Harvey CEO Winston Weinberg: Why You Should Reinvent Yourself Every 4 Months

Harvey CEO Winston Weinberg：为何每4个月就要重塑一次自己 Winston Weinberg分享了带领Harvey经历超高速增长时真实而混乱的现实，详细讲述了公司如何在拥有500名员工的情况下将运行率（run rate）规模拓展至1.9亿美元。他探讨了以“偏好行动”为核心的招聘理念、每四个月重塑自身角色的结构性必然，以及高管招聘与授权的具体战术。Weinberg还透露了Harvey如何通过率先攻克难度最大的律所、采用高度个性化的“痛点工厂”推介策略拿下首批企业级客户，以及为何创始人主导的销售对产品开发依然至关重要。 Harvey 联合创始人兼 CEO Winston Weinberg 将“行动偏好”（bias for action）列为最核心的招聘准则，强调快速决策与迅速纠偏远胜过耗时数月、最终却脱离现实的犹豫不决。 (0:00) 尽管公司年化营收突破 1.9 亿美元，但超高速增长的内部管理极为混乱；Weinberg 指出团队每 6 个月都必须重新证明自身价值，创始人自身也需每 4 个月重塑一次角色以突破规模瓶颈。 (0:48) 在授权与高管招聘上，Weinberg 严格遵循“三个月原则”——坚信自己必须亲自执行某项业务职能满 3 个月，才能精准评估并招募到最合适的接管人选。 (13:18) 明确的单一责任人对组织至关重要；将项目分给多个人往往会导致无人负责，就像托付两个朋友帮忙浇花，植物最终只会因过度浇水或被遗忘而枯死。 (16:34) Harvey 获取前五家客户时，通过爬取 Pacer 上的联邦法院真实诉状，利用早期 AI 挑出律师辩词中的漏洞进行精准推销，以直击要害的“痛点工坊”策略成功签下首批顶级律所。 (34:28)

Harvey co-founder and CEO Winston Weinberg attributes the legal AI company's scaling to extreme internal urgency, single-owner accountability, and aggressive upmarket positioning. Managing through hypergrowth, Weinberg argues that founders must routinely reinvent their leadership structures and operating habits to prevent organizational bottlenecks from stalling momentum. Hypergrowth and the Four-Month Reinvention Cycle (0:48) Harvey operates at a $190 million run rate (up 4x year-over-year) with 500 employees (up 2x year-over-year), yet Weinberg describes the internal environment as chaotic and constantly lacking basic machinery across product and go-to-market functions. To maintain momentum, he operates under the belief that both the company's market standing and each employee's role must be re-earned every six months. For Weinberg, accumulating operational friction creates a recurring bottleneck roughly every four months, requiring structural reorganizations, executive hires, or scope reductions to unlock further scale. Weinberg identifies a "bias for action" as his primary hiring criterion, arguing that making a fast mistake and correcting it within a week is far better than agonizing for months while market conditions change. However, he acknowledges that an obsession with fixing broken processes can backfire by neglecting high-performing departments, missing opportunities to allocate resources and accelerate successful teams tenfold. Absolute Accountability and Organizational Structuring (9:26) To avoid execution stalls, Weinberg insists on strict single ownership for every project, citing an analogy shared by an employee: asking two friends to water a houseplant guarantees it either dies of thirst or is overwatered to death. He views shared accountability and the tendency to define multiple top priorities—such as an employee inventing a "P00" priority—as fear-driven hedges designed to diffuse blame when an initiative fails. When evaluating executive candidates, Weinberg requires them to map out their department's organizational chart across 3, 6, and 12 months to verify they can build leverage rather than getting bogged down in execution. He also adheres to a personal rule of performing any job function for at least three months before hiring someone to manage it. When "layering" executives above high-performing employees, Weinberg stresses transparent communication and inclusion in hiring, noting that about 50 percent of layered employees depart within a year—an outcome he considers preferable to retaining a disengaged former star. Executive Leverage and Tiered Communication (25:47) Harvey structures executive responsibilities to maximize leverage. Chief Operating Officer Katie Burke oversees cross-functional initiatives across engineering, product, design, marketing, and sales, reflected in her sending roughly 81,000 Slack messages in a year compared to Weinberg's 85,000. Meanwhile, a chief of staff aggregates data feeds into strategic dashboards, and an administrative assistant manages logistics—a role Weinberg resisted hiring for two and a half years until an all-day flight booking issue forced delegation. Daily execution revolves around Weinberg’s master document, "the list," which tracks quarterly goals, product roadmaps, UX/UI priorities, and metrics like renewal rates and DAU/MAU ratios. Communication is tiered: Weinberg rejects traditional anti-skip-level rules, sharing 95 percent of his raw strategic thoughts with an extended leadership group of roughly 30 leaders and key individual contributors, while shielding the broader company from speculative ideas to avoid distraction. He has also shifted away from founder heroics toward clearly defining goals and stepping back to let teams execute. Upmarket Enterprise Sales and Application-Layer Focus (36:36) Departing from startup playbooks that target mid-market clients first, Harvey targeted Fortune 100 corporations and global law firms from inception. Weinberg reasoned that commoditized legal tasks would quickly be absorbed by generic foundation models, whereas the demanding workflows of major firms handling cross-border M&A served as a rigorous proving ground. Clearing bank-grade compliance hurdles early streamlined subsequent sales, with corporate legal teams accounting for 42 percent of recent quarterly revenue. To secure early customers, Weinberg used a "pain factory" tactic: downloading federal court filings via Pacer, using AI prompts to identify flaws in an attorney's active motion, and presenting the weaknesses during cold pitches. Weinberg treats sales as essential to product design, noting his best product decisions happen during active customer calls. Furthermore, enterprise requirements like cross-organizational collaboration and persistent document memory ("Memory") represent immediate priorities for legal workflows that frontier AI labs have yet to prioritize, giving specialized applications a durable edge. While Weinberg's high-intensity operating model and enterprise focus have driven rapid growth, the discussion highlights ongoing operational tensions around founder micromanagement, a 50 percent turnover rate when layering early talent, and the personal sustainability of navigating continuous four-month reinventions.

Harvey联合创始人兼CEO Winston Weinberg指出，在企业超速增长的复杂环境中，决策迟缓的代价远高于试错成本；创始人必须建立明确的单一责任制，并在业务剧烈演变中每四个月重塑一次自我定位与组织结构。 超速扩张的混乱现实与四个月重塑周期 (0:48) Harvey目前年化运行率（run rate）达1.9亿美元，较上年实现4倍增长，团队规模翻倍至约500人。尽管联合创始人被媒体赋予纸面财富光环，但Weinberg透露公司内部极度混乱，不仅产品与商业化底层机制仍显匮乏，团队还长期承受极高的赶超压力。在Weinberg的评估体系中，市场地位与员工个人价值每六个月就必须重新证明一次；而在决策层面，花费三至六个月反复纠结的危害远大于一周内犯错后迅速修正，因此“行动导向”（Bias for Action）成为其最重要的招聘标准。 然而，创始人长期执着于解决业务缺陷也带来了反面效应：Weinberg常因过度关注问题而忽略了运转极其出色的部门，错失了为绩优团队追加资源、实现数倍加速的机会。由于各种未决问题通常会在四个月内积聚至心理饱和点，他被迫将“每四个月重塑自我”作为释放系统压力的固定节奏，通过高管招聘、架构重组或业务剥离为下一阶段的扩张清空障碍。 单一所有权原则与高管杠杆架构 (16:34) 针对规模化过程中的执行阻滞，Weinberg指出核心症结往往是权责不清。他引用了“枯死植物”的比喻：若委托两个人同时照顾一盆植物，植物必然会因双方推诿而干死或因重复浇水而被淹死。类似地，团队在排定优先级时经常因恐惧承担失败后果而设定多个P0甚至造出“P00”，从而分散问责风险。CEO的核心职责正是打破这种推诿机制，为每个关键项目明确指定唯一的直接负责人（DRI）。 为了有效委派与管理，Weinberg执行一套严格的领导力标准：在将某项职责交接给新负责人前，他坚持自己必须先亲自从事该岗位至少三个月；在面试副总裁及高管时，他要求候选人现场绘制未来3个月、6个月及1年的组织架构图，以考察其迅速构建杠杆团队、避免陷入微观执行的能力。此外，他提拔前HubSpot高管Katie Burke出任COO，由其统筹跨部门复杂项目（如重构试点流程），通过其极高的沟通协同体量（年发送Slack消息约8.1万条，仅次于CEO的8.5万条）为CEO创造业务杠杆。在空降高管对原有骨干进行“分层管理”（Layering）时，Weinberg提醒这类老员工有一半会在一年内离职，管理者必须保持过程透明、提前吸纳其参与决策，以防引发负面内耗。 逆向获客路径与创始人主导的销售循环 (36:36) 不同于从低合规门槛中小客户起步的常规SaaS路径，Harvey在创立初期便直接瞄准顶尖律所与财富100强企业。Weinberg解释称，2022年通用大模型的快速迭代让他确信低端法律事务（如基础租约审查）会迅速被大模型商品化，唯有直面大型跨国并购等极端复杂的合规与安全挑战，才能建立真正的软件壁垒。在最近一个季度中，来自大型企业（主要为财富100强）的收入占比已达42%，攻克顶级金融机构严苛安全审查的经验也反向降低了后续向下渗透市场的阻力。 在获客手段上，Weinberg早期通过联邦法院系统Pacer检索目标诉讼律师最新的公开立案文件，借助AI提示词挑出其法律论点中的漏洞，再通过LinkedIn冷启动并在Zoom演示中当场展示模型对律师原案文书的批驳。这种挑起对方胜负欲的“痛苦工场”策略大幅提高了初期的成交转化率。Weinberg同时强调，创始人主导销售是产品定义最有效的途径，他在2023年中期曾因短暂脱离客户一线而导致产品路线图出现偏差；在面对多方演示中的尖锐质疑者时，主动安排一对一沟通往往能将潜在阻力转化为企业内部的核心拥护者。 垂直应用护城河与创始人心态的转变 (48:20) 在探讨垂直AI软件与底层大模型实验室（如OpenAI、Anthropic）的竞争边界时，Weinberg认为巨头在产品路线图上难以优先顾及复杂的跨组织B2B协同工作流（例如一家银行与十家外部律所在统一平台协同），而这恰恰是垂直应用初创公司的核心P0级机遇。此外，为了契合律师将文档视为独立且持久实体的业务习惯，Harvey工程团队自主研发了文档级持久化系统“Memory”，通过攻关基础模型尚未解决的前沿应用层架构来吸纳顶级工程人才。 面对早期成长中因学业不顺带来的自我证明诉求，Weinberg承认自己长期靠极度的“高压执念”驱动运转，甚至每周刻意安排两到三件让自己极度不适的事来迫使自我进化。但伴随公司组织体量的扩大，他开始意识到依靠创始人单打独斗的“个人英雄主义”存在明显天花板，正在学习克制持续介入细节的冲动，转向在明确战略目标后放手由专业团队自主执行。 Harvey的高速扩张验证了以极端复杂企业级工作流构筑AI应用壁垒的策略，但随着单一主导者向多元管理层过渡，创始人如何在保持极高敏锐度与行动力的同时，完成从微观操控到体系化放权的心理跨越，依然是其面临的持续考验。

[0:48] Winston Weinberg: every six months you have to like reearn your position in the market [16:34] Winston Weinberg: It's like a plant. If you're going away for a month and you ask two friends to water your plant, the plant's definitely going to die. [18:00] Winston Weinberg: You built the wrong thing. [19:44] Winston Weinberg: Too many big things is the answer. [19:44] Winston Weinberg: No, no, no, we can do more. [25:47] Ben Horowitz (quoted by host): Why and what are you doing? [32:25] Winston Weinberg: the dumbest thing ever [32:25] Gabe: What the [__] have you been doing all day? [34:28] Winston Weinberg: I saw that you filed XYZ motion to dismiss... do you want to see Harvey analyze it and say why it's a bad argument? [34:28] Winston Weinberg: it was over [36:36] Winston Weinberg: insane still crazy bullish [36:36] Winston Weinberg: didn't know what ARR was [41:12] Winston Weinberg: I don't want to go through tens of thousands of documents and do this massive closing checklist. [44:12] Winston Weinberg: The times that I'm best at product is when I'm selling the most [46:50] Winston Weinberg: if we don't solve that, our product goes to zero [49:10] Winston Weinberg: overwhelming obsession and excitement to get up every morning [51:41] Winston Weinberg: I just want to see how far I can go and like keep going until... I break.

