The Method Doesn’t Care What Industry You’re In

A framework for governing specialist work you can’t personally grade is not a financial-services framework — it is a general one, and the fundamentals hold for any firm.

The people who audit the decision are rarely the people who have lived with it. I went to the LSE AI Leadership Accelerator expecting to sharpen a financial-services argument. I came back with something larger — the realisation that the discipline I had built inside regulated FS was never really about FS at all.

This series has spent twelve parts and a next chapter arguing that AI in a regulated firm is a leadership and operating-model decision first, and a technology decision second. Fund decisions, not tools. Govern proportionately to materiality. Build the workforce competence the paperwork assumes. Every instalment has been grounded in Jersey — the JFSC guidance, the Jersey AI Playbook, the island’s compact professional community — because specificity is what separates practitioner writing from commentary.

But sitting in a room at the LSE with leaders from healthcare, manufacturing, logistics, energy and public administration, I kept hearing my own problem described back to me in other people’s vocabulary. A hospital commissioning a diagnostic model it cannot itself validate. A manufacturer outsourcing a control system whose failure modes it will only understand after go-live. A public body procuring an algorithm whose bias it is accountable for but did not design. Different industries, different craft, identical governance gap.

That is the argument of this instalment, and the through-line the whole collection has been building toward. The Sequence Method™ — the four-phase, twelve-instrument framework at the centre of this practice — is not a financial-services artefact. It is a horizontal governance discipline, and the reason it travels is precisely the reason it works at home: it never claims to out-expert the specialist. It governs the decision to commission them, and that decision has the same anatomy everywhere.

The gap is structural, not sectoral

Three forces have converged in regulated financial services, and I used to think they were ours alone. Accountability has outrun capability: boards are answerable for outcomes delivered by systems, suppliers and AI they did not build and cannot personally judge. The judge and the judged are the same party: the expertise needed to evaluate specialist work usually sits with the specialist being evaluated, so trust quietly substitutes for governance. And failures surface late — at go-live, at exit, or under regulatory challenge, which are the most expensive moments to discover them.

None of those three forces is financial in nature. They are properties of any organisation that has grown more accountable for specialist work than it is capable of grading. The JFSC’s own framing makes the point without meaning to: its guidance creates no new requirements and instead asks firms to apply existing obligations proportionately, holding that a firm remains accountable for AI outcomes exactly as it would for any other technology (JFSC, 2026). Strip the word “financial” from that sentence and it is simply a statement about accountability under delegation — true for a bank, a hospital trust, or a manufacturer alike.

The gap, in other words, is structural. It appears wherever commissioning outpaces comprehension. Financial services is where I happened to meet it, and where it is most visibly regulated — but it is not where it lives.

What the Method actually governs

The core claim of this practice is deliberately narrow: a firm that cannot out-expert its suppliers can still out-govern them. You will never know more about your custody platform than your platform vendor, more about your model than your data-science supplier, more about your infrastructure than your cloud provider — and you do not need to. What you need is a repeatable discipline for commissioning, evidencing and controlling work you cannot personally perform. That discipline is the Method.

Notice what is being governed. Not the craft — the decision around the craft. The Method never assesses whether a model is well-built or a migration well-engineered; it assesses whether the decision to commission either was named, tested, evidenced and made reversible. That is why it is horizontal. The substance of a custody migration and the substance of a hospital’s diagnostic-AI rollout have nothing in common. The governance sequence around them is the same discipline wearing different clothes.

Those twelve instruments are not abstractions. They are the concrete tests a commissioning firm can run — among them: whether specialists are governed by evidence or taste; whether the brief names the decision or the deliverable; whether vendor exit has been proven, not just contracted; whether the data model was designed or inherited; whether management information steers the next decision or merely reports the last one; and whether a multi-part change is sequenced by dependency or by wishful go-live date. Each is a question, not a capability — which is why none of them requires the firm to out-expert anyone.

This is the honest boundary of the whole practice, and it is also the source of its portability. The moment a method claims domain expertise, it is trapped in that domain. The moment it claims only to govern the commissioning decision, it is free to travel — because every regulated firm, in every sector, commissions work it is accountable for and cannot grade.

Four phases every commissioning decision passes through

The Method holds that every high-stakes specialist decision — in any industry — moves through the same four phases, and that most failures are traceable to a phase that was skipped rather than a phase that was done badly. The instruments within each phase are financial-services-shaped in their examples, but the governing question at the head of each phase contains no sector-specific word at all. That is the test of a genuine invariant: remove the industry, and the question still stands.

Table 1 — The four phases of the Sequence Method, with the governing question a reader can apply to any commissioning decision in their own firm, whatever the sector.

Read those four questions again and try to name an industry they do not fit. A pharmaceutical firm commissioning a trial-analysis model faces the Brief question exactly as a wealth manager commissioning a custody migration does. An airline outsourcing a maintenance-scheduling system faces the Select and Govern questions in the same shape. The content is unrecognisable across sectors; the sequence is identical. That is not a coincidence — it is what makes it a method rather than a checklist.

