How leaders can use generative AI to accelerate strategic decision-making

Generative AI tools like Claude, Gemini, and GPT have slashed the time needed to prepare board meetings, analyze scenarios, and create executive presentations from days to hours. The key difference between those who leverage this advantage and those who don’t often lies in the quality of the instructions given to the machine—what’s known as a prompt. In an executive’s daily routine, a prompt functions like a briefing: the more context, criteria, and delivery format you specify, the more useful the response will be.

Prompts for decision-making

1. Market diagnosis

Take on the role of a senior strategy analyst. Using the data below about [sector], identify the three most significant changes that occurred in the past 12 months, assess their likely impact on a company of [size and business model], and list the questions that should be presented to the executive committee. Distinguish between facts and hypotheses.

2. Devil’s advocate

Imagine I’m presenting a decision I plan to make, along with the arguments supporting it. Challenge it: highlight the weakest assumptions, the risks that haven’t been considered, and the scenario in which it would prove wrong in 18 months.

3. Comparison of alternatives

Compare options A, B, and C for solving [problem] in a table that includes estimated cost, timeline, execution risk, reversibility, and impact on [indicator]. At the end, indicate what additional information would change the recommendation.

Prompts for communication

4. Summary for the board

Transform the attached report into a one-page document for the board: context in three lines, decision to be made, alternatives, recommendation, and risks. Use clear, jargon-free language.

5. Presentation test

Read this presentation as a skeptical advisor with a finance background. List the ten toughest questions they would ask and point out which slides have weak or missing answers.

6. Communication to leadership

Write a message to directors and managers about [change]. Use a firm and transparent tone. Explain the reason before the action, anticipate the three main concerns of the teams, and conclude with the next steps.

Prompts for planning

7. Scenario building

Create three scenarios for the next 24 months of [business], considering factors such as interest rates, exchange rates, and demand. For each scenario, describe the signals that would indicate it’s materializing and the decisions that should be anticipated.

8. Budget stress test

Based on the budget assumptions below, identify the three that, if wrong by 20%, would have the greatest impact on the result. Estimate the effect of each and suggest review triggers.

9. 90-day plan

Structure a 90-day plan to achieve [goal], divided into biweekly periods, with area managers, verifiable deliverables, and the three risks that could delay execution.

Prompts for AI agenda governance

10. Prioritization of use cases

Rank the processes below by potential gain, implementation complexity, data risk, and time to first result. Indicate which two should become a pilot and what criterion defines a successful pilot.

11. Risk map

Evaluate the AI initiative described below from the perspectives of LGPD, information security, intellectual property, and reputation. For each risk, indicate probability, impact, and the question the board should ask the responsible party.

12. Conversation with suppliers

Prepare the questions I should ask a supplier who promises an AI solution, to distinguish a good demonstration from a measurable result.

What differentiates a generic prompt from an executive prompt

An open-ended question like “How is the retail market doing?” generates a lengthy and generic answer. An executive prompt should define four elements: the role the AI should play, the business context, the analysis criteria, and the delivery format.

Before applying any of these commands, remember that confidential information should only be entered into environments approved by the company’s data policy.