Jump-To technically allows jumping to a UserTask, but it is not a good fit for a proper rejection / send-back scenario. It does not guarantee backward-only steps, can cause timeline inconsistencies, and is risky in workflows with branches.
The recommended approach is to use a Decision activity instead of Jump-To. After each important UserTask, add a simple Approve / Reject decision. The Reject path should be explicitly connected to a clearly defined previous step in the workflow.
With this design, the workflow does not “jump back” unexpectedly. It continues in a controlled and logical way, so the timeline and audit history stay clean and easy to understand.
This approach also works better with branching workflows, because the rejection follows the actual execution path. For Mendix 10.24.x, designing rejection flows with Decision activities is the simplest and most accepted best practice.
Are you saying that after Reject, we should place a “Content Revision” user task and then use a Jump activity to return to the first approval step?
Actually, I’m not suggesting “Reject → Content Revision task → Jump back to the first approval step”.
What I mean is: don’t use Jump-To at all for send-back/rejection. Instead, model the “send back” as a normal, explicit path in the workflow.
A simple example: