The Board Pack That Turns AI Into Evidence
Measuring value and risk together, in a format that survives scrutiny
Most firms measure AI adoption like a marketing funnel: users acquired, sessions completed, use cases launched. Regulated financial services firms must measure it like a control system: value and risk together, in a format that a board can interrogate, a risk committee can act on, and a regulator can review without needing a briefing first.
The difference between these two measurement approaches is the difference between an AI programme that sustains board investment across multiple budget cycles and one that accumulates spend, complexity, and risk without producing accountable evidence of either value or control.
The Problem Gartner Keeps Raising
Gartner's research across 2025 and 2026 documents a persistent gap: a small minority of CFOs can demonstrate measurable AI ROI, while most CEOs continue to increase AI investment based on strategic conviction rather than evidence. Most proofs of concept never reach production. Spend grows faster than accountable outcomes.
PwC's 2025 Channel Islands survey reinforces the local pressure: firms need to target technology investment where it creates measurable impact while managing cost bases that vary markedly across the market. Targeted investment that reduces breakeven mandate sizes creates structural competitive advantage. Untargeted investment that cannot demonstrate its contribution to the ledger does not.
Blue Money vs Green Money: The Framework Boards Need
The most operationally useful framing for AI value measurement comes from Gartner's research on AI ROI:
Blue money is productivity and time-saving that does not appear clearly on the ledger—drafts produced faster, meetings shortened, research compiled more quickly. Blue money is real and has value, but it is not P&L impact and will not sustain board investment over multiple years.
Green money is measurable impact on revenue or cost lines, created only when three things happen simultaneously: workflows are redesigned around the new capability; metrics are restructured to reflect the new model; and business owners—not technology teams—take accountability for the outcome.
A practical executive rule: if you cannot explain how a specific KPI connects to a revenue or cost line and name the business owner accountable for it, you do not yet have an ROI story. You have a productivity narrative that will eventually come under pressure in a budget review.
A Real Business Case: What Green Money Looks Like
The following is a generalised, anonymised illustration of how a rigorous AI business case should be structured—drawn from a Channel Islands financial services context. No client, employer, or firm details are disclosed.
A firm conducted a detailed analysis of its client onboarding process: twenty cases, ten sub-processes, full cycle time decomposition using a Lean Six Sigma approach. The findings:
- Average end-to-end cycle time: 99 days mean
- Client-controlled portion: 28 days (outside the firm's direct control)
- Internal processing portion: 71 days (entirely within the firm's control)
- Sample revenue leakage: approximately £9,500 per case from delayed fee start dates
- Sample salary cost attributable to rework: approximately £37,000 across the sample
- Total sample leakage across twenty cases: approximately £47,000
Root cause analysis (Pareto approach) identified that three sub-processes—document receipt, initial enquiry handling, and application documentation—drove 69% of total internal processing time. The dominant causes: incomplete data at submission creating rework loops, manual validation steps unable to scale, and fragmented ownership of handoffs.
The firm's AI business case set a target of 18 internal processing days—a 75% reduction—through governed AI augmentation within its regulated system of record, with human-in-the-loop controls retained for all high-risk steps. The financial model:
| Scenario | Realisation Rate | Expected Year 1 Value | Payback Period |
|---|---|---|---|
| Downside | 40% | £150,000 | 7 months |
| Base case | 60% | £225,000 | ~5 months |
| Upside | 80% | £300,000 | 3.5 months |
| Gross opportunity | Pre-realisation | £375,000 | 2.8 months gross |
This is what a green money business case looks like: a specific workflow, a specific KPI, a specific financial mechanism (fee acceleration from shorter cycle time), and a realistic model of how benefits convert to ledger impact under different adoption scenarios.
Portfolio Discipline: Defend, Extend, Upend
Gartner's portfolio logic provides the framework for managing an AI programme as a managed portfolio rather than a collection of concurrent experiments:
- Defend: Parity with the market, compliance hygiene, essential risk reduction. Cost-centre discipline. The baseline capability your firm must maintain.
- Extend: Measurable competitive advantage with traceable ROI connected to revenue or cost lines. Profit-centre discipline. The cases that justify continued investment.
- Upend: Higher-risk strategic bets with genuine option value. Venture-style governance with explicit uncertainty acceptance and defined review gates.
The board question this framework enables: what is our current balance across these three categories, and what evidence do we have that justifies the allocation? For most Jersey FS firms, the portfolio is predominantly Defend cases. The strategic opportunity lies in identifying and rigorously governing the first Extend cases.
