A Decision Tree turns a branching support process into something your reader can click through. You ask a question, they pick an answer, and the tree shows them only the next thing that applies. It is the right block when the answer to "why isn't this working?" genuinely depends on what the reader tells you.
See a Decision Tree in Action 🌳
Here is a decision tree doing its job. Click through it and watch the chrome along the top: a Back button appears once you're a step in, and Start over returns you to the beginning.
This is an interactive decision guide. Please view it online.
Pick the closest match.
Your options are parallel, so put each one behind its own tab and let the reader choose. See Using the Tabs Block.
The path forks on something only the reader knows, which is exactly what this block is for. Keep reading below.
Each step gets a title, a description, and its own screenshot slot, and the numbering looks after itself. See Using the Steps Block.
A plain ordered list is lighter and easier to scan. Save the Steps block for when the screenshots are doing real work.
That tree is six nodes. Two of them ask a question, four are outcomes, and one answer leads to a second question rather than straight to an answer. The rest of this article is how to build one.
Adding a Decision Tree
Type / anywhere in an article and pick Decision tree. /decision, /tree, and /troubleshooter all find it.
You start with a tiny tree that already works: a starting question with two empty answers, each wired to a node of its own. Everything else you build from there.
How a Tree Fits Together
A tree is made of nodes. Every node has three parts, and only the first is required:
Part | What It's For |
Question | The decision point, on a single line. This is what the reader sees at the top of the card |
Description | Optional supporting detail underneath, for context the question alone doesn't give |
Answers | The branches. Each answer is a button the reader clicks, and each one points at the node it leads to |
The important idea: a node with no answers is an outcome. That is how a path ends. You don't mark a node as final, you simply don't give it anywhere else to go. Those are the four cards badged Result in the tree above.
Building Your First Tree
- Write your opening question in node 1, the one marked Start
- Type a label for each of its two answers, for example "Yes" or "The page won't load". They already lead to nodes 2 and 3
- Write the question or outcome in nodes 2 and 3
- For more branches, click + Add answer on a node, click + Add node at the bottom of the block to create the node that answer should lead to, then open the answer's Choose destination… dropdown and pick that node
- Repeat until every answer points somewhere, and every path ends on a node with no answers

Setting the Starting Question
One node is the entry point, and it is marked Start. To move it, click Set start in the header of the node you want. The node marked Start is the one your reader sees first, no matter where it sits in the editing list.
This matters more than it sounds. The editor shows your nodes as a flat, scrollable list in the order you created them, which is almost never the order a reader travels through them.
Managing Nodes and Answers
Control | What It Does |
Drag handle | Reorders the node in the editing list. This is presentation only, it does not change the tree's logic |
Set start | Makes this the first question the reader sees |
+ Add answer | Adds a branch to this node |
Duplicate node | Copies the node, handy for outcomes that repeat similar wording |
Delete node | Removes it. Any answer that pointed here goes back to Choose destination…, so re-point it. Until you do, that answer won't appear as a button on the published page. A tree always keeps at least one node. |
Remove answer | Each answer has its own destination dropdown and a × to remove it. |
Keyboard Shortcuts
The fields inside a tree have their own navigation, so you can build one without reaching for the mouse:
Keys | What It Does |
| Walks a node's question, description, and each answer in turn, then moves on to the next node. At either end you step out of the block |
| Drops into that node's description |
| Adds another answer and jumps straight to it, so you can type a run of branches quickly |
| Selects just the field you're in, not the whole article |
Limits
A tree holds up to 30 nodes, with up to 8 answers on each. If you're pushing either ceiling it is usually a sign the tree is doing two jobs, and splitting it across two articles will read better anyway.
Once your tree is built, have a look at Understanding Block Types for the rest of the toolbox, or Shortcuts in the Block Editor for the full keyboard reference.