制度 · 永续 · 机构深度档案

Sequoia Capital

Founded1972
LeadershipValentine → Botha → Lin & Grady
Last Updated2026-05-29(2026-05-29 完整性扫描 Sweep:折入 Turpentine VC/Alfred Lin、Peak XV/Pat Grady、Sourcery/Alfred Lin、StrictlyVC/Botha;consolidation fold 再折入 Axios/Botha-Primack;1 期 partner-side 内容仍 pending)
Web Sources13
Podcast Sources27

范围说明(务必先读):本档案的研究对象是机构,不是某一个人。重点放在偏软性的组织维度——治理、文化、对内部投资人的要求与评估、传承、审美。需要警惕的是:Sequoia 是出了名"会讲自己故事"的机构,其官网 "Our Ethos"、"Crucible Moments" 播客都带有明显的自我塑造(self-mythology)成分。因此全文严格区分三类信息:事实(有来源)、推断(标 [推断])、PR/自我叙事(标 ⚠️)。最厚的一手材料来自合伙人在 on-record 长访谈里的表述,以及 Sebastian Mallaby 的《The Power Law》。引用尽量保留逐字英文原文(英文在上、中文紧随、来源在末)。
Part 01

摘要

Sequoia Capital 由 Don Valentine1972 年创立,是风险投资史上唯一一家成功完成多代领导权交接的合伙制机构:Valentine →(1996)Michael Moritz & Doug Leone →(2022)Roelof Botha →(2025-11,据报道)Alfred Lin & Pat Grady。对一个想"将来自己管一支基金"的人来说,Sequoia 最值得研究的不是它投中了 Apple/Google/Stripe,而是它把自己设计成了一台反人性弱点的制度机器

  • 治理上,它把合伙企业改造成"永续私有合伙"(private partnership in perpetuity)——这一代不向上一代买断、也不向下一代收费,从结构上消灭了"把基金卖掉套现"的可能,强制了 stewardship(守护者)而非自肥;
  • 决策上,一套反羊群效应的流水线(周五备忘录 → 周一会 → 匿名 0–10 预投票、禁投 5 分、指定 devil's advocate、资深者最后发言);
  • 评估上,是 process-over-outcome(看过程不看单次结果):没人会因一笔失败投资被开除、也没人因一笔成功被表彰;
  • 经济上,2021 年首创常青结构(The Sequoia Capital Fund),2023 年把全球版图拆成三家独立机构(Sequoia / Peak XV / HongShan)。

它的底色是一种刻意维持的不安全感(underdog / paranoia)——"我们只和下一笔投资一样好"。

Part 02

一、机构小传(时间线)

Sequoia 的故事线,可以围绕"如何让一家机构活得比创始人更久"来读。

  • 1972:Don Valentine 创立。他刻意取名 "Sequoia Capital"(红杉)而非 "Valentine Ventures"——一个明确的信号:这家机构应当比它的创始人活得更久。早期押注 Atari、Apple。
  • 1970s–90s:Cisco、Oracle、Electronic Arts 等奠定地位。
  • 1996:Valentine 把整个 firm 直接交给 Moritz & Leone——没有买断、没有支票。这是 VC 史上第一次成功的代际移交(详见维度五)。
  • 1999–2010s:Google、PayPal、YouTube、WhatsApp、Airbnb、Stripe……同时建立全球版图(中国 2005 由沈南鹏/Neil Shen 主理,印度其后)。
  • 2008-10-07:金融危机当口,Moritz & Leone 召集所有被投 CEO,做了著名的 "R.I.P. Good Times" 演示(56 页,末页 "Get Real or Go Home")——成为机构"危机预警"文化的标志性文物。
  • 2009–2010:首创 Scout program(侦察员计划),是第一家这么做的 VC。
  • 2017:Botha 成为美国业务的 steward(他与 Jim Goetz 自 2010 起共管美国创投)。
  • 2021-10-26:发布 The Sequoia Capital Fund——开放式 / 常青的永久资本结构(2022 年初注册为 RIA)。
  • 2022-07-05:Botha 出任 senior steward(据报道,刻意安排在 Leone 65 岁生日次日)。
  • 2022-11:FTX 投资 $214M 全额归零,致 LP 道歉信、电话会,deal owner Alfred Lin 公开担责(详见维度七)。
  • 2023-06-06:宣布把全球拆为三家完全独立的机构——Sequoia Capital(美/欧,Botha)、Peak XV Partners(印度/东南亚)、HongShan / 红杉(中国,沈南鹏),2024-03 完成。
  • 2025-11-04(据报道):Botha 交棒,Alfred Lin & Pat Grady 出任 joint stewards——交接机器仍在运转。[此条为最新报道,第一手 color 暂缺]
Part 03

二、七维剖析

维度一:合伙人结构与治理

Sequoia 治理的内核是一个词——stewardship(守护 / 受托),而且它被写进了"宪法"、用法律锁死。这一代人既不花钱从上一代手里买下 firm,也不能把 firm 卖掉变现给下一代,于是"把基金做大套现退出"这条路被结构性地堵死了。

"We structured ourselves to be a private partnership in perpetuity to the extent possible under California law... We didn't have to pay to get the partnership from the previous generation and nor will we charge the next generation. That's our motto."
"我们把自己构造成一家在加州法律允许范围内尽可能永续的私有合伙……我们不必花钱从上一代手中拿到这个合伙企业,也不会向下一代收费。这就是我们的座右铭。"
Roelof Botha, Why Venture Capital is Broken (Podwise id 5420267), 12:35
"We have changed the constitution of Sequoia Capital to make it virtually impossible to take it public, [or] to sell any part of the GP to a third party. So it's a private partnership to the limit of the law in perpetuity."
"我们修改了 Sequoia Capital 的'宪法',使得几乎不可能把它公开上市,或把 GP 的任何一部分卖给第三方。所以它是法律极限之内的、永续的私有合伙。"
Roelof Botha, View From The Top (Stanford GSB, id 2182978), 27:51

这个 stewardship 的基因,被追溯到创始人取名的那一刻——firm 是用来"一起共事"的,不是某个人的个人品牌:

"Don recruited people to work with him, not for him... That culture of stewardship continues to this day."
"Don 招人是来'和他一起'共事的,不是'为他'打工……那种 stewardship 的文化延续至今。"
Roelof Botha, Goldman Sachs Exchanges (id 1922941, 2024-09-20), 05:32

Sequoia 对治理的自我定位,是一组刻意"全都要"、不做取舍的张力:

"We want innovation and performance. We want individualism and teamwork. We want to be demanding and supportive. Most organizations make a trade-off between those dimensions... At Sequoia, we want to do both."
"我们既要创新、又要业绩。既要个人主义、又要团队协作。既要严苛、又要支持。多数组织会在这些维度之间做取舍……在 Sequoia,我们两者都要。"
Roelof Botha, Goldman Sachs Exchanges (id 1922941), 05:32

这套"全都要"的价值观背后,有一个明确的优先级排序,而且把自己排在最后——它是 Sequoia 一切评估与利益分配的隐性宪法:

"We've always thought that it's founders first, then LPs, then Sequoia, then the team, then yourself."
"我们一直认为:创始人第一,然后是 LP,然后是 Sequoia,然后是团队,最后才是你自己。"
Pat Grady, Uncapped #36(Pat Grady & Alfred Lin from Sequoia), 2025-12-09, 03:57

2023 全球三分,是现代最干净的一个"放弃共享品牌、换取去中心化自治"的治理案例。官方理由(联合信)聚焦运营:

"It has become increasingly complex to run a decentralized global investment business."
"经营一个去中心化的全球投资业务,已变得越来越复杂。"
Botha / Neil Shen / Shailendra Singh 联合信, 2023-06-06
"As each entity's portfolio has expanded to include companies that are becoming global leaders, we've seen growing market confusion due to the shared Sequoia brand as well as portfolio conflicts across entities."
"随着各实体的投资组合扩张、纳入了正在成为全球领导者的公司,我们看到因共享 Sequoia 品牌而日益加剧的市场混淆,以及跨实体的投资组合冲突。"
联合信, 2023-06-06

⚠️ 这封信是经过 PR 打磨的:它把好听的"运营理由"放在台前,对真正的驱动力——中美地缘政治——轻描淡写。Botha 后来在播客里说得更直白:

"We all believed that it [China] would integrate into the global economy. That premise proved wrong... life just got too hard for that, honestly. And we saw just more division between the two countries."
"我们都曾相信中国会融入全球经济。这个前提被证明错了……老实说,那件事变得太难了。我们看到的是两国之间越来越深的分裂。"
Roelof Botha, Why Venture Capital is Broken (id 5400127), 09:34

