unfinished.