Campaigns define the program. Passes package the experience. Hops are its stops; drops are the value unlocked there. I use those relationships to decide what an operator needs to see, what the system should handle, and when a choice needs to be explicit.
Decision 01
Make the next step obvious.
Problem
Campaign setup spans content, passes, channels, and launch timing. A complete form does not necessarily mean a campaign is ready.
Approach
The dashboard keeps setup progress beside campaign health and publish controls. Missing configuration becomes an actionable item in the workflow.
Consequence
Operators can distinguish unfinished setup from a launch-blocking issue. Readiness feedback supports the decision; publishing remains an explicit action.
Decision 02
Put the result beside the input.
Problem
A title, image, or description is hard to judge in a field alone. Its length and crop matter in the experience a visitor will actually see.
Approach
The pass builder pairs structured editing with a live mobile preview, so content can be evaluated in the same place it is authored.
Consequence
The preview makes hierarchy and content fit visible during editing. Device-specific interactions still need validation in the native app.
Decision 03
Keep publishing deliberate.
Problem
Editing an experience and releasing it have different consequences, especially once someone owns a pass.
Approach
Draft editing, readiness feedback, and publication are separate concerns. Versioned content gives published experiences a stable reference.
Consequence
The architecture supports continued authoring while preserving the version associated with an owned pass. Updating existing owners requires a separate migration decision.
What I would evaluate next.
Follow an operator from initial setup to publication. Observe time to readiness, configuration errors, and recovery after an interruption. These checks can show where an interface that looks clear still leaves room for uncertainty.
Keep the evidence close.
Screenshots establish what the interface presents. Code excerpts explain a specific behavior. Neither replaces observing someone use the product to complete a real task.