首页  >>  来自播客: David Senra 更新   反馈  

David Senra - Tobi Lütke: 21 Years of Building Shopify

发布时间:   原节目
以下是Toby在对话中表达的观点和故事的全面总结: **一、公司作为技术和社会建构** * **公司是社会技术:** Toby将公司视为一种“社会技术”,它允许个人(例如,每天14小时)将巨大精力专注于某项事业,而这在教育领域之外,在社会上是不可接受的。它为深度探索提供了一个“绝佳借口”。 * **运行反事实:** 公司允许人们建立“本应存在”的事物,并通过市场进行检验。如果市场认可,能量(资金)就会回流,从而推动愿景的进一步实现。这是一种自我供资,也是可以利用的“不可思议的智慧”。 * **现代公司历史:** 现代公司的概念相对较新(只有500年历史,源于东印度公司等准政府机构)。这表明当前的公司结构是解决社会和法律问题的“路径依赖解决方案”,不一定是最佳的。 * **公司“都很糟糕”:** Toby认为“所有公司都很糟糕,包括我自己的”。他预言20年后,我们会回看现在的公司构建实践,并感到“尴尬”,就像程序员看到旧代码时会感到不适一样。这种视角推动了持续改进,防止了停滞不前。 **二、Shopify的演变与Toby的领导力历程** * **“角色扮演”时期:** 在Shopify首次公开募股(大约2015年)之后,Toby承认他“扮演”(cosplay)着一个严肃的上市公司CEO,试图模仿一个传统的60岁、西装革履的高管。他听从他人的建议,并进行大量授权,这差点“毁掉公司”。 * **疫情催化变革:** 疫情迫使他进行了彻底的重新评估。当他发现“无用项目”(boondoggles)——即随机的、互不相关的项目——缺乏战略一致性时,Toby意识到自己的方法失败了。 * **“重新推导一切”原则:** 疫情使核心假设失效。Toby的工程思维让他相信,如果任何一个公理发生改变,就必须“重新推导整个系统”(就像修剪决策树一样)。 * **高管更迭:** 他亲自审查了每个项目,取消了其中60%,并在一年内更换了所有高管,意识到信任被破坏,核心技能(如寻找高效解决方案)缺失。 * **“创始人闭门会议”的洞察:** 在危机时刻,他转向一个由被收购公司创始人组成的Slack频道,认识到他们“对员工生计负责”的共同经历。他观察到,创始人在大公司中常常是“麻烦制造者”(irritants),因为他们不会“妥协”,并且说出绝对的真相。 * **聘用创始人:** 他将创始人提拔到高管职位,意识到他们强大的自主能动性(high agency)和坚定的信念是无价的,尤其是在危机中。 * **Shopify OS:像工程一样构建公司:** Toby将公司建设视为一个工程项目。 * **用于公司结构的Python代码:** 他真的用Python代码构建了配置文件(例如,“一个经理应该管理多少人”),并使用一个“集合求解器”(set solver)来建模Shopify的最佳结构(部门、层级、人员数量)。 * **期望状态系统:** 受UI开发(例如React)的启发,他为公司实施了“期望状态系统”(desire state system)。模型代表理想状态,人力资源部门的工作是以最小的步骤将当前状态与期望状态协调一致。 * **减少政治斗争:** 该系统带来可理解性,消除政治。现在可以对招聘15名销售人员的决策进行建模,以根据商定的约束条件显示权衡(例如,损失工程师)。 * **精通/个人贡献者职业路径:** 该模型揭示了不一致之处,例如工程部门之外缺乏强大的个人贡献者(IC)职业发展路径。Shopify现在强调“精通系统”,在该系统中,专家可以获得与副总裁相当的薪酬,让有才华的工程师能够专注于编码而非管理。 * **薪酬理念:** 尽管基本面强劲,但在疫情期间Shopify的股价下跌了80%,这促使Toby彻底改革薪酬体系。 * **股价下跌的“松一口气”:** 他对股价下跌感到“松了一口气”,认为之前50倍营收估值是不可持续的“投机市场”通胀,而非真实价值。 * **员工在薪酬方面的自主权:** Shopify现在允许员工完全自主选择他们的薪酬组合(工资、股票、限制性股票单位、"Shop Cash"),并每季度调整。这有助于根据当前价值重新平衡股票期权,并赋予员工控制权。 **三、个人哲学与成长** * **书籍是“作弊码”:** 阅读书籍是“真实生活作弊码”最接近的事物,它让人能在几小时内了解整个职业生涯。他强调广泛阅读并经常更换主题。 * **对商业书籍的失望:** 他发现许多商业书籍都是由“有时间的人”撰写的,不一定是由那些真正创建公司的人撰写的,这导致了偏颇的建议(例如,一个销售人员写道“所有问题都可以通过销售来解决”)。 * **积极的自我肯定与身份改变:** Toby相信积极的自我肯定通过主动改变“神经前额皮质”来发挥作用。他通过每天写10分钟“我喜欢公开演讲”,坚持一周,从而克服了他对公开演讲的恐惧。 * **“瓶中信”:** 他给自己写定期的信息(例如,在主题演讲之后),记录见解,反思过程,并设定未来意图,强调思想的“间隔重复”(spaced repetition)。 * **斯多葛主义:** 他认同斯多葛哲学:“控制你所能控制的,了解你所不能控制的”,并专注于尽力做到最好。 * **批判过去的工作/“抨击文章”:** Toby喜欢在Shopify当前状态中发现问题,即使是他自己构建的。他称之为撰写“针对过去的抨击文章”。这不是自我厌恶,而是对改进的兴奋,将问题视为通向更美好未来的“显而易见的蓝图”。他有意“贬低”过去的工作,以确保不盲从,并鼓励创造性解决方案。 * **避免沉没成本谬误:** 他强调克服沉没成本谬误的重要性,不要对过去的工作“感到自豪”,因为它会阻碍进步。“企业掠夺者”(corporate raider)的心态有助于与过去的决策脱钩。 * **“大家是为了看起来好还是为了做得好?”:** 相比于“性格一致性”,他更看重根据更好的信息改变想法,认为这是“永远正确的作弊码”。 * **拥抱矛盾:** 他承认自己的言论可能包含矛盾,解释说不同的“层面”或“参照系”(例如,战术决策与战略决策)需要不同的操作系统和方法。 **四、人才、环境与文化** * **“公司婴儿防护”:** Toby避免“公司婴儿防护”(corporate baby proofing)(过多的政策/流程),他认为这会限制优秀员工,以保护那些表现不佳的员工。相反,他的目标是创造一个“做正确的事是直觉性的”环境,并招聘拥有敏锐直觉的人。 * **为创造力设立“盒子”:** 他将问题定义为团队探索的“盒子”,提供工具和空间,但允许在该空间内自主行动。在项目从“原型设计”阶段过渡到“构建”阶段时,他承担风险。 * **“值得最优秀、最聪明的人”:** 他的工作是使Shopify成为一家“值得”顶尖人才加入的公司,专注于尽量减少“废话”(政治、糟糕的流程),并创造一个大公司里也能感受到小公司的敏捷和影响力的环境。 * **招聘有“专长”(Spikes)和高自主能动性的人:** Shopify不优先考虑资历(Toby是高中辍学生)。他们通过考察候选人过去在“出现问题”时的反应来寻找“高自主能动性”。他们寻找“创业冲动”。 * **“钦佩”作为产品:** 公司能为员工提供的最好产品就是让他们与自己“深感钦佩”的人为伍,并能从中学习。 * **地理共识(而非硅谷):** 他有意将Shopify建在硅谷之外,以避免“交叉授粉”(cross-pollination)和模仿行为。这使得Shopify能够发展自己独特的文化,并通过成为东海岸“最佳雇主”来吸引人才。 * **“不是上市,而是创造一个公共版本”:** Shopify将其IPO视为将其身份延伸到公开市场,而非仅仅遵守规范,它寻求在美国证券交易委员会(SEC)的限制内寻找差异化的方法(例如路演视频)。 * **一级方程式赛车规则手册类比:** 他将自己(以及他所寻求的人才)视为工程师,他们阅读“规则手册”(例如SEC规定)不是为了遵守,而是为了“击败它”,找到以不同方式运作的创新方法。 * **办公室设计(Daniel Weinand的影响):** 他的联合创始人Daniel Weinand根据“没有人能比他们周围的空间更有创造力”的理念设计了办公空间。 * **小组(Pods):** 他们为5人团队创建了“小组”(pods)——既足够私密又开放,让团队能直观地自组织到最佳规模。 * **偶遇与流动:** 咖啡区的位置旨在鼓励偶遇,楼层之间通过楼梯连接,以促进基础设施和产品团队之间的互动。 * **没有“诺曼门”:** Toby对“诺曼门”(Norman doors)(误导用户的门)“根本上感到过敏”。这反映了他对细节一丝不苟的更广泛原则,确保环境促进预期行为并强化质量标准。 * **“我们办公室里的一切都来自Shopify商家”:** 他们使用他们商家群体的产品,甚至让供应商入驻Shopify平台,以采购特定的办公用品。 * **“专注度极高的公司”:** Shopify不是一个“最高薪酬策略”,而是承诺通过技能发展带来快速的职业复利回报。员工被期望“全身心投入”,因为他们将获得巨大的能力和心智模型。 **五、未来:人工智能与创业的使命** * **电子游戏作为模拟:** Toby认为优秀的电子游戏,尤其是像《星际争霸》这样的策略游戏,是模拟器,它们教授决策、资源管理、不完全信息和注意力管理——这些技能直接适用于商业。 * **AI是“星际争霸模拟”:** 他发现当前AI的发展(智能体循环、自主模型)感觉就像在玩《星际争霸》,智能体相互协调并需要他的“注意力管理”。他强调了令人难以置信的快速变化(“世界每三周都在变化”)。 * **未来艰难而有趣的岁月:** Toby认为,由于人工智能,2026年及以后将是Shopify历史上“最艰难也最有趣”的几年,衡量个人将通过他们的适应能力和重新推导一切的速度。 * **公司期望:使用AI:** Shopify内部期望员工能本能地利用AI解决问题,他把这个概念写在一份备忘录中,他相信这很快就会显得是显而易见的。 * **个人使命:促成更多的创业:** Toby的个人使命在2014年左右明确,是“促成更多的创业”。他认为自己的工作是建立事物——从代码到产品,再到构建产品的公司——并为他人加速这一过程。 * **创业是自我实现:** 他认为创业是人们实现自我价值的一种“非常真诚”的方式。 * **“信息鸿沟”:** Toby赞扬David(采访者)通过分享创业故事弥合了“信息鸿沟”。他从过去的调查中意识到,接触创业(例如,有人可以提问)是成功创业的因果因素,而不仅仅是结果。 * **AI时代的创业未来:** 他设想了一个未来,AI、即时制造和3D打印使“制造事物”变得更加简单、普遍和令人兴奋,使更多的人能够成为企业家。他认为自己和David的工作通过使这个概念易于理解和接受,为这个未来做出了贡献。

Here's a comprehensive summary of Toby's opinions and stories from the conversation, covering all the points discussed: **I. Companies as Technologies & Social Constructs** * **Companies as Social Technologies:** Toby views companies as a "social technology" that allows individuals to dedicate intense focus (e.g., 14 hours a day) to a single pursuit, which is otherwise socially unacceptable outside of education. It provides a "perfect excuse" for deep exploration. * **Running Counterfactuals:** A company allows one to build something that "ought to be there" and test it against the market. If the market agrees, energy (money) flows back, enabling further pursuit of the vision. This is self-financing and an "incredible intelligence to tap into." * **Modern Company History:** The concept of the modern company is relatively new (500-year run, derived from quasi-governments like the East India Company). This suggests that current company structures are "path-dependent solutions" to social and legal problems, not necessarily optimal. * **Companies are "Terrible":** Toby believes "all companies are terrible, including mine." He foresees that in 20 years, we'll look back at current company building practices with embarrassment, much like a programmer cringes at old code. This perspective drives continuous improvement and prevents stagnation. **II. Shopify's Evolution & Toby's Leadership Journey** * **The "Cosplay" Period:** After Shopify's IPO (around 2015), Toby admits he "cosplayed" as a serious public company CEO, trying to emulate a conventional 60-year-old, suit-wearing executive. He trusted others' advice and delegated heavily, which almost "killed the company." * **COVID as a Catalyst for Change:** The pandemic forced a radical re-evaluation. Toby realized his approach was failing when he discovered "boondoggles" – random, disconnected projects (e.g., a supermarket feature) lacking strategic alignment. * **The "Re-derive Everything" Principle:** COVID invalidated core assumptions. Toby's