Building Custom Negotiation Playbooks

Learn how to build comprehensive negotiation playbooks with ChatGPT that capture your firm's expertise, risk tolerance, and negotiation strategies.

Learning Objectives

By the end of this tutorial, you will:

• Build a negotiation playbook that encodes your firm's expertise
• Apply the playbook to a contract and flag red lines
• Escalate issues using a clear approval matrix
• Implement your playbook in a Custom GPT

Part 1: What Is a Legal Playbook?

Definition

A playbook is a structured document that captures your organization's:

• Standard positions on key contract terms
• Acceptable ranges for negotiable items
• Red lines that cannot be crossed
• Escalation triggers requiring senior review
• Approved fallback language for common pushback
• Procedural workflows for document analysis and negotiation tracking

Why Playbooks Matter

Part 2: Playbook Architecture

Core Components

PLAYBOOK STRUCTURE
|
+-- 1. SCOPE & APPLICATION
|   +-- Contract types covered
|   +-- Deal size thresholds
|   +-- Industry-specific considerations
|
+-- 2. PARTY IDENTIFICATION
|   +-- Our typical role (vendor/customer/licensor/etc.)
|   +-- Counterparty profiles
|   +-- Risk tier classification
|
+-- 3. CLAUSE-BY-CLAUSE POSITIONS
|   +-- Standard position
|   +-- Acceptable alternatives
|   +-- Fallback language
|   +-- Red lines
|   +-- Escalation triggers
|
+-- 4. RISK ASSESSMENT FRAMEWORK
|   +-- Severity ratings
|   +-- Cumulative risk scoring
|   +-- Deal-breaker thresholds
|
+-- 5. NEGOTIATION GUIDANCE
|   +-- Priority ranking
|   +-- Trade-off guidance
|   +-- Tactical advice
|
+-- 6. APPROVAL MATRIX
    +-- Authority levels
    +-- Escalation paths
    +-- Documentation requirements

Part 3: Building a SaaS Customer Playbook

Starter Template

Use this as your starting point. Customize scope, your role, and add clause positions for each term you care about:

# PLAYBOOK: SaaS Customer Agreements

## 1. SCOPE
Contract types: SaaS, cloud subscriptions, data processing addenda.
Deal tiers: TIER 1 (<$50K), TIER 2 ($50K-$250K), TIER 3 (>$250K).

## 2. OUR ROLE
We are the CUSTOMER. Adjust risk tolerance by counterparty tier.

## 3. SAMPLE CLAUSE - Limitation of Liability
Standard: Mutual cap at 12 months fees; uncapped for indemnity, IP, data breach, gross negligence.
Red lines: Customer liability uncapped; no carve-outs for vendor misconduct.

Part 4: Risk Assessment Framework

Issue Severity Ratings

Cumulative Risk Score

Calculate overall contract risk:

• Count RED issues (x3 points each)
• Count YELLOW issues (x1 point each)
• Total score determines required approval:

Part 5: Negotiation Guidance

Priority Ranking (What to fight for)

1. Data Security and Breach Liability -- Always priority number 1
2. Limitation of Liability Cap and Carve-outs -- Number 2 for all deals
3. Indemnification Scope -- Critical for IP/data
4. Termination Rights -- Important for flexibility
5. Service Levels -- Important but negotiable

Trade-off Guide

Part 6: Approval Matrix

Authority Levels

Escalation Requirements

When escalating, provide:

1. Summary of issue and vendor position
2. Your recommended response
3. Business context (deal importance, relationship)
4. Cumulative risk score for entire contract

Part 7: Implementing Your Playbooks in ChatGPT

Step 1: Create a Playbook Custom GPT

Create a Custom GPT called "Playbook: SaaS Customer Agreements" (or separate GPTs for different specialties).

Step 2: Add Playbooks as Custom Instructions

Copy your complete playbooks into the Custom GPT instructions field. ChatGPT will apply all playbook guidance to every conversation in that GPT.

Step 3: Upload Reference Documents

Add to knowledge base (via file upload in GPT configuration):

• Your firm's standard agreement templates
• Example approved agreements (with sensitive info redacted)
• Playbooks for related matter types

Step 4: Calibrate with Test Cases

Run 3-5 contracts you've already reviewed through the Custom GPT. Compare ChatGPT's analysis to your original review. Adjust playbook instructions where ChatGPT's output differs from your expectations.

Calibration Prompt:

I'm uploading a contract I previously reviewed. Analyze it against
our playbook and provide your clause-by-clause assessment with
risk ratings. I'll compare your analysis to my original review
to calibrate our playbook instructions.

Step 5: Refine and Iterate

Based on calibration:

• Add clarity where ChatGPT misunderstood positions
• Add edge cases you want handled specifically
• Adjust risk ratings if ChatGPT is too aggressive/conservative
• Add examples where helpful
• Expand guidance for common issues your firm encounters

Do This Now

• Create a playbook for one contract type you handle (scope, role, clause positions)
• Add it to a Custom GPT as custom instructions
• Run 3+ real contracts through it and calibrate
• Apply to a live contract and escalate red flags per your approval matrix

Appendix A: Playbook Template Quick-Start

Copy this skeleton and fill in your firm's positions:

# PLAYBOOK: [Contract Type]

## 1. SCOPE
What contracts this covers: [description]

## 2. OUR ROLE
Typical party we represent: [vendor/customer/licensor/etc.]

## 3. CLAUSE-BY-CLAUSE POSITIONS
   ### [Clause 1]
   - Standard: [Position]
   - Acceptable: [Variations]
   - Red Lines: [Must escalate]
   - Language: [Approved text]
   ### [Repeat for all key clauses]

## 4. RISK ASSESSMENT FRAMEWORK
   - RED criteria
   - YELLOW criteria
   - GREEN criteria
   - Scoring methodology

## 5. NEGOTIATION PRIORITIES
   [Ranked list of fight-for items]

## 6. APPROVAL MATRIX
   [Authority levels by issue type]

Appendix B: Sample ChatGPT Prompts for Playbook Use

Contract Review Prompt

Review this [contract type] against our firm's playbook.
Provide a clause-by-clause analysis with risk ratings
(RED/YELLOW/GREEN). Flag any deviations from playbook positions
and recommend negotiation strategy.

Risk Dashboard Prompt

Create a one-page risk dashboard for this contract.
Include overall risk score, category breakdown, red flags,
and executive summary suitable for board presentation.

Issue List Prompt

The counterparty just sent redlines. Extract all substantive
changes and create an issue list with impact analysis, our
position, their likely response, and middle ground options.

Negotiation Tracker Prompt

Update the negotiation tracker. We're now in Round [X].
Show what's resolved, what's outstanding, and recommend
next steps for the remaining open items.

Related family pages

• Claude Playbooks - Same concepts with Claude
• Core Concepts - Platform-neutral legal workflow model

Sources

• OpenAI Pricing
• OpenAI API Documentation
• Prompt Engineering Guide
• Harvey
• Legora

Additional Reading

• ABA issues first ethics guidance on a lawyer's use of AI tools

Next Steps

• Playbooks Advanced - Litigation briefs, memos, term sheet comparison, dashboards, ROI
• Tutorial 06: Legal Workflows with ChatGPT - Custom GPTs and legal tools

Previous: Tutorial 04: Using Projects