diligence-ddr

Due Diligence Request List (DDR) Skill

Generates or customizes comprehensive, investment-grade DDRs for PE buyout and M&A sell-side transactions. Output should feel like it was written by an experienced deal team — precise, specific to the business model, and structured to surface real investment risks.

---

Step 1: Gather Context

Before writing anything, collect the following. If the user has already provided some of this, extract it from context and only ask for what's missing.

Required:

• Company name and one-line description (what do they do, who do they sell to)
• Business model (SaaS, marketplace, services, product, hybrid — subscription vs usage vs transactional)
• Deal type (PE buyout vs M&A sell-side) — this affects framing and emphasis
• Sector (e.g., B2B SaaS, retail tech, healthcare, fintech, logistics) — drives which risk areas matter most

Helpful but optional:

• Revenue range / scale (e.g., "$50M ARR" or "pre-revenue")
• Known deal issues or investment thesis (e.g., "platform roll-up", "international expansion story", "margin improvement")
• An existing DDR to customize rather than build from scratch
• Any specific sections to prioritize or deprioritize

If the user provides a document (DDR, CIM, pitch deck), read it first to extract context before asking questions.

---

Step 2: Choose Mode

Mode A — Generate from Scratch User provides company/deal context. Build a full DDR tailored to the business model and sector.

Mode B — Customize Existing DDR User provides an existing DDR. Tailor every section to the specific company: replace generic placeholders with company-specific terminology, add business-model-specific requests, remove irrelevant sections, and sharpen questions to reflect known deal risks.

---

Step 3: Structure and Output

DDR Structure

A high-quality DDR has 8–12 numbered sections. Always include the core sections below, and add sector-specific sections as appropriate (see references/).

Universal Core Sections (always include):

1. Financial Performance, Planning, and Quality of Earnings
2. Revenue Model and Pricing Strategy
3. Customer Base, Scale Metrics, and Cohort Health
4. Go-to-Market Engine (Sales cycle, pipeline, win/loss, land & expand)
5. Product Strategy and Roadmap
6. Technology, Infrastructure, and Data
7. Customer Success, Implementation, and Operating Model
8. Strategic Positioning and Expansion
9. Key Risks, Compliance, and Legal
10. Diligence Materials / Data Room Request (always last — the "what to send us" section)

Sector-specific additions: See references/sector-modules.md for add-on sections by vertical (e.g., fraud/risk for fintech, carrier/logistics for supply chain, clinical/regulatory for healthcare).

Output Format

• Use numbered top-level sections with descriptive headers
• Use a subtitle under each header that captures the "so what" (e.g., ## 3. Customer Base, Scale Metrics, and Cohort Health → subtitle: Logos, Volumes, and Retention)
• Each section contains bullet-point questions — typically 6–12 per section
• Questions should be specific, operational, and multi-part — not generic ("describe your pricing") but precise ("provide the distribution of ACV by segment, including median, quartiles, and top-decile values, and describe how pricing has evolved over the last 24 months")
• The Data Room section (always last) should be a concrete artifact request list — name specific files, views, and data exports needed
• Include a header block: Company name, "Due Diligence Request List", subtitle "Comprehensive Information Request | Confidential", and a CONFIDENTIAL notice

Writing-Style Review — Narrative Content Only

Before generating any code, run the writing-style self-review on the DDR's narrative elements: the cover block text, any section framing sentences, and any introductory language above question lists.

Does NOT apply to: Bullet questions, sub-questions, or the data room request list. These are operational requests, not prose — applying prose standards to them would make them worse, not better.

Applies to: Cover subtitle text, any section-level framing paragraph if used, and any executive summary or context paragraph added at the top of the document.