但 Sequoia 这套"结构兜底"的治理本能也有一个值得注意的例外——当被投是 xAI 这类无董事会的公司时,Botha 把治理直接押在"对个人的信任"上,并主张创始人内驱力本身就是比正式董事会更强的问责机制:

"It comes down to trust and faith in the individual. Elon Musk does not want to be encumbered by the bureaucracy of traditional boards. He is still open to advice—we text frequently—and he actively seeks out people who will tell him things that are uncomfortable to hear because his primary goal is to build a successful business."
"归根结底,这取决于对个人的信任与信念。Elon Musk 不想被传统董事会的官僚程序所束缚。他仍然乐于接受建议——我们经常发短信——而且他会主动去找那些愿意对他说出难听话的人,因为他的首要目标就是把生意做成功。"
Roelof Botha, Axios(与 Dan Primack 对谈), 2025-11-20, Q7 区段(转述)

⚠️ [推断] 这是一处和 Sequoia 主叙事("把治理写进宪法、结构上无法作恶")形成张力的口径:对自家 GP 它信奉结构兜底,对极端创始人它却愿意用"高信任 + 短信"替代正式董事会——可读成"信任顶级创始人"的延伸,也可读成在最稀缺的 deal 面前对治理标准的弹性放宽(FTX 之后尤其值得警惕地并列阅读)。

给"管基金"的启示:Sequoia 用一份"宪法"把治理从"取决于当代人的品德"变成了"结构上无法作恶"。[推断] 对一支新基金而言,最可迁移的不是"也搞永续",而是这个思路——把你最怕未来某代人做的坏事(套现、稀释下一代、品牌冲突),在文件层面提前锁死。

维度二:决策机制与决策权

这是 Sequoia 最"可抄作业"的部分——一套刻意对抗认知偏差的流水线。

机制细节(多位合伙人交叉印证):备忘录周五传阅,所有人带着"准备好的头脑"到周一会;sponsoring partner(发起合伙人)先介绍公司;然后是匿名"预投票",0–10 分、禁止打 5 分(不许骑墙);之后才进入讨论。

"We do an anonymous vote, pre-vote we call it... you can't vote five, it's zero to 10, five is not allowed. You can't sit on the fence. You need to make a decision."
"我们做一次匿名投票,我们叫它'预投票'……你不能投 5 分,是 0 到 10 分,5 分不允许。你不能骑墙。你必须做一个决定。"
Roelof Botha, Goldman Sachs Exchanges (id 1922941), 15:25

如果预投票一边倒地乐观,Botha 会暗中指派一个 devil's advocate(唱反调者)

"If it's too positive, I'll quietly assign a devil's advocate behind the scenes for somebody to argue the opposite case, to make sure we have rigor in the investment conversation."
"如果太正面了,我会在幕后悄悄指派一个唱反调的人,让某人去论证相反的情形,以确保我们的投资讨论有足够的严谨度。"
Roelof Botha, Goldman Sachs Exchanges (id 1922941), 15:25

为了不让资历压倒判断——资深者最后发言

"The people who have been around for a while tend to speak last so that they don't end up swaying the conversation... influence is a function of expertise, not a function of tenure or power or some other corrosive dynamic."
"在这儿待得久的人往往最后发言,这样他们就不会左右整场讨论……影响力是专业度的函数,而不是资历、权力或其他某种腐蚀性动态的函数。"
Pat Grady, Relentless Application of Force (Invest Like the Best EP.378, id 1403590), 06:24

而 Sequoia 自 2014 年起把每一笔投资的投票数值都记录下来,复盘后得出一个反直觉的结论——共识与否不重要,"信念是否在场"才重要

"We have actual data on this going back to 2014 where we started recording the votes numerically on every single investment... whether something is consensus or contentious actually doesn't matter. The thing that matters is presence of conviction."
"我们对此有真实数据,可追溯到 2014 年,那时我们开始把每一笔投资的投票都数值化记录……一件事是共识还是有争议,其实并不重要。重要的是——信念是否在场。"
Pat Grady, 20VC (id 1733110), 1:04:03
"The only thing that mattered [in] a fund that works is the conviction of the sponsor."
"一只成功的基金里,唯一重要的,是发起人的信念。"
Doug Leone(由 Julien Bek 转述), 20Europe (id 1101062), 29:32

也正因为信念高于共识,Sequoia 主动"要"分歧——真相不在人人都同意的中间地带:

"We need that volatility, because the truth is not somewhere in the middle where everybody agrees."
"我们需要那种波动(分歧),因为真相并不在'人人都同意'的中间地带。"
Pat Grady, Uncapped #36(Pat Grady & Alfred Lin from Sequoia), 2025-12-09, 36:01

而挡在"做出正确决定"前面的,往往不是信息不足,而是两种恐惧:

"Fear of missing out and fear of looking stupid are the two fears that prevent people from making the right decisions."
"对错过的恐惧、和对显得愚蠢的恐惧,是阻止人们做出正确决定的两种恐惧。"
Pat Grady, Uncapped #36, 2025-12-09, 41:49

Leone 对"讨论的目的"有一个很 Sequoia 的说法——找的是"桌子中央的真相",不是你的真相或我的真相:

"We're looking for the truth. Not your truth, not my truth, the truth in the middle of the table that helps the founder... I hear one of our young partners making a point, I say, hold on a second, I didn't think of that. His point is better than my point. I changed my mind."
"我们寻找的是真相。不是你的真相,也不是我的真相,而是桌子中央那个、能帮到创始人的真相……我听到一位年轻合伙人提出一个观点,我说,等一下,我没想到这点。他的观点比我的好。我改变主意了。"
Doug Leone, Sequoia Capital Part II (Acquired, id 39328), 52:19

Botha 在 2025-11(卸任前几天)那场近乎告别的访谈里,把"为什么要禁止 sidebar 联盟、要求会前都读完材料"的目的讲到了最干净——辩论不是为了赢,是为了投得对;也正因为是共识制,一旦投了就是"我们的"投资,全 firm 都欠这个创始人一份支持(事实):

"We start every partner meeting with an anonymous poll... to ensure we aren't just following the loudest voice... the goal is not to win an argument, but to make a great investment. Because we operate on a consensus basis, once we invest, it becomes 'our' investment, not an individual's."
"我们每次合伙人会议都以一次匿名投票开始……以确保我们不是在跟随那个嗓门最大的人……目标不是赢得一场争论,而是做出一笔伟大的投资。因为我们是共识制运作,一旦投了,它就成了'我们的'投资,而不是某个人的。"
Roelof Botha, StrictlyVC Download(days before stepping down), 2025-11-05, 11:43

Pat Grady 把"投资里真正需要人来做的事"压缩成两件,其余都可被流程/工具替代:

"There are only two things that you need a human being to do in the world of investing... number one, build the relationship with the founder, and number two, make the decision."
"在投资的世界里,只有两件事真正需要一个人来做……第一,与创始人建立关系;第二,做决定。"
Pat Grady, Invest Like the Best EP.378 (id 1403590), 1:04:49

个人层面的决策哲学(Botha):逆向且正确多问少断言

"To be successful in business, you have to be contrarian and right."
"要在商业上成功,你必须既逆向、又正确。"
Roelof Botha, LDV, 2024
"Be very careful about making statements. It's far better to ask questions. Be Socratic."
"对下断言要非常小心。提问要好得多。要像苏格拉底那样。"
Roelof Botha, LDV, 2024

维度三:经济结构

两件大事:常青基金反自肥的薪酬哲学

(1) The Sequoia Capital Fund(2021/2022,常青结构)。Botha 的发布文章直接开炮于 1970 年代沿用至今的"10 年封闭期":

"Our industry is still beholden to a rigid 10-year fund cycle pioneered in the 1970s."
"我们这个行业仍受制于一个 1970 年代开创的、僵硬的 10 年基金周期。"
Roelof Botha, The Sequoia Capital Fund, 2021-10-26
"Investments will no longer have 'expiration dates."
"投资将不再有'到期日'。"
Roelof Botha, 2021-10-26
"The best founders want to make a lasting impact in the world. Their ambition isn't confined to a 10-year period. Neither is ours."
"最优秀的创始人想在世界上留下持久的影响。他们的雄心不被限制在 10 年之内。我们的也一样。"
Roelof Botha, 2021-10-26

结构是:一只开放式母基金,作为各个限时子基金(种子/创投/成长)的唯一 LP;公司上市后 6/12/18 个月,可把赢家"挪进"母基金长期持有。其经济收益相当可观:

