VC Due Diligence Simulator — 从 VC 视角模拟 DD 过程

定位： - 不是「创始人应该准备什么材料」（那是 DD checklist） - 而是「VC 分析师在 3 周内到底会挖哪些问题、call 哪些 reference、审哪些数据、从哪个角度 break 这个 deal」 - 创始人用这个 simulator 做 "Red Team"：先让自己被虐，再补足弹药去 pitch

姊妹 skills：

• founder-fundraising-wargame：融资整体战略
• term-sheet-negotiator：TS 条款博弈（post-DD）
• vc-pitch-killtest：pre-pitch 自杀测试

核心认知：DD 不是 formality，是 kill test

创始人的 misconception：「TS 签了就基本是 done」 现实：20-40% of signed TS fail during DD（根据 PitchBook 2023-2024 数据）

DD failure 的三大 root cause：

1. 创始人讲的 story 和数据对不上（TAM 吹大、retention 美化、churn 隐藏）
2. Reference call 暴露 founder 或员工 red flag
3. Technical / legal 发现之前不 visible 的 skeleton

DD 是 VC 在「confirm 自己 thesis」和「find an excuse to back out」之间摆动。每个 DD 信号都有权重——signal strong enough 就 close、signal weak 就 walk。

3-Week DD Timeline 解构

Week 1：Structured Questions + Data Room

VC 目标：建立数据 baseline，对比 pitch 时的声明

具体行动：

• 发 24-48 个 structured questions（见下文）
• 请求 data room access（所有 legal docs、financial records、employee list、customer metrics）
• 分析师独立核算 unit economics（不信你的 LTV/CAC 数字，重算）
• Run 公开检查（litigation history、domain registration、trademark search）

创始人准备：

• Data room 必须 already 100% ready（如果你 Week 1 才开始 scramble 组织，VC signal -10%）
• Structured questions 要在 48 小时内返回（delay > 72 小时 = signal weak）
• 所有数据三角验证（GA4 matches Stripe matches Mixpanel）

Week 2：Reference Calls

VC 目标：triangulate founder + team + product claims，识别 red flag

具体行动：

• Founder reference calls（3-5 个）：past employer、co-founder、mentor
• Customer reference calls（3-5 个）：paying customers、churned customers
• Employee calls（如果有 exits）：前员工
• Competitor intelligence（lightweight）：找 competitor 的 investor 或员工 chat
• Back-channel calls（你不知道的那种）：VC 的 network 里找 6-7 个 random contacts who might know you
• Domain experts（技术 / market / regulatory）

创始人准备：

• 提前 prime references（告诉他们 VC 可能来 call，what deal）
• Prepare list of "inviters" references（happy customers）和 "tough" references（hard but fair customers）
• 对 churn customers，提供 context（why they churned, what's different now）
• Key：VC 的 back-channel calls 是 wild card——你无法控制。这就是为什么你过去的行为 compounds，现在到 founder 阶段才想修复 reputation 来不及

Week 3：Deep Dives + Closing Checklist

VC 目标：针对 Week 1-2 发现的 issues 做 deep dive，最终形成 investment memo

具体行动：

• Technical deep dive（CTO 和 VC 带来的 technical advisor / operating partner 深聊）
• Financial audit-style review（accountant 或 operator 帮 VC 看数据）
• Legal review：cap table、IP assignment、customer contracts、employee agreements
• Final founder meeting：2-3 小时，基于 Week 1-2 findings 深 probe
• IC meeting：VC 内部投委会 approval

创始人准备：

• 任何 Week 1-2 提出的 issue 在 Week 3 必须有 response plan
• Final meeting 要带上 cofounder + 1-2 key team members（让 VC 看 team depth）
• 不要 over-commit 在 closing timeline（保留灵活空间）

VC 会问的 24 个刁钻问题（分 8 类）

Class 1：Market / TAM Kill Questions

Q1：「Walk me through your top-down TAM calculation. Then show me the bottom-up. Why do they differ?」

• Trap：创始人的 top-down 通常用 industry report 数字（膨胀）；bottom-up 更 reliable
• Answer：承认 gap，explain why bottom-up is more reliable，show the delta analysis

Q2：「If your thesis is right, what's the 10-year TAM? What's the serviceable TAM? What's the realistic obtainable TAM for you?」

