Progress
Where am I, and what is complete?
And yet the finance teams responsible for keeping the cycle moving still coordinated it through email. One task stalled upstream held twenty-four regional leads downstream. The only mechanism for detecting that stall was somebody asking.
The data was all there.
What was missing was any way to act on it.


The brief said make it more visual, show the workflow. So I mapped the whole budget process into a single visual flow — every phase, every step, in sequence.
But it showed where the stops were. It could not show you where your train was. A user could trace the process end to end and still not answer: where am I right now, and what is waiting on me?
The problem was never that the process was hard to see. A user’s own state inside it was invisible.
BriefImprove consistency and navigation.
InsightUI inconsistency was a symptom.
ChallengeMake workflow state visible.
Users held the operating model in memory. Mapping made the gap concrete.
Where am I, and what is complete?
Who approves this? Who owns next?
What is blocking me?
What should I do now?
The persistent status bar survives every screen — the coordination layer that was missing.
A coordination layer follows the user through the entire process.
The system surfaces the work with the greatest downstream impact.
Every blockage reveals its owner, consequence and recovery path.
The state would disappear as users moved through the workflow.
It explained location, but not what was blocked or who owned the next move.
Required coordination would become optional. I traded flexibility for a guarantee.
National Finance, regional controllers and agency users walked through the same dependency chain. Each role could identify current state, owner and next action without explanation.
The rule library connected workflow principles, screen behaviour and acceptance criteria — giving future projects proven patterns instead of a blank page.
Once decisions were expressed as roles, states, dependencies and exceptions, they could constrain AI-assisted exploration instead of relying on prompts alone. I later applied the same principle in TAP Mindset: stable rules provide the context that makes generated work reviewable, repeatable and safer to ship.
See how this principle evolved in TAP Design System ↗
Three roles tested the same hand-off, dependency and exception scenarios.
55+ limitations were documented as constraints, not presented as outcomes.