Add questions or instructions to a decision tree

  • Release version: Australia
  • Updated March 12, 2026
  • 1 minute to read
  • Add questions you want to ask your customers or instructions you want to provide to them and configure answer types in a start node in Decision Tree Builder.

    Before you begin

    Role required: admin, sn_gd_core.decision_tree_author

    About this task

    The start node is the first node in a decision tree. It contains a task input and a minimum of one question or instruction. You can configure one or more questions to ask customers or instructions to provide to them.

    Procedure

    1. Navigate to All > Guided Decisions > Decision Trees.
    2. Select the decision tree from the list.
    3. Select Open in Builder.
    4. In Decision Tree Builder, select New node.
    5. In the Node name field, enter a name for the start node.
    6. Optional: In the Reference table field, select a table for the record type.

      The table that you select for the task input does not necessarily have to be a task table. For example, you can select a case, incident, or interaction table for the record context to recommend this decision tree in a workspace.

      Note:
      You must select the same table as the reference table for the context record in Recommended Actions.
    7. In the Question 1 section, enter a question or instruction for customers.
      You can add text formatting, images, or links using the controls at the top of the Question 1 section.
    8. Optional: Make this question a required question by selecting the Required question field.
    9. Select the type of answer for your question.

      For a description of the answer types, see Answer types for questions.

      The answer the customer provides to a question can be passed to other nodes and paths via input mapping.

    10. Optional: Link inputs to reuse answers from prior nodes.
      For more information, see Reuse answers at decision points.
    11. Optional: Add more questions or instructions.
    12. Select Save and close.

    What to do next

    Determine the next node displayed in a decision tree