The DecisionGrade™ engagement method

A decision you can stand behind in front of a board or a regulator, reached without weeks of silence. You see the first outputs at the end of week one, and you can see the picture firm up from there.

Book a conversation (opens in a new tab)

A reasonable question should not cost the organisation a quarter

A senior person asks what happens if a programme is paused. They ask others, who ask others, and everyone involved is pulled off delivery to help answer. The answer arrives weeks later, the moment has often passed, and few fully trust it. The organisation is harmed by the act of asking.

The engagement is designed around the opposite experience: a small first step, outputs you can see and use early, and a record that shows at every point what is firm and what is not.

What you see, and when

  1. Before

    a 25-minute conversation

    What happens: We test whether the pattern fits the decision you are facing.

    What you have at the end: A view on whether Phase 1 is worth doing, and what the decision question might be.

  2. Phase 1

    week 1

    What happens: We agree the decision question with the people accountable for it, and establish what data genuinely exists and who owns what.

    What you have at the end of week one:

    • A Decision Question Statement. The question, written down and agreed: what is being decided, by whom, by when, and what a defensible answer has to cover.
    • An initial dependency hypothesis. A first picture of what the decision touches, with every item marked as firm, resting on an assumption, or still to be validated. You get it even if data access stalls.
    • A plain account of the data position. What exists, what does not, and who would need to confirm what.

    Phase 1 stands alone. It is fixed in scope, one week long, and its outputs are yours to use whether or not you go further.

  3. Phase 2

    weeks 2–6

    What happens: Iterative analysis of the available data, with your experts asked only for the judgements that are theirs to give: minutes to confirm what they own, not weeks trawling the estate.

    What you see as it runs: At each review, more of the picture has moved from "still to be validated" to "confirmed", each confirmation bound to a name and a date. Gaps are flagged, documented and traceable.

    What you have at the end: An executive decision pack: a recommendation, the evidence behind it, the assumptions it rests on, and a record of who confirmed what.

  4. Phase 3

    optional, commissioned separately

    What happens: DecisionGrade stays in place, so the confirmed picture is kept and maintained.

    What you have: The next question starts from what has already been confirmed, by whom and when, rather than from scratch.

A sponsor, a question and minutes of expert time

  • A sponsor who owns the decision and will agree the question in week one.
  • Access to the data that genuinely exists, in whatever observation tooling, repositories and systems of record you already run. Where access stalls, the gap is recorded and the work carries on around it.
  • Short, specific requests to the experts who own what the decision touches. Judgement stays with the people accountable for it.

Advice from the same practice

Viner Advisory Ltd also advises on the questions that surround the decision itself:

  • Governing AI-assisted decisions. Governance frameworks, human-in-the-loop oversight and audit trail requirements for decisions that draw on AI, aligned with the regulatory frameworks that apply to you.
  • Organisational change management for AI adoption. Helping the people who will be accountable for AI-assisted decisions to work with them confidently, and to keep their judgement on the record.
  • Technical consultancy for AI integration. How AI-assisted analysis fits into the enterprise architecture and the systems of record you already run.

Fixed scope, defined outputs

Each phase is a fixed-scope piece of work with defined outputs, agreed before it starts. Pricing is discussed in the first conversation.

The fair comparison is the internal alternative: senior people pulled off delivery for weeks to produce an answer that few fully trust.

Start with Phase 1

A 25-minute conversation is enough to find out whether the pattern fits. If it does, Phase 1 stands alone: one week, one clear decision question.