engineering mindset led him to believe that if any axiom changes, one must "re-derive the entire thing" (like pruning a decision tree). * **Executive Turnover:** He personally reviewed every project, canceled 60% of them, and replaced every executive over a year, realizing trust was broken and core skills (like finding efficient solutions) were missing. * **The "Founder Offsite" Insight:** In a moment of crisis, he turned to a Slack channel of founders from acquired companies, recognizing their shared experience of "being responsible for people's livelihoods." He observed that founders were often "irritants" in larger companies because they don't "settle" and speak absolute truths. * **Hiring Founders:** He promoted founders into executive roles, realizing that their high agency and deep conviction were invaluable, especially in a crisis. * **Shopify OS: Engineering the Company:** Toby treats company building as an engineering project. * **Python Code for Company Structure:** He literally built Python code with configuration files (e.g., "how many people should report to a manager") and a "set solver" to model Shopify's optimal structure (departments, levels, headcount). * **Desire State Systems:** Inspired by UI development (e.g., React), he implemented a "desire state system" for the company. The model represents the ideal state, and HR's job is to reconcile the current state with the desired state with minimum steps. * **Reducing Politics:** This system brings legibility and removes politics. Decisions like hiring 15 salespeople can now be modeled to show the trade-offs (e.g., loss of engineers) based on agreed-upon constraints. * **Mastery/Individual Contributor Track:** The model revealed inconsistencies, such as the lack of a strong individual contributor (IC) track outside of engineering. Shopify now emphasizes a "mastery system" where specialists can earn as much as VPs, allowing talented engineers to code rather than manage. * **Compensation Philosophy:** Shopify's stock price dropping 80% during COVID, despite strong fundamentals, led Toby to revamp compensation. * **Relief at Stock Drop:** He felt "relieved" by the stock drop, viewing the previous 50x revenue valuation as unsustainable "betting market" inflation, not real value. * **Employee Agency in Compensation:** Shopify now offers employees full agency to choose their compensation mix (salary, stock, RSUs, "Shop Cash") and adjust it quarterly. This helps rebalance stock options against current value and gives employees control. **III. Personal Philosophy & Growth** * **Books as "Cheat Codes":** Reading books is the closest thing to "cheat codes for real life," allowing access to entire careers in hours. He emphasizes reading broadly and changing subjects frequently. * **Disappointment in Business Books:** He found many business books written by "people who have time," not necessarily those who built companies, leading to biased advice (e.g., a salesperson writing that "every problem can be solved with sales"). * **Affirmations & Identity Change:** Toby believes affirmations work by actively changing the "neurofrontal cortex." He used this to overcome his fear of public speaking by writing "I love public speaking" for 10 minutes a day for a week. * **"Message in a Bottle":** He writes scheduled messages to himself (e.g., after keynotes) to document insights, reflect on process, and set future intentions, emphasizing "spaced repetition" of ideas. * **Stoicism:** He aligns with stoic philosophy: "control what's under your control, know what is," and focus on doing the best possible job. * **Critiquing Past Work / "Hit Pieces":** Toby enjoys finding problems in Shopify's current state, even if he built it. He calls this writing "hit pieces on the past." This isn't self-loathing but excitement for improvement, viewing problems as "obvious blueprints" for a better future. He intentionally "roasts" past work to ensure no deference and encourage creative solutions. * **Avoiding Sunk Cost Fallacy:** He stresses the importance of overcoming sunk cost fallacy and not being "prideful" about past work, as it hinders progress. The "corporate raider" mindset helps disconnect from past decisions. * **"Is everyone here to look good or be good?":** He values changing one's mind with better information over "character consistency," believing it's the "cheat code to always being right." * **Embracing Contradictions:** He acknowledges that his own statements might contain contradictions, explaining that different "layers" or "frames of reference" (e.g., tactical vs. strategic decisions) require different operating systems and approaches. **IV. Talent, Environment, & Culture** * **"Corporate Baby Proofing":** Toby eschews "corporate baby proofing" (excessive policies/processes), which he believes limits high performers to protect against low performers. Instead, he aims to create an environment where the "right thing is intuitive" and hires people with well-honed intuition. * **Creating a "Box" for Creativity:** He defines problems as "boxes" for teams to explore, providing tools and space, but allowing autonomy within that space. He assumes the risk as projects transition from "prototyping" to "build" phases. * **"Worthy of the Best & Brightest":** His job is to make Shopify a company "worthy" of top talent, focusing on minimizing "bullshit" (politics, bad processes) and creating an environment where a large company feels small and impactful. * **Hiring for "Spikes" & High Agency:** Shopify doesn't prioritize credentials (Toby is a high school dropout). They look for "high agency" by examining how candidates reacted when "something went wrong" in their past. They seek "entrepreneurial impulses." * **Admiration as a Product:** The greatest product a company can offer employees is to be surrounded by people they "deeply admire" and can learn from. * **Geographical Consensus (Not Silicon Valley):** He intentionally built Shopify outside Silicon Valley to avoid "cross-pollination" and mimetic behavior. This allowed Shopify to develop its own distinct culture and attract talent by being "the best company to work for" on the Eastern seaboard. * **"Not going public, creating a public version":** Shopify approached its IPO not as conforming to norms but as extending its identity to the public markets, looking for ways to differentiate within SEC constraints (e.g., the roadshow video). * **Formula 1 Rulebook Analogy:** He sees himself (and the talent he seeks) as engineers who read the "rulebook" (e.g., SEC regulations) not to comply, but to "beat it" and find creative ways to operate differently. * **Office Design (Daniel