A bounded population
Write the inclusion rule in one sentence and pull the list from a trusted system.
For your team / Strategy and innovation
A first workflow becomes useful when its population, measure, owner and expansion criteria are settled before launch.
A first workflow enters an evidence review. The governed decision can be to stop, revise, keep the work local, standardize it, or expand it.
The dossier represents decision gates, not promised dates. No outcome is automatic, and expansion is not visually or operationally privileged.
A first workflow enters an evidence review covering its bounded population, named failure, pre-existing measure, and owner who can approve change. The review produces one governed decision: stop, revise and retest, keep the workflow local, or standardize and expand.
Selection
If a candidate fails two of these tests, it is a poor first choice.
Write the inclusion rule in one sentence and pull the list from a trusted system.
Staff can describe the break without prompting, and a result would matter in a budget conversation.
Its definition and denominator predate Rely. Inventing the measure with the intervention makes the result unfalsifiable.
Someone can change the workflow, not merely observe it. Learning that cannot be released is only a report.
Standard or local
The first workflow should reveal what is genuinely common across the system and what belongs to one department's local reality.
Expansion gates
These are evidence gates, not a promised timeline.
Scope, permitted actions, the clinical boundary, security, legal and tests are signed before launch.
Navigators map the routes the real population takes. Expect discovery, not steady-state performance.
Service levels have history, exceptions are understood, and reporting reconciles against the source system.
Operating evidence decides what becomes standard, stays local or deserves an adjacent transition.
Expansion is a decision against criteria written before launch, never an automatic next phase.
The ask
You leave with one recommended first workflow, its measurement design and initial expansion criteria. If none meet the four tests, we will say that in meeting one.
We will tell you which one is a credible first proof, and which one would produce a result nobody can act on.