Prompt · Strategic decisions

Stress-Test a Proposed Action

Surface the strongest objections, hidden assumptions, and reputational downsides before the board commits.

Version 1.0 Tested with Claude, GPT-4-class Category: Strategic decisions

When to use it

When a proposed board action is on the table and the room is tilted toward approval. The prompt forces the strongest contrary case to be articulated before the decision is taken, which is often when the contrary case would have been most useful.

The prompt

You are serving as a dissenting board member with deep experience in corporate governance, asked to stress-test a proposed action before the board commits to it. The proposed action is: [DESCRIBE THE PROPOSED ACTION IN NEUTRAL, REDACTED TERMS. INCLUDE: what is being done, by whom at a role level (not by name), at what scale, on what timeline, for what stated purpose, at what stated cost/risk.] The stated reasons for taking this action are: [LIST THE REASONS AS THE BOARD CURRENTLY UNDERSTANDS THEM.] Do the following, in this order: 1. State the three strongest objections a well-prepared dissenting director could raise. For each, identify the hidden assumption the proposal depends on, and say how it could fail. 2. Identify the three most likely ways this decision looks in hindsight if it turns out badly within 18 months. Write each as a headline in a financial or trade publication. 3. Identify one assumption that, if wrong, renders the entire proposal unsound — and say how the board could test that assumption before deciding. 4. Identify the most significant reputational downside that is not currently being discussed. Say why it is not being discussed. 5. Steel-man the proposal: give the strongest possible case for going ahead, which engages with all of the above rather than ignoring it. Do not hedge. Do not offer a final recommendation. Be specific, concrete, and direct.

What to supply

Keep the inputs redacted. Replace company names, counterparty names, executive names, and specific figures with neutral descriptors. "A listed Swiss industrial group," “the proposed acquirer,” “a sum in the low tens of millions” will produce output almost as useful as named equivalents, without the confidentiality cost.

What to expect

A well-framed dissenting response and three plausible failure narratives. The headline exercise is specifically valuable: it surfaces reputational and optics risks that pure financial analysis misses. The steel-man at the end avoids an artificially negative read and forces the model to engage with its own objections.

Before using

Read Privilege and confidentiality before using prompts. Do not paste privileged material; redact names and figures.

Related