Dear Negotiation Explorer,
Welcome to issue 25 of our NegoAI series.
Last week we explored building trustworthy AI negotiation agents. This week, I'm sharing something more immediately practical: a single metaprompt that generates a complete negotiation preparation framework—your interests, their interests, BATNA analysis, scenarios, and creative options to enlarge the pie.
The concept is called metaprompting, and it changes how you use AI for negotiation.
What is Metaprompting?
Most people write prompts that ask AI to complete a task directly. Metaprompting is different—you're asking AI to generate a structured prompt that then produces higher-quality output.
Think of it this way:
Regular prompt: "Help me prepare for this negotiation"
Metaprompt: "Create a comprehensive negotiation analysis framework acting as a Harvard professor"
The difference? A metaprompt forces the AI to think systematically before responding. It produces a structured, reusable prompt instead of a one-off answer.
Why This Works for Negotiation
Negotiation preparation requires analyzing multiple dimensions simultaneously:
Your interests vs. their interests
Your alternatives vs. their alternatives
Multiple scenarios for how they might approach the deal
Creative options neither party has considered
A metaprompt captures all of these requirements and generates a professional-grade prompt you can use repeatedly.
The Metaprompt
Here's the metaprompt I use with ChatGPT, Claude, or Gemini:
Hi, I'm preparing for a negotiation. As attachment you have the full information and context.
What I would like you to do is to act as Deepak Malhotra, the renowned Harvard professor, and provide me with:
My interest ranking - what matters most to me, ordered by priority
My BATNA - my best alternative to a negotiated agreement
The plausible ZOPA - the zone of possible agreement
The other party's BATNA and interests - across 2-3 distinct scenarios
Understanding what is very important for them - ranking their interests in each scenario
Their emotional drivers - what motivates them beyond the rational
Creative options to enlarge the pie - provide more than 2-3, some even out-of-the-box
Quality criteria:
Ensure the BATNA analysis includes both explicit information from the context AND implicit information that could be valuable
Provide different scenarios that are non-overlapping
Include at least one out-of-the-box scenario
Revise your output before providing it to me with a quality self-assessment
Output format: Present the final output with clear headings and subheadings for each section.
The Structured Prompt (What AI Generates)
When you run the metaprompt, the AI generates this structured prompt for you to use with your negotiation context:
ROLE:
You are Deepak Malhotra, renowned Harvard Business School professor
and author of "Negotiating the Impossible." Apply your expertise in
negotiation strategy, deal-making, and conflict resolution.
CONTEXT:
I am preparing for a negotiation. Attached is the full background
information including parties involved, history, constraints, and
objectives.
INSTRUCTIONS:
Analyze the attached context and provide:
1. MY INTERESTS
- List and rank my interests by priority
- Include needs, objectives, motivations, constraints, and concerns
2. MY BATNA
- Identify explicit alternatives stated in the context
- Infer implicit alternatives from available information
- Assess consequences if no agreement is reached
3. ZOPA ANALYSIS
- Estimate the plausible zone of possible agreement
- Identify where interests may overlap
4. OTHER PARTY ANALYSIS (Provide 2-3 distinct scenarios)
For each scenario:
- Profile description
- Their likely interests, ranked
- Their probable BATNA
- Their emotional drivers (fears, motivations, pressures)
Ensure scenarios are non-overlapping.
Include at least one out-of-the-box scenario.
5. CREATIVE OPTIONS
- Generate 4-6 options to enlarge the pie
- Include at least 2 unconventional/out-of-the-box ideas
- Focus on creating value for both parties
QUALITY CRITERIA:
- Extract insights from both explicit and implicit information
- Scenarios must be distinct, not variations of the same profile
- Creative options should go beyond obvious trade-offs
- Conduct a self-review before delivering final output
- Note assumptions made and confidence levels
OUTPUT FORMAT:
Present using clear headings and subheadings for each section.
Use bullet points for rankings and lists.
End with a brief quality assessment of your analysisWhy the Structured Prompt Matters
Notice what the AI-generated prompt includes that your original request might miss:
Clear role definition - Not just "act as Deepak Malhotra" but specifying the expertise to apply
Explicit structure - Each section has clear deliverables
Quality controls built in - Self-review, assumptions noted, confidence levels
Completeness checks - Non-overlapping scenarios, both explicit and implicit information
This is why metaprompting works: the AI thinks through what a comprehensive analysis requires before producing it.
How to Use This
Step 1: Copy the metaprompt above
Step 2: Paste it into ChatGPT, Claude, or Gemini
Step 3: The AI generates the structured prompt
Step 4: Take that structured prompt, attach your negotiation context (documents, emails, briefs), and run it
Step 5: Receive your complete negotiation preparation analysis
Customizing the Metaprompt
You can modify the metaprompt for specific negotiation types:
For salary negotiations:
Add: "Include market benchmarks and timing considerations"
For M&A deals:
Add: "Consider due diligence findings and integration risks"
For vendor contracts:
Add: "Analyze switching costs and competitive alternatives"
The metaprompt adapts to your context—just specify what additional dimensions matter for your situation.
This Week's Exercise (20 minutes)
Step 1: Select an upcoming negotiation
Step 2: Run the metaprompt in ChatGPT, Claude, or Gemini
Step 3: Review the structured prompt it generates—does it capture everything you need?
Step 4: If not, refine your metaprompt and regenerate
Step 5: Use the final structured prompt with your negotiation context
The goal is understanding how metaprompting creates better preparation frameworks than asking AI directly.
One metaprompt. One structured framework. Complete negotiation preparation.
