Bloomreach scenarios are built from three kinds of nodes: triggers, which decide when customers enter the flow; operators, which control how they move through it (waits, conditions, A/B splits); and actions, which execute when customers reach them. In everyday speech “action node” often gets used loosely for any node, but reading a scenario correctly means knowing which kind each node is, which is why clear naming and structure matter.

In practice

Keep each node doing one clear thing and label operators by what they test. A scenario full of unlabeled condition nodes is where automations quietly send the wrong message to the wrong person.