Former Twitter CEO Costolo Urges Scale-Ups to Dump Cross-Team Vetoes

Twitter前CEO Costolo呼吁成长期企业废除跨团队否决权 超高速增长企业如果靠多部门流程解决问题，而非向直接责任人问责，其运转速度便会放缓。 在本次对话中，Dick Costolo深入拆解了带领Twitter度过极度超高速增长期的运营打法。他阐述了自己如何通过以“倾向于赞同”（bias to yes）取代群体共识、摒弃冗余流程转而倚重直接责任人（DRI），以及通过深夜巡楼重塑文化，从而全面提升公司的运转速度。此外，Costolo还坦诚反思了自己的领导力失误，包括解雇人员不够果断、未能在与Facebook竞争收购Instagram时更积极地提高报价，以及如何在持续不断的敌意媒体猛烈抨击中挺过来。 Dick Costolo 最初受邀以临时 COO 身份加入正经历疯狂文化爆发且毫无收入模式的 Twitter，随后接任 CEO，迅速破除全员低效投票与群体外交式妥协的文化，全力提升决策与执行速度。 (4:43) 为消除跨部门官僚阻碍，Costolo 推行“偏向肯定（bias to yes）”原则，规定仅直接主管有权否决提案；同时主张以直接责任人（DRI）和运营管控取代冗长的上线核对清单，由管理层快速纠错而非限制犯错。 (15:45) 随着团队突破 150 人导致自发沟通失效，Costolo 借鉴史蒂夫·乔布斯在皮克斯绕过层级直接对话一线员工的做法，以掌握未经滤网过滤的组织真实运作状况并确保战略认知一致。 (21:16) 针对团队管理与行业竞争，Costolo 告诫切莫因拖延解雇而陷入部门间转嫁低绩效员工（trans-firing）的陷阱，并坦言任期内最大遗憾是曾出价约 7 亿美元股权却最终未能抢先 Facebook 收购 Instagram。 (47:30)

Former Twitter CEO Dick Costolo argues that scaling a hypergrowth company requires replacing consensus-driven cultures with decisive operational discipline, rapid feedback loops, and decentralized execution. Rather than mimicking celebrity founder archetypes or layering on bureaucratic processes, effective operators sustain execution speed by providing transparent organizational context and enforcing individual accountability. Replacing Bureaucracy with the Bias to Yes (15:45) When Costolo stepped into leadership at Twitter, the organization operated like a collective, debating minor product adjustments like native retweet buttons for weeks and letting managers invoke authority by claiming "Dick said do it." To dismantle cross-functional bottlenecks and internal fiefdoms, Costolo instituted a strict "bias to yes." Under this operating rule, only an employee's direct manager holds the authority to reject an initiative, and outside departments cannot block access to tools or veto experimentation unless a project violates legal or privacy policies. Engineers were encouraged to bypass design delays by self-authoring basic wireframes and launching experiments on 1% user cohorts. Costolo warns that when mistakes occur, leadership must resist the urge to invent defensive checklists—citing Google's launch checklist, which swelled to an unwieldy 17 pages before Larry Page intervened. Instead of treating organizations like the FDA or TSA by managing to rigid procedural steps rather than outcomes, scaling companies should limit themselves to a couple of high-level organizational processes. Decision-making authority must remain pushed down to Directly Responsible Individuals (DRIs), paired with a leadership culture focused on correcting errors rapidly rather than punishing honest experimentation. Scaling Context and Unfiltered Information Flow (19:14) As headcount expands past 150 employees, internal communication inevitably breaks down without deliberate architecture. Drawing on Ben Horowitz’s Loudcloud curriculum, Costolo stresses the rule: "Make sure everybody understands what you understand." Workers perform best when they understand broader strategic priorities and why their immediate output matters, resembling the difference between laying bricks and constructing a cathedral. To preserve alignment and momentum, Costolo advises against shifting corporate priorities more than once or twice a year, warning that frequent strategic pivots breed employee apathy. Because positive news travels fast while negative news travels slowly and distorts as it moves up the chain, Costolo adopted a technique Steve Jobs used during Pixar’s production of Toy Story 2. Costolo met directly with front-line teams without their managers present to ask what was failing, evaluating whether executive reporting matched ground-level reality. He paired this with active "management by walking around," touring the office around 9:00 p.m. to discover late-night projects and publicly elevating those initiatives at all-hands meetings. For strategic leaders, he frames this as "forestry management": mapping territory and building systemic firebreaks rather than exhausting bandwidth reacting to localized fires. Direct Feedback and Ending Trans-Firing (33:01) Costolo rejects the passive-aggressive communication common in tech hubs, emphasizing that managers must be straightforward without acting like jerks. For difficult performance discussions, he cites Bill Campbell's advice: write down the core message beforehand, deliver it verbatim without softening filler, and simply sit with the ensuing silence. In Twitter's management training, Costolo found venture capitalist Elad Gil to be the rare operator able to hold that silence effectively until argumentative pushback subsided. While tech executives frequently repeat the mantra to "hire slow and fire fast," Costolo admits he personally fired slowly and regretted delaying terminations within six weeks every time. Delaying action mirrors Jeff Weiner's baseball analogy of the hesitant manager who talks to a tiring pitcher instead of removing him before a home run. Furthermore, hypergrowth companies often fall prey to "trans-firing," where inexperienced managers offload underperformers onto other teams. To halt this cycle, Twitter prohibited employees rated below an acceptable threshold from switching teams until their performance demonstrably improved. Hypergrowth Discipline, Competition, and Market Realities (47:30) In managing hypergrowth and capital efficiency, Costolo observes that massive funding rounds—such as modern $100 million Series B rounds—often erode fiscal discipline and operating leverage. When competing against relentless incumbents like Mark Zuckerberg's Facebook, Twitter relied on out-hustling rather than out-spending. Sales leader Adam Bain routinely gathered intelligence on Facebook's pitches prior to customer meetings at events like CES, allowing Twitter to differentiate itself by focusing on customer listening rather than overwhelming product feature lists. Costolo also defends choosing an advertising business model over subscriptions, maintaining that advertising yields superior average revenue per user (ARPU). Reflecting on strategic missteps, Costolo identifies Twitter's failure to acquire Instagram for roughly $700 million in equity (over 10% of Twitter's valuation at the time) just weeks before Facebook purchased it for $1 billion as his single biggest regret. Losing Instagram permanently boxed Twitter out of dominating rich media social feeds. Finally, Costolo argues that companies should delay public offerings longer, as the public market's quarterly volatility—where a 1% revenue miss triggers panic and stock drops—unnecessarily distracts teams from executing long-term strategy. Costolo concludes that leadership cannot be copied from biographical archetypes of figures like Steve Jobs or Elon Musk, noting that inauthentic executive personas quickly alienate teams. While professional operating executives brought in to scale founder-led startups have become culturally unfashionable, Costolo predicts the pendulum will inevitably swing back as high-profile scale-up failures mount.

前 Twitter 首席执行官 Dick Costolo 认为，在企业高速增长期，依赖集体共识与僵化流程会扼杀组织动能；职业管理者应当推行去中心化决策、直接责任人制（DRI）和极度透明的上下文沟通，以执行速度替代官僚内耗，而非盲目复制所谓的“创始人模式”。 破除共识陷阱与推行“倾向于同意” (15:45) Costolo 接任 Twitter 首席执行官时，公司像一个“公社”般运作，团队甚至会耗费数周时间外交式地讨论是否将用户自创的“RT”文本转变为原生的转发按钮。面对这种效率低下，Costolo 将首要任务定为提升决策速度，坚决终止投票表决机制，并在更换状态栏提示语等改动中明确告知基层员工不再事事寻求全体共识。 为了瓦解部门地盘意识与官僚阻力，Costolo 引入了受即兴喜剧和亚马逊哲学启发的“倾向于同意”（bias to yes）原则。在该规则下，唯一直属经理有权否决员工的想法；除法务和隐私政策合规外，安全、信任等平级团队无权设置卡点或阻止工具使用。这促使工程师无需等待设计团队制作精美原型，即可自行开发粗糙版本并向 1% 的用户灰度发布，极大释放了实验迭代速度。 依靠责任人而非流程化解危机 (17:00) Costolo 告诫管理者“切勿试图用流程来解决问题”。多数企业在遭遇客户流失或广告主撤单时，管理层往往产生“绝不能再发生类似问题”的下意识反应，进而不断在发布清单中堆砌审批步骤。Costolo 以谷歌为例，指出其发布清单曾一路膨胀至荒谬的 17 页，直到拉里·佩奇出面削减。这种做法如同安检过度纠结水瓶却忽略了根本目标，导致组织为了流程本身而非业务结果而管理。 为了避免组织官僚化，企业应严格推行直接责任人（DRI）机制，并将核心组织级流程限制在少数几个之内。与决策下沉相配套，领导层必须建立容错机制：领导的职责并非阻止犯错，而是在错误发生后迅速纠正。如果领导层在下属犯错时施加惩罚，去中心化决策和快速试错的心理安全感将荡然无存。 穿透层级的信息流动与走动式管理 (19:14) 随着团队规模突破 50 人乃至 150 人，组织沟通会自然退化为“无人知晓全局”。借鉴本·霍洛维茨的培训纲领，Costolo 强调管理的核心在于“确保每个人理解你所理解的事情”，向员工清晰解释战略背景与长远意义，严禁管理者以“Dick 说的”等借口关闭讨论。同时，主要业务优先级每年最多只能调整一到两次，频繁变更方向会引发员工对项目的漠视与倦怠。 为了防止基层信号在向上传递时被严重粉饰与失真，Costolo 借鉴了史蒂夫·乔布斯在皮克斯的跳级沟通机制：避开部门主管，直接召集一线员工交流哪些环节运转失灵。此外，受比尔·坎贝尔“下午五点半扔手榴弹只能炸到保洁员”的警示启发，Costolo 坚持在晚间重返办公室开展“走动式管理”，通过在全员会上公开表彰并优先推进夜间加班员工的项目，自下而上重塑了工作文化。 极度直接的反馈机制与淘汰低效人员 (33:01) 在团队管理中，硅谷常见的被动攻击式沟通往往导致反馈失真。Costolo 遵循比尔·坎贝尔的训诫，要求管理者在提供困难反馈前写下要点，当面明确陈述后“保持沉默”，而不是用冗长委婉的套话稀释严重性。他以埃拉德·吉尔（Elad Gil）为例，指出优秀的管理者能在给出尖锐反馈后安然承受长久沉默，直到对方消化事实并停止辩解。 在人员淘汰方面，Costolo 坦承自己同许多管理者一样落入了“解雇过慢”的陷阱，几乎每次拖延解雇都会在六周内后悔莫及。在超高速增长期，新手管理者往往为了逃避开除下属而进行“跨部门转移”（trans-firing），将表现不佳的员工甩给其他部门。为此，Costolo 确立硬性制度：绩效评分低于特定标准的员工被严禁内部转岗，强行打破管理者掩盖业绩问题的逃避路径。 商业模式取舍与过早上市的代价反思 (44:53) 在应对巨头竞争时，Costolo 强调必须正视对手。面对马克·扎克伯格统领的 Facebook，Twitter 商业负责人亚当·贝恩在国际消费电子展（CES）上采取“倾听客户痛点而非单向灌输产品”的差异化策略争夺广告预算。Costolo 坚信广告模式比订阅制更具优势，因为其单用户平均收入（ARPU）表现更强。但他职业生涯最大的遗憾是未能更激进地收购 Instagram：Twitter 虽出价约 7 亿美元股票（占当时逾 10% 股权），但仍错失交易，导致 Twitter 在富媒体社交信息流领域彻底被封锁。 回顾资本运作，Costolo 指出当前大量融得巨资的初创企业正丧失资本效率与经营杠杆，且所谓“只招 A 级人才”在规模扩张时是不切实际的幻想。更重要的是，Costolo 明确表示如果能重新选择，他会选择让 Twitter 保持更长时间的私有化状态。上市公司因单季度收入哪怕 1% 的微小偏差就会引发股价大跌和全员恐慌，严重侵蚀了管理层着眼于 24 个月乃至更长期的战略定力。 Costolo 的复盘表明，规模化运营的核心在于建立透明的制度边界、去中心化权责与直面冲突的勇气，而非盲目模仿乔布斯或马斯克的个人管理风格。然而，Twitter 错失 Instagram 带来的产品封锁，以及上市后在短期财报压力与长期战略之间的被动拉锯，也揭示了职业经理人在面对资本市场与技术范式转移时难以回避的系统性挑战。

[0:00] Dick Costolo: The Twitter bird had hanged itself and exploded on two different magazine covers. [0:00] Dick Costolo: Next year's goal, no dead birds on magazine covers. [9:13] Dick Costolo: This is why the [__] site crashes... because everything's just kind of [__] like, oh well, someone else will clean it up. [9:13] Bret Svedberg: We'll make better mistakes next time [15:45] Dick Costolo: I don't want to hear about something being stalled because organization X said you have to go to get approval from organization Y [18:14] Dick Costolo: It's not the job of leadership to prevent mistakes from happening. It's the job of leadership to correct mistakes quickly when they happen. [19:14] Ben Horowitz: Make sure everybody understands what you understand. [24:34] Jeff Bezos: Oh, yeah? Well, I like to do everything. My team has to talk me out of stuff. [27:46] Bill Campbell: You could roll a grenade into Twitter at 5:30 p.m. and only take out the cleaning people. [30:41] Dick Costolo: It seems like you don't like working here. You should quit. [33:01] Bill Campbell: Write down what you want to say and then say that and don't say anything else. [33:01] Dick Costolo: Shut the [ __ ] up. Like the person knows you're about to like hit them over the head. [38:43] David Rosenblatt: 6:00 a.m. wake up in a cold sweat. [46:05] Dick Costolo: Advertising is undefeated [56:37] Jeff Bezos: Jack, there are lots of different ways to be successful.

