Capability Heat Map That Nobody Acts On
You produced a capability heat map for a retail bank: 62 level-2 capabilities scored on a 1-5 maturity scale by 9 stakeholders. It was well received in the workshop. Six months later, no investment decision has referenced it. The CFO says the scores are 'opinion, not data'. The transformation office uses its own project list, which maps to nothing. A vendor has offered a benchmarking dataset that would replace your scores with industry percentiles. You have one quarter to make the heat map decision-grade or it will be quietly retired.
1. Is the CFO right that the scores are opinion? What would make them decision-grade?
Read the worked answer
The CFO is substantially right, and arguing otherwise loses credibility. A 1-5 maturity score averaged across 9 stakeholders is a perception index. It has real value - it surfaces disagreement - but it cannot bear an investment decision alone.
To become decision-grade, each hot capability needs at least one observable, sourced indicator alongside the perception score, chosen to fit the capability:
- cycle time and touch count (for a process-heavy capability),
- cost-to-serve or unit cost,
- defect/rework or error rate,
- system-of-record count and integration count,
- for customer-facing capabilities, an external measure (NPS driver, complaint volume).
The test to apply: could a skeptical CFO reproduce this number from a system? If not, it is a perception input, and should be labelled as such rather than presented as a score.
Also fix the aggregation: report the spread across stakeholders, not just the mean. A capability scored 1 by four people and 5 by five people is a disagreement about scope, and that disagreement is itself the most actionable finding on the map.
2. The vendor benchmarking offer: how do you decide, and how do you connect the map to the transformation office's project list?
Read the worked answer
On the vendor data: accept it only with conditions. Benchmarks are useful for relative positioning and for challenging internal consensus, but they carry a definitional trap - the vendor's capability definitions will not match yours, and percentile position on a differently-scoped capability is worse than no number, because it looks authoritative. So: map vendor definitions to your level-2 set explicitly, publish the mapping, and use the data as a second opinion on the top 10 capabilities only. Do not replace your scores wholesale; that discards the internal disagreement signal you just paid for.
On the transformation office: the disconnection is the real failure, and it is yours to fix, not theirs. Take their existing project list as-is and do a reverse mapping - every project to the capability it changes. Two outcomes follow immediately:
- Projects that map to no capability are either missing from your map or unjustified; both are findings.
- Hot capabilities with no project are unfunded risks, which is exactly the sentence a CFO will act on.
Deliverable for the quarter: one page per top-10 capability - perception score with spread, one sourced indicator, benchmark position where credible, and the funded/unfunded project count. That is a document that can be argued with in an investment committee, which is the only definition of decision-grade that matters.