Weinand's Influence):** His co-founder, Daniel Weinand, designed office spaces based on the philosophy that "no one can be more creative than the space around them." * **Pods:** They created "pods" for 5-person teams – private enough but open, making it intuitive for teams to self-organize to the optimal size. * **Serendipity & Flow:** Coffee areas are placed to encourage chance encounters, and floors are cut through with stairs to foster interaction between infrastructure and product teams. * **No "Norman Doors":** Toby is "fundamentally allergic" to Norman doors (doors that mislead users). This reflects a broader principle of meticulous attention to detail, ensuring the environment facilitates the intended behavior and reinforces the quality standard. * **"Everything in our offices is from Shopify stores":** They use products from their own merchant base, even going so far as to onboard suppliers onto Shopify to acquire specific office materials. * **"Unshared Attention Company":** Shopify is not a "highest compensation play," but promises rapid career compound returns through skill development. Employees are expected to be "all in" because they will gain immense capabilities and mental models. **V. The Future: AI & The Mission of Entrepreneurship** * **Video Games as Simulation:** Toby sees good video games, especially strategy games like Starcraft, as simulations that teach about decision-making, resource management, imperfect information, and attention management – skills directly applicable to business. * **AI as a "Starcraft Simulation":** He finds current AI development (agentic loops, autonomous models) feeling exactly like playing Starcraft, with agents coordinating and requiring his "attention management." He highlights the incredibly fast pace of change ("world changes every three weeks"). * **Hard & Interesting Years Ahead:** Toby believes 2026 and beyond will be the most "hard and interesting" years in Shopify's history due to AI, measuring individuals by their adaptability and speed of re-deriving everything. * **Company Expectation: Use AI:** Shopify has an internal expectation for employees to reflexively use AI to solve problems, a concept he put out in a memo, which he believes will soon seem like stating the obvious. * **Personal Mission: Cause More Entrepreneurship:** Toby's personal mission, clarified around 2014, is to "cause more entrepreneurship." He sees his job as building things – from code to products to companies that build products – and accelerating that process for others. * **Entrepreneurship as Self-Actualization:** He views entrepreneurship as a deeply "honest" way for people to self-actualize. * **The "Information Gap":** Toby credits David (the interviewer) for bridging the "information gap" by sharing entrepreneurial stories. He realized from past surveys that having exposure to entrepreneurship (e.g., someone to ask questions) was a causal factor in successful entrepreneurship, not just a consequence. * **Future of Entrepreneurship with AI:** He envisions a future where AI, just-in-time manufacturing, and 3D printing make "making things" much simpler, more common, and more exciting, enabling far more people to become entrepreneurs. He sees his and David's work as contributing to this future by making the concept accessible and understandable.