[0:00] I just felt like Twitter was such, such [0:02] a drama. Twitter was like the drama [0:05] queen of hyper growth companies. [0:09] >> Yeah. You know, it was like when I took [0:11] over a CEO, that year the Twitter bird [0:14] had hanged itself and exploded on two [0:17] different magazine covers. So, I went up [0:19] to our head of communications, Gabriel [0:21] Stricker, and I was like, "Gabriel, [0:22] great news. Next year's goal, no dead [0:25] birds on magazine covers." You know, low [0:28] bar. [0:29] >> Yeah. [0:37] >> [music] [0:47] >> Everybody, we have Dick Costolo on the [0:49] show. He's a rare bird in Silicon Valley [0:51] these days. Uh he was CEO of Twitter, [0:54] but he was the third CEO. The first one [0:56] was Jack Dorsey, who you probably met on [0:59] the pod a couple weeks ago. Uh the [1:01] second was Ev Williams, the other [1:03] founder. [1:04] And Dick took over as COO, and then [1:07] ultimately as CEO of Twitter, and he ran [1:09] it for five years. It was a really [1:11] extremely fast-growing organization. [1:14] They created a category. Uh didn't have [1:16] a revenue machine, and was famously [1:18] dysfunctional. I think Dick did a really [1:20] nice job, and he was sort of the adult [1:22] in the room. [1:24] And he kind of ran the scale-up CEO [1:26] masterclass. He talks all about it in [1:27] this podcast. He talks about increasing [1:29] velocity of decisions and execution, [1:31] which is tricky as you scale. He talked [1:34] about how he managed by walking around. [1:36] This is something I did. I think it's a [1:37] really good idea. [1:39] He talked about how do you give very [1:41] tough feedback to very tough people. [1:43] There's some gems in there. [1:45] Dealing with the press, particularly [1:46] press that doesn't like you. [1:48] Um he talks about focus [1:50] and how much focus you really need need [1:53] in the era of AI where you can build [1:54] stuff so quickly. [1:56] He talks about how do you manage the [1:57] forest instead of fighting all the [1:59] fires. He's got a lot of wisdom in here. [2:01] He also talks about his regrets, [2:03] including a whopper, his regret in not [2:05] acquiring Instagram. That one hurt. If [2:08] you're scaling your 100 employees, [2:09] you're ripping, growing fast, you've got [2:11] a lot of growing pains, maybe you feel a [2:12] little lot of control, [2:14] this episode's for you. [2:16] Welcome to the show. [2:17] Thanks for having me. Appreciate it. I [2:19] want to wind the clock back to Twitter, [2:21] of course. Of course. [2:23] I watched the documentary recently. [2:26] And [2:28] Was it the one I was in? Yes. I thought [2:30] you were really good in it. [2:30] >> Oh, thank you. Great. You're great. You [2:32] were great in the documentary of [2:34] yourself. [2:35] >> came off as rational and and [2:37] knowing what the hell you were doing. [2:39] But you're in a long line of interesting [2:41] CEOs of Jack Dorsey and then Ev Williams [2:43] and yourself and Jack Dorsey and Parag [2:45] and Elon Musk. [2:47] >> and don't forget um, you know, [2:49] that Elon's hired hired hand, Linda [2:51] Yaccarino. [2:52] >> Okay. Of course. Of course. Um, you [2:55] joined as COO. What was the pitch? [2:58] What was the pitch to me? [2:59] >> Yeah. Ev pitched you. [3:00] >> yeah, Ev pitched me. The pitch was The [3:02] initial pitch was I'm taking um, [3:05] paternity leave for a couple weeks. I, [3:07] Ev Williams, am taking paternity leave. [3:09] Need someone to come in and like just [3:11] hold down the fort for a couple weeks [3:13] while I'm, you know, out. Interim? Yeah, [3:15] interim. Would you Would you do that? [3:17] And but then while we were having that [3:18] discussion, um, he's like, "Well, what [3:21] if it was permanent? You know, what if [3:23] you were COO?" And I was going to go [3:24] start another company. So, [3:26] rewinding the clock a little bit [3:27] further, I'd sold my previous company [3:29] Feedburner to Google. It's where I [3:31] rebumped into the um, Twitter guys cuz [3:35] they were all there from Blogger and [3:36] getting ready to leave and start Odeo, [3:38] which became Twitter. Anyway, [3:41] uh, I'm leaving now. I'm leaving Google [3:43] a couple years later. Ev texts me that. [3:45] I was like, "Ah, I'm going to go start a [3:46] I'm going to go start another company [3:47] again." [3:48] Um, with one of my Feedburner one or [3:49] more of my Feedburner co-founders. And [3:51] then Ev "What if it was What if it was [3:54] permanent? What if you were COO?" [3:57] And Twitter was already So, it's it's [3:59] two It's summer 2009. It was already [4:02] like had hit the you know, Ashton [4:04] Kutcher and Oprah Winfrey are, you know, [4:07] fighting for the Who's going to get a [4:09] million followers? [4:11] I can see why you wanted to do it. [4:12] >> But zero zero revenue model. Zero [4:15] business model. All the articles were, [4:17] "Sure, Twitter's interesting, but it'll [4:18] never They don't have a business model. [4:20] They'll never make money." Yeah. [4:21] And I just kind of had the Man, you [4:24] don't get an opportunity to go into [4:25] these to get on these rocket ships. And [4:28] when you get a get the opportunity, you [4:29] should do it. So, I did it. So, it was [4:31] famously dysfunctional and chaotic. I [4:34] assume you knew that going in. Was it [4:37] much more so than you thought? Or was it [4:41] about what the press was saying? [4:43] It's always both, you know. Um it was It [4:46] was The things that were different were [4:49] um [4:50] It was very much um a collective, by [4:53] which I mean like decisions were kind of [4:55] made as a group. Yeah. It wasn't Okay, [4:58] well, you're responsible for [4:59] engineering. I don't know. Make it [5:01] figure it out and, you know, figure out [5:03] what the right answer is. Or [5:04] >> Should we, you know, have Should we make [5:07] retweet a native thing or should it be, [5:09] you know, just you write You had to [5:10] write RT at the time and then rewrite [5:12] the tweet? I remember that. And And [5:14] there was a discussion about, "Well, [5:16] should we just have retweet be a thing [5:17] you can do in Twitter?" [5:20] And it would be kind of sit around the [5:21] table and talk about it and decide. [5:24] Okay. [5:25] >> You know, so there So, things [5:28] moved methodically as a diplomatic way [5:30] of putting it. I see. And decisions were [5:33] reached over a period of time. It was [5:35] chaotic and then it was also just um It [5:38] was just We just didn't It just didn't [5:40] move very quickly. It was didn't move [5:41] quickly at all. [5:43] In the In the documentary, it went [5:44] through the [5:47] you know, pushing out and giving you the [5:49] CEO job. It sounded awful, the whole [5:53] process. It was awful. [5:55] >> What's your side of it? What was What [5:56] was going on? What's your [5:57] >> really have a side of it other than it [5:58] was an awful like it's it's, you know, [6:00] these things in hindsight are all It [6:02] happened very very quickly. It happened [6:04] super slowly and painfully and and [6:07] um [6:08] you know, I was informed like in the [6:11] driving rain on the road home by, you [6:14] know, Evan and Jason that the board has [6:16] decided that you should be the CEO, but [6:19] I hadn't been told. And then Bill [6:21] Campbell called me and like [6:23] you should just quit, you know, and then [6:25] and then I was like, "Well, why am I [6:27] quitting?" you know, and like anyway, it [6:29] went and then this went and then like [6:31] the board decided actually no, Ev was [6:32] going to remain CEO, [6:34] um but you've been great and we're going [6:36] to vest a bunch of your options and so [6:38] long. [6:38] >> Okay. And then I mean, no kidding, over [6:41] many many days and weeks. And finally [6:44] [clears throat] I was like, "Listen, [6:45] just write up the separation agreement [6:47] and you know, I'm good. I'm good. Like [6:50] whatever. I'll go find some I'll go do [6:52] something else. And then, you know, to [6:54] make a long story short, the board got [6:56] back together was like, "Oh, no, no, no, [6:57] no. Remember Bijan Sabet from Spark [7:00] texted me, 'Don't quit.'" Okay. Like I [7:03] was like, "I'm not quitting. Like [7:05] >> [laughter] [7:06] >> Okay. [7:06] >> You guys told me you wanted me to leave [7:08] cuz Ev was going to stay. Anyway, it was [7:10] super crazy. [7:11] >> Yes. Um and there was obviously a ton of [7:13] stuff going on. And anyway, um long [7:16] story short, they're finally like, [7:17] "Look, you got to do it." Anyway, and [7:19] you know, a give really long story [7:21] short, I was like, "All right, I'll I'll [7:23] stay and do it." Okay, you take over. [7:26] Um [7:27] you're the adult. What did you do [7:29] initially? Yeah. Like it's very tricky [7:32] in this situation. Yeah. And I think the [7:35] current the existing founders are [7:36] hanging around probably [7:38] it looking over your shoulder if not [7:39] breathing down your neck. [7:40] >> Well, Jack was the chairman of the [7:42] board, still. Ev had made Jack chairman [7:44] of the board when he made himself CEO. [7:46] Um Ev's still on the board. So, two of [7:48] the founders, Ev and Jack, are still on [7:51] the board. And Biz was in the company. [7:53] >> And Biz is in the company, and Jason [7:55] Goldman's in the company running [7:56] product, and you know, and a bunch of [7:57] other people. Um [7:59] the the [8:01] the first thing that I felt, and I've [8:03] I've sort of foreshadowed this, the [8:05] first thing that I [8:07] just felt we had to do was like velocity [8:09] velocity velocity. And and that remained [8:12] a concern of mine throughout my time [8:14] there. Like, faster faster faster [8:17] cadence cadence cadence. Like, how can [8:18] we go faster? Why do things take so [8:20] long? If this takes 6 weeks, what could [8:22] have What would have to be true for it [8:24] to be 2 weeks? Yeah. We just [8:25] did stuff like that constantly cuz I was [8:28] always like, man, this is like dental [8:30] surgery. But it was starting from a [8:32] position of [8:33] we're not going to make we're not going [8:35] to sit around the table and vote on [8:36] things anymore. I remember vividly um [8:40] the statu- like the status line changed [8:42] from like, what are you what's going on [8:44] or what's happening to something else? [8:46] Or maybe it changed to what's happening. [8:48] I don't even remember. I wasn't even [8:49] involved in it. [8:51] And one of the front end engineers, Bret [8:53] Svedberg, came into my office and was [8:54] like, "Hey man, what's going on?" I was [8:56] like, "What are you talking about?" He's [8:57] like, "The status bar changed to what's [8:59] happening." Mhm. I was like, "Okay." [9:01] Mhm. He's like, "I didn't know one asked [9:03] me, you know, what I thought about [9:05] that." And I was like, "Oh man, yeah, [9:06] we're not doing that anymore." [9:08] >> [laughter] [9:08] >> Mhm. Like, it's going to start being [9:10] There's going to start being a lot more [9:11] of that. So, that was thing number one. [9:13] Okay. Thing number two was there were a [9:16] set of operating principles um that were [9:19] written on a whiteboard that Biz had [9:21] been involved in with either either on [9:23] his own or with other people. But they [9:27] were like [9:29] You know like sometimes you'll get on a [9:30] United Airlines flight or whatever and [9:32] they'll say, "Here at United Airlines, [9:33] we believe in integrity and honesty." [9:35] I'm like, "Okay, who cares?" Like, that [9:38] doesn't help me. That tells me nothing [9:40] about you. [9:41] >> company sounds exactly the same. [9:42] >> whatever. And also, it doesn't help me [9:44] make any decisions and these things are [9:46] meaningless and they're usually [9:47] weaponized against management by like, [9:50] you know, well, it says right here. [9:52] Anyway, so I was like, all right, we got [9:53] to get rid of these things. They're like [9:55] they're like not useful. They're costly. [9:58] >> Uh, no, they didn't really. And so, got [10:01] rid of those. I like there was there [10:03] were just [10:04] Um, I remember there were just like [10:07] everything was just