One Register, Two Scales
MetroLink was formed eighteen months ago by merging a city bus operator and a regional rail company. The bus business uses a 4 x 4 probability-impact matrix with financial impact bands in pounds; the rail business uses a 5 x 5 matrix with a separate safety-impact dimension because of its safety-critical operations. The board risk committee has asked for a single top-ten risk report for its first combined meeting in six weeks. The interim head of risk proposes converting bus scores to the rail scale by multiplying them by 1.25, then ranking all risks by the product of probability and impact. Operations directors object that this will distort safety risks. Two risks currently score identically: a cyber risk to signalling and a driver-shortage risk. The CEO says, 'Just give me the list; we can refine later.' There is no agreed risk taxonomy and appetite statements still refer separately to each legacy organisation.
1. Critically evaluate the proposed score-conversion method against M_o_R principles. Which principles are most exposed, and what should the interim head of risk recommend instead?
Read the worked answer
Why the conversion fails
Multiplying a 4-point score by 1.25 manufactures a number; it does not create comparable information. The most exposed principles are:
- Informed by best available information - the two legacy scales measure different things. A rail score may already include a safety dimension that the bus score does not.
- Tailored - one scale cannot be tailored to both a safety-critical railway and a bus operation unless the impact categories are agreed first.
- Transparent - decision-makers cannot see what a converted score of 12 means, so they cannot challenge it.
- Supports decision-making - ranking by probability x impact ignores appetite, velocity, proximity and control effectiveness. Two identical scores can justify different responses.
- Involves stakeholders - the operations directors' objection is a signal that the method will not be accepted.
What to recommend instead
Agree a common impact taxonomy covering financial, safety, regulatory, reputation and service impact, with explicit 5-point bands for each. Map the existing top risks into that framework through facilitated workshops rather than arithmetic. Keep safety as a separate impact category with a board-escalation threshold, so a high safety impact cannot be diluted by an aggregate score. Record data confidence and assumptions on every entry. Present the first report as transitional: top risks by strategic objective, appetite breaches, and a portfolio view, not a single false-precision league table. The CEO's request for speed is legitimate, but a list built on an invalid conversion will not survive the first challenge.
2. The CEO insists on a list in six weeks. Set out a pragmatic six-week sequence, and state two things you would explicitly refuse to do.
Read the worked answer
Six-week sequence
Week 1 - Agree with the CEO and risk committee the purpose of the report and the decisions it must support. Agree the impact categories and scale with both leadership teams. Identify 20-25 candidate risks from both legacy registers without publishing converted scores.
Week 2 - Run two workshops covering bus and rail operations, finance and safety. Rescore only the candidates, capturing rationale, proximity and confidence.
Week 3 - Validate with named risk owners. Test each entry against the agreed appetite thresholds and record control effectiveness.
Week 4 - Aggregate and deduplicate. Look for concentration and shared causes, such as signalling cyber and driver shortage both affecting service continuity.
Week 5 - Challenge the draft with the executive and operations directors. Agree owners and immediate actions.
Week 6 - Present the board report: top risks by objective, appetite breaches, confidence ratings and a plan to extend the assessment to the full register.
Two refusals
- I would refuse to convert legacy scores arithmetically and present them as equivalent.
- I would refuse to publish a single ranked list without appetite context, confidence and a note that it is a transitional view. Both refusals protect the credibility of the first report rather than delaying it.