Chapter 70 — Scoping & Proposals

Overview

Go from discovery to a clear scope, plan, and outcome definition with risks and assumptions. Effective scoping transforms exploratory findings into actionable engagements with measurable success criteria, clear boundaries, and shared accountability.

Why It Matters

Clear scope builds confidence and avoids churn. Good proposals set boundaries, articulate value, and make risks explicit. In AI consulting:

• Ambiguity is expensive: Poorly scoped projects lead to scope creep, budget overruns, and stakeholder friction
• AI projects have unique uncertainties: Model performance, data quality, and integration complexity are harder to predict than traditional software
• Proposals establish trust: They demonstrate understanding of the problem, feasibility of the approach, and professionalism of delivery
• Change control depends on baselines: Without clear scope, every request becomes a debate

Structure of a Comprehensive Proposal

1. Scope and Out-of-Scope

Define what will and won't be delivered with precision:

Best Practice: Use a three-column format to explain boundaries explicitly. The "Rationale" column prevents misunderstandings.

2. Deliverables and Milestones

Structure work into clear phases with concrete outputs:

graph LR
    A[Phase 1: Foundation<br/>Weeks 1-2] --> B[Phase 2: Build<br/>Weeks 3-5]
    B --> C[Phase 3: Validate<br/>Weeks 6-7]
    C --> D[Phase 4: Deploy<br/>Week 8]

    A --> A1[Data assessment report<br/>Architecture blueprint]
    B --> B1[Working RAG prototype<br/>Eval results v1]
    C --> C1[Production-ready system<br/>User acceptance testing]
    D --> D1[Deployed system<br/>Runbook & training]

    style A fill:#e1f5ff
    style B fill:#e1f5ff
    style C fill:#d4edda
    style D fill:#d4edda

Milestone Template:

3. Assumptions and Dependencies

Document what must be true for the project to succeed:

Assumptions:

• Client provides access to knowledge base in structured format (PDF, Markdown, or HTML)
• Approximately 80% of documents contain relevant, high-quality information
• Stakeholders available for weekly review sessions
• Budget approved for OpenAI API usage (~$500/month during development)
• Security review process takes <2 weeks

Dependencies:

• Client IT team provisions cloud infrastructure (Azure/AWS account with appropriate permissions)
• Legal approval for using third-party LLM provider
• Access to 5-10 domain experts for evaluation query creation
• Slack workspace admin permissions granted by Week 3
• Test user group identified and available for UAT in Week 6

Risk: If assumptions prove false or dependencies are delayed, timeline and scope must be revisited.

4. Risks and Mitigations

Build a RAID (Risks, Assumptions, Issues, Dependencies) log:

5. Team, Timeline, and Governance

Team Structure:

graph TD
    A[Steering Committee<br/>Monthly] --> B[Project Manager<br/>Full-time]
    B --> C[Tech Lead<br/>80% allocated]
    B --> D[ML Engineer<br/>60% allocated]
    B --> E[Integration Specialist<br/>40% allocated]

    C --> F[Architecture & Eval]
    D --> G[RAG Implementation]
    E --> H[Slack Integration]

    I[Client Product Owner<br/>Weekly reviews] -.-> B
    J[Client IT Team<br/>As needed] -.-> C

    style A fill:#fff3cd
    style I fill:#d4edda
    style J fill:#d4edda

Governance Model:

6. Acceptance Criteria

Define "done" for each major deliverable:

Example: RAG System Acceptance Criteria

• Functional:

  • Answers 90% of test queries with relevant information
  • Response time <5 seconds for 95th percentile
  • Cites sources accurately in 100% of responses
  • Handles "I don't know" gracefully when information is unavailable

• Non-Functional:

  • Passes security scan with no critical vulnerabilities
  • API uptime >99% during testing period
  • Scales to 50 concurrent users without degradation
  • Monitoring dashboard shows key metrics (latency, error rate, usage)

• Documentation:

  • Architecture diagram with data flow
  • API documentation with examples
  • Runbook for common issues
  • User guide with screenshots

The Scoping Process: Step-by-Step

Step 1: Gather Inputs (Week -1)

Collect from discovery phase:

• Business context: Strategic goals, pain points, success metrics
• Use case prioritization: Which scenarios to tackle first
• Technical landscape: Existing systems, data sources, infrastructure
• Constraints: Budget, timeline, compliance requirements, resource availability
• Stakeholder map: Decision-makers, users, technical contacts

Tool: Discovery synthesis template

Step 2: Draft Initial Scope (Week 0, Days 1-2)

Create first version of:

