Challenge a proposal before you commit.
Bring a business proposal, its constraints, and the alternatives you considered. Ask reviewers to make the strongest case against it and identify what would change the recommendation.
Two free Refinement reviews each month · No credit card required
A proposal should explain when it would stop making sense.
Include the expected benefit, costs, dependencies, and decision deadline. State which assumptions are supported by evidence and which remain estimates.
Ask reviewers to compare a credible alternative and explain what information would settle the choice. Use the review to improve the proposal and prepare for stakeholder questions.
What to look for in the review.
Expose a dependency
Ask which assumption the recommendation depends on most.
Compare alternatives
Consider another option using the same goals and constraints.
Prepare the next step
Turn unresolved objections into specific questions, experiments, or evidence requests.
Replace generic pushback with a decision-grade challenge
A useful Devil’s Advocate does not predict disaster for every plan. It states mechanisms, evidence, impact, and the condition that would make the objection irrelevant.
| Decision area | Weak objection | Decision-grade challenge |
|---|---|---|
| Customer | Users may not like it | Which segment loses value, how will behavior change, and what signal appears first? |
| Economics | It could be expensive | Which unit, sensitivity, or cash constraint breaks the expected return? |
| Execution | The team may struggle | Which dependency, capability, sequence, or owner creates the critical path? |
| Competition | A rival may respond | What response is rational, how quickly can it occur, and which advantage erodes? |
| Incentives | Stakeholders may resist | Who loses status, money, control, or optionality and how can they block adoption? |
| Reversibility | We can change later | Which commitments, data changes, contracts, or reputational effects cannot be unwound? |
| Evidence | We need more research | Which specific fact would change the decision and how will it be obtained? |
Prepare a useful review.
Give reviewers a clear question and enough context to examine it.
State the decision and default
Write what is being decided, by whom, by when, against which alternatives, and what happens if no action is taken.
Expose the load-bearing assumptions
List the few beliefs that carry the outcome: adoption, price response, hiring speed, partner behavior, conversion, cost, timing, or regulatory path.
A Devil’s Advocate can create paralysis if opposition has no decision rule.
Every plan has risks and every model can invent more. Define materiality, evidence standards, time horizon, and risk tolerance before the debate. Require challengers to identify mechanisms and reversal conditions. The goal is not maximum pessimism; it is a decision that has confronted its strongest credible failure case.
Models do not know your culture, customers, cash position, contracts, or political constraints unless you provide them, and they can over-index on generic business patterns. Validate material findings with owners and primary data. Humans remain accountable for judgment, ethics, employment consequences, and external commitments.
What to know before you run the panel
Which business decisions benefit most from a Devil’s Advocate?
Use it for consequential, contestable choices with uncertain assumptions: launches, pricing, hiring plans, partnerships, acquisitions, market entry, strategic bets, vendor concentration, policy changes, and irreversible operating commitments. Skip it for routine low-risk choices.
How do I stop the panel from producing generic business risks?
Provide the decision, alternatives, business model, customer evidence, constraints, economics, owners, timeline, and materiality threshold. Require every objection to include a mechanism, evidence, impact, leading indicator, and condition under which the objection would not matter.
Does agreement mean the answer is correct?
No. Models can repeat the same mistake or rely on the same weak source. Read the objections and check important claims against evidence.
Why use other models instead of asking one to critique itself?
Self-critique can help. Other models and specific review roles provide additional responses to compare. The format determines whether reviewers begin independently or see earlier answers.
Does this replace human review?
No. Use the findings to improve a draft or prepare questions for a responsible reviewer. The models do not approve a decision or verify every claim.
Bring your next draft.
Create a free account to draft or import your work. Choose the reviewers and settings before starting a review.