[0:00] I think people feel like you can't make [0:01] mistakes and that is actually the [0:03] opposite of how I feel. I would much [0:06] rather people just try and make a [0:08] decision and then it's wrong and a week [0:11] later they adjust and change than they [0:14] spend like 3 months not making a [0:16] decision and by the time that that [0:18] decision was made the entire market has [0:20] changed. Our internal or has changed it [0:23] doesn't make sense. And so right now I [0:26] think my number one hiring criteria is [0:28] literally just bias for action. Like [0:30] that's it. [0:32] [music] [0:41] [music] [0:48] Hey everybody, this is an episode with [0:50] Winston Weinberg, the co-founder and CEO [0:53] of Harvey. It's a particularly good time [0:55] to interview Winston. Uh the company's [0:58] at 190 million run rate and that's 4x [1:03] from last year. They're at about 500 [1:05] employees which about 2x from last year. [1:07] So the company is absolutely ripping. [1:10] He's at the head of a real rocket ship. [1:13] And what I was interested in [1:14] interviewing him for is I just wanted to [1:17] see what it was like when you're in a [1:19] company growing that fast and scaling [1:20] that fast. Like what's working, what's [1:22] not, how does it scale? We got into some [1:26] good stuff and he's got some real [1:28] counterintuitive takes on scaling, on [1:31] hiring, on building organizations, [1:34] on go to market. Uh he is a very [1:38] thoughtful fellow and I think you're [1:41] going to enjoy the episode. At the end [1:43] I'll come back and I'll give you some of [1:45] my personal takes on it. Hope you like [1:47] it. How does it feel to be the world's [1:51] newest at least on paper billionaire? [1:55] >> Is it is it newest? I actually not even [1:57] sure if that's true. [1:58] >> Who became a billionaire after you? [2:00] >> I'm sure like five to six people from [2:03] application layer companies. Not sure. [2:06] Um, no. I mean, I I I know we we had a a [2:10] thing in the New York Times and there's [2:11] something in the New York Post about [2:12] this recently. I mean, like Gabe and I [2:15] still live in the same apartment. Like [2:16] our lives like haven't changed much at [2:19] all. Um, and I know like I think like [2:21] some of those outlets kind of tried to [2:23] make it be like, "Oh my god, like even [2:24] billionaires can't afford like housing [2:26] in San Francisco, etc." or something [2:28] like that. You know, I mean, I I think [2:30] like there is definitely a housing [2:32] crisis in San Francisco. I don't think [2:34] like my co-founder and I are exactly [2:36] like the people that are massively [2:37] struggling with it. Um, but I will say [2:39] like we have lived in the same apartment [2:41] for like 3 years. I think it's more just [2:43] like why move? Like we like the [2:45] apartment, we like living together. Like [2:47] I don't know. [2:48] >> I think what's interesting to other CEOs [2:51] who are listening is I mean you're the [2:54] head of a rocket ship. Um, [2:58] and [2:59] it's an interesting time in the history [3:00] of the company to interview you. [3:03] What's it like? What is surprised you? [3:06] Like what is being a CEO in a company [3:09] like this that people would be surprised [3:11] about? [3:12] >> Yeah, I mean definitely is how chaotic [3:15] it is internally. That's for sure. Um, I [3:18] think that's like I think there's [3:19] there's two things I probably hear the [3:21] most is one [3:23] >> I think that we hopefully have done a [3:26] lot of the very like hard things right. [3:28] Um, and then a lot of the machine [3:29] building we it's still not there like [3:31] the machinery is still not there right. [3:33] So like the direction I think of travel [3:36] has been really good. We've made some [3:37] good product decisions etc. But like a [3:40] lot of the just like normal business [3:41] machinery is is not in place on the [3:44] product side and also on GTM. And so so [3:47] what I find is maybe two things when [3:49] people join is one they go oh I thought [3:52] this would be here like you guys are at [3:54] such a large scale you know [3:56] >> by the way that never goes away. [3:57] >> Yeah that probably never goes away right [3:59] like oh [ __ ] I thought this would be [4:00] here. Um that happens a lot and then I'd [4:03] say the other thing too is like I hold [4:06] ourselves and and Gabe does too and the [4:07] rest of the leadership team does to [4:09] extremely high expectations. Like the [4:11] reality is I think we're executing well, [4:13] but also this is just like a once- [4:15] ina-lifetime time period for everyone. [4:19] And so I think that definitely something [4:21] that happens internally is you will get [4:23] folks that from the outside they'll say, [4:24] "Oh my god, like Harvey's doing so well, [4:26] all these things." And then internally [4:28] there'll be more, oh no, we're like not [4:31] going fast enough or we're not doing [4:32] right. There isn't tons of everyone sits [4:36] around every Friday and you know claps [4:38] etc. Um, and even at the end of the [4:40] year, right? Like my memo to start the [4:42] year was, "Hey, we did a good job last [4:44] year, but these are all the things that [4:46] we really now need to catch up to." Um, [4:49] and I really believe that like it it [4:51] feels like every six months you have to [4:53] like reearn your position in the market [4:56] and that trickles down through the whole [4:58] company where you have to reearn your [4:59] position at the company, including [5:01] myself like and every six months you [5:03] absolutely have to change and if you [5:06] don't, I think you just break. Okay, I [5:08] got a bunch of questions on that. When I [5:11] was CEO, I was always criticized [5:14] that I was very good at finding the [5:18] problems. [5:19] Uh very good at the worries [5:24] and fi in like obsessing over what was [5:27] going wrong and was not very good at [5:29] celebrating the successes. Um I got [5:33] every one of my reviews every year and I [5:35] tried to get better. I just wasn't good [5:36] at. Are you the same? [5:39] >> Yes. And I I think like part of it is [5:44] good, I think, and part of it is bad. [5:45] And I'll give maybe a couple examples of [5:47] it being good and bad. I think on the [5:49] the good side, [5:52] if if something's going super well, I [5:54] just don't think about it, right? And [5:55] I'm going to go spend time on something [5:57] else, right? [5:58] The bad side is sometimes that thing [6:01] that's going super well would be doing [6:04] 10 times better if you spent more time [6:05] on it, right? And so there have been [6:07] instances in the past where I've said, [6:10] you know, I've kind of someone at the [6:12] company is doing incredibly well running [6:13] an org, right? And I spent a lot of time [6:15] with them, but I don't spend tons of [6:17] time like hands-on running that part of [6:18] the org, right? And then I realized [6:21] after like a month or so, wait, this org [6:23] could just be 10 times higher if we were [6:25] spending more resources on it, if we [6:26] promoted this person. and if we gave [6:28] them more under them, etc., right? And [6:30] so I think the the positive of it is it [6:33] really is you should spend the most [6:35] amount of your time unplugging or fixing [6:38] the biggest problem at the company, [6:39] right? And as we scale, a lot of what [6:41] I'm doing is just figuring out how do I [6:43] get bandwidth to actually do more and [6:44] more of those, right? The bad side of [6:47] that is sometimes there is something [6:49] that it could be accelerating way more [6:52] and you're overlooking it because you're [6:54] it's going really well, right? That I [6:56] think is the the thing I've gotten it [6:58] wrong is when I'm only paying attention [6:59] to the negative. So, it's not about [7:01] celebration. It's more like, wait, this [7:03] is working or this person is incredible [7:05] at their job. They should go to try [7:08] something even harder. [7:09] >> I've been on the journey with you for [7:10] the last 18 months or so. A lot has [7:13] changed [7:16] in in in very fast scaling like you're [7:18] doing. What has broken? Like what breaks [7:21] along the way? like at HubSpot just like [7:24] as we [laughter] feel like everything [7:26] broke like what are the things that [7:27] founders can learn [7:29] >> from your journey over the last couple [7:32] years. So, one of the first ones is [7:35] because you're and it ties back to what [7:37] you were saying because you're only [7:38] paying attention to what is going like [7:40] wrong right at the at the time at the [7:43] company there. Every four months I'd say [7:47] I get this like mental block that is [7:50] there's too many things going wrong on [7:52] at the company and I can't tend to all [7:54] of them. that happens every month. [7:57] >> And what that means, so far we've [7:59] survived them every time, [8:01] >> but what that means is I need to there's [8:03] a different leadership hire I need. [8:04] There's a different structure in the [8:06] company that I need or there's something [8:07] I need to cut from the company, right? [8:10] >> And so it almost is like you have to [8:13] just reinvent yourself as a founder [8:16] every like four months or so, otherwise [8:19] you are not going to be able to fix all [8:21] of the things that are going wrong at [8:22] the company. Um, and I feel that [8:24] pressure like it's weird like it I'm [8:26] like f and then once I do it I feel very [8:28] good for like the first month and the [8:30] second month I start feeling worse and [8:32] then the third and then the fourth and [8:34] all of a sudden it is like this pressure [8:35] system and you have to figure out some [8:37] way to like relieve that pressure system [8:39] and then it unlocks the next thing of [8:41] scale. So [8:42] >> okay for me [8:45] it was it often I don't know if this is [8:47] I'm just kind of testing if it feels the [8:49] same. For me