The idea is:
The Time Saved node should act as a passive observer that only tracks metrics without affecting the workflow’s data flow.
Currently, it behaves like any other node: it counts as an incoming trigger for downstream nodes and propagates items forward. This causes unintended side effects when used as a parallel branch.
My use case:
I have a security monitoring workflow with an existing chain:
Node A ──→ Node B
I added a Time Saved node as a parallel branch to track that each run saves 5 minutes:
Node A ──→ Node B
│ ↑
└→ Time Saved ┘
Node B now has two incoming connections and executes twice — once from Node A directly and once via the Time Saved node. This doubled all downstream output (duplicate Telegram messages, duplicate emails, duplicate database writes).
The only workarounds are:
- Place the Time Saved node inline in the main chain — this couples tracking logic to business logic.
- Use it as a dead-end branch with no outgoing connection — this looks like an incomplete workflow and is not intuitive.
I think it would be beneficial to add this because:
The Time Saved node is purely a metrics/tracking node. Users expect to add it anywhere in a workflow without side effects — similar to how you would add analytics to a web page without changing its behavior. The current behavior forces users to restructure their workflow to accommodate what should be a passive measurement.
A passive mode would also prevent accidental data duplication, which can have real consequences (duplicate notifications, duplicate database entries) that are hard to debug because the Time Saved node does not visibly transform data.
Any resources to support this?
- n8n docs on Time Saved / Insights: https://docs.n8n.io/manage/insights/
- Related concept: observer pattern — metrics collection should not alter the observed process