kind of like messy, [10:09] you know? I remember um, there was this [10:11] kitchen and there was like it was just [10:12] kind of a mess. There were like dirty [10:14] dishes in the sink and everything. And I [10:16] think I sent out an email. I was like, [10:18] this is why the this is why the [ __ ] [10:20] site crashes. Like [10:22] you may you don't you may not get it, [10:24] but like this this is why the site [10:25] crashes cuz everything's just kind of [10:27] [ __ ] like, oh well, someone else will [10:29] clean it up. And I was like, this is [10:30] ridiculous. [10:32] And it was just try I just tried to get [10:33] this sort of like, we're not going to do [10:35] this like [10:36] oh well, we'll get them next year stuff [10:38] anymore. There was a poster [10:40] on one of the floors that was like, [10:42] we'll make better mistakes next time or [10:44] something like that, you know? And I was [10:45] like, you get that poster out of here. [10:48] It's not I don't really care about it, [10:49] but it's just sort of indicative of, oh [10:52] well, you know? And then the site would [10:53] crash again and people would go like, oh [10:55] well, we'll we'll figure it out, you [10:56] know? We'll hopefully it won't happen [10:58] again next you know, next year, [11:00] whatever. So, it was a lot of cleaning [11:02] that up and a lot of velocity velocity [11:04] velocity. All any CEO wants to talk [11:07] about now is speed. In particularly like [11:09] companies seem to slow down as they are [11:11] in hyper growth. [11:12] >> Yeah, because you [11:13] >> It's like you get bigger and get more [11:14] less done. Yeah. You because barnacles [11:17] build up on your you get like [11:19] organizational barnacles. [11:21] >> Yeah. You know, so a couple things. [11:23] Initially, when we were smaller, this is [11:25] sub sub 200 people, [11:27] what I did to speed things up was one, [11:30] we're not going to make decisions as a [11:32] group. We're going to push decisions [11:33] down the stack. By the way, I don't have [11:34] to be like don't come to me and ask me [11:36] what you should do about this unless you [11:38] have no idea and want my opinion. Like [11:41] figure it out. You know, and and and [11:42] don't tell your team what to do. Tell [11:44] them to figure it out and push as many [11:46] decisions down the stack as you can. And [11:48] I don't want to hear, you know, [11:50] I remember I would talk to Jack Dorsey [11:52] at these dinners we would have from week [11:54] to week when he was chairman and was CEO [11:56] of Block down the street from us. We [11:58] would get together on Tuesday nights and [12:00] have [12:01] And I remember telling him, "God, I hate [12:04] hearing Dick said in the company as a [12:06] reason for things." Cuz what would [12:08] happen was let I'll give you an example. [12:11] I remember this example [12:14] really well. The Android app [12:18] started taking longer and longer and [12:19] longer to boot. And so we had a we would [12:21] have a metrics meeting. I was like, "Hey [12:23] man, the you can see the usage numbers [12:26] going down almost linearly with Android [12:29] boot time going up. People are firing up [12:31] Twitter and it's taking too long and [12:32] they're going to move on to something [12:33] else." Like over here on iOS, [12:36] that doesn't happen. It's got to be the [12:38] same boot. It's got to take just as long [12:40] to boot Android as it does iOS. Okay. [12:42] So, [12:44] two, three nights later, week later, I'm [12:46] walking around like one of the [12:47] engineering floors after hours [12:49] and this Android engineer comes up to me [12:51] and goes, "Hey, can I ask you a [12:52] question?" Like, "Yeah." He goes, "Hey, [12:54] did you say um remove the JSON file from [12:57] the Android boot system?" I was like, [13:00] "I didn't. I know what a JSON file is, [13:02] but I couldn't tell you what's in the [13:03] Android boot JSON." And he's like, [13:05] "Okay, because I was told Dick said [13:08] remove that." And I and I was like, [13:10] "Aha, I think I know what's going on [13:11] here." Right? And you this would happen [13:13] throughout the company. Some manager [13:15] director decides, "Here's a way to solve [13:18] this problem." And the engineer pushes [13:21] back against them. And they just go, [13:23] "You know what? Dick said do it." [13:25] >> Yep. And it became like this reason for [13:28] doing things. So, part of the speed was [13:31] now we're not going to have someone so [13:33] said. It's push the decisions down the [13:36] stack. And by the way, if your team [13:37] makes a decision and you can't defend it [13:39] to your manager, like that's [13:42] don't go well so-and-so said. Like have [13:44] the engineer talk to the director, [13:45] whatever. [13:47] The second thing I did as we got bigger [13:49] and bigger [13:50] is I started realizing [13:53] oh, the organization is slowing down [13:56] because there are all these fiefdoms in [13:58] the organization that want that want [14:00] like ball control over their stuff. Mhm. [14:03] So, Very common. Very common. So, So, [14:06] here's here was here was how it [14:08] surfaced. Um we ran an experiment and I [14:11] was like, "Oh, this is actually really [14:13] super interesting. What are the [14:15] hypotheticals about what this experiment [14:16] would do in production?" The engineer [14:19] runs through it in an experimentation [14:20] meeting. Like great, let's go ahead and [14:22] launch this to, you know, X percent of [14:24] users and see what happens. [14:25] >> Yeah. [14:26] One week, two week, three weeks go by [14:28] and I'm like, "Hey, where's the I'm like [14:29] looking at the app. I'm like, "Where's [14:31] the you know, where's this engagement [14:33] experiment we're running?" And I find [14:35] the engineer he's like, "Oh, um I'm [14:37] waiting for um you know, trust and [14:39] safety to sign off on I'm like, "What?" [14:42] It's like, "Oh, yeah. Well, I asked a [14:43] so-and-so if we could launch the if [14:46] ready launch [14:47] and they didn't say no, but they said [14:49] you have to go get approval from years [14:51] of security and I started realizing that [14:55] that when people would try to go do [14:57] something [14:59] they wouldn't be told no, you can't do [15:01] that. [15:02] They It was something much, much worse [15:04] and slower than that. [15:05] >> You need approval from security [15:06] >> to get from [15:07] You got to go ask these four people. [15:09] Go ask these 14 people and if they all [15:12] say yes Okay, everyone's got this [15:14] problem. How do you fix it? I So, the [15:16] what I what I did was um [15:18] it's very similar to what uh Bezos used [15:20] to do at Amazon with biased action. I [15:22] had to It was from my improv background. [15:25] I was like, we're going to have a bias [15:26] to yes. And bias to yes is only your [15:30] direct the person you report to [15:33] is allowed to tell you you're not [15:34] allowed to do that. Or or legal can tell [15:36] you, if you do that you'll be fired [15:38] because it's illegal and a violation of [15:40] the privacy policy. But no other [15:42] organization, period, is allowed to tell [15:45] you you can't do that or you need to go [15:48] ask these people for permission. Just [15:49] your direct manager. And that means if [15:52] you have an idea for an experiment and [15:53] your manager signs off on it and you [15:55] have time to do it because you're done [15:56] with whatever else you're working on, [15:58] no one can tell you you can't do that. [16:00] Or, [16:01] you know, the marketing team's not [16:03] allowed to tell you you can't use the [16:05] Mailchimp software we license cuz that's [16:08] the marketing team's, right? And so bias [16:10] to yes was [16:12] I don't want to hear about something [16:14] being stalled because organization X [16:17] said you have to go to get approval from [16:18] organization Y. And that was great. [16:20] Experiments started flying out the [16:22] flying out um [16:25] in fact, engineers started going like, [16:27] wait, I don't have to necessarily get [16:29] design to build all these wireframes for [16:31] something. I can just do it myself since [16:33] we're only launching it to 1% and see if [16:35] it works. And so that was just great. It [16:38] worked [clears throat] throughout the [16:38] organization and I would repeatedly get [16:41] on stage at all-hands meetings and like [16:43] talk about like bias to yes and what it [16:45] means and what it doesn't mean. Doesn't [16:46] mean you can go do stop doing what [16:48] you're supposed to do because it's [16:49] priority and do whatever you want to do [16:51] uh cuz your manager's going to go like, [16:53] if you do that, I'm going to fire you [16:54] cuz, you know, the you're being [16:56] >> Okay. No group decisions. No group [16:58] decisions. DRIs. [17:00] >> And bias to action. Anything else that [17:02] all these CEOs I work with are like, [17:04] we're slowing down. How do we speed up? [17:06] Yeah. Don't solve problems with [17:08] processes. [17:09] So, it happens all the time in [17:11] organizations. Like, some customer [17:13] problem happens, the customer churns [17:16] out, or the advertiser quits yeah, and [17:18] the CEO goes, "That can never happen [17:20] again." You know, and so what happens? [17:23] Some step gets added to some process [17:26] that's like, "Before you do" I remember [17:29] Google famously the launch checklist got [17:31] to be like 17 pages long or something [17:33] and Larry was like, "This is insane." [17:35] [laughter] I was like, "Yeah, it's cuz [17:36] you know, it's like the FDA never [17:38] another thalidomide or or the TSA like [17:41] make sure there's no water bottle on the [17:42] plane." And pretty soon you're not [17:44] managing to outcomes, you're managing to [17:46] the processes and you got to go through [17:49] nine steps before you, you know, [17:51] Yeah, instead instead of just managing [17:53] to the outcome of make sure there's not [17:54] a bomb on the plane. So, the way we did [17:57] that you way you do that is you solve [17:59] problems with [18:00] DRIs, operating control, and not [18:03] processes. [18:04] You should probably only really have a [18:07] a couple organizational processes and [18:09] everything else should be span of [18:11] control and and [18:12] decision making. [18:14] When you do that, you can't [18:17] and this was a this was something I had [18:19] to [18:20] under come to understand and and [18:23] adapt to. When you do that, [18:26] you can't punish people when things go [18:28] wrong. [18:29] You you want them to take risks. You [18:31] want them to be bold. You want them to [18:33] have this [18:34] bias to action and doing things and not [18:37] going around and asking for tiptoeing [18:38] around and asking for permission. [18:40] But, one of the implications is you [18:42] can't punish people for making mistakes. [18:43] And I would got to the point where I was [18:45] I would tell the team [18:48] it's not the job of leadership to [18:49] prevent mistakes from happening. It's [18:51] the job of leadership to correct [18:52] mistakes quickly when they happen. So, [18:54] if you screw something up, like by all [18:55] means tell us so we can help you fix it, [18:58] but no one should get in trouble for [18:59] screwing something up or making a [19:00] mistake. And I I didn't come to that [19:02] immediately, but I [19:04] came to realize that that was an [19:05] important artifact if you will of [19:08] pushing decisions down the stack and [19:10] bias to yes and we're not going to have [19:12] lots of processes. [19:13] >> Love it. [19:14] Um so many of the CEOs I work with, I [19:17] just was at a [19:18] I won't say the company, but here in New [19:20] York City and spoke at their company [19:22] meeting. And here's how I would just [19:23] describe every company when it's call it [19:26] under 50 employees, everybody knows [19:28] what's going on all the time. When it's [19:30] over 150 employees, nobody knows what's [19:32] going on anytime. [19:34] >> Yeah. How did you [19:36] fix How did you scale communication, [19:39] keep people on page? [19:40] Tell me a little bit about the [19:41] management by walking around. Yeah, [19:44] well, I did two things. Um first of all, [19:47] I taught this management class at [19:49] Twitter. It was actually based on Ben [19:51] Horowitz teaching a management class at [19:53] Loudcloud. Fascinating. [19:55] >> Um [19:55] and Ben had in his outline from his [19:59] management class a statement that I like [20:02] highlighted and underlined and circled [20:03] three times. It was make sure everybody [20:06] understands what you understand. [20:08] Okay. So, I would always tell my [20:09] managers that. Like, your job is to make [20:12] sure everybody understands what the [20:14] priorities are [20:15] and [20:16] what it means for them and their team, [20:19] how they fit into those priorities, and [20:21] and more importantly for your team, what [20:24] it means for them personally and [20:25] professionally if we achieve what we [20:27] have to achieve. Those Those kinds of [20:29] things are great. I remember somebody [20:30] telling me a story of taking a factory [20:33] tour at a Tesla shop and seeing some guy [20:35] who's like screwing a widget into the [20:37] car and asking the guy like, "What What [20:39] are you doing?" And And the guy didn't [20:41] just say screwing the widget into the [20:42] car. He said, "Well, if I do this and [20:45] we're able to build these kinds of cars [20:47] and these kinds of cars can become more [20:48] popular in America." I mean, the guy [20:50] understood the story and I thought that [20:53] was great. It's It's like the analogy [20:55] somebody asked a bricklayer what he's [20:56] doing and Yeah. the bricklayer says, [20:58] "Well, I'm laying bricks." And the other [20:59] bricklayer says, "I'm building a [21:00] beautiful cathedral." I love that. [21:03] >> Yeah, like you want people to under I [21:05] mean, how much more motivating is it to [21:07] work in a company where you know if I'm [21:10] successful in my work, I understand what [21:12] the implications are for the for both me [21:14] personally and the company. [21:16] >> and context. [21:17] >> Priorities and context. Um Good news [21:19] travels fast, bad news travels slow. And [21:21] it's slow as the org gets bigger it [21:23] slows down. And the and the signal gets [21:25] distorted. Yeah. How does CEO deal with [21:27] that? [21:27] >> Yeah. I said earlier, [21:29] it's not the job of leadership to [21:31] prevent mistakes from happening. It's [21:32] the job of leadership to help correct [21:33] them quickly when they happen. So I [21:35] would always communicate, you got to [21:36] tell leadership if something screwed up. [21:38] If you're if your situation is screwed [21:40] up or your team or some decision is a [21:42] mess, you have to communicate that [21:44] constantly. Um and then secondly, [21:48] I would regularly communicate the [21:50] priorities to the team over and over and [21:52] over at all hands meetings. Here's what [21:54] the priorities are that we're working [21:55] on. This is how we're going to measure [21:57] them. This is why they're important. And [21:59] this is what they mean for us as a [22:00] company if we're successful. You always [22:02] wanted peo- I always wanted people to [22:04] understand not just what we had to do, [22:06] but why it was important to do it. And [22:09] and I think that scaling communication [22:10] architecture is like one of the most [22:13] important things in any company. And [22:16] part of it is [22:18] never saying because so-and-so said. [22:21] That's not context. That doesn't help [22:22] your team understand context for why [22:24] some decision was made or what or what [22:26] they should do or how they should go do [22:27] it. Um so that combination of pushing [22:30] decisions down the stack, making sure [22:32] that everybody understands what you [22:34] understand, and that a lot of that is [22:36] context, and always communicating [22:38] problems or issues up the stack as a [22:40] means of helping get helping get it [22:42] resolved quickly were the ways we dealt [22:44] with that. One last thing that I picked [22:46] up from Ste- a Steve Jobs trick. [22:48] >> Okay. So [22:49] >> a Steve Jobs trick. [22:50] >> Yeah. So when Steve was managing both [22:52] Pixar and Apple. So at the So he's CEO [22:56] Pixar still, but now he's back in the [22:58] he's back in the seat at Apple. So now [23:00] he's only at Pixar one day a week [23:02] because there's all this stuff I got to [23:04] go fix at Apple. [23:05] So, they're making like Toy Story 2. And [23:08] what Steve would do, so he's got one day [23:09] to come up to speed on what the heck's [23:10] going on at Pixar. So, he'd go go into [23:12] the office on Friday and he would say, [23:14] "All right, today I want to sit down [23:15] with like the um illustrators, you know, [23:19] without Lasseter, the director of the [23:20] movie, John. Without Lasseter in the [23:22] room. I just want to sit down me and [23:23] the, you know, illustrators." And he [23:25] would go around, there's like eight or [23:26] nine or whatever of them. He'd say, "You [23:28] know, Janet, um tell me something that's [23:29] not working well at Pixar." [23:31] >> Yep. [23:32] And then, you know, "Bob, do you agree [23:33] with that?" You know, taking notes like, [23:35] "Okay, Phyllis, tell me something that's [23:37] working well at Pixar. Not on the movie, [23:39] not about your job, you know." [23:41] And at the end of the meeting, [23:43] 45-minute, 1-hour meeting, he'd say, [23:44] "Okay, [23:45] here's what I heard. Um you know, [23:47] people are having these kinds of issues. [23:49] This is going really well. I'm going to [23:51] go talk to John about it and he'll deal [23:53] with it on he'll deal with the rest of [23:54] this on Monday. You know, you can follow [23:56] up with him." So, he's not saying, "Oh [23:58] yeah, that sucks" or [24:00] or, you know, "No, you're wrong" or [24:02] whatever. He's taking notes, [24:04] making sure he hears everything that was [24:05] said, and then going talk to Lasseter [24:06] about it. I started doing that with [24:08] teams. You quickly find out if that team [24:11] knows what the priorities are, is in [24:14] sync with what you're hearing from the [24:16] leadership of that team. And some of the [24:18] leaders of the company hated it. And I [24:19] was like, "Well, now I'm concerned about [24:21] why you why you hate it cuz yeah." But, [24:24] I wouldn't, you know, you're not in that [24:26] meeting doing things like throwing the [24:27] director under the bus. You're going, [24:29] "Okay, here's what I heard. I'm going to [24:30] go talk to my management team about it [24:32] on Monday and, you know, Chris or [24:33] whoever will follow up with you guys." [24:34] Let me switch to you on priorities. So, [24:36] all these CEOs I coach, they have their [24:38] list of priorities, it's January 1, but [24:40] holy my god, there's so much [24:41] opportunity, so much interesting stuff, [24:43] and this employee has this thing going [24:44] on. [24:45] How often did you rejigger those [24:47] priorities? Were you were you [24:48] disciplined about it over the course of [24:51] whatever that period was? [24:52] >> You can rejigger them once in a while. [24:54] But, once in a while? know, like, you [24:56] know, if you've got four and huge [24:58] priorities for the first four to five [25:01] months of the year, you can't you can't [25:03] get to to February and go these two are [25:05] stupid and we're you know, once or twice [25:08] a year you can go at this one is let's [25:10] let's actually change that one to this [25:12] other thing. [25:13] If you start doing it regularly, people [25:15] are like, well, we can get to work on [25:17] the direct messaging architecture, but [25:19] Dick's probably going to you know, 50/50 [25:21] he's going to change you just can't do [25:22] stuff like that. By the way, that's [25:24] another way companies slow down is [25:26] because people are just like, look, I'm [25:27] not going to bust my ass on this thing [25:29] when there's a 50% chance Dick's going [25:30] to throw it out the window in in a [25:32] month. [25:33] >> Yes. Let me test something on you. A lot [25:34] of these hyper-growers [25:37] it used to be that there were a lot of [25:39] one-way doors in the software industry. [25:40] Stuff took a long freaking time to get [25:43] out the door even to test. [25:44] >> Yeah. And you had to kind of do things [25:46] serially. Now you can do so much so fast [25:51] and so they're just doing more and it's [25:52] more parallel and a lot of those one-way [25:54] doors are two-way two-way doors. Is [25:56] focus overrated now? [25:57] No. [25:58] Focus is focus I don't think focus is [26:01] ever overrated. However, [26:03] I however, the fact that you can do more [26:06] quickly and since they're two-way doors, [26:08] you're like, great, no I mean, this is [26:10] the old Jeff Bezos [26:12] only take a long time to make decisions [26:14] that are existential. Like, we might die [26:16] as a company if we make the wrong [26:17] decision. There's fewer and fewer of [26:19] those. This decisions that can be like, [26:21] if it doesn't if we screw it up, we can [26:23] go back and do the other one. [26:25] Do those right away. So, focus is is [26:27] still important, but it's also true that [26:31] you should feel you should feel like you [26:33] can do a lot more now with less because [26:35] you can. There was a point in 2011 when [26:38] we wanted to try to get Jeff Bezos to [26:39] join the board of the company. [26:41] Like, we have an independent seat open, [26:43] who would be the best possible [26:44] independent director? Jeff Bezos. So, [26:47] Jack and I go meet with Jeff Bezos. I [26:50] think it was like at the Allen & [26:51] conference or something like that. And [26:53] Jeff asks a question about strategy. [26:56] And [snorts] Steve Jobs was still alive [26:58] at the time, and Jack said, "Well, Steve [27:00] says the most important part about being [27:03] a CEO is focus. And you should say no to [27:06] almost everything. And it's in fact it's [27:08] the it's the CEO's job to be the editor [27:11] in chief, and you should say no to [27:13] almost everything and do very, very few [27:15] things." [27:16] And Bezos looks at Jack and goes, "Oh, [27:18] yeah? Well, I like to do everything." [27:21] >> [laughter] [27:22] >> And he does the big Bezos laugh. And he [27:25] goes, "My team has to talk me out of [27:27] stuff." [27:28] >> [laughter] [27:29] >> He goes, "You know, there's lots of ways [27:30] to be successful." And I remember that [27:32] like it was yesterday, and he's right. [27:35] Um there's many ways to climb the [27:37] mountain. I think it's true that today [27:39] he would say, "Great, I can do even more [27:41] now." We can still We can still be [27:43] focused, but I can do even more things. [27:45] I think those I think those things can [27:46] both be true. Okay, I just want to get [27:48] back to some of your management [27:49] techniques cuz I think they were great. [27:51] This management by walking around was a [27:53] very old school Hewlett-Packard concept. [27:56] Just talk about It sounds like you did [27:58] the same time every week, and okay. I [28:00] did it at all hours. [28:02] >> Okay. [28:03] Mostly late night. I It started because [28:07] um Bill Campbell, uh for people who [28:09] don't know or listening to podcast, Bill [28:11] was on the board of Apple and famously [28:13] coached Larry, Sergey, and Eric at [28:15] Google. I wrote a book about him called [28:17] Trillion Dollar Coach. He the the [28:19] Benchmark folks had also brought him [28:20] into Twitter before I got there. [28:22] Um and Bill um famously said um when I [28:26] took over as CEO, like you could um [28:29] roll a grenade into Twitter at 5:30 p.m. [28:31] and only take out the cleaning people. [28:33] >> [laughter] [28:33] >> So, and he was and he was and he was [28:35] right. So, I was like, "Okay, how do you [28:39] You can't tell people You just can't [28:41] tell people work harder." You know, [28:43] they're like, "Okay, you work harder." [28:45] You know? So, what I did was I just I [28:48] mean it's so such a stupid thing, but I [28:51] was like, all right. I'll walk around