"Since we launched this three and a half years ago, we've accumulated another $6.7 billion in gains by doing nothing except being patient. $6.7 billion in gains that our LPs would not have seen if we had just distributed those shares outright."
"自三年半前推出至今,我们仅仅靠'保持耐心'、什么都不做,就又累积了 67 亿美元的收益。如果我们当初只是把那些股票直接分配出去,LP 是看不到这 67 亿的。"
Roelof Botha, Why Venture Capital is Broken (id 5400127), 16:30

这个"耐心"的最干净的单点证据,是 Square(Block)——常青结构存在的全部理由都浓缩在这一个数字里:上市时退出是 9 倍,长期持有变成 90 倍(事实):

"With Square, we made a 9x return from our initial investment to the IPO, but by being patient and holding the shares for several years post-IPO, that return grew to 90x."
"在 Square 上,从初始投资到 IPO 我们获得了 9 倍回报,但靠着保持耐心、在 IPO 后又持有了数年,那个回报增长到了 90 倍。"
Roelof Botha, Sequoia's Crucible Moment (Invest Like the Best EP.250, id 2185154), Q1

这套结构背后是一句对"传统 VC 把 IPO 当终点"的直接否定——IPO 是里程碑,不是(投资人的)终点站

"We tell the founders that the IPO is a milestone. Why should the IPO be a destination for the investor?"
"我们告诉创始人,IPO 是一个里程碑。那为什么对投资人来说,IPO 就该是终点站呢?"
Roelof Botha, Sequoia's Crucible Moment (id 2185154), 05:11

而支撑这种"超长持有"心智的,是 Botha 把自己精算师(actuary)出身当成一种认知优势——会计向后看一年,精算师向前看二十年;他甚至把自己多数的差投资归因于"视野太短"(事实):

"Accountants are trained to think a year in arrears, and actuaries are trained to think 20 years into the future... Most of my poor decisions were the result of having a horizon that was too short."
"会计被训练成向后想一年,精算师被训练成向前想二十年……我大多数糟糕的决策,都是因为视野太短造成的。"
Roelof Botha, Sequoia's Crucible Moment (id 2185154), 1:02:05 / Q9

关键的反"资产管理规模冲动"设计——这只母基金只有跑赢 NASDAQ 才收业绩费:

"We only earn an incentive fee if we can outperform the NASDAQ over a rolling three-year period. And to us, that felt like a wonderful design constraint... The last thing we want to do is be a partnership that focuses on asset gathering and fee income."
"只有当我们能在滚动三年期内跑赢 NASDAQ,我们才赚取业绩费。对我们来说,这是一个绝妙的设计约束……我们最不想成为的,就是一家专注于'攒资产规模'和'吃管理费'的合伙企业。"
Roelof Botha, Goldman Sachs Exchanges (id 1922941), 48:53

这种"反规模"在 Botha 那里有一个更尖锐的底层论证——对绝大多数 LP 而言,把 VC 当成一个固定配比的资产类别其实是"无回报的风险"(return-free risk):多投钱进硅谷不会多出伟大公司,反而稀释人才;把前 20 家机构从历史回报里剔除,整个行业跑输指数基金(事实——这正是 Sequoia 刻意"少而精、不扩规模"的经济学依据):

"There are only a small number of companies that truly matter. Throwing more money into Silicon Valley does not yield more great companies; it actually dilutes talent... If you remove the top 20 venture firms from the industry's historical results, you actually underperform an index fund."
"真正重要的公司只有极少数。往硅谷里砸更多钱并不会产出更多伟大公司;它反而稀释人才……如果你把前 20 家风投机构从行业的历史回报里剔除,你其实是跑输一只指数基金的。"
Roelof Botha, StrictlyVC Download(days before stepping down), 2025-11-05, Q1

这条"反规模"哲学的最具体落地,是 Alfred Lin 给出的一个罕见公开的口径——Sequoia 根本不追踪 AUM,只盯单只基金的具体规模(这一段也顺带填上了此前缺失的"具体基金规模"事实):种子基金约 $200M、创投基金约 $700–800M,并以此把投资团队人数刻意保持得很小(事实):

"We don't actually track AUM, as it is a metric that doesn't help us build better companies. Instead, we focus on the specific size of the fund we are investing out of—whether it is our $200 million seed fund, our $700–800 million venture fund, or our growth and expansion funds. This structure allows us to maintain a small, focused team of investors."
"我们其实并不追踪 AUM,因为它是一个帮不上我们'建更好公司'的指标。我们盯的是我们正在投资的那只基金的具体规模——无论是我们的 $200M 种子基金、$700–800M 创投基金,还是成长与扩张基金。这种结构让我们得以保持一支小而专注的投资团队。"
Alfred Lin, Turpentine VC(An Inside Look Into Sequoia's Strategy & Operations), 2023-11-01, Q3

Ben Thompson 给了一个最好的分析框架——Sequoia 在从"金融资本"转向"生产资本":

"Sequoia is transforming itself from financial capital to production capital."
"Sequoia 正在把自己从金融资本,转变为生产资本。"
Ben Thompson, Stratechery, 2021-10

LP 的选择本身就是价值观——大头是大学捐赠基金与基金会("我们为这些'伟大的事业'工作"):

"At Sequoia, we work for what we call great causes... the Ford Foundation, the Wellcome Trust, Stanford University, MIT, Harvard, Dana-Farber Cancer Institute, Boston Children's Hospital, these are the people we work for."
"在 Sequoia,我们为我们称之为'伟大事业'的对象工作……福特基金会、Wellcome Trust、斯坦福大学、MIT、哈佛、Dana-Farber 癌症研究所、波士顿儿童医院——这些就是我们服务的对象。"
Roelof Botha, Pre-Mortems and Pre-Parades (id 2927142), 42:37

(2) 反自肥的薪酬哲学。Leone 对 VC 的经济结构有一段近乎"自我批判"的话——正因为下行无责、上行抽成,这门生意容易让人"松懈":

"Think of the economics of venture capital. They should be illegal... If I lose it, too bad. And if I make it, I keep 30% or 20% or 25%... that's the root of all ills because... it affords people to take it easy."
"想想风险投资的经济结构。它们简直应该是非法的……我亏了,算我倒霉;我赚了,我留下 30%、20% 或 25%……这是万恶之源,因为……它让人有本钱去松懈。"
Doug Leone, Divot #1 (id 5562652, 2025-03-05), 22:44
"The moment you build enterprise value... what you're really asking is to monetize the enterprise value... Selling a piece of that firm means that the people in the building today are getting richer, but the pie to be shared for the next generation is smaller."
"当你开始积累企业价值的那一刻……你真正想要的是把这份企业价值变现……卖掉 firm 的一部分,意味着今天在这栋楼里的人变得更有钱了,但留给下一代去分的那块饼,变小了。"
Doug Leone, Lessons from a Titan (Invest Like the Best, id 3136669), ~41:34

⚠️ 反面证据(必须给读者):PitchBook / TechCrunch 指出常青结构的 GP–LP 错位风险——"一只表现不佳的子基金可能被允许继续存在,靠兄弟基金供养、为管理人产生管理费,却未必为 LP 创造好回报"。永久性既能锁定耐心,也可能庇护平庸

证据薄弱处:具体的 carry / comp 数字与"Sequoia Fund"的法税细节未公开(同行估 carry ~30%);合伙人把 carry 捐给慈善更多是个人行为(Botha/斯坦福、Moritz/Crankstart)而非 firm 政策。

维度四:人才模型

核心三件:斜率优先、学徒制、不为"制造 picker"搭脚手架

招人看"斜率而非截距"——成长速度比已有积累更重要:

"Favor slope over intercept. It's the rate of growth. It's your willingness to absorb new information and to keep getting better."
"看重斜率,而非截距。看的是成长的速率,是你吸收新信息、持续变得更好的意愿。"
Roelof Botha, LDV, 2024

而且招人这件事被当作"微缩版的传承"——招进来的人,要能想象 ta 有朝一日领导这家 firm

"When I'm interviewing people to join Sequoia, it's not to work for me, it's to work with me. And I evaluate each candidate wondering, could I imagine you leading this firm down the road?"
"当我面试想加入 Sequoia 的人时,不是来'为我'工作,而是'和我一起'工作。我评估每一个候选人时都在想:我能想象你将来领导这家 firm 吗?"
Roelof Botha, View From The Top (id 2182978), 27:51

Alfred Lin 把硬标准压成三个 high:

"A high IQ, high EQ and high hustle."
"高 IQ、高 EQ、高 hustle(拼劲)。"
Alfred Lin, Sequoia's Investment Playbook (Acquired, id 39305), 52:25

招"投资人"这一侧,则有一句被两代合伙人共用的招聘 spec——"高度好胜,但有一颗金子般的心"(highly competitive with a heart of gold)。Alfred Lin 解释这是一对刻意的张力:好胜驱动个人业绩,金子心让人能在团队里活下去、并把"为创始人做对的事"放在第一位(事实):