• Problem statement and success criteria
• High-level approach (architecture, methodology)
• Scope boundaries (in/out/deferred)
• Effort estimate (rough order of magnitude)

Tool: Scope canvas (1-page visual summary)

Step 3: Build Detailed Plan (Week 0, Days 3-5)

Expand into full proposal:

• Break work into phases and milestones
• Identify risks and dependencies
• Define team structure and governance
• Estimate effort and cost with ranges
• Draft acceptance criteria

Tool: Proposal template (see below)

Step 4: Internal Review (Week 0, Day 6)

Review with internal team:

• Technical feasibility (Architect)
• Resource availability (Delivery Manager)
• Commercial viability (Finance)
• Risk assessment (Risk Manager)

Outcome: Refined proposal with confidence levels

Step 5: Client Validation (Week 0, Days 7-10)

Present to client stakeholders:

• Walk through proposal deck
• Discuss options and tradeoffs
• Align on priorities and constraints
• Negotiate scope, timeline, and budget
• Document decisions and action items

Tool: Proposal presentation deck

Step 6: Finalize and Sign-Off (Week 0, Days 11-14)

• Incorporate feedback
• Create final Statement of Work (SoW)
• Review legal and commercial terms
• Obtain signatures
• Kick off project

Scoping Workflow Diagram

flowchart TD
    Start([Discovery Complete]) --> Inputs[Gather Inputs:<br/>findings, constraints,<br/>stakeholders]
    Inputs --> Draft[Draft Initial Scope:<br/>problem, approach,<br/>boundaries]
    Draft --> Detail[Build Detailed Plan:<br/>milestones, risks,<br/>estimates]
    Detail --> Internal{Internal<br/>Review}
    Internal -->|Issues| Draft
    Internal -->|Approved| Client[Client Validation:<br/>present, discuss,<br/>negotiate]
    Client -->|Changes Needed| Draft
    Client -->|Aligned| Options[Offer Options]
    Options --> Finalize[Finalize SoW]
    Finalize --> Sign[Sign-Off]
    Sign --> Kickoff([Project Kickoff])

    style Start fill:#e1f5ff
    style Kickoff fill:#d4edda
    style Internal fill:#fff3cd
    style Client fill:#fff3cd

Commercials and Pricing Options

Offer multiple options to give clients flexibility:

Variable Scope Controls:

• Time-boxed sprints: Fixed duration, flexible scope within priorities
• Capacity-based: Allocate % of team capacity, deliver what fits
• Gated funding: Release budget phase-by-phase based on results

Change-Order Policies:

• Minor changes (<10% effort): Absorbed within contingency
• Moderate changes (10-25%): Require written approval, may adjust timeline
• Major changes (>25%): Trigger formal change request with cost/schedule impact analysis

SLAs and SLOs:

• Response to client questions: <24 hours
• Defect resolution (critical): <48 hours
• Defect resolution (minor): <2 weeks
• Milestone delivery: On-time or 3-day advance notice of delay

Proposal Templates

Proposal Value Architecture

graph TB
    A[Proposal Foundation] --> B[Business Case]
    A --> C[Technical Approach]
    A --> D[Commercial Structure]
    A --> E[Risk Management]

    B --> B1[Problem Statement]
    B --> B2[Success Metrics]
    B --> B3[ROI Projection]

    C --> C1[Architecture Design]
    C --> C2[Delivery Methodology]
    C --> C3[Technology Stack]

    D --> D1[Pricing Options]
    D --> D2[Payment Terms]
    D --> D3[Value Realization]

    E --> E1[Risk Register]
    E --> E2[Mitigation Plans]
    E --> E3[Contingency Strategy]

    style A fill:#fff3cd
    style B fill:#e1f5ff
    style C fill:#e1f5ff
    style D fill:#d4edda
    style E fill:#f8d7da

Template 1: Executive Summary (1-Page)

Visual Framework:

┌─────────────────────────────────────────────────────────────┐ │ PROJECT: [Name] │ │ CLIENT: [Company] | SPONSOR: [Name, Title] │ │ PREPARED BY: [Your Company] | DATE: [Date] │ ├─────────────────────────────────────────────────────────────┤ │ PROBLEM: │ │ [2-3 sentences describing the business challenge] │ ├─────────────────────────────────────────────────────────────┤ │ SOLUTION: │ │ [2-3 sentences describing the proposed approach] │ ├─────────────────────────────────────────────────────────────┤ │ OUTCOMES: │ │ - [Measurable outcome 1] │ │ - [Measurable outcome 2] │ │ - [Measurable outcome 3] │ ├─────────────────────────────────────────────────────────────┤ │ TIMELINE: [X weeks] | INVESTMENT: $[Range] │ │ NEXT STEPS: [Sign SoW by Date] → [Kickoff Date] │ └─────────────────────────────────────────────────────────────┘