it was kind of cranking [8:52] along. Problems would start, dissension [8:56] would start, things would get stuck. And [8:59] like the thing that would unstuck it for [9:02] me is a lot of the slowness or a lot of [9:04] the things that were stuck was because [9:07] decisions hadn't been made. Trade-offs [9:10] hadn't been made. And a lot of that was [9:12] my being unsure in delaying that [9:15] decision. So every once in a while, I [9:16] don't know if I was four months, be [9:17] like, I just need to make some [9:19] decisions. I might be wrong, but I need [9:21] to make decisions so the damn war can [9:23] move forward. Is that how you kind of [9:24] feel about it? [9:26] >> Yes, I the the only thing I would say is [9:29] for better or for worse, I make [9:30] decisions insanely fast. Um, which is [9:33] definitely can be a a pretty big [9:35] problem. What ends up happening is I [9:37] haven't decided who is the decision [9:39] maker, [9:40] >> right? And then that doesn't scale. [9:42] That's a huge problem. like almost every [9:45] problem that I run into, we have si like [9:47] six DRIs on and it's just it's [9:50] impossible. Like no one was going to [9:51] make a decision about this in the first [9:53] place. [9:54] >> Um and then I also think that one thing [9:56] that I I definitely messed up on it in [9:58] the beginning and I'm trying to get [9:59] better at is when you're a really fast [10:02] growing company and everyone feels like [10:04] there's a lot of pressure on you, I [10:06] think people feel like you can't make [10:08] mistakes. And that is actually the [10:10] opposite of how I feel. Like if I would [10:13] much rather people just try and make a [10:16] decision and then it's wrong and a week [10:19] later they adjust and change then they [10:21] spend like 3 months not making a [10:24] decision. And I think something I did [10:26] wrong and I am improving it is [10:29] >> sometimes people would think that they [10:31] were penalized or that I was upset about [10:33] something or leadership was upset about [10:35] something because they made the wrong [10:37] decision and it failed. [10:39] >> Not true. it's because they took six [10:41] months to make the decision and by the [10:43] time that that decision was made, the [10:45] entire market has changed. Our internal [10:47] orc has changed. It doesn't make sense. [10:49] And so right now, I think my number one [10:52] hiring criteria is literally just bias [10:55] for action. Like that's it. It is just [10:57] are you willing to take action and learn [10:59] from those actions and then make a [11:00] decision based off of that or iterate. [11:02] >> Okay, let's talk about that. You've [11:04] hired a lot of people. the exec team has [11:07] gone through a lot of change for you, [11:09] which I think is typical if you're [11:10] growing that fast. Just kind of walk [11:13] through some of those changes and what [11:16] you've learned about hiring and that [11:18] exec team. Like what are some of the [11:19] ahas that have hit you on the way on the [11:21] team? Like every founder I'm talking to, [11:23] they want to talk about their team and [11:25] it's like should I hire this person? [11:26] Should I hire someone from the outside? [11:28] Should I shoot this person? [11:30] >> Um what should the or structure look [11:31] like? You what what have you learned on [11:34] all that? [11:35] the okay so I can say the number one [11:37] thing that I've learned from like a [11:39] hiring perspective which is the main [11:41] thing that I ask seuite now to do if I'm [11:43] going to hire them or or VP as well is [11:46] can you draw out what your org would [11:48] look like in six in 3 months 6 months [11:51] and a year [11:52] >> it is insane how many people can't do [11:55] that like it is wild to me that that is [11:58] really hard for people to do [12:00] >> um and there's different levels of your [12:02] ability to do that right like there is [12:04] the ability of can you describe like oh [12:07] yeah I need a VP of this region and then [12:09] a VP of that region and things like that [12:11] whatever no it's actually like can you [12:14] map out every single one of those hires [12:16] what they do and can you think of who [12:19] would be a good person or who would be a [12:20] good fit for it right [12:22] >> that is like the number one thing that I [12:24] see when I hire leaders and they don't [12:27] scale is they're they do the same thing [12:30] that I do poorly or did poorly and I [12:32] think I'm getting better at which is [12:34] what is your leverage team around you, [12:36] right? Like when you join like what and [12:38] I think that there's a little bit of [12:41] founders are told like you don't want to [12:43] hire somebody who immediately wants to [12:45] go hire more people, right? Like that's [12:47] like oh they just want to build their [12:48] org. Totally. But at the same time, you [12:51] want them to create leverage as quickly [12:53] as humanly possible. And if they don't [12:55] create leverage for themselves, [12:56] especially in the like glue parts of [12:58] their function, right, [13:00] >> they're not going to be able to do the [13:02] the core competency of their function. [13:03] it's not gonna happen. [13:05] >> And I'd say that a lot of my time right [13:08] now is actually spent with my leadership [13:10] team figuring out who those lieutenants [13:12] are for them and helping us all hire [13:14] those. I had a problem with trust [13:18] uh building HubSpot. Like there's always [13:20] a very small group of people who I was [13:22] comfortable with like just go do it and [13:24] figure it out and I'm not really even [13:26] going to worry about it. And it [13:28] frustrated my team that I didn't have [13:31] deeper trust. Do you have trust issues? [13:34] >> Massive trust issues. [laughter] [13:37] Working on them. Working on them. Um, [13:40] definitely. I mean, it's it is hard to [13:42] let go. I also will say one thing I the [13:46] hires that I've done a good job on are [13:48] the ones where I had I was forced to do [13:50] the role for like three months. Yeah. [13:52] >> Um, I mean, I've never done this before. [13:54] Like, I've never started a company [13:55] before. And so if you have no idea, like [13:58] if you have spent zero time doing that [14:01] role, you are going to hire the wrong [14:03] person 100% of the time. Like there's [14:04] just no way, right? [14:06] >> Um [14:07] >> it's hard for me to then get out of [14:09] that. Right. [14:10] >> Right. Like it's hard for me to like, [14:11] okay, that was my job and now I have to [14:14] delegate part of that job. [14:16] >> It's hard for me. Um, I'd say I kind of [14:19] go back and forth on this of hire more [14:22] leaders and give them more [14:23] responsibility and then also expand the [14:26] group that I talk to daily, right? So, [14:29] one thing I'm trying to do a lot more of [14:31] is we have like a in this is really [14:33] simple things like in Slack we have a [14:35] leadership like seuite and then we have [14:37] VPs and heads of and some other folks [14:40] that are really good IC's etc. Right? [14:42] The leadership one is like seven people. [14:45] The other one is almost 30. Right? All [14:48] of my thoughts about the market, general [14:51] thoughts about the company, product [14:52] direction, all of those that I'm [14:54] constantly putting in there, I used to [14:56] put them into only the seven people. Now [14:59] I try to put 95% of it into the 30. [15:02] >> Why not the whole company? [15:04] >> So for the whole company, I I thought [15:07] about this too. I thought about just [15:08] constantly posting in general chat. The [15:11] problem with that is it just gets too [15:13] distracting. [15:14] >> Like we did that in the beginning of the [15:16] company and it it just creates too many [15:18] silos and and the reality is like some [15:20] of the stuff I post in there is like [15:22] crazy talk. Like it's like hey in like [15:25] two years we should do XYZ or like this [15:26] is a crazy idea and you don't want too [15:29] many like you want people that [15:30] understand that that doesn't mean okay [15:32] let's go build that from scratch, right? [15:34] Um you want to make sure that it's like [15:36] a little bit tighter of a group than [15:37] that. But I am trying to expand that [15:40] group as much as I possibly can. Um cuz [15:43] I don't want that. I don't believe in [15:44] the whole idea of you can never talk to [15:46] a skip level and you should never do [15:48] that. Like I think that's [ __ ] Like [15:50] I I don't believe in that at all. Okay. [15:52] Let me just ask a couple of very [15:53] tactical questions that CEOs ask me all [15:55] the time. [15:56] >> Yeah. [15:57] >> Uh [15:58] you just brought up heads of as a title. [16:01] What the hell is that? Why do you have [16:03] them? What's wrong with director or VP [16:05] or manager? What the hell is a heads up? [16:07] Like look at a LinkedIn profile. [16:09] Somebody's a head of like what is that? [16:12] >> Um I like head of actually more than [16:14] director. Um I I think as simple as [16:17] possible [16:19] director makes it feel like you aren't [16:21] as responsible for the one thing in the [16:24] box basically and head of I feel like I [16:28] know it's a subtle difference but it [16:30] gives you more this person. [16:32] That's a good explanation. I get that [16:34] >> it's all ownership. Like I think of all [16:36] titles as just ownership because the [16:38] thing that everyone likes to do is just [16:40] kind of squirm them out of ownership, [16:42] right? And so if you can and it seems [16:45] like all of the problems are just it's [16:47] unclear who's in charge of