"My partner Pat Grady likes to say we look for people who are 'highly competitive with a heart of gold.' Being highly competitive means you are willing to 'take the shot' when there are three seconds left on the clock... The 'heart of gold' part is equally essential because you cannot do this job effectively if you don't care deeply about your founders."
"我的合伙人 Pat Grady 喜欢说,我们找的是'高度好胜、但有一颗金子般的心'的人。高度好胜意味着——当比赛只剩三秒时,你愿意'出手投那一球'……'金子般的心'同样不可或缺,因为如果你不深切在乎你的创始人,你根本做不好这份工作。"
Alfred Lin, Turpentine VC, 2023-11-01, Q5

Pat Grady 把"高度好胜"进一步拆成两件事——drive(拼劲)+ killer instinct(杀手本能),并点出"金子心"的真正功能是防止一屋子好胜的人互相残杀(事实):

"We decompose hyper-competitive into drive and killer instinct. Drive is the willingness to show up and put in the work. Killer instinct is recognizing the one opportunity that really matters when it presents itself... but you need the heart of gold to keep hyper-competitive people from killing each other."
"我们把'高度好胜'拆解为 drive 和 killer instinct。drive 是愿意到场、把活儿干了;killer instinct 是当那个真正重要的机会出现时,能认出它……但你需要那颗金子般的心,来阻止一群高度好胜的人互相残杀。"
Pat Grady, Peak XV(Building Enduring Companies, the Sequoia Values), 2025-04-23, ~05:24

漏斗之严苛——一个数字说明问题:

"It was about the most rigorous hiring process we've ever had, like top of funnel was 9,000 and what came out the bottom of the funnel was Andrew Reed and Matt Huang."
"那大概是我们有史以来最严苛的招聘流程,漏斗顶端是 9000 人,从漏斗底端出来的,是 Andrew Reed 和 Matt Huang。"
Pat Grady, 20VC (id 1733110), 44:24

培养是学徒制——刻意"不搭脚手架"去批量制造 picker,因为顶尖的判断者是 one-of-one:

"The answer on getting better at picking is not to erect a bunch of scaffolding around people... be true to this as an apprenticeship business, have experienced investors in the field doing the work alongside the people who are earlier in their career."
"想让人更会'挑项目',答案不是在人周围搭一堆脚手架……要忠于这是一门'学徒制'的生意,让有经验的投资人在一线,和那些职业生涯更早期的人,并肩做事。"
Pat Grady, 20VC (id 1488894), 26:07

Scout program(2009/2010 首创)的起源,是 Botha 对 Jason Calacanis 的一句反问——后来 Uber、Stripe(经 Sam Altman 之手)都来自这个网络:

"What if we'd just given you some money to make those investments?"
"要是我们直接给你一笔钱、让你去做那些投资呢?"
Roelof Botha 对 Jason Calacanis(Scout program 起源,约 2009)

一个易被忽视、但对"未来基金长什么样"极有启发的事实——Sequoia 里工程师/PM/设计师比投资人还多

"We have more engineers, product managers, and designers at Sequoia than we have investors."
"在 Sequoia,我们的工程师、产品经理和设计师,比投资人还多。"
Roelof Botha, Goldman Sachs Exchanges (id 1922941), 15:25

而"为什么要刻意凑一支背景各异的团队",Pat Grady 给了一个很形象的对照——别人把 VC 当高尔夫球队(各打各的),Sequoia 当板球俱乐部(不同位置、不同特长,但为同一个客户——创始人——协同):

"Michael Moritz once said, 'We never forget that the competition is outside the building, not inside the building.' Most venture capital firms approach it as a golf team, but we approach it more as a cricket club. We try to hire people who have different backgrounds and different spikes... then our customer, the founder, gets the benefit of all of those skills."
"Michael Moritz 曾说:'我们从不忘记——竞争在大楼外面,不在大楼里面。' 多数风投把它当成一支高尔夫球队,但我们更像一个板球俱乐部。我们尽量招那些背景不同、长板(spike)不同的人……这样我们的客户——创始人——就能享受到所有这些技能的好处。"
Pat Grady, Peak XV, 2025-04-23

而且他们愿意要"不好相处但卓越"的人(一个 2×2)——这也呼应了 Valentine 对创始人的口味:

"We invest in a two by two matrix of people, exceptional, not exceptional, easy to get along with, not always easy to get along with."
"我们按一个 2×2 矩阵来投人:卓越 / 不卓越,好相处 / 不总是好相处。"
Roelof Botha, LDV, 2024
"the best founders are the ones who are the most difficult"
"最好的创始人,往往是最难相处的那些。"
Don Valentine 对 Roelof Botha(via Mallaby, The Power Law

维度五:传承与文化延续

这是 Sequoia 最强、也最稀缺的部分——对想建立"活得比自己久"的机构的人,这一节是金矿。

起点是 1996 年那个近乎传奇的"交付"动作——Valentine 白送,不要一分钱:

"Don didn't ask for any money whatsoever. He just handed it over. And in doing so, he imbued the partnership with this sense of stewardship... The only thing that's been asked of us in return is to make sure that we leave it in a better place than we found it."
"Don 没有要任何钱。他就这么把它交了出去。正因如此,他给这个合伙企业注入了一种 stewardship 的意识……作为回报,唯一被要求我们做的,就是确保我们把它留在一个比接手时更好的地方。"
Pat Grady, Invest Like the Best EP.378 (id 1403590), 03:47

交接现场的"三件事"(Leone 的回忆,由 Acquired 转述),把"交权"与"交责任"分得很清楚:

"Don, one day in 1996, invited Mike and Doug into a conference room and... said, I'm giving this firm to you. And there are three things. One, you're going to run the firm... Two, you get to decide what I do... Three... one of the things I'm not willing to do is run the firm."
"1996 年的某一天,Don 把 Mike 和 Doug 叫进一间会议室……说,我要把这家 firm 交给你们。有三件事。第一,由你们来经营这家 firm……第二,由你们来决定我做什么……第三……我唯一不愿意做的事,就是经营这家 firm。"
Ben Gilbert(转述 Doug Leone), Sequoia Capital (Part 1) (Acquired, id 39336), 1:15:14

Leone 自己反复强调"白给"这个事实,并把它转译成对下一代的义务:

"Sequoia was given to us by Don Valentine. Given... Mike and I didn't write a check to Don Valentine. Zero... Our job is really to make it a better place... and leave it a better place for the next generation."
"Sequoia 是 Don Valentine '给'我们的。是'给'……Mike 和我没有给 Don Valentine 开过一张支票。零……我们的工作,真的就是把它变成一个更好的地方……并把它更好地留给下一代。"
Doug Leone, Lessons from a Titan (id 3136669), ~41:34

交接被"职务化"而非"个人化"——刻意用 steward 这个词,而非 "managing partner";交班前几年就铺垫(Botha 接手时几乎没有直接下属)。Botha 的时间线:

"I became the steward actually of the U.S. business in 2017. Jim Goetz and I have been running the U.S. venture business since 2010... And then... in 2022, I became senior steward."
"我实际上在 2017 年成为美国业务的 steward。Jim Goetz 和我自 2010 年起就在管美国创投业务……然后……在 2022 年,我成为了 senior steward。"
Roelof Botha, Uncapped #28 (id 5441971, 2025-10-15), 01:21

退下来的人怎么"在场而不越位"——Leone 给了一个模板:

"[I'm] a model retired partner, which means I'm there when asked to be there but not someone who volunteers opinions."
"(我是)一个模范的退休合伙人,意思是——被需要时我在,但不会主动塞给你意见。"
Doug Leone, 2022

⚠️ Leone 在宣布交班时的这句话既是事实陈述、也是自我定位

"The spirit of growth and renewal drives our culture of generational transfer. We are proud to be the only venture capital partnership to have successfully navigated multiple leadership transitions over five decades."
"成长与更新的精神,驱动着我们'代际传承'的文化。我们自豪于:我们是唯一一家在五十年里成功穿越了多次领导权交接的风险投资合伙企业。"
Doug Leone(宣布 Botha 接任), 2022-04-04

2025-11 这次 Botha → Alfred Lin & Pat Grady 的交接,终于有了 Botha 本人的第一手 color。他把"会交班"直接抬到与"会投资"并列的核心能力——这是机构自我叙事里极罕见的、把"传承"当作一项可练习的 craft 来说:

"There are a few things we know how to do well at Sequoia. One is to make great investments; the other is transitioning leadership."
"在 Sequoia,我们有几件事是知道怎么做好的。一件是做出伟大的投资;另一件是交接领导权。"
Roelof Botha, Axios(与 Dan Primack 对谈), 2025-11-20, 00:01:05