at [28:53] night. I'm going to start [28:54] I'm going to go out and get dinner. I'm [28:56] going to come back at like 9:00 p.m. [28:57] 9:30 p.m. I'm going to walk around and [28:59] whoever's here, I'm going to talk about [29:01] the stuff they're doing at the next [29:02] all-hands meeting and I'm going to like [29:04] I'm going to like bump up the [29:05] prioritization of that. [29:06] >> Yeah. And I remember like Marcel Molina [29:09] @noradio on Twitter. That's a That's a [29:11] little promotion for Marcel there. [29:13] Marcel is always working late on stuff [29:15] and I would like, "What are you working [29:16] on?" And he'll be like, "I'm working on [29:17] native retweet or whatever." And I would [29:19] get up in front of the company and go [29:20] like, "What Mar- the work Marcel is [29:22] doing." And people are like, "How come [29:23] How does he always know what Marcel is [29:24] doing?" [29:25] >> Okay. And the other people would go [29:26] like, "Oh, Dick comes back after dinner [29:27] at night and walks around and people [29:29] slowly slowly slowly started [snorts] to [29:33] change the culture of Okay, the people [29:35] who are around when Dick's walking [29:37] around are getting their stuff [29:38] prioritized." [29:40] I I did I I copied you. I walked around [29:43] Friday afternoons. [29:44] >> Yeah. And I I learned a lot. [29:46] >> Yeah. And you Yeah, you So, I would just [29:48] go, you know, up to engineers and go [29:50] like, "What's going on? What are you [29:51] working on?" Yeah. [29:53] Um you have an interesting expression [29:55] that I don't understand about forest [29:58] fires in forestry management. [30:00] >> Yeah. [30:01] >> What the hell are you talking about? [30:02] >> Yeah, I I had I was trying to help um a [30:05] director in the company [30:09] be more strategic. They were super [30:11] tactical. Okay. And I was finally like, [30:13] "Your like your job as a director as you [30:15] grow grow in this company is [30:18] you can't just say like, 'Well, I put [30:20] out this forest fire and I put Like your [30:21] job is to map the territory and start to [30:24] do forestry management. Like, what are [30:26] we going to do so that next year if [30:29] there's a fire over here, that highway [30:33] is not going to be in trouble?' Like [30:35] you can only work so many hours a day [30:37] and stamping out fires isn't going to [30:39] work as we scale. You need to push those [30:41] down the stack. Your team should be like [30:43] fighting the fire that's over there [30:45] because it might, you know, [30:47] burn the house down or something. You [30:49] need to be focused more on like thinking [30:51] about this whole pie and what what's [30:54] going to be true what has to be true 24 [30:56] months from now so that this whole pie [30:58] is in this whole territory is in better [31:00] shape. And I just That was the way I [31:02] That was the metaphor I had for helping [31:05] people try to think strategically versus [31:07] like, well, I'm I'm doing, you know, I'm [31:10] chopping down the trees. Look, I'm Look [31:11] at all the trees. What That's not like [31:13] leadership and strategy. That's tactics. [31:15] >> that. Okay, now I get it. One of the [31:17] people I connect you connected to [31:20] connected with before interviewing [31:22] you said something interesting about [31:24] you. Oh boy, I can hardly wait. I'm sure [31:26] most people did, but anyway. [31:27] >> And And both sides of this are [31:28] interesting. He's Both Both sides. [31:31] >> He said, [31:32] "You know, Silicon Valley, San [31:34] Francisco, everyone's [31:35] passive-aggressive. [31:36] >> [snorts] [31:37] >> And [31:38] >> [laughter] [31:39] >> So, I'm curious your take on that. [31:41] Um cuz I'm I'm I I don't live in Silicon [31:43] Valley. And the second thing he said, [31:45] "Dick's not passive-aggressive. Dick's [31:47] aggressive." [31:47] >> Yeah. [31:48] What's Give me both sides of that. [31:50] >> Uh it's just like I'm super impa- I know [31:53] this is a personal flaw, but I'm super [31:55] impatient. Like people would come in and [31:57] go, "Here's why I'm, you know, why I [32:00] think or whatever. I'm sad." You know, [32:02] like And they would start talking to me [32:04] like, "Okay, stop." Like [32:06] It seems like you don't like working [32:07] here. You should quit. Okay. [32:09] >> like, "No, I want to tell you why I'm [32:10] sad." You know, like "Okay, I don't like [32:13] I don't care. [32:14] Who cares? [32:16] I'm sorry. I'm not trying to be a jerk, [32:17] but like we have a million things to do. [32:20] I remember and then, [32:22] you know, I just like [32:24] This is I just direct and blunt with [32:26] people. I'm not a jerk. You don't have [32:28] to You don't have to People think like [32:31] think a lot of times, especially maybe [32:33] engineering leaders who, you know, [32:35] come up in the organization like coding [32:38] and programming and doing stuff on my [32:39] own and I have to manage people. They [32:41] think when you tell them you have to [32:43] like look, you have to manage [32:44] performance in your team and have to [32:45] manage poor performance out. They think [32:47] you have to be a jerk when you do that. [32:48] You don't have to be a jerk about it. [32:50] You just have to be direct about it and [32:52] forthright and not, you know, [32:54] no one's sadder than this about than me, [32:56] you know, don't say stupid stuff like [32:57] that. You know, you can be empathetic [32:59] and still be straightforward. Okay, a [33:01] lot of these CEOs I coach [33:04] are in their they call it in their mid [33:06] late 20s. [33:07] Um [33:08] and like in their whole life they [33:10] haven't given [33:11] they don't go to like mom, I got some [33:13] feedback on dinner or like you're [33:15] playing you're hogging the basketball. [33:17] You've just never given a homo sapien [33:19] feedback. It's unnatural. [33:21] And advice for that CEO? Yeah, my advice [33:24] for that CEO, Bill Campbell actually [33:26] gave me this advice early on [33:28] is like write down what you want to say [33:31] and then say that and don't say anything [33:33] else. So, the night before if you write [33:35] back down the the honest feedback that, [33:37] you know, I would give Brian especially [33:39] if have to give Brian a bad performance [33:40] review, the night before if you write it [33:42] down [33:43] I guarantee you 99 out of 100 people if [33:46] they go into the meeting the next day, [33:47] they don't say that. They're like, [33:49] listen Brian. Instead they're like, you [33:51] know, it's really hard time here for [33:54] everybody. We're all working, you know, [33:57] shut the [ __ ] up. Like the person knows [34:00] you're about to like hit them over the [34:02] head or something. And you also don't [34:03] have to say like [34:05] Hey Brian, that's just the way it is, [34:06] you know, too bad if you don't like it. [34:08] You just just write down what you want [34:10] to say [34:11] and be forthright and honest and, you [34:14] know, you can say like I understand it's [34:16] like you can't don't want to deal with [34:17] it anymore. I get it. [34:19] And say those things and don't add a [34:21] bunch of fluff around it. I thought that [34:22] was great feedback and I would do that. [34:25] Sit with the silence. [34:26] >> Yeah, sit with sit with the silence. [34:27] >> Sit with silence is hard. Yeah, I so in [34:30] unnatural it's unnatural. The best [34:32] person so I used to do a thing in my [34:34] management class where it was a 6-hour [34:37] course. There was a ton of stuff in it. [34:39] But one of the things I would do, it was [34:40] about feed giving people feedback. And [34:42] like, "Okay, [34:44] like Brian, you're going to give Dick [34:46] feedback and the feedback is you've been [34:48] working on this thing for 6 weeks and [34:50] busting your ass and you've been here [34:52] until midnight every night and we have [34:53] to cancel the project cuz it's no longer [34:55] a priority. Yep. For what something [34:57] horrible that no one wants to hear. Yep. [35:00] You know, it's a made-up made-up of [35:01] course. The only person who did the role [35:03] play that was ever good at sitting in [35:05] silence was Elad Gil, current venture [35:08] capitalist. Elad sat in the chair and [35:10] gave the person the feedback and and I [35:13] and I say to the employee that I see [35:16] before the role play, I kind of pull [35:17] them aside and go, "You're You're this [35:19] is [ __ ] and you're going to tell [35:21] Dick and you don't want to hear this and [35:24] it's it shouldn't be canceled. They need [35:25] to look at the data. Whatever." [35:28] And Elad was great at just sit there. [35:30] It's hard. And then eventually the other [35:33] person runs out of gas and stops, you [35:36] know, talking. And Elad was great at [35:38] that and the only person who ever did [35:40] it. Okay. [35:41] One of the things I want to ask you [35:42] about is you had some very talented [35:44] folks, some of which you brought in, [35:45] probably others brought in like Adam [35:47] Bain and Kevin Weil, Anthony Noto, Elad. [35:50] Yep. How did you guys recruit? What did [35:53] you look for? Are you looking for slope [35:55] or experience? Like what was your [35:56] filter? [35:57] It's a great question. I always I [35:59] actually tried to promote with from [36:01] within as much as possible because you [36:03] want people inside the company to feel [36:05] like [36:06] we're not like going to get capped out [36:08] at which point, you know, I can only get [36:10] to this level because anything above [36:11] that Dick just goes out and finds [36:13] someone. So Kevin Weil, for example, was [36:15] promoted from within from an IC when I [36:17] got there in 2009. Um well, there were [36:20] only a few people in the 40 people or so [36:22] in the company when I got there. Um to [36:24] eventually running all of all of product [36:26] um after running all all of that [36:28] product. Um, so he was promoted from [36:30] within. Noto was [36:33] I just felt like Twitter was such such a [36:35] drama. Twitter is like the drama queen [36:39] of hypergrowth companies. [36:42] You know, it was like constantly in the [36:45] news. When I took over as CEO, that year [36:48] the Twitter bird had hanged itself and [36:51] exploded on two different magazine [36:53] covers. So I went up to our head of [36:55] communications, Gabriel Stricker, and I [36:57] was like, "Gabriel, great news. Next [36:59] year's goal, no dead birds on magazine [37:02] covers." You know, low bar. Uh, anyway, [37:05] but it was just constant drama and Noto [37:09] was so good at Goldman Sachs where he [37:11] ran technology and media at dealing with [37:14] investors and communicating with them, [37:16] etc., that I had been I had been trying [37:19] for a while to bring him into Twitter [37:21] and I was I was finally able to and he [37:23] was, of course, he's he's fantastic. How [37:25] about Adam? [37:26] Adam was funny. Adam was a non, um, what [37:29] we what you call a non-consensus hire [37:32] when when I first recruited him because, [37:35] uh, um, it was while I was COO and I [37:38] went to, um, the board and Evan said, "I [37:40] got the perfect person to run revenue at [37:43] Twitter. [37:44] It's this guy, Adam Bain. He's in LA and [37:48] he works for Fox Interactive." And [37:51] everyone [laughter] was like, [37:53] "You're going to bring in someone who [37:55] works [37:55] >> selling ads. So you're going to want to [37:56] bring in someone who works at Fox in LA [37:59] and bring them to San Francisco to run [38:02] our, like, you know, revenue side of the [38:03] business?" But he was the perfect person [38:05] to do it. I must have flown down there [38:08] to LA [38:10] for five or six months before I I could [38:12] convince him to do it. Um, but I'd met [38:15] him while we were in the process of [38:17] selling FeedBurner and, um, his his [38:19] employer, News Corp, was one of the [38:21] interested parties before Google bought [38:23] it and I got to know him and I was just [38:24] like, man, this guy is like tireless. [38:27] And we're going to be in like a constant [38:30] dog fight