• Trap：大部分创始人 only know total TAM
• Answer：有明确的 TAM → SAM → SOM 框架 + 数据支持

Q3：「Who is NOT your customer? And why?」

• Trap：创始人说 "everyone is our customer"（表示 PMF 弱）
• Answer：清晰的 exclusion criteria + 为什么这些 segment 不是 fit

Class 2：Customer / Product Kill Questions

Q4：「Please share 3 customer interviews — both positive and negative. Raw transcripts or recordings」

• Trap：创始人只 share 正面 feedback
• Answer：提前 prepare 3 balanced interviews (positive + cautionary + churn)

Q5：「Your churned customers — have you talked to them post-churn? What did they tell you?」

• Trap：创始人没做过 churn post-mortem
• Answer：systematic churn interview program + learnings

Q6：「Show me your top 10 customers' actual usage data, not self-reported NPS」

• Trap：NPS 可能被 biased（只看 engaged users）
• Answer：Product analytics data show real usage depth + frequency

Class 3：Unit Economics Kill Questions

Q7：「Your CAC payback — walk me through the calculation. What discount rate? Does it include ramp time? Free trial cost?」

• Trap：简化的 CAC payback 通常低估真实成本
• Answer：fully-loaded CAC（包含 sales 招聘、training、tools）+ proper discount rate

Q8：「LTV — what retention assumption are you using? What's the actual churn cohort data?」

• Trap：LTV 常用 aggressive retention assumption
• Answer：Cohort-based LTV calculation + sensitivity analysis

Q9：「Gross margin breakdown — what goes into COGS? Hosting, payment processing, support, customer success, what else?」

• Trap：创始人算 gross margin 时忽略 customer success (technical 属于 COGS or OpEx)
• Answer：Transparent breakdown + justify each category

Class 4：Competition / Moat Kill Questions

Q10：「If [Incumbent] decided to build this tomorrow with their distribution advantage, what's your defense?」

• Trap：创始人 handwave「they're too slow」
• Answer：Specific structural advantages (data network effect, specialized talent, timing, regulatory moat)

Q11：「Your competitor [X] raised [Y] 6 months ago. They have [Z] employees. Why do you win?」

• Trap：创始人 say "we're better"（太主观）
• Answer：Specific differentiation in 2-3 dimensions + evidence

Q12：「What's the best piece of evidence that your competitors are failing or will fail?」

• Trap：创始人没做 competitor intelligence
• Answer：Specific signals (review trends, churn customers' feedback, competitive wins, talent movements)

Class 5：Team / Founder Kill Questions

Q13：「Tell me about a decision you made 12 months ago you now regret. What did you learn?」

• Trap：创始人 perfectionist trap——say "I don't have regrets"
• Answer：Specific, significant regret + concrete learning + how it changed behavior

Q14：「Your last founder you worked with — walk me through why that ended. Who called it? What's your relationship with them now?」

• Trap：模糊回答 = 可能有 toxic pattern
• Answer：clear story，own responsibility，relationship status

Q15：「Your co-founder — where do you two disagree? Tell me a real conflict, how you resolved it」

• Trap：创始人 say "we never disagree"（假的，或 dysfunctional）
• Answer：Real conflict + resolution process + what you learned about working together

Class 6：Strategy / Execution Kill Questions

Q16：「What are the 3 things that would have to go right for you to hit your 5-year plan? Rank them by risk」

• Trap：创始人 only list things they control
• Answer：Include external dependencies (market timing, regulatory, partner behaviors)

Q17：「If I gave you $50M today instead of the $10M you're asking — what would you do with the extra $40M?」

• Trap：创始人 struggle to answer (意味着他们 asking amount wrong)
• Answer：Specific playbook for bigger check + why not asking for it now

Q18：「If I gave you $2M today instead of the $10M you're asking — what would change?」

• Trap：创始人 say "we wouldn't hit milestones"（implies 10M is just-right，no flexibility）
• Answer：Show how plan adapts to smaller check + what milestones shift

Class 7：Financials / Fraud Kill Questions

Q19：「Your revenue recognition policy — walk me through. Do you book multi-year deals upfront or ratably?」

• Trap：创始人 loose with revenue recognition 可能 inflated ARR
• Answer：Clear policy aligned with GAAP (or equivalent) + applied consistently

Q20：「What percentage of your ARR is from customers with concentration risk? Top 3 customers = % of ARR」

• Trap：高 concentration = fragile
• Answer：Transparent + 如果有，show diversification plan

