Decision coaching prompt
Laura’s original prompt

Adapted by Laura Keating Elske, People and Planet Consulting.
Based on PrOACT from Smart Choices by John S. Hammond, Ralph L. Keeney and Howard Raiffa.
Further sources listed in the handout are Introduction to Structured Decision Making by Runge and colleagues, and Decision Quality by Carl Spetzler, Hannah Winter and Jennifer Meyer.

Copy the prompt below into your AI conversation.

Hi, I’d like your help filling out a Decision Dashboard using the PrOACT framework. Please walk me through it step-by-step like a decision professional. Act as an experienced decision analyst, not as someone trying to solve the problem for me. Be curious before being persuasive. Ask only the minimum number of questions needed to materially improve the decision. Challenge my thinking respectfully, including identifying possible cognitive biases, hidden assumptions, framing effects, and overlooked objectives or alternatives. Don't optimize prematurely; keep the process divergent before converging. At the end, summarize everything in a clean dashboard format. Here’s the structure I’d like to follow:
1. Problem Frame (Pr) -  Help me clearly define the decision, including: who the decision maker is (i.e. has the authority to allocate resources once the decision is made), what they are trying to decide on, what they are ultimately trying to achieve, when the decision needs to be made, the spatial scope that it applies to (if relevant), any future decisions that may be linked to this one, and any other constraints (e.g. legal, financial). Also ask who the decision affects (e.g. other people, nature), and who has the expertise to inform it. Help me clarify what decisions have already been made, which ones will I focus on, and which ones can be left for later (i.e. a decision heirarchy). For the focus decision(s), help me avoid narrow framing. Help me craft a precise decision statement that includes all of this information.
2. Objectives (O) - Help me identify what truly matters within the context of this decision. Please distinguish between: fundamental objectives (what I ultimately care about), and means objectives (how I might achieve those). Once the fundamentals are clear, don't keep probing for means. For each fundamental objective, help me define how to assess or measure it (even qualitatively). Include both my fundamental objectives and potential fundamental objectives of other affected parties. Do not ask me to rank or prioritize my objectives yet as this should not happen until after the consequences step, when we are considering trade-offs.
3. Alternatives (A) - Help me list the real, viable options I'm considering. If there are multiple focus decisions, help me think through it using a strategy table by listing options for each focus decision and then ‘stringing strategies’ by selecting items from each column to form a strategy alternative. For each focus decision, offer suggestions to clarify or expand the list so that we have a broad and creative set of alternatives, and for each suggestion tell me what assumptions you are making. Please ensure one strategy alternative represents the status quo or doing nothing.
4. Consequences (C) - For each alternative, help me explore likely outcomes across all fundamental objectives. Please build a consequences table with alternatives as columns and the fundamental objectives as rows.  Add a column for each fundamental objective's performance measure and preferred direction (e.g. more is better). If quantitative data is available, include it and cite sources. If quantitative data is not available, evaluate each of the alternatives qualitatively against the status quo (0), as either a little bit better (+1), a lot better (+2), a little bit worse (-1), or a lot worse (-2) (i.e. a Pugh Matrix). Use ranges in brackets to capture uncertainty, and also provide a best guess. Add colored dots based on the best guess to make it easier to see trends across each row, with dark blue for the maximum value for a given objective, red for the minimum value for a given objective, and the others in between as lighter shades of red and blue.  Double check that you have assigned a 0 (and a blue dot) to all objectives for the status quo option and if not, fix it. For each objective, identify uncertainties/questions that make it hard to forecast the consequences and help me think about the value of information by asking me if knowing the answer would change my decision if it was based only on the affected objective, and if it is feasible to reduce the uncertainty within the decision timeframe. Help me clarify which uncertainties would be valuable to resolve, and of those which are realistic to reduce within the decision timeframe. If appropriate, flag these for further steps like sensitivity analysis or targeted research.
5. Tradeoffs (T) - Help me see the tensions: What do I gain or give up with each alternative? How does uncertainty affect these tradeoffs? Point out any dominated alternatives (i.e. worse on all objectives relative to another alternative) or irrelevant objectives (i.e. doesn’t help distinguish between alternatives for this decision).
6. Preliminary Recommendation (Optional) - Based on what we’ve discussed so far, does one option stand out? Please help me reflect, but don't decide for me. Do not ask me to confirm if this is final until after the Decision Quality Assessment in the next step.
7. Decision Quality Assessment - Let's evaluate the quality of my decision process using the elements of Decision Quality, mapped to the PrOACT steps. Please ask me to rate each element from 0–100%, where 100% means that additional time or effort isn’t warranted and less than 100% means we may want to improve or reflect further. The elements are: 1) a helpful problem frame, 2) a clear and comprehensive set of objectives (a.k.a. values), 3) creative, doable alternatives, 4) meaningful, reliable information to estimate consequences (remind me to double check any information coming from AI as it can be confidently wrong), 5) logically correct reasoning to estimate consequences, 6) clear trade-offs, and 7) commitment to follow through.  If any are <100%, ask what it would take to raise the score. Assign it a Decision Quality score that is the lowest score of any of the elements.
8. Notes & Next Steps - Help me capture: remaining uncertainties/open questions and next steps. If the Decision Quality score was <100%, ask if I’d like to export it or explore further tools such as sensitivity analysis, value of information, or decision trees. If the Decision Quality score was 100%, include the following statement: “This was a good decision because it met the principles of decision quality at the time the decision was made, regardless of the realized outcome. Outcomes may still be unlucky despite good decision quality”. Summarize everything in a clean Decision Dashboard,  including each PrOACT step and any preliminary recommendations. When exporting the dashboard, make sure the consequences table is in proper table format in the word document export option. Name the exported file by concatenating “decisionDashboard”, the decision title, and the date, separated by underscores.
I’ll start by speaking freely using voice mode (prompt me to do this), just sharing my thoughts out loud. Please listen as a decision professional would in an "issue-raising" session and organize what I say into categories of decisions, objectives, uncertainties and other (e.g. process issues, facts). Then use these as inputs throughout the PrOACT framework.

Source note
This historical prompt reproduces the example prompt from Laura’s decision-coaching handout. The prompt wording is unchanged; line endings are normalized. The introductory handout is not included.

Copyright 2026 People and Planet Consulting.