this, right? [16:49] >> Um and even again like going back to the [16:52] I very much don't penalize mistakes, I [16:55] penalize in action. If there isn't a [16:58] head of for a particular thing, then [17:01] you're in trouble. By the way, I have an [17:02] analogy for you. The DRRi thing hit me [17:05] square in the head because one of my [17:06] employees said, you know, it's like it's [17:09] like a plant. If you're going away for a [17:11] month and you ask two friends to water [17:12] your plant, [17:14] >> the plant's definitely going to die. [17:15] It's either going to get no water or [17:18] just constant water and it's dead. Do [17:21] you feel is that the way you feel about [17:23] it? [17:24] >> Yeah, 100%. That's how I feel about it. [17:26] And then it takes it takes a long time. [17:28] And so I'll give you an example of I [17:30] think the job of a CEO is like actually [17:35] figuring out who is that person, right? [17:37] And so we were we were doing this pilot [17:39] process. There were a bunch of people [17:40] and I started with okay, should I like [17:44] should I say, hey, someone needs to be [17:45] in charge of this, right? That was [17:47] option one. That usually doesn't go that [17:49] well. Instead, it is this person should [17:52] be in charge of this, right? And [17:54] >> why why can't you get feedback on who [17:56] should be in charge? And [17:58] >> so you can do that, but sometimes that [18:00] takes a really long time, right? And and [18:02] what I've found is is a lot of times [18:06] it's the same as a product road map to [18:07] be honest. Like people when you say what [18:11] is the P 0? What does almost everyone do [18:14] on a product road map? Okay, so here's a [18:16] P 0, but then also there's a P1 and then [18:18] there's a P2 and actually let's have [18:21] three P 0. And actually recently, last [18:23] quarter, uh, someone I definitely made [18:26] fun of them for doing this. They made a [18:27] P00, [laughter] [18:29] >> okay? [18:30] >> Because they didn't want the P. They're [18:32] like, it's too hard to choose, right? [18:34] And and the reason people do this, and I [18:36] do this too, is you don't want to be [18:38] wrong. And if you say there's three [18:40] things that are P 0, you can hedge your [18:41] bets. And if you built the wrong thing [18:43] and you said, "Hey, this is a P 0." and [18:46] then a quarter goes by and no one bought [18:47] it or no one used it, you can be like, [18:50] you know, sorry, if you have three P 0, [18:52] you can be like, oh well, we built all [18:54] three and like some people kind of like [18:55] this one and some people kind of like [18:56] this and it's like very diffuse. If you [18:59] said at the beginning of the quarter, [19:00] this is the P 0. You built that and no [19:02] one used it. I mean, [19:04] >> you built the wrong thing. And it's the [19:06] same as the DRRi where I think people [19:09] get uncomfortable with having the DRRi [19:11] because then there's shared there's [19:12] shared ownership if you don't do it. and [19:14] you can say there's three people in [19:16] charge of it and then they can all go [19:17] like this and point at who whose fault [19:19] it is, right? [19:21] >> If you just say that person's in charge, [19:23] they're in charge. And it could, you [19:25] know, it might not be their fault, but [19:27] like at the end of the day, like when we [19:28] put up quarter numbers, when we have [19:30] product numbers or anything like that, [19:31] it's my fault no matter what. Like no [19:33] matter what, our performance as a [19:35] company is my fault no matter what. And [19:37] if you have a DRRi in each spot, there [19:39] is somebody who is like that and feels [19:41] that about each thing. [19:43] >> Yeah. [19:44] One of the things I noticed about being [19:46] a CEO is as the company got bigger, we [19:49] could get less things done. Yeah. [19:52] >> But we get them done very well. Have you [19:54] noticed that as well? And how many big [19:56] things are you working on right now? [19:59] >> Too many big things is the answer. [20:00] >> Okay. How many you got? [20:01] >> Um, [20:03] >> we would have four and I would try to [20:06] get it down to four. [20:07] >> Uh, like productwise, we have way too [20:10] many things. And I keep doing this. I [20:11] mean part of it is we're building a [20:12] platform and I do think you can build a [20:14] lot faster now and you can build these [20:15] things in parallel. [20:16] >> Um but we still I am by far the main [20:19] problem here of every quarter I end up [20:21] being like no no no we can do more we [20:22] can do more we can do more [20:24] >> and then as like company initiatives we [20:27] also do too many things. Um and like [20:29] we're in 60 countries operating across [20:32] we have in-house customers and then we [20:34] have law firms they have different needs [20:36] >> and we are honestly still trying to do [20:38] kind of it all. Um, and so I do think we [20:43] are getting better at saying no, this is [20:46] the P0 and that we do really well. [20:50] >> I don't think it's we're not at the [20:52] stage where we do four if we pushed it [20:54] down to four things. I think we do [20:55] incredibly well at that. We're still at [20:57] the stage where I think the P 0 will [21:00] always be done extremely well as long as [21:02] there's repetition and not changing the [21:04] P 0 every week, which I definitely do [21:06] sometimes. Okay. [21:07] >> Um, for sure. Um, for sure. Um, as long [21:11] as there's consistency about that P 0, I [21:14] think that we can do a P 0 well and then [21:16] we'll do everything else pretty well and [21:18] we'll do them well enough to eventually [21:20] they'll become a P 0, then you jump onto [21:22] that. And I do think you can take more [21:23] bets right now. Like you can you can [21:25] build really fast. [21:26] >> More [ __ ] done now. [21:27] >> You can. Yeah. [21:28] >> Okay. I have a lot more questions. [21:30] There's two on the org. [21:33] There's two of the CEOs I work with that [21:35] are in the process of layering a very [21:37] good employee. [21:39] Um, [21:40] and there's a lot of stress that CEO is [21:43] feeling a lot of stress about that [21:45] layering about the signal and sense of [21:47] the company about that individual [21:49] employee losing that employee. You've [21:52] done a ton of this type of thing. [21:56] >> Mhm. [21:56] >> Advice to those CEOs who are in that [21:58] process. tell them about it. Uh, and [22:01] I've done this wrong too. Like I have [22:04] this is I mean it's it's actually [22:06] weirdly similar to the P 0 problem of I [22:09] think the reason people do this wrong is [22:10] because they're scared, right? Where [22:12] like they're scared of making the wrong [22:13] decision and they're scared of I mean a [22:16] lot of things, right? Like it takes time [22:17] to hire somebody really well like to [22:19] hire someone who's very good at this job [22:22] and do you want to bring them along for [22:24] that hiring process, right? And in some [22:26] instances I have, some instances I [22:27] haven't, right? And you have to [22:30] basically say this is how I do am doing [22:33] it going forward and this is my plan is [22:36] I am going to bring that person along [22:37] for the hiring process. And this maybe [22:40] goes back to what I said before which is [22:42] I think there are almost everyone that [22:44] is doing their part of the company can [22:46] do a better job than me. And if they [22:48] can't I am hiring someone who can. Like [22:50] I should not be the best person at every [22:52] single role at the company. God know [22:54] that's terrible. You're a terrible [22:55] leader if that's what you're doing. But [22:57] I do know how to do all of the roles to [23:00] at least a the company doesn't go to [23:02] zero state, right? And so number one is [23:07] make sure you understand every part of [23:09] your company so that if that person did [23:11] leave that superstar, you could take [23:13] their place while you hire the new [23:15] person. You need to be able to do that. [23:17] That will reduce your fear. Two, bring [23:19] that person along. If you really think [23:21] they're that good, it is worth the [23:24] chance that they will stay. And the only [23:27] way they will stay is if they trust and [23:29] respect you. And they have to trust that [23:31] their career growth can be really high [23:33] even with a new person on. And so you [23:35] just have to take that risk. Um, and [23:37] I've done I've done an okay job of this. [23:39] There's been instances where I've done a [23:40] really bad job. [23:42] >> Um, and the reason I've done the bad job [23:44] is because [23:45] >> I'm scared that that person is just [23:48] going to leave and we're going to have a [23:48] huge gap at the company. But at the same [23:50] time, I don't think that person is the [23:52] right for 3 months from now or 6 months [23:55] from now. we're going too fast. And you [23:56] you just you have to feel a confident in [23:59] your own ability to fill and b you just [24:01] have to trust them and bring them along [24:03] for the ride. And if you do bring them [24:04] along and they do want to stay, they're [24:07] going to even work. They're going to be [24:08] better at their job and they're going to [24:09] learn from that person. They're going to [24:10] act like they're going to treat them [24:11] like a mentor. I got layered twice in my [24:14] care. So I'm old. I got layered twice in [24:16] my career. Uh once [24:21] there was a lot of heads up and a lot of [24:22] feedback that I I I wasn't quite ready [24:26] and that we might bring someone over me [24:28] and then they did it and I was involved [24:30] in the interviewing process and I stayed [24:32] and I was happy and I learned a lot from [24:34] the person they brought over me. [24:35] >> Yeah. [24:36] >> In another instance I kind of got happy [24:39] talk and I thought I was doing pretty [24:41] well. Maybe didn't think I was killing [24:43] it but pretty well. And out of the blue [24:44] they said, "Hey, we're going to bring [24:45] someone over you." I I left in that [24:48] case. And my sense on this stuff is just [24:52] surprises are never good. Feedback's [24:55] underrated. [24:56] Um and overall, if I look at the history [24:59] through HubSpot and PTC on my whole [25:01] career, like when layering happens, [25:03] about 50% of the time that person who [25:05] gets layered a year later is gone. [25:08] >> Yeah. Well, it could be worse. gone is [25:12] actually potentially better than what [25:14] could happen, which is they're there but [25:16] they're not they're dragging everything [25:18] down. Yeah. Right. And and especially if [25:20] they were a star because that like if [25:24] they were a star then that's an [25:25] incredibly important part of your [25:26] business and they will really block the [25:28] person from joining, right? Like they [25:30] massively block that person. Um I don't [25:33] know. I I we've we're going through more [25:35] of those transitions right now. And [25:38] >> to me, if I could redo some of the [25:40] stuff, it would be bring them along even [25:42] more. I see. [25:43] >> Like even more. Bring them along and [25:45] just risk it. [25:47] >> Okay. Uh [25:49] you hired Katie Burke, HubSpot's chief [25:51] people officer. [25:52] >> Yep. [25:53] >> Uh [25:55] tell me about her, why you hired her, [25:58] and then you just promoted her to COO. [26:01] What does a COO do inside a company like [26:05] yours and should companies like yours? [26:06] Ben Horowitz wrote a really good book [26:08] called The Hard Thing About Hard Things [26:10] >> and he was very negative on the idea of [26:13] a COO [26:14] >> and I read his book about a month before [26:18] I pitched him at Andre and Horowitz and [26:22] I had just hired a COO Sherman great guy [26:26] and I needed a COO. I needed someone to [26:28] keep my trains on time and we went into [26:30] the pitch. my co-founder Dharmash, [26:32] myself and JD and we sat in the order JD [26:34] first then me and then Darmash and I had [26:37] forgotten that Ben didn't like the idea [26:40] of COOs [26:42] and we and we got to intros and the [26:45] first person with JD who had just [26:47] started the COO and Ben just was like [26:51] why and what are you doing and like was [26:53] kind of really negative on that idea. it [26:56] turned out to work for us. And Ben may [26:58] have changed his mind about this. [26:59] Actually, he's coming on the pod. I'm [27:00] gonna ask him about this. And he's [27:02] obviously very wise. Uh [27:05] what does your COO actually do? [27:08] >> Yeah. I mean, so a couple things here. [27:11] One, uh there are one nice thing about [27:14] GTM is you have objective measures, [27:17] right? So you can be like, okay, did [27:18] they hit these quotas, etc. Each and you [27:20] can kind of test, right? Harder on [27:22] engineering like sure you can do like [27:24] code shipped, etc., But not an insanely [27:27] good metric, right? One maybe metric for [27:29] Katie. Um, last year we just did like [27:31] our year in review or whatever today and [27:33] I sent I think it was like 85,000 slacks [27:36] or something like sent that that many [27:37] slacks. [27:38] >> She was she was I think second and it [27:41] was like 81,000 [27:44] or something. Third place was in like [27:46] the 60s something like that. So there is [27:49] maybe a little bit of an objective [27:50] measure. Not that Slack scent is like [27:52] performance, but it does have some [27:56] indicator of like how involved you are [27:58] in the company, right? And she Katie [28:03] basically was was the COO probably 6 [28:06] months ago, I would say, in all but [28:07] title, right? And what that I think it's [28:10] different for every company, right? So I [28:12] am definitely one of the founders who I [28:14] like being knowing about every single [28:16] part of the company. So like I like [28:18] thinking about operating cadences. I [28:20] like working on product. I also like [28:22] working on GTM. I like working in [28:24] partnerships. I work like working on [28:25] hiring. I like working on everything. So [28:27] it there are some COOs where what you do [28:30] is basically, hey, you go run GTM, you [28:32] go run every part of the business and [28:34] I'm only going to work on product and [28:35] that's what I'm doing. And that is [28:36] awesome and it works for a bunch of [28:38] folks. [28:38] >> There's a bunch of companies doing that. [28:40] >> Tons. And and at that point, I think the [28:42] role of the COO is quite literally just [28:44] like GTM. Like a lot a lot of it is GTM. [28:47] Like trains on time, sure. I think a lot [28:50] of it is they just they kind of [28:51] basically run GTM [28:54] >> that that's not what Katie's doing. What [28:56] Katie is doing is basically every other [28:58] part of the business um combined. And so [29:02] she her role is like a lot closer to a [29:06] partner like throughout the entire [29:08] process. So you know I have like an [29:11] operating cadence personally. I have a [29:13] it's called like the list. Um, and it [29:15] basically has all of the things like all [29:17] of my docs, all the things I'm tracking, [29:19] etc. And then it has like a daily and [29:21] then I copy paste and refresh it every [29:23] single day. Um, and you go through. [29:25] >> Wait, talk about a little more detail. I [29:26] want to hear that. [29:27] >> Yeah. Um, so I have a list that is [29:29] basically like I have focus or operating [29:31] cadence of the company, which is like [29:33] all of our quarterly goals, things like [29:35] that. And then I have uh like product [29:37] road maps, PRDS, everything like that. [29:40] And then I have things that I want to [29:41] remember. You're actually on there. I [29:43] have basically a quote from every [29:45] person. Yours is no. Uh it's just Brian [29:47] and then dash no is what it says [29:49] actually. [laughter] There's some other [29:51] ones on there, but I'll I'll tell you [29:53] you're actually on there. Um and and [29:56] then I have basically like metrics that [29:58] I'm tracking. So like renewal rates, [30:00] like product usage, DAU over MAU, and [30:02] then I have like star focus, which is [30:05] basically like one thing I care about a [30:07] lot right now is improving our the UX UI [30:10] of the company. So I'll put like every [30:12] single doc that is relevant to that and [30:13] to tracking that and any of the ideas, [30:15] right? And then I have my daily list, [30:18] right? And on that daily list, every [30:22] morning like when I wake up, I go [30:24] through that list and there's multiple [30:26] things on there that I send to Katie [30:29] >> and Katie is able to go through and [30:31] these aren't like small things, there's [30:33] big things and they're all [30:34] crossunctional [30:35] >> like what [30:36] >> um like we need to revamp our our pilot [30:38] process. Yeah. [30:39] >> Right. So that's a a really good example [30:41] where revamping pilot process revol is [30:44] engineering, product, design, GTM, post [30:47] sales. I mean it's literally every [30:49] marketing. It's every single part of [30:52] >> all of it. Right? [30:54] >> And I will like originally it was I [30:57] would just try to do all of that myself [30:58] and wrangle all of that. Right? Now it's [31:01] hey could you help me with this? Come up [31:02] with a decision and then bubble up three [31:05] decisions to me. And what ended up [31:07] happening from that process is like a [31:08] lot of people were involved. my chief [31:09] business officer was involved, CTO's [31:11] involved, had a product VP of product, [31:12] all of those people, right? [31:14] >> And then I still just made a decision at [31:16] the end of the day, right? [31:17] >> And if you can scale that, like if you [31:20] can get a COO who can start doing that, [31:22] you're basically just duplicating all of [31:25] the crossf functional initiatives that [31:27] you can handle. Like earlier on you [31:29] said, hey, how do you pick your [31:31] priorities? How do you make sure you're [31:32] doing all of these things? This is how [31:34] you increase leverage. Like this is how [31:35] you have someone that can help with [31:37] this. The other thing that she's [31:39] incredible at, because she did come from [31:40] being a chief people officer, is the [31:43] number one thing at our company is [31:45] hiring. And I don't mean hiring new [31:47] people necessarily. I mean hiring, [31:49] retention, making sure that there's one [31:51] person in charge of each role and making [31:53] sure that the people are really good are [31:55] not being pushed down by people above [31:57] them that are not as good. Right? In the [32:01] number one most important thing at the [32:02] company, nothing else is even close, [32:04] right? [32:05] >> She's really good at that. And so she's [32:07] very very helpful at thinking about what [32:09] is the org chart, what should this look [32:11] like, and who is that person, who should [32:12] be in charge of that, who should be [32:14] working across from XYZ. And she's very [32:17] good at also hiring, which is again like [32:19] the number one thing that I'm trying to [32:20] do. [32:21] >> Okay, let me ask you just a superh [32:22] tactical question. Do you also have some [32:25] kind of admin assistant? And what does [32:28] that