Mục lục bài viết11 mục
- 1.Step 0: Pick your variant in 2 minutes
- 2.Step 1: Lock numbers and disclosure before writing anything
- 3.Step 2: Draft the executive summary first
- 4.Step 3: Build each section by variant
- 4.1A. Board deck — 12 to 18 slides
- 4.2B. Diligence packet — 12 to 25 pages
- 4.3C. RFP analysis — 4 to 10 pages
- 5.Step 4: Build the evidence spine
- 6.Step 5: Risks, scenarios, and model tie-in
- 7.Step 6: Pre-submission cross-check
- 8.Appendix: prompts ready to paste
Three variants. One evidence standard. No prose filler.
Step 0: Pick your variant in 2 minutes
| Signal | Variant |
|---|---|
| Tomorrow’s board or VP meeting | Board deck |
| Investor checklist with data room | Diligence packet |
| Page limits and required headings from a buyer | RFP analysis |
Deliverable shape:
- Board deck: 12–18 slides, visuals first, 1-page executive summary up front
- Diligence packet: 12–25 pages, narrative + appendix, explicit evidence IDs on every quantitative claim
- RFP analysis: 4–10 pages within page limits, headings mapped 1-to-1 to RFP, compliance matrix inline
Step 1: Lock numbers and disclosure before writing anything
Rework after drafting costs 2–4 hours. Lock these first.
Version register — one line, paste everywhere:
Model v[XX.YY.ZZ], [currency], data through [YYYY-MM-DD]. Revenue retention: NRR by activated-month cohorts. CAC payback: months to gross margin recovery on blended CAC. Pipeline coverage: [X]x next-quarter bookings, win-rate trailing 3 months. Customer identities redacted per NDA; revenue in bands.
Assumption decisions to make now:
- Cohort logic: activated date or first invoice date?
- Retention metric: logo or revenue? Which lookback window?
- CAC: blended or channel-split? Which cost lines included?
- Forecast drivers: pipeline coverage rule, win-rate source, ramp time, seasonality toggle on or off?
- Confidentiality: which customers are under NDA? Use “Top 10 fintech client” format or revenue bands?
Place the finalized version on slide 2 (Board) or page 1 (Diligence/RFP). Cross-reference it by name everywhere else — never restate the definition mid-document.
When the model isn’t ready at deadline:
- Use last locked version with a visible version tag and cut-off date
- Add one line on the assumptions page: “Model pending [specific update]; forecast bridge uses [version] as proxy”
- Flag which numbers will change and by when — do not silently use stale data
Step 2: Draft the executive summary first
Write this before any other section. It forces clarity on what the document is actually arguing.
5-sentence formula:
- Situation and decision window
- Numbers at a glance (3–5 KPIs, actual vs plan)
- What drives the numbers (1–2 root causes)
- Key risks and mitigations
- The ask or next decision with specific owners and dates
Filled example — SaaS payments platform:
We are entering Q3 with a bookings gap against plan and must choose between accelerating mid-market expansion or deepening enterprise focus before tomorrow’s board session. YTD ARR is 18.4M vs plan 19.6M, NRR 116%, CAC payback 13.5 months, gross margin 71%. Shortfall is concentrated in enterprise where sales cycle extended to 142 days and win-rate dipped from 26% to 19%; mid-market self-serve grew 28% QoQ with 43% lower CAC. Primary risks are discount pressure in mid-market and potential churn from two enterprise accounts; mitigations include pricing guardrails, expanded onboarding playbooks, and dedicated save plans. Decision requested: approve Option B to prioritize mid-market for two quarters, reallocate 3 AEs and 600k OPEX, target ARR exit of 25.8M with NRR 118% and CAC payback under 12 months.
For RFP: after sentence 2 or 3, add one clause mapping to a scoring criterion — “addresses Section 3.1 Understanding and Section 3.3 Methodology.”