而"为什么要交得这么快",Botha 给出的理由是规避 interregnum(权力真空期)——他把"快"本身当成一种刻意的反停滞设计:

"We have a secret at Sequoia: we transition leadership quickly to avoid 'interregnum,' a period between administrations where organizations suffer stasis and decision-making grinds to a halt... after 23 years, it was time for the next generation—Alfred Lin and Pat Grady—to take the lead."
"我们在 Sequoia 有一个秘密:我们很快地交接领导权,以规避'interregnum'——两届之间那段组织陷入停滞、决策戛然而止的真空期……在 23 年之后,是时候让下一代——Alfred Lin 和 Pat Grady——来掌舵了。"
Roelof Botha, Axios(与 Dan Primack 对谈), 2025-11-20, Q1 区段(转述)

⚠️ 这与 WebSearch surfaced 的反叙事("被部分合伙人请下台")并不相容——Botha 把交接讲成主动、deliberate、为机构着想的"我们擅长的事",而外部报道暗示存在被动成分。两面并存,按"官方口径可能经过自我美化"打折阅读。

证据薄弱处:2025-11 Botha → Lin & Grady 交接的"快速/反 interregnum"动机现已有 Botha 第一手口径(Axios + StrictlyVC + Uncapped #36);但Lin 与 Grady 本人对接任的第一人称 color 仍少(多为他人转述其联合掌舵早/成长两队多年)。

维度六:自我更新机制

Sequoia 的"防衰老"靠的是一种刻意维持的不安全感。这是跨发言人、被重复最多的主题。Moritz 的版本:

"We've always been scared of going out of business. We've always been paranoid. We've always thought that what we've earned yesterday is not guaranteed tomorrow."
"我们一直害怕倒闭。我们一直偏执。我们一直认为,昨天挣到的,并不保证明天还在。"
Michael Moritz(由 Morgan Housel 引述), Tim Ferriss Show #576 (id 115997), 1:27:10

Botha 把它落到一句机构口头禅——"我们只和下一笔投资一样好":

"There was a hunger and an underdog mentality at the partnership that wasn't apparent from the outside... There was a sense of, we're only as good as our next investment."
"这个合伙企业里有一种从外面看不出来的饥饿感和'弱者'心态……有一种感觉是:我们只和我们的下一笔投资一样好。"
Roelof Botha, Goldman Sachs Exchanges (id 1922941), 05:32

Pat Grady 的版本更具画面感——每一笔都是"肉搏":

"One of our core beliefs is that anybody can beat us on any given day... It is hand-to-hand combat, and we have to be at our absolute best if we have any hopes of making that investment."
"我们的核心信念之一是:任何人在任何一天都可能打败我们……这是肉搏战,我们必须处在绝对最佳状态,才有希望拿下那笔投资。"
Pat Grady, Invest Like the Best EP.378 (id 1403590), 20:05

这种不安全感有事实根基——Botha 入职时做过的功课:

"If you look at the top venture firms in Silicon Valley in 1990, the majority of them no longer exist... So there's this insecurity that we have at Sequoia that really drives us."
"如果你看 1990 年硅谷顶尖的那些风投机构,它们中的大多数已经不复存在了……所以我们在 Sequoia 有一种不安全感,是它真正驱动着我们。"
Roelof Botha, Uncapped #28 (id 5441971), 04:49

一个被投 CEO 的外部视角(HubSpot 的 Brian Halligan),印证这不是表演:

"They work much harder than any other VCs I've come across, and they're absolutely paranoid that they're going to lose it... They don't celebrate their successes. They beat themselves up for their failures."
"他们比我见过的任何 VC 都更拼命,而且绝对地偏执,怕自己会失去这一切……他们不庆祝成功,却为失败痛打自己。"
Brian Halligan, 20VC (id 176537), 54:42

最深的"更新"观,是 Moritz 的——firm 本身才是最重要的那个产品

"When people write about the venture business, they're always writing about the startups we back. They never write about the most important investment that we make, which is in the business."
"当人们写风投这门生意时,他们总在写我们投的那些创业公司。他们从不写我们做的最重要的那笔投资——对'这家 firm 本身'的投资。"
Michael Moritz(via Mallaby, The Power Law

文化文物:2008 年 "R.I.P. Good Times" 的末页一句,至今被引用:

"Get Real or Go Home."
"要么动真格,要么回家。"
"R.I.P. Good Times" 收尾页, Sequoia, 2008-10

以及把"重大不可逆决策"概念化的工具——Crucible Moments:

"Crucible moments is the term we use at Sequoia for things that are basically very important type one decisions... these decisions change the trajectory of a company."
"Crucible moments(熔炉时刻)是我们在 Sequoia 用来指那些极其重要的'第一类决策'的词……这些决策会改变一家公司的轨迹。"
Alfred Lin, How to Think Like a World-Class Investor #212 (id 2755503), 1:12:27

维度七:基金运营与绩效评估(新增维度)

这是本次新增、也是材料最厚、对读者最直接有用的一维。 Sequoia 评估自己人的方式可以一句话概括:看过程、不看单次结果(process over outcome)

最核心的一段(Pat Grady)——没人因一笔失败被惩罚、也没人因一笔成功被表彰;要看的是"一串":

"You will never see anybody at Sequoia reprimanded, fired, chastised for a failed investment... nor promoted, lauded, praised for a single great investment... If you have a string of bad investments, there's probably something wrong in your process. If you have a string of good investments, there's probably something right in your process."
"在 Sequoia,你永远不会看到任何人因为一笔失败的投资而被斥责、开除、责难……也不会因为单独一笔伟大的投资而被提拔、赞誉、表彰……如果你有一连串糟糕的投资,那很可能是你的过程出了问题;如果你有一连串好的投资,那很可能是你的过程有对的地方。"
Pat Grady, 20VC (id 1488894), 11:57

而且评的是行为,不是运气——好行为配坏结果 = OK;坏行为配好结果 = 不 OK:

"If the activities and the behaviors are good, but one of the outcomes happens to be bad, that's okay. If the activities and the behaviors are bad, and one of the outcomes happens to be good, that's probably not okay."
"如果活动和行为是好的,但其中一个结果恰好是坏的,那没关系。如果活动和行为是坏的,而其中一个结果恰好是好的,那很可能就不行。"
Pat Grady, 20VC (id 1488894), 11:57

Pat Grady 把"为什么必须看过程"压成一句更锋利的话——这是一门终极的"长反馈周期"生意,最大的错误是去"扑结果"而非"以输入的质量为荣":

"We're in the ultimate long feedback cycle business and I think one of the mistakes that people make is they try to lunge for outcomes instead of just taking pride in the quality of the inputs."
"我们身处一门'终极的长反馈周期'生意,我认为人们常犯的一个错误是——他们想去'扑'结果,而不是单纯地以输入的质量为荣。"
Pat Grady, Peak XV, 2025-04-23, 04:06

那么具体"评什么、何时评"? Uncapped #36(Pat Grady & Alfred Lin,2025-12)给出了 Sequoia 罕见公开的评估节奏:一年两次、刻意把两类东西分开评——6 月评 values(价值观,如"aggressive but humble"),12 月评 capabilities(能力:sourcing / picking / winning / harvesting / building);并且刻意回避个人化的量化指标,以免激励出"为了让数字好看"的反效果行为。(事实,转述自该集 QA。)其底层逻辑是:投资的"产出"要十年后才见分晓,当下漂亮的结果可能只是幻象——所以只能评"输入/行为":

"In venture, the outputs are like 10 years into the future. So you have outputs that look like their outputs, but those can be mirages."
"在风险投资里,产出是十年之后的事。所以你会有一些'看起来像产出'的东西,但它们可能只是幻象。"
Pat Grady, Uncapped #36(Pat Grady & Alfred Lin from Sequoia), 2025-12-09, 10:03

Alfred Lin 把这套评估的具体抓手讲得更细——既有半年一次的全组合复盘(对照创始人自己定的目标),也有 sourcing / picking / winning / company building 四象限,并特意提出一个定性的领先指标:"创始人出事时第一个打给谁",外加 founder NPS;而对"看起来很美"的 markup(账面估值上调)则大幅打折,因为它常常不反映真实市场或长期健康(事实——直接补强了"为什么不能看单次输出"的论点):