and [38:32] and fighting from behind against [38:34] Facebook for social ad dollars. [38:37] And Adam was just he's just like one of [38:39] those people who's got a motor that [38:40] doesn't stop and he was the perfect [38:42] person for the role. There's an old [38:43] adage [38:45] uh fire fast [38:46] >> Yeah. hire slow. [38:48] >> Yeah. I always said that out loud. I [38:49] didn't do that. [38:50] >> [laughter] [38:50] >> I didn't I fired slow and almost every [38:53] time I was like, man, I should have been [38:55] 6 weeks I knew. [38:56] >> Yeah. Yeah. It was [38:58] >> about that? Yeah, we I tried to hold [39:00] myself accountable to that and it's [39:01] horrible. Like there's nothing there's [39:03] nothing worse than having to do [39:04] something like that. I remember I one of [39:06] my board members David Rosenblatt You [39:09] know, [39:10] I was [39:10] I [39:12] I was um I'm a little bit like, come on, [39:14] when people are like, oh, I like firing [39:15] people because you know, it's it's the [39:17] right thing for them. Like that's just [39:19] sounds like you're an [ __ ] cuz it's [39:21] horrible. [39:22] >> Um but you have to do it. Yeah. And um I [39:25] remember asked David Rosenblatt cuz I [39:27] had to terminate someone with global [39:28] operating responsibility. I was like, [39:30] wait a minute. [39:31] If I fire him in the morning, the people [39:33] in Japan [39:35] are going to be in bed, but if I fire [39:37] him in the afternoon, then the you know, [39:39] what whatever. And so I was like, I got [39:41] to figure out and and I was talking to [39:43] my board about it and David Rosenblatt, [39:45] who had been the CEO of DoubleClick [39:47] before it sold to Google, was like, oh, [39:48] I had to do that. I actually saved the [39:50] calendar date from that cuz I had a [39:52] whole series of communications I had to [39:53] roll out. This is where now a couple [39:55] thousand person company, etc. [39:57] with like 30 offices around the world. [39:59] So, I get his day I get his day of [40:02] communications and the only thing I [40:04] remember from it it's a whole you know, [40:06] it's a whole layout of how to go about [40:07] it. The only thing I remember from it is [40:10] 6:00 a.m. wake up in a cold sweat. You [40:12] know, it's just horrible. Uh you know, [40:14] it's not fun. You can't wait till it's [40:16] over. Um and you just have to be but you [40:18] have to do it. And most Jeff Weiner's [40:21] old analogy from he was LinkedIn CEO [40:23] when I was running Twitter. Jeff Weiner [40:25] would call it the um, [40:28] for those of you who are baseball fans, [40:30] he was like, you know, sometimes you're [40:31] watching a baseball game and you see the [40:32] pitching coach or the manager walk out [40:34] to the mound and talk to the pitcher and [40:37] the pitcher's like, "No, no, no, I [40:38] Totally. Don't worry, coach. I still got [40:39] it." And the manager walks back to the [40:41] dugout and next guy cranks one out of [40:43] the park again. And you can see the [40:45] manager walking out like, I knew it. I [40:47] knew it. And Jeff was like, "I never I [40:50] always tell myself, don't be the manager [40:53] having the first conversation with the [40:54] person when you know what you should be [40:56] doing is having the second conversation [40:57] with the person." As a Red Sox fan, that [40:59] really hits. [laughter] Really hits [41:01] home. [41:02] Uh, I think his example is actually [41:03] Pedro Pedro Martinez from the Red Sox, [41:05] but anyway. [41:05] >> great little didn't take him out. Yes. [41:07] Um, I work with all these hyper-growth [41:09] companies. I think you do, too. What [41:13] What are the pitfalls when you're in [41:15] hyper-growth and what are your tips for [41:17] the CEOs listening today who are growing [41:20] incredibly fast? One of the real [41:22] pitfalls, especially for these [41:23] hyper-growth AI companies that are going [41:25] from, you know, a [41:27] single-digit revenue million to 10 to [41:29] 100 million, um, is [41:32] it's very, very easy to take your eye [41:35] off the ball when you raise too much [41:36] money. And by taking your eye off the [41:38] ball, I mean we're thinking about [41:40] operating leverage and operating [41:41] efficiency. Because once you've once the [41:43] money's there, you're like, "Don't have [41:44] to worry about that. I'll go worry about [41:46] this thing over here." And next thing [41:47] you know, the you know, burn is like, [41:50] "We have We know We have What are you [41:52] worried about? We have It happens all [41:54] the time. And Ryan wrote this great post [41:56] saying, "Look, I thought I was like Mr. [41:58] Efficiency. We're going to be super Not [42:01] It's not going to be the same for us. [42:02] We're going to be super capital [42:03] efficient." But when you don't have to [42:05] worry about it at all, [42:07] and there are a million other things to [42:08] do, you stop worrying about it and you [42:10] start getting people [42:11] >> too much money? [42:12] I think so and many many cases. I don't [42:14] know that the models are the models [42:16] probably need all the money for compute. [42:18] But but like an app company going for [42:19] $100 [42:20] >> I think people are raising too much [42:21] money. I think a lot of companies are [42:23] raising too much money because you know, [42:25] raising a 100 million series B is the [42:27] you know, is the new you know, raising [42:29] is the new $20 million series B. What [42:32] else in hyper growth? Like what breaks [42:34] in hyper growth? [42:37] Uh hiring all hiring almost I mean my [42:40] favorite thing is you talk to you know, [42:43] you talk to like a CEO who's like I [42:45] don't know whatever pick pick some [42:46] number, you know, first company they [42:49] were lead product manager at awesome [42:51] fintech company X and now they're CEO [42:55] and they're like we're only going to [42:57] hire a players. You know, I'm like yeah, [42:59] PS [43:00] guess what? [43:02] Everyone says that. [43:04] Yeah. Yeah, come back to me when you're [43:06] when your [43:08] customer success team has to hire twice [43:12] as many people next quarter because you [43:14] just turned a bunch of customers because [43:16] no one checked in with them last quarter [43:18] because you're growing like a weed and [43:20] then tell me, you know, and you just you [43:21] just end up not doing that. And so one [43:23] of the things that you have to do [43:26] when you come to the realization that we [43:29] don't only only hire A plus players is [43:32] you have to get like religious about we [43:37] have to you have to manage out low [43:38] performers. [43:40] But as you move down the stack and [43:42] you've got inexperienced managers who [43:43] are first time managers, first time [43:45] directors, man, there's nothing that [43:47] they want to do less than let someone [43:48] go. Yeah. They'll do anything to not do [43:50] it. [43:51] You prevent them from trans firing [43:53] people, you know, trans firing is like [43:56] Janet would be great on the ads team. [43:58] She loves ads. You need people. [44:01] >> Yes. [44:02] And then you're like no no no. So I had [44:05] a you know, we had we had sort of [44:06] performance reviews. [44:08] I was like, if your review is below a X [44:10] number, you can't switch teams. Yeah. [44:13] >> Like, that's a good idea. [44:14] >> No, you know, uh whatever, let's say [44:16] it's one through four. Four is amazing, [44:18] one is you're got to go. If you're a [44:19] two, you can't move teams until you're [44:21] no longer a two. PS, if you go two to [44:24] three to two, I'm going to talk to your [44:25] manager again and go, don't let me [44:27] guess, this person could be a three [44:28] again this next term so you don't have [44:29] to put them on a performance plan. So, [44:31] we would, after performance reviews, [44:34] always go through as a senior team, [44:37] like, okay, [44:38] I can see from the way these peop- these [44:40] couple people or these eight people are [44:42] bouncing from two to three to back to [44:45] two to back to three, like, you're just [44:46] trying to not fire people. [44:48] >> Mhm. And I can see that your curve is [44:50] inflated and blah blah blah. So, we [44:51] would just be super hardcore about that. [44:53] So, you can't do that, you know? Let's [44:55] talk about competition for a second. You [44:57] had some [44:58] I mean, you're competing with with [44:59] Facebook. [45:00] >> Yeah, they were the juggernaut that we [45:02] just had the hardest time like [45:03] >> And like tomorrow I'm speaking at [45:04] another company meeting. I'm not going [45:06] to say the name of the company, but [45:07] they're [45:08] they're going to be competing with [45:10] OpenAI and Anthropic. Almost every [45:12] boardroom I'm in is talking about that. [45:14] Everyone's anxious about that. Their [45:15] employees are anxious about that. Advice [45:18] about that. First of all, you have to [45:20] acknowledge it. And then you just have [45:21] to I mean, one of the things I would [45:23] always do is like [45:24] they're like huge and Mark is the most [45:28] competitive person on the planet. Trust [45:30] me when I tell you like [45:32] they don't like they don't say to [45:34] themselves, oh, whatever, Twitter won [45:35] that thing, that's fine. They're like, [45:38] that can never happen again. So, we have [45:40] to be like [45:42] and we're we're way smaller than them [45:45] and have far fewer resources and can, [45:47] you know, and so it's just like we just [45:49] have to be we have to we have to think [45:51] about them, we have to talk about them, [45:54] have to understand what they're doing, [45:55] and have to try to figure out how to [45:56] out- out-hustle them. And Bain Adam Bain [45:59] on the revenue the was great at that. He [46:01] would come into company meetings at [46:05] Consumer Electronics Show at the [46:06] beginning of the year and go to one of [46:09] the first big customer meetings like [46:10] let's say it's a Unilever [46:12] and he would sit down with the CMO from [46:14] Unilever before the meeting [46:16] and say, "How'd the Facebook meeting [46:17] go?" You know, "Oh, it great. blah [46:19] blah." What are you like what are they [46:20] What are they psyched about? What are [46:21] they psyched to show you this year? [46:23] You know, and then he would go into the [46:25] he would learn all that and then he [46:27] would go into the meeting with the [46:28] Unilever and say [46:29] you know, even if we had a bunch of [46:30] things to show them he'd say, "We have [46:31] some things we want to show you, but we [46:33] really want to spend the next 2 hours [46:34] today just understanding your business [46:36] and what you guys are concerned about [46:38] what your problems are." I mean, knowing [46:39] that she just told him that the Facebook [46:42] people came in and showed us this and [46:44] this and this and weren't asking those [46:46] kinds of questions. Now, Facebook wised [46:47] up to that over time, etc. But, Adam had [46:49] just come up with a great way of [46:51] thinking about how to compete with this [46:53] juggernaut. And the juggernaut's rolling [46:55] out new things every quarter. This is [46:57] really good salesmanship. He's like, "I [46:58] can't compete with their eight new [47:00] things and they're microtargeting and [47:02] blah blah blah. So, I'm going to totally [47:04] come at it from a different angle." And [47:05] he was just smart about learning how to [47:07] do that by talking to the customer [47:09] beforehand. [47:10] You guys picked the advertising model to [47:12] make money. As you look back at that [47:15] good call? Yes. [47:17] On advertising is undefeated. The ARPU [47:19] is higher. [47:20] It just is on versus subscription [47:23] revenue, etc. I think we would do the [47:24] same thing over again. Any other as you [47:27] look back like, "Man, [47:29] I wish I had to do over that one." [47:30] [laughter] [47:31] Oh, yeah. Uh we tried to buy Instagram [47:33] the month before month and a half before [47:35] Facebook bought it. And by the way, for [47:37] a much higher percentage of the company [47:38] that you know, Facebook's offer is [47:40] famously a billion dollars and I think [47:42] it was 700 of stock and 300 million of [47:44] cash. They only had eight or nine [47:46] employees at the time. [47:47] Um [47:48] uh Mike Krieger and Kevin Systrom. [47:51] But, we knew well before that like these [47:54] guys are [47:55] Kevin is an extraordinary product [47:57] manager. Kevin and Evan Spiegel, I [48:00] think, are probably two of the best [48:01] product thinkers in Silicon Valley. And [48:03] so, I knew that. And well before Mark [48:06] made his offer to them, I pulled Kevin I [48:09] went out to dinner with Kevin and Mike [48:11] and told them we wanted to buy the [48:12] company and would end up giving them [48:15] basically [48:16] a well over 10% of Twitter and, you [48:19] know, something equivalent to 700 [48:21] million in stock. I think we were maybe [48:22] valued at uh [48:24] 5 billion dollars at the time. [48:26] And, you know, I would have gone And And [48:29] could I go back there and talk to Kevin? [48:30] Oh, I'd say like, [48:31] "You can run the company when I leave. [48:33] Like, congrats. You'll get to be see I [48:35] mean, may still not have done it." And I [48:36] would also say, "Also, I'll go borrow [48:38] 300 million dollars from JP Morgan and [48:40] give it to you." [48:41] >> I see. [48:42] Um and you can have it all and you know, [48:44] we'll pay them back later. Um that [48:46] because that was a game changer. Once [48:48] they had Instagram, I was like, "Oh, [48:51] man. We're like boxed out on [48:53] that is that is the future of of rich [48:57] media social media feeds. We have Vine, [49:00] but it's much much smaller and they're [49:02] not going to come at us with [49:05] on on Instagram with video, which of [49:07] course they did. And that And that was [49:10] That was sort of checkmate. Other [49:11] regrets? Any regrets on Man, the content [49:14] moderation stuff has been in the news [49:16] over the last 5 years. As you It wasn't [49:18] as hot while you were there. Any As you [49:20] look back at that, regret [49:21] >> was always like It was all The thing [49:23] I'll say about content moderation that [49:25] people get wrong [49:26] >> Yeah. [49:27] all the time is like, "This is real [49:29] easy. Just prevent people from doing [49:31] that." And like, "Okay. Well, that is a [49:33] lyric in a Kanye song. Can you not tweet [49:35] his lyrics?" Or whatever. These hard and [49:37] fast rules, [49:39] there has to be again, pushing, you [49:41] know, [49:42] uh having operating control and somebody [49:44] with an understanding of the situation [49:46] like our GC, you're going to have just [49:48] have a bunch of judgment calls and [49:49] you're going to get a bunch of them [49:50] wrong and that's life. [49:52] In like after your tenure [49:56] and like I don't know what to believe, [49:58] but it looks like there was a lot of [50:00] conversation between the Biden [50:01] administration and that group and then [50:05] a lot of accusations from Elon that that [50:07] group was just very left of center in [50:09] their orientation. Any reaction to that? [50:11] Was that the case when you were there? I [50:14] mean, yeah, probably it was left of [50:16] center, but I don't think that I don't [50:18] think when I was there [50:19] that [50:22] I don't think it impacted the way we [50:23] thought about abuse. I mean, it wasn't [50:26] like um you you know, we we didn't have [50:29] any like [50:30] for example, we didn't have any like oh, [50:32] you're not allowed to dead name trans [50:34] people. And like that stuff was all post [50:36] me being there. At the time it was [50:38] mostly what we were dealing with was [50:40] >> kind of happened after you left. [50:41] >> Yeah, the the yeah, the stuff we were [50:43] dealing with and I what when I was there [50:45] were ridiculous amounts of Russian bots, [50:48] which at the time were not election [50:50] stuff. It was maybe they were preparing [50:52] for to preparing the groundwork for [50:53] 2016, but it was mostly like [50:56] phishing attacks and just massive bot [50:59] farms that were doing phishing attacks [51:01] on on people and trying to hack into [51:02] their accounts or get credit card [51:04] information, etc. Most of the CEOs who [51:07] have been around a while have got a [ __ ] [51:08] ton of bad press. You got a metric ton [51:11] of bad press. [51:12] >> Yeah. [51:13] It must I got a bunch of bad press. It [51:15] impacted me. I tried to show a face like [51:17] it didn't, but it it it was soul [51:19] crushing to me. [51:20] >> Yeah. How did you handle it? [51:22] >> know. I like I had been a I had had the [51:24] fortunate opportunity to be a performer [51:27] and do improv comedy and stuff and [51:30] audition for SNL a couple times um [51:32] before I went back into [51:34] business and I had been like you know, I [51:38] was at I remember I was at the Adelaide [51:40] Comedy Festival with my improv group and [51:42] we had this midnight show at the Opera [51:45] House, the Adelaide Opera House. [51:46] And the crowd Yeah, and the crowd is [51:49] hammered drunk. Maybe a thousand people. [51:52] And it's a like an hour and 10 minute [51:54] show, completely improvised. We get on [51:57] stage and like 5 minutes in, the crowd [52:01] starts yelling like, "You suck. Get [52:03] off." [52:05] >> [laughter] [52:05] >> And like, "Man, we got 80 more minutes [52:08] or whatever." And [52:11] and uh [52:12] So, I don't know. I just became [52:14] resilient to like oh well. You got to [52:15] like Did it hurt your inside? [52:18] I don't know. Like we got good pre- I [52:20] remember my daughter called me. [52:22] >> The good press never hit for me. [52:23] >> No, no. I'll tell you I'll tell you my [52:26] daughter um called me. This is at the [52:28] end of 2014. So, 2013, we go public, IPO [52:34] successful, everyone's super excited, go [52:37] yay us. And there's some, you know, [52:39] there were these stupid awards that come [52:41] out come out. And like CEO of the year, [52:43] Dick Costolo. Under 2014, my daughter [52:46] texts me. She's like, "I have bad news [52:48] and good news." [52:50] And I was like, "What's the bad news?" [52:52] She's like, "Yahoo Finance says you're [52:54] one of the five worst CEOs of 2014." [52:58] >> [laughter] [52:59] >> I was like, "What's the good news?" She [53:01] goes, "Nobody reads Yahoo Finance." [53:03] Okay, got it. [53:05] Okay. I just felt like I don't know. [53:07] It's like you're like, "Okay." And I'll [53:09] say one last thing on it. I've said this [53:11] a bunch, but but it's I think it's [53:13] important. [53:14] >> [snorts] [53:14] >> Um I got invited to go to like the [53:16] Vanity Fair Oscars party one of the [53:18] years. [53:18] >> way, that must have been cool. Well, I [53:20] got invited to go. I didn't go. [53:21] >> Oh. [53:22] And [snorts] I told my daughter [53:25] Good question. So, I told my daughter [53:27] "I don't want to start going to a bunch [53:28] of stuff that I get invited to because [53:30] I'm the CEO of Twitter, because then [53:32] I'll be I'll be like, 'Oh, I don't want [53:34] to not be the CEO of Twitter cuz I won't [53:36] get invited to the' So, I went to the [53:38] NBA All-Star game because I knew Adam [53:39] Silver and he invited I felt like it was [53:41] a Dick Costolo invitation. I was like, I [53:44] want to I'll go to things I get invited [53:45] to cuz I'm Dick Costolo. I don't want to [53:47] go to things that that the CEO of [53:48] Twitter gets invited to because that'll [53:50] be like [53:51] >> Wasn't it part of your job? [53:53] No, I mean [53:54] I mean going to the White House, sure. [53:57] Like, hey the president wants to talk to [53:59] you about, you know, Edward Snowden and [54:01] abuse on the platform and what are we [54:03] going to do about terrorists using the [54:04] services? Yeah, you got to go to that as [54:06] part of the job. But going to the like [54:08] going to these things, you know, like [54:10] the Oscars or whatever. Like, I don't [54:12] know. I don't know any of these people. [54:13] They don't care about me. [54:14] >> would have done it. Yeah. I [laughter] [54:16] totally would have [54:16] >> Well, you would have been standing over [54:17] in the corner going like, no one's talk [54:19] Tom Cruise won't talk to me. Tom Cruise [54:21] isn't talking to me. [54:22] >> [laughter] [54:23] >> Uh [54:25] it's exceptionally rare these days that [54:27] somebody a founder-led company is [54:30] replaced or [54:31] >> Yeah. [54:33] Just that they don't hire the adult [54:35] external CEOs. Is that Should it be [54:37] happening a lot more? I think it goes in [54:39] waves. And it's like not doesn't happen [54:42] at all these days. [54:43] >> Yeah, I know. I think it goes in waves [54:45] and I think it will continue to be there [54:48] will something will happen and people [54:49] will start going, you know, we really [54:51] need for these companies that are [54:53] making, you know, $4 billion in revenue [54:56] after 3 years and the 26-year-old [54:58] founder is they'll just something will [55:00] happen and it will there will be a wave [55:01] of um you know, operating executives [55:04] coming back in and run some of the [55:05] companies. It happens a little bit like [55:07] Slutman and [ __ ] you know, stuff like [55:09] it. Pretty rare. [55:10] >> Yeah, yeah, rare. Even the public [55:12] companies like MongoDB but [55:14] I mean Workday recently but like maybe [55:16] one a year. Yeah, if it's it's true. I [55:19] do think it goes in waves. We'll see if [55:20] it we'll see if it happens. [55:21] >> a VP at blabbity blah company and you [55:23] want to be a Dick [55:25] Costolo. [55:26] What should you do? [55:28] You want to be that adult. [55:30] >> You you Oh, you want to be that adult. [55:33] >> to be you when I grow up. I'm a VP [55:35] >> Do Anthony Noto did, you know, when he [55:37] was CFO of Twitter he started it like [55:39] telling making it clear to [55:41] you know, executive recruiters like I [55:42] want a CEO role. Like start you know, [55:44] when you start seeing CEO roles out I'm [55:46] I might not get one right away but like [55:48] I want to I want to talk to you about [55:49] CEO roles when you get them. Get your [55:51] name out there and just I never [55:54] and Anthony would tell you this too [55:56] about his own career. I never took like [55:58] a risk that I didn't you know, that I [56:00] wasn't super psyched I took. Even though [56:02] I was like trying to get on SNL and [56:04] wasn't successful. It ended up [56:06] benefiting me down the road cuz it was [56:07] like all right, like like live TV on [56:10] CNBC are you okay doing that? I'm like [56:12] Look at the camera? Like I got booted [56:15] midnight in an Australia by a thousand [56:16] people. I can look in the camera and [56:18] talk about it. Should all CEOs Take a My [56:20] co-founder took a comedy class in [56:21] business school. Yeah, it really helped [56:23] him. [56:23] >> Yeah, being on stage and you know, being [56:25] uncomfortable being on stage make I mean [56:28] you got to you stand up at all hands [56:30] meeting where like I know exactly what [56:31] I'm going to say at the all hands [56:32] meeting. I'm just going to talk about [56:33] the priorities. Yeah. [56:35] >> That's easy after doing that. Okay. Let [56:37] me give you a fun one. Yeah. Hit me. Is [56:39] San Francisco passive aggressive? Is [56:41] everyone passive aggressive? Are CEOs [56:42] passive aggressive? I think there's a [56:44] lot of passive aggressiveness. I think I [56:45] know who I know the person you talked to [56:47] and I think and I know the other [56:48] companies he's been at. I think he's [56:50] generally right but there are some great [56:52] exceptions. [56:53] >> Okay. I think the Collison brothers at [56:56] Stripe are [56:58] are both [57:00] the rare combination [57:02] of [57:04] really really [57:06] widely read and and and and polymaths [57:09] and understand have a broad perspective [57:12] on humanity and biology and the human [57:14] condition and it and are and are [57:17] forthright communicators [57:20] and are [57:22] you know, neither passive aggressive nor [57:24] hyper