Frame the advisory decision · Lesson 1

From a broad question to a bounded decision

4 min read

Lesson objectives

  • Frame an advisory engagement with a decision owner, alternatives, horizon, system boundary, evidence needs and decision date, while separating the adviser’s inference from the owner’s choice.
Reading tools

A client may ask, ‘What will transport look like in the future?’ That question alone cannot produce usable advice. It names neither the person with authority to choose nor the live options nor the period in which the choice can still change. Start by asking for a concrete decision: fund a limited bus trial, expand a fixed fleet, or defer spending. Separate the foresight horizon, perhaps the next decade, from the administrative decision date, perhaps a budget meeting next month. The former determines which changes to explore; the latter limits the evidence that can be gathered before advice is due. A scenario is not a certain forecast. Foresight methods use several plausible futures to challenge choices made now, so the client's decision and the conditions that might change it remain central.

Write a seven-line decision frame. First, name the person or body entitled to choose, rather than only the person who requested slides. Second, describe the action to decide now: choose an option, start a trial, or stop a commitment. Third, identify a feasible baseline; comparing a new proposal with nothing gives it an artificial advantage. Fourth, state the exploration horizon and approval date separately. Fifth, bound the system: which users, services, costs, and decisions are inside the analysis, and what is excluded and with what consequence. Sixth, state the criteria, such as access, cost, and flexibility, and ask the decision owner to rank them instead of burying the adviser's preference in a score. Seventh, list available and missing evidence, including who can close each gap before the decision. When demand cannot be estimated credibly, a designed trial and review threshold may be the responsible product rather than a numerical forecast.

Test the frame with three questions. Could any plausible analysis change a choice that is genuinely open? If an investment is already irrevocably approved, do not sell a fictitious option study; the question may instead be how to adapt implementation or when to halt it. Are the alternatives judged on the same basis? Define total cost and effects on affected groups consistently across choices. Can the owner name evidence that would reverse a current preference? That is a test of recommendation sensitivity, not a promise to know the future. Record stakeholder differences instead of merging them into ‘the organisation's view’. Listen to users, implementers, and funders; then distinguish observations, value judgments, and assumptions needing a test. The legitimate owner may weight criteria differently. The adviser's obligation is to expose the consequences of that weighting.

Lesson summary

Test the frame with three questions.