Workbench
Inspect the reasoning, not just the answer.
Finished systems are useful. They also hide the difficult part: deciding what the problem really is, which evidence deserves trust, where uncertainty matters, and what a manager can actually do next. The Workbench keeps that middle visible.
Good analysis has to survive contact with the meeting.
The work rarely moves in a neat sequence. Evidence changes the question. A sensible option fails because nobody has authority to choose it. A dashboard settles the colour and starts an argument about the meaning. These are the recurring tests underneath the work.
What actually needs deciding?
Start with the choice, constraint, affected people and consequence. If solving the stated problem would change nothing a manager does, the problem probably needs another look.
Would we trust this number in a difficult meeting?
Check definitions, freshness, ownership, traceability and conflicts between sources. A precise number with an unstable definition is still unstable evidence.
What gives if we choose this?
Compare plausible options, dependencies, capacity and consequences. Make uncertainty visible when it could change the choice; do not manufacture precision merely because a dashboard has room for another decimal.
Who can say yes, and what happens next?
Clarify the choice, authority, owner and next action. “Continue monitoring” is sometimes prudent. It is also a remarkably convenient place for an unresolved decision to hide.
What happens when people actually use it?
Implementation is part of the reasoning. Watch where the workflow breaks, what people ignore, what they work around and whether the original management problem changes. Then revise the system rather than defending the diagram.
One risk problem, worked far enough to require a decision.
From project risk to executive decision.
The case follows project and portfolio risk through evidence QA, normalization, interdependency analysis and escalation until an executive decision is actually required. Inspect the management logic and evidence discipline—not a claimed client result.
Different artifacts expose different parts of the work.
Choose the artifact that matches the question. A synthetic case exposes reasoning without pretending to be a client result. A decision aid exposes an instrument. A reference architecture shows how the management machinery fits together. They are different kinds of evidence, not levels in a maturity model.
The dashboard is green. The delivery forecast is not.
Use this when status looks healthy but the forecast does not. It tests forward signals, decision latency and risk-to-delivery integration against the question management actually needs answered.
Every project is a priority. The portfolio cannot move.
Use this when every initiative is defended as essential. It works through scarcity, capacity, dependencies and the uncomfortable requirement to displace something.
Executive Decision Brief
Use this when the analysis is good enough but the meeting still lacks a decision. The brief keeps evidence, uncertainty, options, authority and commitment visible without turning a one-page decision into a small bureaucracy.
Integrated management and decision system
Use this when the problem is not one bad dashboard but disconnected management machinery. It traces how authority, information, controls, evidence and decision flow have to connect.