admin do? [32:29] >> Yeah, so I have a chief of staff and the [32:32] the biggest difference [32:32] >> So not an assistant, a chief of staff. [32:34] >> Yeah, chief of staff. And she's also [32:36] incredible. And the biggest difference [32:38] is she manages everything that I need to [32:40] do. Like my entire life is personal, but [32:44] everything that I need to individually [32:45] do myself and Katie manages everything [32:48] that is crossf functional that I need to [32:50] do, right? So fix the pilots, right? [32:54] That is not something that my chief of [32:56] staff would do. That is something that [32:57] Katie would do, right? But if it is, [33:00] hey, I need to figure out how to change [33:02] my schedule and how to basically monitor [33:05] all of these different things. I want to [33:06] get these data feeds from all of these [33:09] different sources of the company to put [33:10] into my dashboard and I want to track [33:12] all those things. That is something [33:14] chief of staff would do. [33:15] >> Okay. What about like admin stuff like [33:18] somebody who books your travel, somebody [33:21] who you know get does all your [ __ ] like [33:23] all your personal like there's so much. [33:26] >> Yeah, I have that as well. I have that [33:27] as well. Um I I took way too long to do [33:31] that. I was still I was book [33:33] >> So you have someone that does? [33:34] >> Yeah, I have someone now. Yeah, but for [33:36] a while I was booking my flights until [33:38] like two and a half years in or [33:41] something. Something like that. It was [33:43] so stupid. I mean it was the dumbest [33:44] thing ever. Like I remember uh I spent [33:48] there was the the day that I broke [33:50] actually and then I decided I needed to [33:51] hire someone for this was I spent 6 [33:54] hours trying to book a British Airways [33:56] flight or something like that and it was [33:59] literally six hours and I it was just [34:01] all day and it was something about like [34:03] my card they thought it was fraud [34:05] because I had never you know it was like [34:07] a $6,000 flight or something crazy like [34:10] that from SF to to UK and I had never [34:13] paid for anything like that on my flight [34:15] or like personally and JP Morgan was [34:18] just like complete fraud and it was like [34:20] six hours and I thought back to that the [34:22] other day and I I went to Gabe and he [34:24] was just like what the [ __ ] have you [34:25] been doing all day? [34:26] >> Yeah. [34:27] >> What is that? [34:28] >> Very interesting. Let's talk a bit about [34:32] let's go back in time [34:35] and your go to market. [34:38] Um you have some interesting ideas about [34:42] go to market uh in some thesis on it. [34:45] Can you talk about like how does a [34:48] random company that almost no one's [34:50] heard of a CEO who's doing it for the [34:53] first time that doesn't have some [34:54] amazing long history in legal? How do [34:58] you go about nailing your first five [35:01] customers? [35:02] >> Yeah. Uh, I mean it is just insane [35:06] repetition and trying things until they [35:08] work. The the thing that AI allows you [35:10] to do is personalize everything to an [35:11] insane degree. Um, and now there's [35:14] incredible tools like Clay and etc for [35:16] like actually doing this at scale, but [35:18] you could do it by yourself, right? So [35:20] the thing that I would used to do for [35:22] all lawyers is you can if you're a [35:25] litigator, everything needs to be filed [35:27] with a federal court if you're if you're [35:28] in federal court, right? And most big [35:30] law attorneys, you have to file [35:31] everything, right? So, it's online. It's [35:33] on this like document repository called [35:35] Pacer. And I would basically download [35:37] the last thing that they submitted to [35:39] court. And then I would try to come up [35:42] with prompts that were like, "This is [35:43] bad." [laughter] [35:44] >> Yeah. [35:45] >> Um and because they're a litigator and [35:47] I'm basically attacking something that [35:49] they just wrote. Y [35:51] >> they would instantly read the screen. Um [35:53] it was risky because sometimes the like [35:56] basically the system, you know, Harvey [35:57] would hallucinate and then it would just [36:00] be over. But the times that they got it [36:02] right, it was over. Like, and and if you [36:05] think about doing a Zoom with a lawyer, [36:06] it's like, I don't know how you got on [36:08] this Zoom. Like, I don't really pay [36:09] attention to this. Half of them were [36:11] doing this, like literally like this on [36:13] the Zoom, right? And then you bring up, [36:15] hey, I saw that you filed XYZ, you know, [36:18] motion to dismiss and in this case, do [36:20] you want to see Harvey analyze it and [36:22] say why it's a bad argument? And they [36:25] would jump at that and they would go [36:27] through everything. You can do on the [36:28] transactional side, too. you just, you [36:30] know, do like a merger agreement or [36:31] something like that on Edgar. Um, and so [36:33] a lot of it was personalized. I think [36:36] the other thing that we did on GTM is we [36:38] wanted to do the hardest ones first. So [36:40] >> tell me about that. [36:41] >> Yeah. So [36:42] >> kind of counterintuitive. [36:44] >> The thesis was kind of twofold. One is [36:47] I'm insane still crazy bullish on the [36:50] models improving and just basically all [36:53] of the underlying systems improving. I [36:56] know it sounds like duh. Now, uh, in [36:59] 2022, it was not at all, right? It [37:02] really wasn't. Um, and so there are two [37:06] thesis here. One is the lower-end work. [37:09] It's just like, why not use CHBT, right? [37:11] For like way way like lower-end legal [37:13] work. It's like I've was concerned that [37:16] the models are just going to get so good [37:17] that it's like, yeah, I can review your [37:18] lease. Like that's just going to happen, [37:20] right? Um, and so building a bunch of a [37:23] solution for that at the end of the day, [37:24] I don't know how much value you would [37:25] add, right? So that's one. Two is if you [37:30] can figure out all of the massive use [37:33] cases that one of these giant firms has, [37:36] you kind of have like a perfect design [37:38] partner for the entire legal industry, [37:41] right? Like the reason these firms exist [37:43] is because when you're doing a, you [37:45] know, a massive M&A that involves like [37:47] 60 businesses that operate in 60 [37:50] countries, the work is insanely complex. [37:52] You need like an antitrust lawyer in [37:54] every single jurisdiction, right? That's [37:56] why big law firms exist. That's why you [37:58] have these crazy amount of specialists [38:00] in every area is because they [38:02] [clears throat] can basically jump onto [38:03] these massive massive transactions or [38:05] litigations. [38:06] >> And so really good design partners. [38:08] >> Their requirements are brutal. [38:10] >> It's compliance requirements. Why not go [38:13] after a mid-tier firm that doesn't have [38:16] to go through a long decision cycle? the [38:18] partner can decide herself and there's [38:21] not all this compliance and the deal [38:22] just happens and then move up. That's [38:24] the conventional wisdom. Why you're [38:26] torturing? [38:27] >> Well, I mean, so a couple things here. [38:30] Weirdly, revenue grow like revenue [38:32] growth like how fast can you gain [38:34] revenue was like not something we were [38:36] thinking about. Like another thing to [38:38] think about here is like I didn't know [38:39] what ARR was in the beginning. [38:40] >> Yep. [38:41] >> I didn't even know what that was. Like I [38:42] didn't even know of like oh company [38:43] health like net new AR quarter quarter [38:45] growth like things like that. And to me, [38:48] I was just thinking about how do you [38:49] build the best product and not just the [38:51] best product, but how do you like [38:52] actually show that this is going to [38:54] transform the industry? And the reality [38:56] is if you can get one of these top firms [38:58] that have just never ever talked about [39:01] adopting technology before, doing [39:03] anything cutting edge to do it, that [39:06] would be a wakeup call, like a wow, this [39:09] is actually going to be a change, right? [39:11] And it's the same on corporates. So now [39:14] like last quarter I think it was like [39:16] 42% of our revenue actually came from [39:19] large corporates um not law firms right [39:22] and of those large corporates the [39:24] majority of them are fortune 100 not [39:26] even fortune 500 fortune 100 right and [39:28] the the reason why is because [39:30] >> if you can figure out how to sell to a [39:32] large bank and if you can get past [39:34] security review on a large bank and do [39:36] all the implementation on a large bank [39:38] you can sell everybody else right um and [39:40] so some of it is just do the hard thing [39:42] >> back to the first ones like so Many of [39:44] the CEOs I'm working with get their [39:45] first accounts because they work for [39:48] that company before or their VC knew [39:50] somebody at that company. It doesn't [39:53] sound like that was the game you played. [39:55] Like usually it's you'reworked into your [39:58] first 10 pilots. It sounds like you kind [40:00] of cold bold your way into your early [40:02] deals. [40:03] >> Uh 100% that. Um I mean the firm that I [40:06] worked at uh no no shade towards them, [40:08] but they were definitely like customer I [40:10] think like 200 or something like that. [40:12] Um, so it was definitely later later [40:14] down the down the line. Um, and I also [40:17] I've been practicing law for like eight [40:18] months. Um, so I didn't I I didn't have [40:21] any connections, right? Um, it really [40:24] was just could you get people that would [40:26] respond to your LinkedIn message? I mean [40:28] just LinkedIn like you message a [40:29] thousand people like you just keep doing [40:31] that like tens of thousands at this [40:33] point, right? Um, and I still get a lot [40:35] of lawyers that will I mean now it's [40:36] like hard to even go through my [40:38] LinkedIn, but like I still get all the [40:40] time like I I message them in 2022. They [40:45] message me back and be like, "Oh, love [40:46] Harvey, right?" or something like that. [40:48] Okay. [40:48] >> Happens all the time. Yeah. Happens all [40:50] the time. And so a lot of it was just [40:51] reach out to as many people as you [40:53] possibly can. If somebody gets on the [40:55] phone, you know, to get them to stop [40:57] doing this while they're on the call [40:58] with you, show them something personal. [41:02] I think a lot of CEOs miss this. The way [41:05] I described HubSpot's go to market was I [41:08] called it the pain factory. [41:10] And the way it worked was we built a [41:13] little application [41:15] called website grader where you put your [41:17] URL in and it gave you a score of 1 to [41:20] 100 on how much mojo you had on the [41:22] internet, how many links you had, your [41:24] SEO power, your blog, your social and [41:28] um if I wanted to get into XYZ company, [41:31] I put XYZ.com in there. I get the [41:33] report. I would email it right to the [41:35] CEO and the CEO would be like, whoa, I [41:39] got a 64 out of 100. [laughter] [41:42] What the hell do I do to get to 100? [41:44] >> Yeah. Yeah. I don't have any marketing [41:46] sales rep like talk to this guy. So, I [41:48] immediately [41:51] got some credibility through it. [41:53] >> Um, and then the sales rep would show up [41:55] and like explain the report and show a [41:57] demo. It's like I you have your tooth [41:59] really hurts. You got a lot of pain in [42:00] there. Would you like me to fix it? [42:02] would you like me to like put a filling [42:03] in there and put the Mcn numb it? Yeah, [42:05] that's HubSpot. [laughter] [42:07] And so it's sort of the pain factory. [42:09] And I think what you did is exactly what [42:11] we did. [42:11] >> It's exactly the same thing. And it's [42:13] interesting because one thing I think [42:15] about a lot is like as we're selling to [42:17] in-house teams versus law firms. I think [42:20] about and I think about other AI [42:22] application layer companies and which [42:24] vertical they're in as like are you [42:27] trying to build something that automates [42:29] the core competency, right? And that is [42:32] going to cause friction. And what's [42:34] interesting is we had less friction [42:36] actually in the beginning because we [42:38] were weren't pitching to firms. We were [42:41] pitching to lawyers, [42:42] >> individual lawyers. Yeah. [42:44] >> And so their pain was I don't want to do [42:46] this particular piece of my job. Like I [42:49] don't want to go through tens of [42:50] thousands of documents and you know do [42:53] this massive like sign off thing, right? [42:55] Like closing checklist. I don't want to [42:56] do that, right? That's not a part fun [42:58] part about being a lawyer. Now you have [43:00] to kind of change your pitch because [43:02] you're not just pitching to the lawyers [43:04] like the users, you're pitching to the [43:06] law firms too and they law firms have [43:10] different pain points especially as AI [43:12] develops than just the lawyers [43:13] themselves do, right? So you have to [43:15] kind of like change your pitch. Um, and [43:18] that also impacts kind of what you're [43:20] building. Whereas on the in-house side, [43:22] you're still actually just pitching to [43:24] the lawyers. Like that's it. Like you're [43:26] you're just pitching to them, right? And [43:28] it's because you know you get like a big [43:30] a big customer like Bridgewwater or [43:31] something like that like Bridgewwater's [43:33] core competency like they're not trying [43:35] like they don't provide legal services [43:38] right that's not like what they are [43:40] doing. Yeah. and that helps the [43:42] business, right? And because of that, it [43:45] is seen as less threatening because it [43:47] is not the core competency of what [43:49] Bridgewwater does, right? Um, and I I [43:51] just find that interesting because you [43:52] kind of have to balance the pitch of [43:54] like this is good for your users and the [43:57] pain point for your users versus what is [43:59] actually the broader ecosystem painpoint [44:01] for the law firm or for the corporation [44:04] or whatever it is. And as AI I think [44:06] gets better and better, you got to think [44:08] about not just the individual user, but [44:10] what is the pain point of that industry? [44:12] >> Yep. Uh you didn't know anybody in these [44:16] big law firms. You also had never sold [44:19] anything. [44:20] >> Yeah. [44:20] >> Or built a go to market engine. [44:23] And so many of the CEOs I'm working with [44:25] are the same. Like by the way, you're [44:27] quite different than most of the CEOs I [44:29] work with because most of them are [44:31] software developers. [44:32] >> Yeah. [44:32] >> And deeply technical and product people. [44:34] and then they built something amazing. [44:36] You're like, I have to sell this thing. [44:38] [laughter] [44:40] Uh just talk a little bit about that and [44:43] how did you learn your craft and how did [44:46] you because my sense is one of your [44:48] superpowers is you can sell. [44:50] >> I I so I think there's a couple ways to [44:53] do it. One one thing I want to say in [44:56] the beginning too, the times that I'm [44:58] best at product is when I'm selling the [45:00] most. [45:00] >> Yes. [45:01] >> It's not even close. Apologize. [45:03] >> It's like actually not even close. Like [45:05] the times I did this in [45:07] >> I think it was like Q3 or Q2 of last [45:10] year. I spent too much time not talking [45:13] to customers and like sitting down in [45:15] San Francisco and being like this is the [45:17] road map for the next like year. [45:19] >> Yeah. [45:20] >> I missed like I missed so many product [45:22] things by doing that. Like I made wrong [45:24] product decisions. There were so many [45:26] things that were wrong. I do best by far [45:29] when I'm talking to customers 247 and [45:33] talk and working on product [45:34] simultaneously. If you do one or the [45:36] other, it actually like doesn't work at [45:38] all. So, I want to start by that. Um, [45:40] but the the the learning to sell thing I [45:43] think the most important thing is you [45:46] have to adjust to who your buyer is [45:49] personally. Like it is people reading. [45:53] Um, and there's like some instances [45:54] when, you know, you'll get on a call [45:56] with a big group, right? A big a big [45:59] group of folks, and it's very obvious [46:02] that one of the people on those calls is [46:04] like trying on purpose to make you sound [46:08] bad for a particular reason. [46:10] >> Like, you don't know why, but it's just [46:11] like very clear. [46:13] >> The number one thing to do is to meet [46:15] with that person one-on-one. [46:17] >> And no one wants to do that cuz [46:19] everyone's like, "Ah, that was [46:20] uncomfortable. I didn't enjoy that." [46:21] Right? But that's actually the person [46:24] that there's something else going on. [46:26] Whether it's like they tried to build [46:27] something in house and it didn't quite [46:28] work and you're kind of threatening to [46:30] them or whatever it is. [46:32] >> And then if you turn them around and [46:34] they're they become champions, it [46:35] completely changes everything. And so [46:37] it's kind of what you said before where [46:39] honestly it's like a lot of it is just [46:40] like find the thing that hurts and like [46:43] you don't want to end up doing almost [46:44] entirely. That's what you want to do. [46:46] Like if you push on the the two green [46:48] field stuff, it actually doesn't work. [46:50] >> Okay. A lot of my CEO founders, some are [46:54] in really sexy businesses and it's easy [46:56] to hire developers to build sexy stuff. [46:58] >> Yeah, [47:00] >> don't take this in any offense, but [47:02] building legal software [laughter] [47:03] doesn't feel that sexy. How are you able [47:06] to attract top tier engineering talent [47:08] on building software for lawyers? Yeah, [47:11] the biggest reason is because we have to [47:14] work. There are a lot of actually like [47:16] frontier things that we need to solve [47:18] that the big labs will maybe eventually [47:20] solve, but they aren't working on it [47:22] right now. I'll give you a really good [47:23] example of this. Um, so we launched [47:25] memory this morning, right? And this is [47:28] going to become super important for us. [47:29] >> There's a lot of ways to do memory. Um, [47:31] but the most important version for us is [47:34] almost everything that lawyers do is [47:36] document-based, right? Right now the way [47:38] that chatbt memory anthropic memory etc. [47:41] there is actually no like understanding [47:43] of these are the documents as artifacts [47:45] and things like that right [47:46] >> super irritating [47:48] >> super irritating but that's something [47:50] that they will solve eventually [47:52] >> if we don't solve that our product goes [47:54] to zero [47:56] >> right and so there's actually a tons of [47:59] frontier type things that I think these [48:02] application layer compan do [48:04] >> okay so they're hard problems [48:05] >> really hard problems that Everyone is [48:08] going