aggressive. They're just they're [57:26] forthright and straightforward and blunt [57:28] and I think those guys are great. And I [57:30] and I feel the same way about Reed [57:32] Hastings, a totally different executive [57:34] at Netskope at Netflix. Reed is always [57:37] like [57:38] this is just how it is, you know, I [57:39] don't like I'm not I don't [57:42] I'm not trying to be mean about it. This [57:43] is just the way it is. And I think it I [57:45] think [57:46] >> with Reed in my head. [57:47] >> Yeah, yeah, yeah. I take that as a huge [57:50] compliment. I love that guy. Um I always [57:52] thought those three people, John and [57:55] Patrick at Stripe and Reed at Netflix, [57:58] are extremely self-aware. I remember I [58:01] was at table with Reed at a dinner once [58:03] and he went around the table and said [58:06] I want to hear everybody's opinion on [58:07] this. How much of your success do you [58:09] think is based on your DNA [58:12] and how much of it is based on like I [58:14] just learned how to kick ass and be [58:16] amazing? [58:17] And I kind of I was like I think I know [58:19] where he's going with this. And of [58:20] course when he got around to himself, he [58:21] was like I think it's basically like 80% [58:24] DNA. Interesting. [58:25] >> Yeah. How would you answer that? [58:26] >> Yeah, I'd I think similarly. And I think [58:29] that's a I think that's a result of him [58:31] being [58:31] >> skill? [58:32] >> Yeah, I think there's certainly a bunch [58:34] of luck. But I don't think he was saying [58:35] luck versus skill. I think he was sort [58:37] of saying like you're not like some, you [58:39] know, exceptional person who came out [58:41] and like I'm going to fight against the [58:43] out like it's basically encoded in you [58:45] to be this way and, you know, it's it's [58:48] ended up helping you. And sure there's [58:50] like 20% luck along the way that put you [58:52] in the position to get this role that [58:53] you might not have otherwise had. But [58:55] you were always going to do something [58:56] like this. Okay, you brought up some [58:58] good CEOs. The the CEO rule book feels [59:00] to me to be changing. [59:03] Your take on how Jensen Huang runs that [59:05] with like 60 direct reports, no [59:07] one-on-ones, publicly criticizing. He's [59:09] got a whole different playbook. The [59:11] thing that surprises me is I work with [59:12] all the CEOs, nobody's copying him, [59:15] which I think is interesting. Your take [59:16] on that. I go back to the Jeff Bezos [59:19] comment to Jack Dorsey. Jack, there are [59:21] lots of different ways to be successful. [59:23] I think that I think that the mistake [59:25] that CEOs make is they read the Walter [59:28] Isaacson book on Elon Musk or they read [59:31] the one on Steve Jobs and they think [59:33] this is the [59:35] archetype. [59:37] And [59:38] Walter's just left out a whole bunch of [59:40] stuff, you know, and that's not [59:42] >> have an archetype? No, I didn't have an [59:44] archetype. [59:44] >> was mine. He tried to copy everything he [59:46] did. I did it quite poorly. [59:48] >> [laughter] [59:49] >> I mean, it's goes again to the like, [59:51] well, Steve says you have to be an [59:52] editor and you said you know, and you [59:54] know, and Jeff saying like, oh yeah, [59:56] well, there's lots of different ways to [59:58] do it. I think that and the reason I say [60:00] that about the books is not just because [60:02] they're more mythology than they are [60:04] what really happened. Some of it's of [60:06] course this happened, but it leaves out [60:08] a bunch of stuff. [60:10] But more importantly, you just if you [60:12] try to be someone you're not, nobody is [60:14] fooled. Like you think people are [60:16] [ __ ] fooled? Nobody is fooled. [60:18] Everyone knows you're spinning them or [60:20] you're playing some role and you're not [60:22] yourself and nothing is more miserable [60:25] than trying to be someone you're not. [60:27] One of the things I used to I used to do [60:28] this in my management class actually [60:30] during one of the NBA finals in back in [60:32] like the 90s. This reporter is in the [60:34] Houston Rockets locker room and it was [60:36] at a time where Charles Barkley did this [60:38] ad for Nike where Charles dunks a [60:40] basketball and looks at the camera and [60:42] goes, I ain't no role model. Something [60:44] like that. [60:45] >> Yeah. [60:45] And um this reporter is in the Houston [60:48] Rockets locker room and asks Hakeem [60:49] Olajuwon [60:50] >> wrong. He was he wasn't [laughter] he [60:51] wasn't he wasn't wrong. He wasn't wrong. [60:54] Although he appears to be on Ozempic or [60:55] something cuz he's got to get have his [60:56] suits retailored. Um [60:58] this reporter is in the Houston Rockets [61:00] locker room and goes up to Hakeem [61:01] Olajuwon and says, you know, Charles [61:03] Barkley says that kids shouldn't, you [61:05] know, idolize look up to NBA players as [61:07] role models. What do you think about [61:08] that? [61:09] And Hakeem Olajuwon goes, I'm happy to [61:11] be a role model. I love being a role [61:13] model for kids. She's like, well, [61:15] what do you And and Hakeem says, well, [61:17] here's the deal. Charles is a different [61:19] person in public than he is in private. [61:22] And so he's got this anxiety and tension [61:25] in his mind about whether he's supposed [61:27] to be public Charles in this moment or [61:28] private Charles in this moment. I'm [61:30] always the same person. So if people [61:32] want to look up to me as a role model, [61:34] that's fine with me. I just think that [61:36] it's nothing but misery if you're trying [61:38] to, you know, look at some written down [61:41] playbook of how Steve Jobs did things [61:43] and I got to go do that when it's almost [61:46] It was impossible for Steve to do it [61:48] almost and you're certainly not going to [61:50] be able to do it and you'll be miserable [61:51] trying. Okay, last question I have is [61:53] like [61:54] the Twitter board, man, there was a lot [61:56] of freaking drama there. [61:58] >> Tons. [61:59] So much. [61:59] >> So much. You're building that your $100 [62:01] million company going to $400 million. [62:03] You're building a board. [62:05] Advice, like should you bring in [62:07] independents early? Should you wait? [62:08] >> I had two board members that were [62:09] immensely helpful to me. [62:11] Peter Chernin, who was brought in from [62:13] NewsCorp as an independent director cuz [62:15] he'd managed tens of thousands of people [62:17] and I needed someone with operating [62:18] experience, global operating experience. [62:21] Um and Peter was great at always like [62:24] always did the sort of the Jeff Bezos [62:27] let's talk about this from the user's [62:28] perspective instead of like how are we [62:30] going to do XYZ? Like he would always [62:31] back up and say, [62:33] you know, wait. If I'm the chief if I'm [62:35] the chief user officer at this company, [62:38] you know, and I'm thinking about this [62:39] from the [62:40] And so that was immensely helpful in [62:43] lots and lots of cases. And then [62:45] secondly, Peter Fenton from Benchmark [62:47] would always help me interview [62:49] candidates and either close them or help [62:52] me with hey, how should I think about [62:54] this candidate? And that having someone [62:56] on your board who's not in the [62:58] day-to-day at the company and who can [63:00] also give the candidate a sort of a [63:01] third-party perspective on why I think [63:04] this is going to be a monster and why I [63:06] think you should do this um was like [63:09] crazy helpful and helped convince [63:11] numerous execs like Adam Bain and [63:13] Anthony Noto at the company. [63:15] >> any Okay, you're at $100 million, you're [63:16] doing your series B, do you bring in an [63:18] independent or you like what happens now [63:19] is that wait right before the IPO? No, [63:21] yeah, I think it's helpful to have [63:23] operating people with operating [63:24] expertise in there much sooner than [63:26] that. [63:26] >> Okay, and just while we're on IPO, you [63:28] took Twitter public. [63:30] Um, what's your It's been incre- It's [63:31] very hard to take companies public. It's [63:33] become very unfashionable to take [63:35] companies public. [63:37] >> It's your turn. I would I would have [63:38] kept it private longer if I could go [63:40] back and do anything over again. [63:41] >> have too. It's just so much easier [63:43] running a private company and not having [63:45] to like [63:46] Okay, well, that's going to mean we're [63:48] going to miss the revenue number by a [63:49] percent. It's the right thing to do, but [63:51] you know, [63:53] the employees are going to be running [63:54] around like what's wrong when the stock [63:56] goes down 8% after hours, you know, and [63:58] as you know, you don't have a daily [64:01] ticker that the employees are freaking [64:02] out about when you're private company [64:04] and it's just a lot more you can be It's [64:06] so much easier [64:08] and stress-free to focus on the long [64:11] term as opposed to Okay, now we have to [64:13] make a long-term versus a short-term to [64:16] have a discussion about this because, [64:18] you know, everything's going to go [64:19] sideways if we do this in Q1 instead of [64:22] blah blah blah. You get it. Okay. It was [64:24] just It's just a lot easier to run a [64:25] private company. I wanted to talk to you [64:26] for a long time because you are kind of [64:28] a rare species these days, the adult CEO [64:31] brought into a hyper-growth company [64:33] and you delivered. Thanks for coming on [64:34] the pod. Thanks for having me. Yeah. [64:36] Okay, I thought Dick was great. It's a [64:38] master class in how do you scale and be [64:40] a terrific scale-up CEO. There's a few [64:43] things he talked about that I think are [64:45] worth reviewing, some of which I did, [64:47] some of which I wish I did when I was [64:50] CEO of HubSpot. [64:51] One thing he did that I totally did was [64:53] management by walking around. I'd wait [64:55] until like 7:00 at night and I'd walk [64:58] the halls of HubSpot through the dev [64:59] organization, support, sales, who was [65:01] around and really hear what what was [65:04] happening. [65:05] That was super useful for me. It was [65:06] great context and it was also good to [65:08] see who was burning the midnight oil. [65:10] Um, I liked his feedback around giving [65:13] tough people tough feedback. I was never [65:16] very good at this. [65:17] And his advice is good. It's, you know, [65:19] the night before you're going to give [65:20] feedback for that tough person, it's not [65:22] great feedback. You got to write it down [65:24] the night before, practice it in the [65:25] mirror. [65:26] I think that's good advice. And when you [65:28] deliver it, you deliver it and you sit [65:30] in silence. [65:33] No fluff. You sit in silence. [65:35] Really hard to do, um, but I think [65:38] that's great advice. He talked about [65:40] killing committees and using DRIs. Uh, [65:43] a lot of the CEOs I've interviewed are [65:44] talking about similar things. Um, [65:47] particularly important as you've got [65:48] silos in your organization and the [65:50] important stuff cuts across boundaries [65:52] of those silos. [65:54] Uh, he gave a great quote from another [65:56] pod guest, Ben Horowitz. He said, "Make [65:58] sure everybody understands what you [66:00] understand. Not just the what, the why [66:03] and what success means for them [66:05] personally. [music] [66:06] Priorities without context are useless." [66:09] I thought that was a gem from Ben and I [66:11] [music] thought it was really cool that [66:12] he repeated it. He accuses Silicon [66:14] Valley of being passive-aggressive. He [66:16] said he's not, he's aggressive. [music] [66:18] I think that's good advice. Like a lot [66:20] of CEOs, he says you should hire fast [66:23] and fire slow. He didn't fire slow. [66:25] Almost no other CEOs don't fire slow. [66:28] Uh, I did not fire slow and so he wishes [66:31] he did that. Some gems in there. Hope [66:33] you enjoyed it and looking forward to [66:35] having you in the next pod.