"We hold semi-annual portfolio reviews where we look at every company in our portfolio and measure their progress against the goals they set for themselves... we evaluate ourselves on sourcing, picking, winning, and company building. A key qualitative metric is whether we are the 'first call' for a founder when they have a critical issue to discuss. We also look at founder NPS. We heavily discount markups because they don't always reflect market conditions or the long-term health of the business."
"我们做半年一次的组合复盘,逐一审视组合里的每一家公司,按它们为自己设定的目标来衡量进展……我们从 sourcing、picking、winning、company building 几个维度评估自己。一个关键的定性指标是:当创始人有要紧事要谈时,我们是不是那个'第一通电话'。我们也看 founder NPS。而对 markup(账面估值上调),我们会大幅打折,因为它并不总是反映市场状况或这门生意的长期健康。"
Alfred Lin, Turpentine VC, 2023-11-01, Q4

对一个新投资人,第一个能看的领先指标,是 ta 怎么分配自己的时间:

"The first leading indicator of good investment judgment for a new investor is how they invest their time."
"对一个新投资人来说,好的投资判断力的第一个领先指标,是他如何投资自己的时间。"
Pat Grady, Uncapped #36, 2025-12-09, 17:43

而"picking(选得好)"也被重新定义——不是"不亏钱",而是"足够高的非对称命中率":

"Good picking isn't not losing money. Good picking is a high enough inclusion rate of asymmetry."
"好的'选'不是'不亏钱'。好的'选',是足够高的'非对称机会'命中率。"
Pat Grady, Uncapped #36, 2025-12-09, 31:33

"严苛"在 Sequoia 被重新定义为"支持"的最高形式:

"We think that being demanding of one another is being supportive and that the best thing we can do for each other is to demand excellence, first of ourselves and then of each other."
"我们认为,彼此严苛就是彼此支持;我们能为对方做的最好的事,就是要求卓越——先要求自己,再要求彼此。"
Pat Grady, Invest Like the Best EP.378 (id 1403590), 12:02

评估工具一:每一笔投资都写 pre-mortem(事前验尸)与 pre-parade(事前庆功)——而且把它用在 firm 自己身上:

"For every investment we contemplate, we write a pre-mortem and a pre-parade... We had a pre-mortem. It literally said 2030 autopsy of Sequoia Capital. And what went wrong?... it really crystallizes first-order issues."
"对每一笔我们考虑的投资,我们都会写一份'事前验尸'和一份'事前庆功'……我们做过一次事前验尸,标题就直接写着'2030 年 Sequoia Capital 的尸检报告'。哪里出错了?……它真的能把第一性的问题结晶出来。"
Roelof Botha, Pre-Mortems and Pre-Parades (id 2927142), 23:43

评估工具二:点名认知偏差、互相纠偏。Botha 最经典的一课,是 Square(Block)上的确认偏差:

"It was really tough for me to swallow the fact that we could have invested a year prior at a quarter of the price [in Square]... but overcoming that confirmation bias of your previous decisions was just an incredible thing for us to get better at investing."
"要咽下这个事实对我来说真的很难——我们本可以在一年前、以四分之一的价格投资(Square)……但克服'对自己先前决策的确认偏差',是让我们在投资上变得更好的一件了不起的事。"
Roelof Botha, Goldman Sachs Exchanges (id 1922941), 12:47

而 Botha 复盘自己"最大的投资错误"时,得出一个对"如何评估一笔投资"很重要的诊断——错的往往不是当时的商业模式判断,而是"想象力的失败"(即低估了公司未来能变成什么)。这恰好解释了 Sequoia 为何如此强调"value creation before value capture"和长期持有(事实):

"The most significant investment errors stem from a failure of imagination regarding a company's future potential rather than the failure of the specific business model at the time of investment."
"最重大的投资错误,源于对一家公司未来潜力的'想象力失败',而不是投资当时那个具体商业模式的失败。"
Roelof Botha, 20VC (id 100421), takeaway(参见 04:33 关于 YouTube 仅 9000 用户时的想象力)

这套"对内复盘"的文化,也定义了 Sequoia 人在被投公司董事会上的行为基准——做减震器,不做放大器(低谷时托住、高峰时施压),且用苏格拉底式提问而非发号施令(事实):

"Board members are meant to be shock absorbers, not amplifiers. When the team is in the doldrums, be there to rally them; when things are going well, challenge them."
"董事会成员应当是减震器,而不是放大器。当团队陷入低谷时,在那儿把他们重新鼓动起来;当一切顺利时,挑战他们。"
Roelof Botha, 20VC (id 100421), 26:27

Alfred Lin 给了同一行为基准的一个更具操作性的版本——低谷时"不补刀、卷起袖子捡碎片",顺境时才做 sparring partner;而且他把"何时该加压"反直觉地放在顺境(因为"公司失败的第一个征兆,是成功带来的傲慢")(事实):

"We partner with the founders and we know to be shock observers during bad times and to be sparring partners during good times. There's no point when the company is down... to beat someone when they're down. The best thing you can do is roll up your sleeves, pick up the pieces and help the company get through that hardship... the first sign of why companies fail is the hubris of much success."
"我们和创始人结成伙伴,我们懂得:坏的时候做'减震观察者',好的时候才做'陪练'。当公司处在低谷时……没有任何意义去对一个倒下的人补刀。你能做的最好的事,是卷起袖子、把碎片捡起来,帮公司熬过那段艰难……公司之所以失败的第一个征兆,正是巨大成功带来的傲慢。"
Alfred Lin, Sourcery(Backing Founders Who Redefine Markets), 2025-10-30, 00:23

评估工具三:对 LP 的极致坦诚——回报放第一页,"低光时刻"放第二页:

"Our culture at Sequoia is not to spit shine things... Slide one, welcome, here's returns [net]... Slide two is probably the lowlights. Not the highlights, the lowlights."
"我们 Sequoia 的文化不是把东西擦得锃亮……第一页,欢迎,这是回报(净额)……第二页,大概就是'低光时刻'。不是高光,是低光。"
Doug Leone, Lessons from a Titan (id 3136669), 1:01:17

评估工具四:技术化的留痕——Botha 手机上的内部 app 能调出任一公司"我们最后见的是谁、打了几分":

"On my phone I can pull up an app that... if you give me any company name, I'll be able to tell you who my team last met at, how we rated it."
"在我手机上我能调出一个 app……你给我任何一个公司名,我都能告诉你我的团队最后见的是谁、我们打了几分。"
Roelof Botha, Why Venture Capital is Broken (id 5400127), 08:28

为什么必须看过程而非结果——因为这门生意的因果极度嘈杂:

"This is a humbling business because you can make money even if you got the investment thesis wrong and you can lose money even though you got the investment thesis right."
"这是一门让人谦卑的生意,因为即使你的投资逻辑错了你也可能赚钱,即使你的投资逻辑对了你也可能亏钱。"
Alfred Lin, Sequoia's Investment Playbook (Acquired, id 39305), 51:05

Leone 把整套评估文化压成一个"神圣三元组",且明确:业绩是第一信条

"Trust is the foundation, performance is the next level, and teamwork is the third. I call that the sacred triad of Sequoia."
"信任是地基,业绩是上一层,团队协作是第三层。我把它叫做 Sequoia 的神圣三元组。"
Doug Leone, Divot #1 (id 5562652), 22:44

而"业绩第一"在 Leone 嘴里没有任何缓冲空间——它是一个"一票否决"的前置条件,缺了它,其余美德全部作废(事实):

"If you're missing performance, the other nine tenets don't matter. If you have performance, you better have teamwork."
"如果你缺了业绩,其他九条信条都无所谓。如果你有了业绩,那你最好还要有团队协作。"
Doug Leone, 20VC (id 100338), 10:57

而"teamwork"这一层在 Sequoia 是用一条具体的社会规范来锁的——消灭"抢功"行为。Pat Grady 的说法把它和 founders→LPs→partnership→team→self 的优先级栈直接接上:你只要反复把对的事做对,功劳自然会来;越不在乎归谁,越能成事(事实):

"The key is to get rid of the credit-seeking behavior and institute this mentality of, you can accomplish amazing things if you don't care who gets the credit. If you just do the right thing and do it over and over again, you'll end up with plenty of credit."
"关键是消除'抢功'的行为,并建立这样一种心态:如果你不在乎功劳归谁,你能成就惊人的事。只要你反复地把对的事做对,最后你会得到足够多的功劳。"
Pat Grady, Peak XV, 2025-04-23

这套"看业绩"的文化有一个被反复引用、却容易被误读为"无情"的注脚——Sequoia 是"团队",不是"家庭"。Leone 的区分很锋利:家庭无论成员是否称职都得留着,高绩效组织不行;但"分手"要带着尊重(事实):