Template 2: Detailed Proposal Deck (15-20 slides)

1. Cover: Project name, logos, date
2. Executive Summary: Problem, solution, outcomes (1 slide)
3. Background: Context from discovery, stakeholder input (2 slides)
4. Objectives: Business goals and success metrics (1 slide)
5. Approach: High-level methodology and architecture (2 slides)
6. Scope: In/out, rationale (2 slides)
7. Deliverables: By phase with acceptance criteria (2 slides)
8. Timeline: Gantt chart or roadmap (1 slide)
9. Team: Roles, responsibilities, governance (1 slide)
10. Risks: Top 5 risks with mitigations (1 slide)
11. Assumptions: Critical dependencies (1 slide)
12. Investment: Options with tradeoffs (1 slide)
13. Next Steps: Decision timeline, kickoff plan (1 slide)
14. Appendix: Detailed estimates, technical notes, case studies (3-5 slides)

Proposal Decision Framework

flowchart TD
    Start{Client Decision<br/>Process}
    Start --> Budget{Budget<br/>Approved?}

    Budget -->|Yes| Options[Review Pricing<br/>Options]
    Budget -->|No| Justify[Build Business<br/>Case]

    Justify --> ROI[Calculate ROI]
    ROI --> Budget

    Options --> MVP{Start with<br/>MVP?}
    MVP -->|Yes| Quick[Quick Win<br/>Approach]
    MVP -->|No| Comprehensive[Full Solution<br/>Approach]

    Quick --> Risk1[Lower Risk]
    Comprehensive --> Risk2[Higher Value]

    Risk1 --> Timeline1[8-10 weeks]
    Risk2 --> Timeline2[16-24 weeks]

    Timeline1 --> Approve{Approve?}
    Timeline2 --> Approve

    Approve -->|Yes| Contract[Sign Contract]
    Approve -->|No| Negotiate[Negotiate Terms]

    Negotiate --> Options
    Contract --> Kickoff[Project Kickoff]

    style Start fill:#fff3cd
    style Budget fill:#e1f5ff
    style Approve fill:#d4edda
    style Contract fill:#d4edda
    style Kickoff fill:#d4edda

Template 3: Statement of Work (SoW)

Structure and Components:

┌─────────────────────────────────────────────────────────────┐ │ STATEMENT OF WORK │ │ [Your Company] and [Client Company] │ ├─────────────────────────────────────────────────────────────┤ │ 1. PROJECT OVERVIEW │ │ 1.1 Background │ │ 1.2 Objectives │ │ 1.3 Success Criteria │ ├─────────────────────────────────────────────────────────────┤ │ 2. SCOPE OF WORK │ │ 2.1 In Scope │ │ 2.2 Out of Scope │ │ 2.3 Deliverables │ ├─────────────────────────────────────────────────────────────┤ │ 3. APPROACH AND METHODOLOGY │ │ 3.1 Phases and Milestones │ │ 3.2 Quality Assurance │ │ 3.3 Change Control │ ├─────────────────────────────────────────────────────────────┤ │ 4. TIMELINE │ │ 4.1 Project Schedule │ │ 4.2 Key Dates │ │ 4.3 Dependencies │ ├─────────────────────────────────────────────────────────────┤ │ 5. TEAM AND GOVERNANCE │ │ 5.1 [Your Company] Team │ │ 5.2 [Client] Team │ │ 5.3 Governance Structure │ │ 5.4 Communication Plan │ ├─────────────────────────────────────────────────────────────┤ │ 6. ASSUMPTIONS AND DEPENDENCIES │ │ 6.1 Assumptions │ │ 6.2 Dependencies │ │ 6.3 Exclusions │ ├─────────────────────────────────────────────────────────────┤ │ 7. ACCEPTANCE CRITERIA │ │ 7.1 Milestone Acceptance │ │ 7.2 Final Acceptance │ │ 7.3 Testing Approach │ ├─────────────────────────────────────────────────────────────┤ │ 8. COMMERCIAL TERMS │ │ 8.1 Fees and Payment Schedule │ │ 8.2 Expenses │ │ 8.3 Change Order Process │ ├─────────────────────────────────────────────────────────────┤ │ 9. INTELLECTUAL PROPERTY │ │ 9.1 Ownership │ │ 9.2 Licenses │ │ 9.3 Third-Party Components │ ├─────────────────────────────────────────────────────────────┤ │ 10. SIGNATURES │ │ [Your Company]: _____________ Date: _______ │ │ [Client Company]: _____________ Date: _______ │ └─────────────────────────────────────────────────────────────┘