Step 3: Build each section by variant
A. Board deck — 12 to 18 slides
Slide order is fixed. One idea per slide. Numbers large, captions short.
| Slide | Content | Prompt to use |
|---|---|---|
| 1 | Executive summary | 5-sentence formula above |
| 2 | Assumptions and data scope | Version register from Step 1 |
| 3 | Objectives and constraints | 3 bullets: objective, timeframe, constraints — under 40 words each |
| 4 | Business model snapshot | 4 bullets: revenue streams, segments, cost drivers + 1 diagram caption |
| 5 | Market and customer | 5 bullets: ICP, segments, TAM/SAM method, top personas, 2 recent customer insights with dates |
| 6 | Performance vs plan | 5 KPIs actual vs plan with % variance and 1-line driver each. State cut-off date and source. |
| 7 | Unit economics | 5 bullets: pricing/ARPA, CAC, LTV, contribution margin, sensitivity |
| 8 | Forecast bridge | 5 steps: actuals → adjustments → plan. Every step has a number. |
| 9 | Strategic options | 2–3 options: benefits, costs, time-to-impact, KPI impact range |
| 10 | Risks and mitigations | Top 5 risks: likelihood, impact, mitigation owner |
| 11 | Decision and next steps | Decision required + 3 next actions with owners and dates |
Unit economics slide — filled bullets:
- Pricing: tiered SaaS + 15 bps payment take-rate; blended ARPA 1,460 USD MRR Q2
- CAC: blended 22.8k USD; payback 13.5 months at 71% gross margin
- LTV: 5.9x CAC based on 116% NRR and 1.3% monthly logo churn
- Contribution margin: 54% after variable processing fees and support
- Sensitivity: 2-point NRR change moves LTV:CAC by ~0.5x
Finish Line 2: VP or chair acknowledges in writing the decision option to be brought to vote. Materials include 1-page executive summary + decision slide with owners and dates. If they defer: add forecast bridge and 3 numbers on the decision slide — budget required, expected ARR impact, payback timeline.
B. Diligence packet — 12 to 25 pages
Every quantitative claim gets a Source: E0xx tag. No exceptions.
| Section | Content | Evidence needed |
|---|---|---|
| 1. Company overview | 1-page overview linking business model to financial model tabs and version. List sheet names and what each drives. | Model version tag + tab names |
| 2. Market and competitive | Market size logic in 6 bullets. Top 5 competitors with positioning statements and validation dates. | TAM/SAM sources with dates |
| 3. GTM and pipeline | Pipeline as of [date]: coverage, stages, win-rates, cycle time, 3 anonymized logos with ARR bands and contract dates. | CRM export file + timestamp |
| 4. Cohort and retention | 12-month revenue retention, logo churn, expansion drivers, 2 cohort charts. | Cohort export with filter logic |
| 5. Unit economics | CAC by channel, gross margin by product, payback, LTV:CAC with definitions. Replication notes included. | Defined per assumptions page |
| 6. Financial model | Revenue build logic, key assumptions, 2 bridges: bookings-to-revenue and actuals-to-forecast. Cite cell ranges or tabs. | Model version + cell refs |
| 7. Legal and compliance | Key contracts, regulatory scope, certifications, open issues with owners and target close dates. | Legal sign-off on disclosure |
| 8. Risks and mitigations | Top 8 risks: likelihood, impact, mitigations cross-referenced to model sensitivities. | Model sensitivity tabs |
| 9. Evidence index | ID, title, file path, timestamp, privacy tag, replication steps for 3 critical charts. | See evidence format below |
DD checklist → section mapping:
| Typical DD item | Section |
|---|---|
| TAM/SAM/SOM with method | 2 Market |
| Churn, retention, cohorts | 4 Cohort |
| CAC, LTV, gross margin | 5 Unit economics |
| Pipeline, win-rate, ARR bands | 3 GTM |
| Model assumptions and bridges | 6 Financial model |
| Contracts and compliance | 7 Legal |
| Risk register | 8 Risks |
| Data room and evidence | 9 Evidence index |
Finish Line 2: diligence lead emails that checklist item is satisfied with no outstanding evidence requests and all claims cited by evidence IDs. If they pause: check that risks are not downplayed and KPI definitions match the assumptions page exactly.