"We are a team, not a family, because families are stuck with members regardless of their performance... We view ourselves like a production company... where everyone from the stars to the lighting crew must perform at a high level. We hate the 'I' pronoun; it is always 'we.'... When we part ways, we do it with respect, forward vesting, and a hug."
"我们是一支团队,不是一个家庭,因为家庭无论成员表现如何都得绑在一起……我们把自己看成一家制片公司……从明星到灯光师,每个人都必须高水平地表现。我们讨厌'我'这个代词;永远是'我们'。……当我们分道扬镳时,我们带着尊重、带着 forward vesting(提前兑现的股权)、带着一个拥抱去做。"
Doug Leone, 20VC (id 100338), Q3

⚠️ [推断] "team not family + 一切皆 we" 既是真实的绩效文化,也明显是一句经过打磨的对外口径——它把"会请人走"这件冷酷的事,包进了"尊重 + forward vesting + 拥抱"的温情叙事里。按"既是事实、也是自我定位"来读。

真实的压力测试(FTX,2022):$214M 全额归零后,Sequoia 给 LP 写信,召开道歉电话会,deal owner Alfred Lin 公开担责,并改了流程(此后即便早期项目也上四大审计)。信里的措辞,是"承担风险"框架下的担责(而非甩锅):

"We are in the business of taking risk. Some investments will surprise to the upside, and some will surprise to the downside. We do not take this responsibility lightly and do extensive research and thorough due diligence on every investment we make."
"我们做的就是承担风险的生意。有些投资会带来向上的惊喜,有些会带来向下的意外。我们不轻视这份责任,对每一笔投资都做广泛研究和彻底的尽职调查。"
Sequoia Capital 致 LP 信, 2022-11

而担责的底气,来自一种把"脆弱"当信任机制的文化:

"One of the best ways to build trust is to show you're not defensive, you're vulnerable."
"建立信任最好的方式之一,就是展现出你不设防、你是脆弱的。"
Roelof Botha, Fortune, 2025-03

这种"脆弱即信任"最具体的一手证据,是 Botha 自曝的一幕——并且它给了"为什么只能押 outlier"一个冷峻的数字底座:Sequoia 的种子/创投投资里约有一半收不回本金,正因如此才要"哺乳动物式"地少生、精养(事实):

"We are more 'mammalian' than 'reptilian'—we do not lay a hundred eggs and see what happens. We make a small number of investments and provide them with significant, focused attention... Even with that approach, 50% of the time we make a seed or venture investment, we fail to fully recover capital, which is humbling. We take it personally; the first time I had a complete write-off, I was so embarrassed I cried in the partner meeting."
"我们更像'哺乳动物'而非'爬行动物'——我们不会下一百颗蛋、然后看会发生什么。我们只做少量投资,并给予它们大量、聚焦的关注……即便如此,我们做的种子或创投投资里,有 50% 收不回全部本金,这令人谦卑。我们把它当成自己的事;我第一次彻底减记一笔投资时,我尴尬到在合伙人会议上哭了出来。"
Roelof Botha, StrictlyVC Download(days before stepping down), 2025-11-05, 25:11

⚠️ 但 FTX 也暴露了这套坦诚文化的边界:他们承认了亏损,但同时也为自己的尽调辩护、强调"被欺骗"。坦诚是真的,无限坦诚是假的。

Part 04

三、投资风格与审美

Sequoia 的投资审美,最古老的一条来自 Valentine——先看市场,再看人

"The size of the market, the dynamics of the market, the nature of the competition. Our objective is always to build big companies — if you don't attack a big market, you're highly unlikely to build a big company."
"市场的规模、市场的动态、竞争的性质。我们的目标始终是建立伟大的公司——如果你不去攻击一个大市场,你极不可能建立一家大公司。"
Don Valentine, Stanford GSB

⚠️ 流传的压缩版("We invest in markets. Not people." / "Great markets make great companies.")是真的,但已被口号化;上面这段 GSB 的完整表述才是承重的原话。

到 Botha 这一代,审美里多了一层极度稀缺意识——一年只有屈指可数的真正机会,因此每一枪都要负全责:

"We only have 35 important opportunities to invest in our venture fund."
"我们的创投基金,一年只有 35 个重要的投资机会。"
Roelof Botha, Fortune, 2025-03

而 Pat Grady 把这种"稀缺意识"翻译成一句口径——目标不是稳定输出,而是抓住每年那两三个 outlier

"Our objective is not consistency. Our objective is to find the two or three or four outliers in any given year."
"我们的目标不是'稳定一致'。我们的目标是找到任何一年里那两个、三个、或四个'异类'(outliers)。"
Pat Grady, Uncapped #36(Pat Grady & Alfred Lin from Sequoia), 2025-12-09, 05:43

在 Valentine"市场优先"的基础上,这一代把"市场"与"人"的分工说得更精确——市场决定一家公司"能"长多大,创始人决定它"会"不会长到那么大

"The market determines how big the company can get and the founder determines how big the company will get."
"市场决定一家公司'能'长到多大,创始人决定它'会'长到多大。"
Pat Grady, Uncapped #36, 2025-12-09, 49:08

以及对"失败"的态度——不是耻辱印记,而是可累积的经验(这直接喂回维度七的评估文化):

"Failure is not a scarlet letter. It is an experience that can accumulate."
"失败不是一枚红字(耻辱的烙印)。它是一种可以累积的经验。"
Roelof Botha, LDV, 2024

到 Botha 这一代,审美里还有一条非常清晰、且和它的常青结构互为表里的判据——先创造价值,再捕获价值(value creation before value capture):极少有公司能给世界创造巨大价值却建不成一门好生意,所以早期他几乎不看营收模型,只看价值主张是否独特(事实):

"I think about value creation before value capture. It is rare that a company delivers an enormous amount of value and fails to build a good business."
"我想的是'创造价值先于捕获价值'。一家公司给世界交付了巨量价值、却建不成一门好生意——这是罕见的。"
Roelof Botha, Sequoia's Crucible Moment (id 2185154), 32:40

对应到"看人",Botha 给了一个比 Valentine"市场优先"更偏人格的过滤器——最难合作的,是 mercenary(雇佣兵)而非 missionary(传教士)型创始人;为钱而非为热爱而建,会在逆境中崩掉:

"The most challenging founders are those who come across as mercenary rather than missionary. If they are building a business for the sake of getting rich rather than solving a problem they are passionate about, they will struggle in the face of adversity."
"最难合作的创始人,是那些显得像'雇佣兵'而非'传教士'的人。如果他们建一家公司是为了发财、而不是为了解决一个自己充满热情的问题,他们会在逆境面前崩溃。"
Roelof Botha, Sequoia's Crucible Moment (id 2185154), 34:33
Part 05

四、对"我怎么管一支基金"的启发

本节是基于以上事实的应用性推断([推断],是把 Sequoia 的做法翻译成"一个新基金的可迁移原则 + 各自的代价/前提"。不是 Sequoia 的原话主张。
  1. 先设计"宪法",再设计投资。 Sequoia 最深的护城河不是某个 picker,而是把"未来某代人最可能干的坏事(套现 GP、稀释下一代、品牌冲突)"在文件层面提前锁死。新基金第一年就该问:二十年后,我最怕这家机构变成什么样?哪些条款现在写下能阻止它? —— 代价:永续/锁死会牺牲灵活性与个人流动性,且高度依赖创始合伙人"愿意不套现"。
  1. 决策机制要对抗的是人性,不是信息不足。 匿名预投票、禁投中间分、指定唱反调、资深者最后发言、把投票数值留痕复盘——这些机制几乎零成本,却系统性地压制了羊群、锚定、确认偏差。一支两三人的小基金也能立刻照搬其中三四条。—— 前提:要有"对事不对人"的文化托底,否则唱反调会变成站队。
  1. 评估看"一串",不看"一笔";评行为,不评运气。 这是 Sequoia 给"如何评估大家绩效"最干净的答案。把它落地:为每个投资人建立决策档案(投了什么、当时的论点、预投票分、pre-mortem),按"过程质量"季度复盘,而非盯单笔 mark。"好行为+坏结果=OK,坏行为+好结果=不OK"应当明文写进评估标准。
  1. 把"严苛"重新定义为"支持"。 "demand excellence is the best support" 是一种可以主动建立的团队规范——但它需要先有信任地基(Leone 的"神圣三元组":信任→业绩→团队)。顺序不能反:先 demanding 后 trust,会散伙。
  1. 常青/耐心资本是结构性的 alpha,但会庇护平庸。 Sequoia 靠"什么都不做、只是有耐心"多赚了 $6.7B,并用"跑赢 NASDAQ 才收业绩费"对冲了"攒规模"的诱惑。新基金即便不做完整常青,也可借鉴这个反 asset-gathering 的设计约束思路。—— 代价:PitchBook 指出的 GP–LP 错位,必须用硬性的业绩门槛(hurdle)对冲。
  1. firm 本身是你最重要的那笔投资。 Moritz 这句话是整套打法的元原则——招聘、传承、文化、工具,都是对"机构"的投资。具体抓手:把工程/产品/设计当 first-class 团队(Sequoia 里这类人比投资人还多),用工具给小团队"超能力"而非靠堆人头。
  1. 传承要"职务化"而非"个人化",而且要提前几年铺路。 用 steward 而非 founder-name 来锚定身份;交接前先把运营层(CFO/COO)搭好;老人"被问才在、不主动塞意见"。—— 这是 Kleiner Perkins(创始人品牌、内部明星竞争)衰落的反面教材。
  1. 刻意维持不安全感。 "1990 年的顶尖 VC 多数已不存在"——把这种机构记忆制度化(duty-to-warn 的 memo、不摆 IPO 纪念品、"只和下一笔一样好")。它是反复利复利的复利。
Part 06