Q21：「Your burn rate — monthly? Walk me through the gross vs net burn. What's your runway without revenue growth?」

• Trap：创始人 optimistic about revenue offset
• Answer：Conservative: net burn with current revenue, no growth assumption, clear runway months

Class 8：Existential Kill Questions

Q22：「In the worst-case scenario where this doesn't work — how does it fail? What's the most likely failure mode?」

• Trap：创始人 refuse 谈 failure (signals overconfidence)
• Answer：Specific failure modes + monitored signals + pivot triggers

Q23：「If you had to shut down tomorrow, what do you think your investors would say about you 5 years later?」

• Trap：Pass the test of "would they back me again"
• Answer：Honest self-assessment + relationship-building practice

Q24：「Why are you doing this? What's the personal reason you'd keep going when every rational person would quit?」

• Trap：创始人 give PR answer (signals lack of real motivation)
• Answer：Personal narrative that is emotionally real, not performative

Reference Call 脚本（VC 会问 reference 什么）

创始人需要 prime reference，让他们有 meaningful answers。

VC 给前员工打电话

VC 的典型问题：

1. 「Tell me about working with [founder]. What was great? What was hard?」
2. 「If you had to point out the founder's weakness / blind spot, what would it be?」
3. 「Why did you leave? Honestly.」
4. 「Would you work for this founder again? Why / why not?」
5. 「What's the craziest thing you saw the founder do during difficult times?」

你如何 prime 前员工（不是 coach them, 是 give context）：

• 告知 raising，important to founder career
• Ask "can I list you as reference"
• Not "please say nice things" — but "be honest, I can take it"

VC 给 customer 打电话

VC 的典型问题：

1. 「How did you find [company]?」
2. 「What was the purchase decision process? Who was involved?」
3. 「What are the 2-3 things that make this product indispensable to you?」
4. 「When have you been frustrated? How did they respond?」
5. 「If [company] went away tomorrow, what would you do? Go back to [old solution] or find alternative?」
6. 「What competitors did you evaluate?」

创始人准备：

• Select references strategically — mix of enthusiastic + pragmatic + critical (critical but fair)
• Pre-call customers to say "VC may reach out, it's OK to share your honest experience"
• Prepare a backup list (if first list doesn't respond)

VC 给 industry expert 打电话

VC 的典型问题：

1. 「Given this thesis, who are the top 3-5 companies we should look at?」
2. 「[This founder] — what's your take? Have you heard of them?」
3. 「What's the most contrarian thing someone could do in this space?」
4. 「Who's going to win this market in 5 years?」
5. 「What are the technical / market / regulatory risks we should worry about?」

创始人影响：

• Industry experts should know you (positive signal if they've heard of you)
• If they're key influencers, get an intro to them before fundraising
• Proactively engage experts in thought leadership (podcasts, conferences, research) — becomes reference by default

Data Room 结构 —— VC 期望看什么

Tier 1：必须完美 (VC first look)

• Cap table（过去 + 当前 + 融后）
• Pitch deck 最新版
• Financial statements (past + forecast)
• Unit economics breakdown
• Customer / revenue breakdown

Tier 2：TS 签后 visible

• Employee agreements + IP assignments
• Customer contracts（redacted top 10）
• Vendor contracts
• Cofounder agreement / equity agreements
• Board materials from past meetings

Tier 3：Pre-close deep dive

• Detailed product roadmap
• Security / privacy audits
• Insurance policies
• Past investor communications
• Tax filings (for older companies)

Red Flag Detection — VC 会 flag 的 10 个信号

Red Flag 1：Revenue concentration

• Top 1 customer > 30% ARR
• Top 3 customer > 50% ARR
• How to mitigate：Show diversification pipeline + long-term contracts

Red Flag 2：Churn pattern

• Churn rate > 25% annually (SaaS)
• Churn concentrated in one segment (shows PMF problem)
• Churn accelerating, not decelerating
• How to mitigate：Explain churn composition + recovery plan + new cohort improvement

Red Flag 3：Founder team dynamics

• Cofounder already left
• Equity imbalance (one founder >> other)
• Unresolved founder disputes
• How to mitigate：Clean, fair equity structure + healthy dynamics + documented agreements

Red Flag 4：IP / Legal issues

• IP disputes or unclear ownership
• Pending litigation
• Regulatory investigations
• How to mitigate：Clean IP assignments + proactive legal audit + transparent disclosure