C. RFP analysis — 4 to 10 pages
Before writing: read the scoring rubric first.
Most RFPs publish scoring weights by section. Before allocating word count, rank sections by weight:
- Highest weight (often Methodology or Technical Approach): maximum detail, specific proof points, measurable outputs
- Mid weight (Staffing, Implementation): concise, owners named, timelines realistic
- Lowest weight (Understanding, Compliance): 1–2 paragraphs, confirm and move on
Do not write equal-length answers for unequal-weight sections. Evaluators score what they weight.
Heading rule: use the RFP’s exact heading text and numbering. Evaluators score against a rubric that matches headings — a renamed section may score zero even if the content is correct.
| RFP section | Content | Length guidance |
|---|---|---|
| [3.1] Understanding of requirements | Restate problem in buyer’s terms. 5 bullets. Map to a KPI the buyer cares about. | Proportional to weight |
| [3.3] Approach and methodology | 6–8 steps, 1 line each: tool, timeline, measurable output | Highest weight section — expand here |
| [3.4] Implementation and staffing | Phased plan with owners and durations. Named individuals if allowed. | Mid weight |
| [3.5] Value and differentiation | 4 differentiators, each tied to buyer outcome + referenceable proof point with date | High weight if scored |
| [3.6] Risk management | Top 5 risks: mitigations + early-warning indicators | Mid weight |
| [3.7] Assumptions and dependencies | What you need from buyer. What is out of scope. | Low weight — be brief |
| [3.8] Compliance statement | Declare compliance. List exceptions. Point to section + exhibit for each requirement. | Required — cannot skip |
Value and differentiation — filled bullets:
- Payment success optimization improves approved transaction rate from 93.2% to 96.1% within 30 days using adaptive retries and issuer routing; evidenced by 280 bps uplift for a mid-market subscription customer Q1 2026
- Go-live in 15 business days with prebuilt connectors to Shopify and Chargebee; last 3 implementations averaged 12 business days from SOW to first transaction
- Cost visibility via margin dashboard surfacing scheme fees and network costs; finance teams received weekly automated cost variance reports Q2 2026
- 24×7 merchant support, SLA first response under 5 minutes chat / 30 minutes email for P1; April–June 2026 average 3 minutes chat first response
Finish Line 2: internal approver confirms every required heading present, within page limits, mapped to scoring criteria. If evaluators score low on methodology: each step needs a measurable output, not just a description of activity.
Step 4: Build the evidence spine
Attach to every quantitative claim. Do this as you write — retrofitting takes twice as long.
Evidence ID format:
E[###]: [filename].[ext] — generated [YYYY-MM-DD HH:MM UTC] — [Internal only / Restricted / Public] — Source: [system/view/report] — Filter: [applied filters]
Three entries ready to copy and adapt:
E014: Revenue_retention_cohort_export.csv — 2026-06-25 19:15 UTC — Internal only — Snowflake view bi.cohort_rr — Filter: activated_date <= 2025-06-30
E022: CRM_pipeline_snapshot_2026-06-24.xlsx — Internal only — Salesforce report Q2 Pipeline by Stage — Snapshot 2026-06-24
E031: Pricing_policy_v2026-03.pdf — Restricted — Legal-approved pricing
Replication notes for critical charts (required for Diligence): For each chart, record: data export used, filters applied, computation steps, and who can re-run it. One paragraph per chart is enough. The test: can someone else reproduce the chart from your notes without asking you a question?