The evidence for that invariance is not just structural — it shows up in the results. Gartner’s own case work contrasts two organisations that could not be less alike: Goldman Sachs, which doubled software test coverage and cut testing time by 90% yet saw no financial conversion because the underlying process never changed; and Mitsui Chemicals, which invested $90m and drove $270m of revenue impact through deep process redesign (Gartner, 2026). An investment bank and a chemicals manufacturer, the same lesson: the value was never in the tool. It was in whether the decision around the tool — the Build and Govern phases — was actually done. The sector changed; the failure mode, and the fix, did not.

Why lived accountability is the thing that generalises

There is a fair objection here. If the Method is industry-agnostic, does it matter that its author spent twenty years inside regulated financial services rather than, say, healthcare? It does — but not in the way you might expect. The instruments generalise; the judgement that applies them well does not. I have found consistently that the value in governing a hard decision is not knowing the answer in advance — it is having been answerable for a decision like it, and having stayed to live with the consequences when the migration slipped or the vendor underdelivered.

That is scar tissue, and scar tissue is domain-specific. I know where the operational workarounds hide in a regulated FS firm because I ran the operations that quietly depended on them. A hospital chief operating officer knows where theirs hide. Neither of us could stand in for the other on the substance. But the discipline we would each bring — refusing to let a specialist’s confidence substitute for evidence, forcing the exit plan into the contract before the demo dazzles the room — is the same discipline. The Method is how you make that discipline transferable without pretending the expertise is.

This is also why the practice is built to work the way the Method preaches. Before it offers a view, it runs an instrument: a structured diagnostic that scores a firm against the twelve, names the gaps, and ranks them by the damage they do and how soon — evidence before opinion, the brief before the craft. The discipline is applied to the practice itself, not just recommended to clients. A method that asked firms to govern on evidence while itself trading on assertion would not survive its own first question.

This is the distinction the large advisory firms tend to blur. Their model, in the main, recommends the decision rather than having been accountable for it — and then the engagement ends and the team moves on. That is not a criticism of their talent, which is considerable; it is an observation about what their model can and cannot carry. Analysis is portable across industries because it does not commit. Accountability is not portable, which is exactly why a method built from it is worth more than one built from observation.

What this means for a Jersey firm reading it

Let me ground the Build phase in something real rather than hypothetical. In one Channel Islands onboarding process I examined, twenty cases broke down into ten sub-processes — and just three of those sub-processes accounted for roughly 69% of internal processing time. The root causes were not exotic: incomplete data at submission, manual validation, and fragmented ownership of who was accountable for the data. No amount of specialist tooling would have fixed that, because the failure was upstream of the tool — in the Brief and Build phases, where the data model and the ownership of it were left accidental rather than designed. That is the invariant made concrete: the sector was financial services, but the failure was a governance one, and it would look identical in a hospital admissions process or a manufacturer’s supplier-onboarding flow.

So if the Method is general, the temptation is to conclude that Jersey’s specifics stop mattering. The opposite is true. The Jersey AI Playbook names skills development as the most consistently cited priority across every major stakeholder group — Digital Jersey, Jersey Finance, the JFSC, the IoD and the Government of Jersey — and holds that the cost of delayed investment in skills exceeds the cost of the investment itself (Government of Jersey et al., 2026). A general method is worthless in a firm that lacks the people to operate it. Universality of framework and specificity of competence are not in tension; they are the two halves of the same requirement.

Jersey’s compact scale sharpens this. In a jurisdiction where the professional community is small enough to align faster and share learning, a governance discipline that transfers cleanly across sectors is more valuable, not less — because the same instruments can raise the standard of commissioning across funds, trust, banking and insurance without being rebuilt for each. The island advantage the Playbook describes is partly this: a method does not have to be reinvented at every firm to travel between them.

And the JFSC’s Materiality Ladder — its low, medium and higher-impact tiers for classifying a use case and scaling governance to match — is itself a horizontal instrument wearing a Jersey label (JFSC, 2026). Its logic is proportionality, and proportionality is not a Jersey idea or a financial one. It is the idea that governs the whole Method: match the depth of control to the consequence of being wrong, whatever the firm, whatever the sector.

For the last high-stakes decision your firm commissioned but could not grade from the inside — can you say it passed all four phases, or only that no one has yet had to live with the phase you skipped?

The next body of work carries these four phases beyond the Jersey FS frame — and tests whether they still hold in a room that has never heard of a custody migration.

Sources: JFSC, Guidance on the use of AI in Jersey’s financial services sector, 2026; Government of Jersey et al., Jersey AI Playbook, 2026; Gartner, AI Is Coming for Inefficiency, Not People, 2026. Comparisons reflect general market positioning and the author’s own experience; they are not statements about any specific firm’s quality or personnel. The onboarding example is a generalised, non-attributable illustration.

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

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