Red Flag 5：Financial irregularities

• Cash / revenue mismatch
• Related party transactions
• Inconsistent accounting policies
• How to mitigate：Clean accounting + optional CFO/controller review + transparent disclosure

Red Flag 6：Market timing concerns

• Industry in cyclical downturn
• Regulatory risks increasing
• Disruption from AI/new tech
• How to mitigate：Frame as opportunity + show specific positioning for changes

Red Flag 7：Technical debt

• Product built on legacy stack
• Performance / reliability issues
• Security concerns
• How to mitigate：Specific roadmap + hire plan + acknowledge trade-offs made

Red Flag 8：Team gaps

• Missing key roles (CFO / VP Sales / VP Eng)
• Over-reliance on founders
• High turnover in senior roles
• How to mitigate：Specific hiring plan with timeline + references for planned roles

Red Flag 9：Go-to-Market uncertainty

• Unclear primary channel
• Unsustainable CAC in primary channel
• Dependency on one marketing tactic
• How to mitigate：Multi-channel strategy + experiments + pipeline validation

Red Flag 10：Founder behavioral signals

• Evasive answers
• Different story for different audiences
• Overly aggressive tone
• Unable to acknowledge weakness
• How to mitigate：Practice transparency + own mistakes + build trust slowly

DD Execution — 4 Phase

Phase A：Before DD Starts (TS signing day to Week 1)

创始人行动：

• Immediately prepare data room structure
• Pre-call potential references
• Schedule all required founder time for DD
• Clean up any loose ends (legal docs, cap table, etc.)

Phase B：Week 1 — Structured Questions

• Return questions within 48 hours
• Triple-check all numbers
• Flag any issues proactively
• Request weekly sync with VC to maintain momentum

Phase C：Week 2 — Reference / Calls

• Be transparent about references
• Don't over-prime (VC can tell)
• Offer tough references alongside enthusiastic ones
• Be available for follow-up

Phase D：Week 3 — Closing Sprint

• Legal review in parallel
• Founder availability for deep dives
• Be prepared for any late-breaking questions
• Set clear expectations on closing timeline

Post-DD Red Team Debrief

After DD, whether it closes or not, do a red team:

1. What did VC focus on that I didn't prepare well for?
2. What signal was weakest in my DD?
3. What would I have changed in my pitch based on DD questions?
4. What references / materials should I upgrade for next raise?

File output：dd-debrief-[vc-name]-[date].md

执行指令（给 LLM 代理）

使用此 skill 的 prompt 模板：

我是 [公司 X]，准备进入 VC 的 DD 阶段。帮我：

1. 基于我的业务类型 [B2B SaaS / AI / ...]，预测 VC 最可能 probe 的 8 个刁钻问题（从 24 题库中选 + 定制）
2. 给我 3 组 reference call 脚本（前员工 / customer / industry expert），我需要 prime 什么
3. 根据我的情况，识别最可能的 3 个 red flag，给我 mitigation 策略
4. 给我一个 3-week DD timeline，具体每天应该做什么
5. 如果我 DD 可能 fail，最可能的 failure mode 是什么，如何 pre-mortem

关键信息：
- 业务：[describe]
- 关键 metrics：[share]
- 已签 TS：[VC name, terms]
- 特殊情况：[any issues that may surface in DD]

与其他 skills 的协作

• 前置：vc-pitch-killtest（已经 kill test 过 idea）、founder-fundraising-wargame（已经 understand 博弈）
• 并行：term-sheet-negotiator（DD 期间同时在 negotiate TS）
• 后置：board-dynamics-wargame（DD 通过 close 之后 board 管理）

校准锚点 (V1.0)

本 skill 的设计假设：

• 2026 年 VC 行为：更严格 DD（post-FTX/Theranos），Tier-1 VC DD period 3-4 周，Tier-2/3 更短
• DD Fail Rate：TS 签后 20-40% fail during DD
• Red Flags 严格度：2023+ LP 压力下 VC 对 red flags 更 unforgiving

changelog

• v1.0 2026-04-17：首版。提出「DD = kill test，不是 formality」核心认知、3-week timeline 解构、24 个刁钻问题（分 8 类）、3 组 reference call 脚本、10 个 red flag 识别 + 缓解、4 阶段 DD 执行流程、执行指令。与姊妹 skill 差异化定位。卫星文件 dd-casebook.md 待后续完成。