Read: /mnt/skills/user/writing-style/SKILL.md

Quality Bar

Calibrate to the Narvar DDR standard:

• Depth: Questions anticipate second-order issues. Don't just ask "what is churn" — ask for GLR/GRR/NRR by cohort, vintage curves, and the operational definition of churn
• Specificity: Name the company's actual products, segments, or business model constructs in the questions wherever possible
• Coverage: Every major value driver and risk area has dedicated questions
• Actionability: Each question maps to something a deal team can actually get — a file, a metric, a table, a management answer

---

Step 4: Sector and Deal-Type Adjustments

PE Buyout vs. M&A Sell-Side

Dimension	PE Buyout	M&A Sell-Side
Emphasis	Operational detail, cost structure, management depth	Strategic fit, revenue quality, integration complexity
Financials	QoE focus, EBITDA bridge, working capital	Revenue recognition, ARR bridge, customer concentration
Forward look	LRP, hiring plan, margin expansion levers	Pipeline quality, NRR trajectory, platform scalability
Risk focus	Key person, retention, leverage capacity	Synergy assumptions, platform dependency, data portability

Adjust question framing and emphasis based on deal type — but the overall structure stays the same.

Business Model Adjustments

SaaS / subscription:

• Heavy emphasis on ARR metrics (NRR, GRR, cohorts, expansion)
• Pricing architecture, contract terms, renewal mechanics
• Product adoption → retention correlation

Usage-based / transactional:

• Volume trends and predictability
• Take rate / monetization efficiency
• Customer commitment levels vs. actual consumption

Services / implementation-heavy:

• Services margin and whether it's a profit center or deployment cost
• Utilization rates, blended billing rates, project profitability
• Dependency between services and software retention

Marketplace / network:

• Supply and demand-side metrics separately
• Take rate trends and competitive pressure
• Network effects evidence and defensibility

---

Step 5: Delivery — Formatting and Output

Deliver the DDR as a Pattern-branded Word document (.docx).

Before writing any code, read both skill files:

Read: /mnt/skills/user/pattern-docx/SKILL.md
Read: {SKILL_DIR}/references/narvar-example.md

The narvar-example.md is the canonical formatting reference. It specifies exact font sizes, colors, spacing, bullet levels, and the orange highlight convention. Do not deviate from it.

Document Formatting Spec

Cover block (Normal paragraphs, run-level formatting — not heading styles):

[Company Name]          36pt, SemiBold, color 1F4E79
Due Diligence Request List    20pt, color 2E75B6
Comprehensive Information Request | Confidential    italic, color 595959
[blank line]
CONFIDENTIAL — FOR AUTHORIZED RECIPIENTS ONLY    9pt, Bold, color C00000

Page layout:

• Size: 8.5" × 11" (US Letter)
• Margins: top=1.0in, left=0.88in, right=0.88in, bottom=0.88in
• Note: DDR uses 0.88in side margins (not the standard pattern-docx 0.5in)

H1 — Section headers:

• 18pt, Bold, color 1F4E79, space_before=18pt, space_after=8pt
• Numbered inline: 1. Section Name
• Page break before: true (except first H1)

H2 — Section subtitles:

• 13pt, Bold, color 2E75B6, space_before=14pt, space_after=6pt
• Descriptive label (3–5 words), not a question

List Paragraph — bullet questions:

• Bullet character (•) at ilvl=0 for top-level questions
• Decimal numbering at ilvl=1 for sub-questions within a question
• space_before=3pt, space_after=3pt

Orange highlight (E97132):

• Apply to 2–4 priority/thesis-critical questions maximum
• Applied at run level — entire question text in orange
• Signals highest-priority diligence items to management

Header and footer:

• Must be injected via the pattern-docx Python patch script (asset pipeline)
• Do NOT attempt to write header/footer in docx-js
• Header: [Company Name] — Due Diligence Request List (left, gray) + CONFIDENTIAL (right, Bold, red)
• Footer: [Company Name] Due Diligence (left) + Page X of Y (right)
• Follow the exact pipeline in pattern-docx SKILL.md Step 3

After Delivering the File

• Note any sections where significant company-specific customization was applied
• Flag the 2–4 orange-highlighted questions and explain why they were prioritized
• Offer to go deeper on any section or add a sector-specific module
• Immediately run doc-quality-checker — do not wait to be asked. State "Running quality check..." and produce the full issue report. If zero issues, state that explicitly.

---

Reference Files

• references/sector-modules.md — Add-on DDR sections by vertical (fintech/fraud, logistics/supply chain, healthcare, consumer/retail)
• references/narvar-example.md — Canonical formatting reference and quality calibration standard. Load before generating any DDR.