The One-Page Board KPI Pack
| Domain | KPI | What It Measures | Cadence |
|---|---|---|---|
| Value | Cycle time reduction | Baseline vs. current in targeted workflows | Monthly |
| Value | Rework / exception rate | Cases requiring manual intervention or correction | Monthly |
| Value | Throughput improvement | Volume without proportional headcount growth | Quarterly |
| Risk & Control | Unsanctioned AI usage | Attempted or confirmed bypass of approved boundary | Monthly |
| Risk & Control | DLP incidents | Data loss prevention events, trend and severity | Monthly |
| Risk & Control | Override rate | Human-in-the-loop overrides in governed workflows | Weekly |
| Adoption quality | Workflow adoption | Usage by workflow, not by login count | Monthly |
| Adoption quality | Training attestation | Completion with attestation, by role | Quarterly |
The Four Quarterly Questions for Boards
- Are we safer than last quarter? Control and risk KPIs trending in the right direction? Incidents and exceptions reducing or stable?
- Are we faster than last quarter? Workflow performance metrics improving? Cycle time moving toward targets?
- Are we seeing green money, or only blue money narratives? Can we connect a KPI to a revenue or cost line? Does the business owner own the outcome metric?
- Should we scale scope, or tighten controls first? Is there evidence that Stage Gate criteria have been met before the next funding commitment?
The Stop/Go Gate That Protects Value
At every funding transition—Proof of Value, Proof of Concept, Pilot, Scale—require three categories of evidence before the next stage is approved:
- Value evidence: Did the target KPI move in the intended direction, even incrementally?
- Control evidence: Did controls function as designed? No critical incidents? Monitoring operating on cadence?
- BAU readiness: Are there runbooks, a named owner, trained users, and a completed handover?
If any of these three categories is absent, pause and remediate before scaling. The cost of pausing at a gate is trivial. The cost of scaling a governance gap is not.
The Medium-Term Question Boards Must Start Asking
Gartner's 'AI shockwaves' research argues that efficiency gains are only the first wave. More material effects will come when AI changes industry economics, pricing power, and client expectations at a structural level. For Jersey's financial services boards:
- Will AI-enabled competitors reduce cost-to-serve significantly enough to compete for mandates that currently sit near your breakeven point?
- Will alternative investment access and tokenisation—already emerging as a priority in the PwC 2025 Channel Islands survey—increase operational complexity your AI governance needs to anticipate?
- Will the next generation of investors expect AI-enhanced transparency and responsiveness as a baseline capability, not a differentiator?
Closing the Series
This final article closes a series covering, in sequence, the strategic foundations, infrastructure requirements, data prerequisites, training obligations, governance architecture, and measurement disciplines that a responsible AI strategy in Jersey's financial services industry requires.
The common thread across all six: AI strategy in a regulated environment is a leadership decision first and a technology decision second. The firms that govern AI well—funding decisions rather than tools, building infrastructure that enforces rather than aspires, preparing data before deploying capability, training for behaviour rather than prompting, building audit-ready operating models, and measuring value and risk together—will have a structural advantage in Jersey's market over those that do not.
The opportunity is real. The governance requirements are clear. The question is whether leadership treats both with the same seriousness.
Closing question: If the market tightened and board scrutiny of AI investment intensified tomorrow, would your AI portfolio demonstrate measurable value and controlled risk—or would it surface a set of disconnected pilots with unmeasured impact and unclear governance?
Sources: PwC Channel Islands Wealth Management Insights 2025; Jersey AI Playbook, March 2026; Gartner research briefings on AI ROI, portfolio management, AI shockwaves, and FinOps disciplines 2025–2026.
Professional disclaimer: The views expressed are my own and are shared for general informational purposes. This content is intentionally non-confidential and does not include proprietary, client-specific, or employer-sensitive information. It is not legal, regulatory, investment, or professional advice; readers should seek appropriate advice for their circumstances.
Financial services leader (20+ years) | Lean Six Sigma Green Belt | London School of Economics (LSE) AI Leadership Accelerator
See where your firm stands
The free diagnostic scores your firm across the twelve operating-model instruments in about eight minutes.
Take the diagnostic →A version of this article was first published on LinkedIn as part of the Sequence body of work. This is the canonical version.
The views expressed are the author's own and shared for general information. This content is intentionally non-confidential and does not include proprietary, client-specific or employer-sensitive information. It is not legal, regulatory, investment or professional advice. © 2026 Paul Soares. Sequence is the trading name of Sequence Advisory Limited (in formation) · hello@sequenceadvisory.com