to have to solve eventually, but [48:10] we have to solve sooner because of just [48:13] the nature of our customer base, right? [48:16] Another example of this is [48:18] >> we are we already have something where [48:20] you can basically collaborate with AI [48:22] systems in our platform and outside [48:25] folks. So not just like multiple people [48:27] from your own corporation working on [48:29] something, but I'm a big bank and I want [48:32] to collaborate with 10 law firms. [48:34] >> Yep. also something the big labs have [48:36] not solved yet right [48:37] >> and we first they haven't [48:39] >> they haven't super irritating and they [48:41] will to be clear I'm just saying there's [48:44] so many of these domains where I think [48:46] it's really exciting to work at the [48:48] company and building software there [48:49] because you're working on frontier [48:51] systems that for sure open AI and [48:53] anthropic are thinking about but going [48:55] back to our P zeros it's like it's way [48:58] down that [49:00] >> and so you get to work on that and you [49:01] might be able to solve that better as a [49:03] company I you should otherwise your [49:05] company's over better than they can [49:07] faster and that I think is really [49:08] exciting for people. [49:10] >> I want to talk about you a little bit. [49:11] >> Okay. [49:12] >> Something I noticed about you and then I [49:14] I I spoke to some people who work for [49:16] you and it's they I would ask when I ask [49:19] people like what tell me about Winston. [49:22] What's his superpower? The word [49:24] intensity is the first word that comes [49:27] to everyone's mouth. And I agree and [49:30] probably people watching this would [49:32] agree. [49:34] Where does that come from? [49:35] >> Um, [49:36] >> you're extremely motivated and extremely [49:39] intense. [49:41] Um, why do you have a chip on your [49:43] shoulder? Something from your childhood. [49:47] >> Um, I think it's a combination of [49:49] things. So, I always let my co-founder [49:51] whenever we do introd like intros, I [49:52] always let my co-founder go first. And [49:54] he has a awesome background. Um, [49:56] >> amazing. Yes. [49:57] >> Incredible background. And then I go I [49:59] did nothing for 27 years is like what my [50:01] well now I'm a little older but that was [50:03] basically my my background. Um I think [50:06] the reality is like I definitely had a [50:09] lot of problems when I was younger kind [50:12] of mental health problems and things [50:13] like that and I I was a terrible [50:15] student. Um didn't got kind of like [50:18] kicked out of a bunch of schools and [50:19] things like that and I didn't find [50:21] something that made me like want to get [50:23] up every morning and like apply myself. [50:25] And when I started this company, I had [50:27] no idea that this would happen. But I've [50:30] never felt this in my life. Like I've [50:31] just [clears throat] never felt I've [50:34] never felt this like overwhelming [50:36] obsession and excitement to get up every [50:39] morning ever. [50:40] >> And it is absolutely terrifying to me to [50:44] lose that. Like it is terrifying that I [50:46] will lose that. [50:47] >> And I think the problem with it is [50:50] that's not like, oh, the company's going [50:52] to go to zero. It's if we don't keep [50:55] moving and we don't basically capture [50:57] the opportunity that we have, I think [51:00] I'll be stuck in a situation again where [51:02] I don't feel that like I don't feel like [51:04] massively challenged. I don't feel like [51:06] what I'm getting up to do sucks. And [51:08] like one example of this is [51:11] >> I literally try I try every week to have [51:14] at least like two to three things that [51:16] make me super uncomfortable that I'm [51:18] doing. Um, and I don't think I feel good [51:22] unless I do those things. Like if I [51:24] >> were you always like that's pretty [51:26] intense and not that fun. Were you [51:27] always that way or like [51:29] >> No, I I I think maybe I might have been [51:32] but I didn't have the opportunities to [51:34] do those things or I maybe let me take [51:36] more ownership of it. I didn't put [51:38] myself in situations where I would find [51:40] those. [51:41] >> What changed inside of you that all of a [51:43] sudden you were [51:45] >> Yeah. Um, I still figuring that out to [51:48] be 100% honest. Like the the part that I [51:50] have figured out is what motivates me. [51:52] And that took me like [51:53] >> I think I figured out that out like last [51:55] year. Like it's pretty recent. Um, and [51:57] the thing that is insanely motivating to [51:59] me is [52:01] I just want to see how far I can go and [52:05] like keep going until I'm like people [52:08] this is just I can't do this. like I [52:09] break like it's just this is a scale or [52:12] this is a type of project or something [52:14] that I am not good enough to do and that [52:17] doesn't mean personally by the way like [52:19] this is something that has changed [52:20] massively for me is I used to think that [52:22] companies were very much like heroics [52:24] and it was like just you know it's like [52:26] the founder can the founder do it can [52:27] the founder do it can the founder do it [52:29] no like not being able to do it is you [52:32] can't hire people that are good enough [52:34] like you can't or you can't convince [52:36] people or you can't lead people well [52:37] enough right you There's so many [52:39] different ways why you can get stopped, [52:41] but it's not physically you can can't do [52:44] it yourself. [52:45] >> But that's what is massively motivating [52:47] to me is I think we have an insane [52:50] opportunity and I'm like terrified of [52:52] missing out on it and I'm terrified of [52:54] going back to how I felt before this. [52:57] >> Got it. Do you think there's a downside [52:59] of your intensity? [53:01] >> I think sometimes. Um I'm trying to get [53:04] better at it. [53:05] >> Yeah. [53:06] >> Um but I think Yeah. I think I mean uh a [53:09] I think I do that sometimes and I think [53:11] that sometimes people say hey that [53:15] actually like I didn't think I could do [53:17] that and I could and that made me feel [53:19] amazing like that is sometimes how it [53:21] happens sometimes I think I take it too [53:24] far um and I think it becomes a problem [53:27] and I think the other thing I need to do [53:28] is there are some situations and it is [53:31] very humbling to feel this as a founder [53:33] but it happens where the chance of [53:35] success is higher If I say this is the [53:38] goal and what we should do and then I [53:40] stop poking [53:41] >> and I let them do it. I'm serious. And [53:43] it is so hard and it's so humbling, but [53:46] once it happens a couple times, it gets [53:49] easier, right? Um but I I do feel that. [53:53] >> Okay. I've known you for a while now. [53:55] I'm really proud of you. [53:57] >> You come so far and matured so much and [54:00] you're doing such an excellent job. I'm [54:02] really happy to know you and happy [54:05] Sequoa investing you. I think they're [54:07] doing great. [54:08] >> Thank you, man. [54:09] >> I'm really proud. [54:09] >> Thank you. [54:10] >> Yeah, [54:10] >> that means a lot. [54:11] >> Appreciate you coming on the pod, giving [54:12] the pearls of wisdom to all the CEOs out [54:14] there and go get them. Keep the rocket [54:16] ship going. [54:17] >> Thank you. [54:17] >> Yeah, [54:18] >> thank you, man. Thanks. [54:19] >> Okay, I hope you like that episode. [54:21] We've done [54:23] six now. I think that's my favorite. For [54:26] some reason, it got a little dusty in [54:28] the room towards the end. You know, that [54:30] can happen. As you could tell, I know [54:32] Winston pretty well. And when I look at [54:35] CEOs these days and evaluate them, I [54:38] have a rubric that is my lock rubric. [54:41] The L is for lovable. The O is for [54:44] obsessed. The C is for chip on the [54:46] shoulder. And the K is for [54:48] knowledgeable. I remember first meeting [54:50] him. And at least to me, he's lovable. [54:53] Like if I were a younger me graduating [54:55] from school would want to go to work for [54:56] him. Um really obsessed. You saw that? [55:00] like he's literally a billionaire and [55:02] he's got roommates. Um I talked to him [55:05] over the holidays. He didn't have [55:07] holidays. He's pretty obsessed with [55:08] completely obsessed with what he's [55:10] doing. Uh you felt the chip on the [55:13] shoulder. Part of it is how he grew up. [55:15] But part of it also is that he's a non [55:19] software developer founder here in [55:21] Silicon Valley. And he brings that up a [55:23] lot. He brought it up a little in this [55:24] pod, but he's got a big chip on the [55:26] shoulder. And the is knowledgeable, very [55:28] knowledgeable guy. He's also very [55:30] humble. He knows he's made a lot of [55:33] mistakes. He faces those mistakes and he [55:35] tries to improve them. That's something [55:37] I love about great CEOs is [55:40] um the problem with being a CEO is [55:42] everyone tells you how great you are all [55:43] the time. And [55:46] >> [music] [55:46] >> um as it scales more that happens, [55:48] particularly if you're growing very very [55:49] fast and he's a bit immune to that. Um [55:53] and he's very self-critical and adjusts [55:56] a lot. And I think you saw that in the [55:58] pod. Anyway, I thought this was pretty [56:01] good. If you are enjoying the pod, I [56:04] think this is our sixth episode. Give me [56:07] a comment. We need some feedback on how [56:09] we're doing. Good, bad, indifferent. [56:10] [music] [56:11] I'd love to hear from you. See you on [56:12] the next one.