Value Realization Timeline

gantt
    title Value Realization and Cash Flow
    dateFormat  YYYY-MM-DD
    section Investment
    Upfront Payment (30%)           :milestone, 2025-01-01, 0d
    Phase 1 Payment (40%)           :milestone, 2025-02-15, 0d
    Final Payment (30%)             :milestone, 2025-03-30, 0d

    section Delivery
    Foundation Phase                :active, 2025-01-01, 6w
    Development Phase               :2025-02-15, 6w
    Deployment Phase                :2025-03-30, 4w

    section Value
    Quick Wins (20% value)          :crit, 2025-02-15, 2w
    Core Value (60% value)          :crit, 2025-03-30, 4w
    Full Value (100%)               :crit, 2025-04-30, 2w
    Continuous Improvement          :2025-05-15, 12w

Business Case Financial Summary

Case Study: Enterprise RAG Implementation

Background: A Fortune 500 financial services company wanted to improve internal knowledge access for 5,000 employees across compliance, legal, and operations teams.

Challenge:

• Unclear scope initially ("make it easier to find information")
• Multiple stakeholders with competing priorities
• Strict regulatory requirements
• Limited AI expertise in-house

Scoping Approach:

1. Discovery (2 weeks): Interviewed 15 stakeholders, analyzed 3 use cases, assessed data landscape

2. Prioritization: Scored use cases on value vs. feasibility; selected compliance Q&A as MVP

3. Options Presented:

   • Option A: Single-use case MVP (10 weeks, $120K)
   • Option B: Two use cases with advanced features (16 weeks, $200K)
   • Option C: Platform approach with extensibility (24 weeks, $350K)

4. Client Decision: Selected Option A with Option B features as Phase 2

Proposal Highlights:

• Clear Scope: 3,000 compliance documents, Q&A via web portal, 95% accuracy target
• Phased Milestones: 4 gates with go/no-go decisions
• Risk Mitigation:
  • Data quality assessed in Week 1 with exit clause
  • Security review started Week 2 to avoid delays
  • Weekly demos to build confidence
• Assumptions:
  • Document access granted within 1 week
  • Compliance SMEs available 4 hours/week
  • Budget for GPT-4 API approved ($2K/month)

Outcomes:

• Approved in 1 week: Clear options and justification
• Zero scope creep: Boundaries respected; new requests logged for Phase 2
• On-time delivery: All milestones met within ±3 days
• Smooth change control: 2 minor changes handled via documented process

Key Success Factors:

• Visual roadmap that executives could understand
• Risk register reviewed weekly with mitigations
• Options gave client control over investment
• Acceptance criteria prevented ambiguity at milestones

Best Practices

Do's

• Start with "why": Connect scope to business outcomes, not just features
• Be specific: "Improve search accuracy to 90%" beats "better search"
• Offer choices: Multiple options show flexibility and let client control risk
• Make risks visible early: Transparency builds trust
• Use visuals: Diagrams, timelines, and tables are easier to digest than text
• Include examples: Show what good looks like (sample queries, mock-ups)
• Plan for learning: AI projects have uncertainty; build in iteration
• Document assumptions: Make the implicit explicit

Don'ts

• Avoid vague language: "Best-in-class," "world-class," "cutting-edge" say nothing
• Don't overpromise: Underpromise and overdeliver is safer in AI
• Don't hide risks: They'll surface anyway; better to own them upfront
• Don't skip acceptance criteria: Leads to endless debates about "done"
• Don't ignore change control: Scope creep kills margins and morale
• Don't propose without discovery: Garbage in, garbage out
• Don't forget governance: Even small projects need decision-making structure

Common Pitfalls and How to Avoid Them

Tools and Templates

ROI Projection Framework

graph TD
    A[ROI Analysis] --> B[Investment]
    A --> C[Returns]
    A --> D[Payback Period]
    A --> E[Risk Adjustment]

    B --> B1["Development: $180K"]
    B --> B2["Infrastructure: $24K/year"]
    B --> B3["Maintenance: $60K/year"]

    C --> C1["Time Savings: $400K/year"]
    C --> C2["Error Reduction: $150K/year"]
    C --> C3["Productivity Gain: $300K/year"]

    D --> D1["Total Investment: $180K"]
    D --> D2["Annual Return: $850K"]
    D --> D3["Payback: 2.5 months"]