Evidence spine failure signals:
- Reused charts with no export timestamps
- KPI definitions differ between sections
- Sources missing privacy tags or file paths
- RFP answers claim compliance without pointing to exact section or exhibit
Step 5: Risks, scenarios, and model tie-in
Scenario setup — 3 columns, fill for your context:
| Driver | Downside | Base | Upside |
|---|---|---|---|
| Win-rate | 19% | 22% | 26% |
| Sales cycle | 142 days | 128 days | 115 days |
| CAC payback | 14.2 months | 12.8 months | 11.6 months |
| ARR exit | 24.0M | 25.1M | 26.6M |
Early-warning triggers (make these measurable):
- If monthly logo churn exceeds 1.6% for 2 consecutive months → invoke hiring freeze for enterprise, reallocate 2 AEs to mid-market
- If pipeline coverage falls below 2.5x → trigger weekly pipeline review with CRO
- If enterprise cycle time exceeds 150 days for 2 months → switch to Option B
Model tie-in — cite exactly:
Downside scenario: Tab GTM_Drivers, cells C18:C22, win-rate adjusted from 26% to 21%. ARR impact: Tab Revenue_Build, cell F44.
Step 6: Pre-submission cross-check
Run this before calling it done. Each item is a yes/no.
Numbers consistency:
- [ ] All KPIs match model version stated on assumptions page
- [ ] KPI definitions are identical across all sections
- [ ] Forecast bridge reconciles actuals to plan with explicit line-item drivers
- [ ] Cut-off dates stated on every data exhibit
Evidence completeness:
- [ ] Every quantitative claim has a
Source: E0xxtag - [ ] Evidence index includes file path, timestamp, and privacy tag for each ID
- [ ] Replication notes written for at least 3 critical charts (Diligence) or key proof points (RFP)
- [ ] Redactions applied and noted where NDA applies
Variant-specific:
Board:
- [ ] Decision slide has 3 numbers: budget required, expected ARR impact, payback timeline
- [ ] 1-page executive summary matches slide 11 decision exactly
Diligence:
- [ ] DD checklist mapped — no open items without a documented constraint and timeline
- [ ] Risks section has likelihood, impact, and mitigation for each item
RFP:
- [ ] Headings match RFP word-for-word and in order
- [ ] Page or word limits respected — checked against highest-weight sections first
- [ ] Compliance statement points to exact section and exhibit for each requirement
- [ ] Each differentiator tied to a buyer outcome and a dated proof point
Common failure modes to catch in this pass:
- Boilerplate claims without a proof point or date
- Contradictions between narrative and model assumptions
- Different timeframes used across charts without a note explaining it
- Claims that cannot be disclosed under current NDAs but weren’t redacted
Appendix: prompts ready to paste
Executive summary — paste with your numbers:
Write a 5-sentence executive summary for a board deck. Model v23.06.27, data through 2026-06-25. KPIs: ARR 18.4M vs plan 19.6M, NRR 116%, CAC payback 13.5 months, gross margin 71%. Options: A — double down on enterprise; B — prioritize mid-market for 2 quarters. Risks: discount pressure mid-market, churn risk 2 enterprise accounts. Ask: approve Option B, reallocate 3 AEs and 600k OPEX, target ARR exit 25.8M, NRR 118%, payback under 12 months.
Diligence cohort section — variable template:
Using E014 cohort export and definitions on the Assumptions page, summarize 12-month revenue retention, logo churn, and top 2 expansion drivers. Provide 2 bullets on methodology. Add Source: E0xx tag after each claim.
RFP methodology section — variable template:
Under heading [3.3] Approach and Methodology, describe an 8-step plan to achieve the buyer’s objective of [specific outcome] within [timeframe]. For each step: name the tool or method, state the timeline, and state the measurable output. Map each step to RFP Section 3.3 scoring notes. Keep under [word limit].
Unit economics section — variable template:
Summarize unit economics using this data: [ARPA], [blended CAC], [gross margin %], [NRR %], [logo churn %]. Compute LTV:CAC ratio. Show a sensitivity: how does LTV:CAC move if NRR changes by 2 points? Format as 5 bullets. State data source and cut-off date.