五、数据来源

网络研究(13 篇,缓存于 cache/web/

  1. 01-wikipedia-baseline.md — 创立、steward 时间线、2021 常青基金、2023 三分、scout 起源、AUM、FTX。
  2. 02-sequoia-fund-botha-essay.md一手:Botha 2021 常青基金文章逐字。
  3. 03-valentine-philosophy.md — Valentine "市场优先于人"、GSB 原话、人才观。
  4. 04-power-law-mallaby-sequoia.md — 《The Power Law》:文化、行为/决策科学、回报、"purge under-performers"。
  5. 05-succession-stewardship.md — Leone→Botha→Lin/Grady;stewardship 机制;交接逐字引用。
  6. 06-botha-talent-decision-philosophy.md — LDV 2024:斜率/截距、去自我/苏格拉底、2×2、"next investment"。
  7. 07-rip-good-times-2008.md — 2008 演示;"Get Real or Go Home";自我更新文物。
  8. 08-2023-global-split.md一手:三分联合信(why);Stripe/Airwallex;地缘政治推断。
  9. 09-scout-program-seed.md — Scout 起源(Botha/Calacanis)、机制、carry 分成、种子管线。
  10. 10-fund-economics-evergreen-analysis.md — Stratechery "production capital";基金规模/流动性;PitchBook 错位批评。
  11. 11-botha-fortune-2025-candor-evaluation.md — Fortune 2025:"own our failures together"、check-in 仪式、beginner mindset。
  12. 12-ftx-accountability.md — FTX 减记;LP 道歉电话会;"business of taking risk";流程变更。
  13. 13-decision-rights-and-ethos.md — 周一会/veto 规范;Sequoia "Our Ethos" 自述价值观;非营利 LP 基础。

播客访谈(~26 期已折入 + 2 期 pending,digest 缓存于 cache/podcasts/

节目主讲集(id)日期重点维度
Goldman Sachs ExchangesRoelof Botha19229412024-09-20治理/决策/经济/评估(最高密度)
Invest Like the Best EP.378Pat Grady14035902024治理/评估/自我更新/传承
20VCPat Grady14888942023评估(过程vs结果)/人才
20VCPat Grady17331102023决策(2014起留痕投票)/招聘漏斗
Invest Like the Best (CLASSICS)Doug Leone31366692023-02-28经济/传承/评估(lowlights)
Divot #1Doug Leone55626522025-03-05经济/"神圣三元组"
Invest Like the BestRoelof Botha29271422024pre-mortem/LP观
Why VC is BrokenRoelof Botha5400127 / 54202672025永续合伙/常青$6.7B/China/scout/app
Uncapped #28Roelof Botha54419712025-10-15传承时间线/不安全感
View From The Top (Stanford GSB)Roelof Botha2182978~2023"宪法"锁死/招聘即传承
Acquired — Sequoia Part 1(Don Valentine 史)393362020创立/1996 交接
Acquired — Sequoia Part IIDoug Leone393282020-02-18"桌子中央的真相"/never lose money
Acquired — Investment PlaybookAlfred Lin393052021-02-01招聘三 high/谦卑的生意
Acquired — Black Swan MemoRoelof Botha393252020duty-to-warn
Tim Ferriss #576(Moritz 名言 via Housel)1159972022paranoia
20VC(Brian Halligan 外部视角)176537外部印证 paranoia
How to Think Like a World-Class Investor #212Alfred Lin2755503Crucible moments
20Europe(Doug Leone via J. Bek)1101062"conviction of the sponsor"
Uncapped #36Pat Grady & Alfred Lin63297842025-12-09评估节奏(6月values/12月capabilities)/决策/审美/优先级栈
Invest Like the Best EP.250Roelof Botha21851542021value creation before capture / missionary vs mercenary
20VCRoelof Botha100421想象力的失败 / 董事会=减震器
20VCDoug Leone100338业绩第一信条 / team not family
Turpentine VCAlfred Lin14729942023-11-01基金规模 $200M/$700-800M、不追踪 AUM(dim3)/ 半年复盘+first-call+NPS(dim7)/ heart-of-gold(dim4)
Peak XVPat Grady37497222025-04-23long-feedback-cycle/inputs(dim7)/ drive+killer-instinct、golf-vs-cricket(dim4)/ 不抢功(dim7)
SourceryAlfred Lin57394642025-10-30shock-observers/sparring-partners + hubris-of-success(dim7)
StrictlyVC DownloadRoelof Botha62341042025-11-05辩论不为赢/共识='我们的'投资(dim2)/ mammalian-not-reptilian + ~50% write-off + "cried in partner meeting"(dim7)/ return-free-risk、剔除 top-20 跑输指数(dim3)
Axios(w/ Dan Primack)Roelof Botha62492642025-11-20交接第一手 color:transitioning-leadership=核心能力、interregnum 反停滞、23 年后交给 Lin & Grady(dim5)/ xAI 无董事会=信任替代官僚(dim1)

已折入处理(含 2026-05-29 Sweep)

  • 2185154100421100338:2026-05-29 refresh 折入(详见 changelog 与各维度)。
  • 6329784(Uncapped #36,用户指定):2026-05-29 折入维度一/二/七与审美。
  • 147299437497225739464:2026-05-29 完整性 Sweep 折入(维度三/四/七,详见 changelog)。
  • 6234104(StrictlyVC,Botha 卸任前几天):本轮 Sweep 中途处理完成,当场折入维度二/三/七(详见 changelog)。
  • 6249264(Axios,Botha & Dan Primack,2025-11-20):2026-05-29 consolidation fold 处理完成并折入——维度五(交接第一手 color:transitioning-leadership / interregnum / 交给 Lin & Grady)+ 维度一(xAI 无董事会=信任替代官僚)。直击 thin_evidence 中"2025-11 Lin/Grady 第一人称 color"缺口。

Pending(已提交处理、本轮未折入)

  • 5610319(Training Data = Sequoia 自家播客,Doug Leone × Cresta CEO Ping Wu,2025-10-14):合伙人在自家 pod 出场,少见的 partner-side 机构内容。仍在 Podwise 异步处理中,处理完成后下一轮 refresh 再折入。

证据薄弱 / 未硬凑的部分

  • 具体 carry/comp 百分比与"Sequoia Fund"法税结构细节(未公开)。基金规模已补上:Alfred Lin(Turpentine 2023)公开 ~$200M 种子 / ~$700-800M 创投 / 另有成长扩张基金,且明确"不追踪 AUM"。
  • 周一会的精确表决规则(正式有 veto,实际按资历加权,机制由 IC 写作 + Mallaby 拼出)。
  • 2025-11 Lin/Grady 交接:Uncapped #36(2025-12)+ StrictlyVC(6234104)+ Axios(6249264, consolidation fold 已折入) 已补上 Botha 第一手 steward 口径(transitioning-leadership=核心能力 / interregnum 反停滞 / 交给 Lin & Grady);剩余缺口:Lin 与 Grady 本人对接任的第一人称 color 仍少。⚠️ WebSearch(Newcomer/Fortune 等,2025-11)出现一条反叙事:有报道称 Botha 是"在部分合伙人对其领导风格提出担忧后被请下台"——与官方"成长与更新驱动的平滑传承"口径相左,已在维度五正文并列,按"事实存疑、两面并存"读。
  • Scout/seed 项目的具体经济分成(高层描述为主;基金规模已补)。

PR / 自我叙事旗标(阅读时打折)

  • "Our Ethos" 页自称"羞于用第一人称单数、不在意见报",与 Moritz/Leone/Botha 高度公开的人设自相矛盾——按"理想"而非"事实"读。
  • "Crucible Moments" 播客 = Sequoia 的内容营销(Botha 主持的创始人故事),非内部文化证据。
  • 2022-09 那篇被撤下的 14000 字吹捧 SBF 的长文,是营销反噬的具体案例——用来区分神话与现实。