    E --> E1["Success Probability: 85%"]
    E --> E2["Expected ROI: 340%"]

    style A fill:#fff3cd
    style B fill:#f8d7da
    style C fill:#d4edda
    style D fill:#e1f5ff
    style E fill:#e1f5ff

Commercial Options Matrix

Scoping Canvas (1-Page Summary)

Visual Template:

┌─────────────────────────────────────────────────────────────┐ │ PROJECT: [Name] DATE: [Date] │ ├─────────────────────────────────────────────────────────────┤ │ PROBLEM │ SOLUTION │ │ [2-3 sentences] │ [High-level approach] │ ├─────────────────────────┼────────────────────────────────────┤ │ IN SCOPE │ OUT OF SCOPE │ │ - [Item 1] │ - [Item 1] │ │ - [Item 2] │ - [Item 2] │ ├─────────────────────────┴────────────────────────────────────┤ │ MILESTONES │ TIMELINE: [X weeks] │ │ M1: [Name] - Week [X] │ INVESTMENT: $[Range] │ │ M2: [Name] - Week [X] │ TEAM: [X FTE] │ ├─────────────────────────┴────────────────────────────────────┤ │ TOP RISKS │ ASSUMPTIONS │ │ 1. [Risk + mitigation] │ 1. [Assumption] │ │ 2. [Risk + mitigation] │ 2. [Assumption] │ ├─────────────────────────┴────────────────────────────────────┤ │ SUCCESS METRICS │ │ - [Metric 1]: [Baseline] → [Target] │ │ - [Metric 2]: [Baseline] → [Target] │ └─────────────────────────────────────────────────────────────┘

RAID Log Template

Effort Estimation Framework

Bottom-Up Estimation Model:

Cost Calculation Model:

┌──────────────────────────────────────────────────────┐ │ COST CALCULATION │ ├──────────────────────────────────────────────────────┤ │ Total Effort: 102-158 days │ │ Blended Rate: $1,800/day │ │ ────────────────────────────────────────────────────│ │ Base Cost:$183,600 - $284,400 │ │ Contingency (10$18,360 - $28,440 │ │ ────────────────────────────────────────────────────│ │ TOTAL PROJECT COST:$202,000 - $313,000 │ │ ────────────────────────────────────────────────────│ │ Recommended Price:$220,000 - $280,000 │ │ (Target Margin: 25-30%) │ └──────────────────────────────────────────────────────┘

Implementation Checklist

Pre-Scoping (Discovery Phase):

• Complete discovery interviews with key stakeholders
• Document business objectives and success criteria
• Assess data landscape and technical environment
• Identify constraints (budget, timeline, compliance)
• Map stakeholders and decision-making process

Initial Scoping:

• Draft problem statement and proposed approach
• Define scope boundaries (in/out/deferred)
• Identify major deliverables and milestones
• List assumptions and dependencies
• Perform rough effort estimation

Detailed Planning:

• Break work into phases with clear acceptance criteria
• Build RAID log (risks, assumptions, issues, dependencies)
• Define team structure and governance model
• Create detailed timeline with dependencies
• Estimate effort and cost with ranges
• Develop 3 pricing options with tradeoffs

Proposal Development:

• Create executive summary (1-page)
• Build detailed proposal deck (15-20 slides)
• Draft Statement of Work (SoW)
• Prepare supporting materials (case studies, technical appendix)
• Review with internal stakeholders (technical, commercial, risk)

Client Validation:

• Schedule proposal review meeting with key stakeholders
• Present proposal and walk through options
• Discuss risks, assumptions, and mitigation strategies
• Negotiate scope, timeline, and budget
• Document decisions and action items
• Address concerns and incorporate feedback

Finalization:

• Update proposal based on client feedback
• Finalize SoW with legal and commercial terms
• Obtain necessary approvals (legal, procurement, finance)
• Execute contract and obtain signatures
• Schedule kickoff meeting
• Set up project infrastructure (communication channels, tools)

Post-Approval:

• Conduct kickoff meeting with full team
• Review scope, timeline, and governance
• Establish communication cadence
• Set up RAID log tracking
• Begin Phase 1 work

Key Takeaways

1. Clarity prevents chaos: Time invested in scoping saves 10x in delivery
2. Options empower clients: Give choices, not ultimatums
3. Risks are features: Transparency about uncertainty builds trust
4. Acceptance criteria = agreement: Define "done" before you start
5. Assumptions need validation: Check them early and often
6. Change control protects everyone: Scope creep hurts client and consultant
7. Governance matters: Even small projects need decision-making structure
8. Visuals communicate better: Diagrams beat paragraphs every time