A16z Partners Urge AI Startups to Capitalize on Platform Shift

A16z合伙人敦促AI初创企业抓住平台转型机遇 企业对AI自动化的需求让创始人能够再次销售高价值软件平台，他们可在高信任度的灯塔客户与快速的中端市场抢占之间做出抉择。 该视频深入剖析了AI初创企业面临的战略抉择：是在规避风险的市场中赢得高信誉度标杆大客户的“灯塔”（Lighthouse）策略，还是积极抢占由明确ROI或市场刚需驱动的中端市场需求的“圈地”（Land Grab）策略。嘉宾借鉴Samsara与Meraki的销售实战经验，探讨了监管变革与AI紧迫感如何为再次销售大型企业软件创造罕见的窗口期。他们强调要设定严格的POC界限，尽早招聘销售运营人员，并通过将1%的时间用于规划、99%的时间用于前线执行，来避免陷入GTM战略瘫痪。 “灯塔与圈地”框架基于两个核心维度评估GTM策略：买方风险暴露（运营/监管风险）和背书传播效应（头部品牌客户具备多大的市场影响力）。 (0:00) 灯塔策略依靠社会认同和信任来赢得高风险或强监管市场；而圈地策略则瞄准低风险市场，在这些市场中，立竿见影且可量化的ROI能促成快速、基于数据核算的销售。 (0:00) 监管或行业转变（如Samsara面临的ELD强制令）会迫使整个市场立即划拨预算，从而创造巨大的圈地机会。 (6:57) 初创企业应首先瞄准中端市场客户以提升推进速度并打磨产品，然后再尝试向大型企业客户销售。 (6:57) 创始人应确立一个能跨越单位经济效益门槛的基准ACV，随后专注于可复制的执行，而非过早追逐复杂的企业级大单。 (16:28)

a16z partners Joe Schmidt and Andy McCall argue that AI startups must choose between two core go-to-market models—a 'Lighthouse' motion driven by marquee social proof or a 'Land Grab' motion driven by rapid quantitative ROI—while prioritizing field execution over theoretical strategy. The Evaluation Matrix: Risk Versus Social Proof (0:00) To structure an AI startup's go-to-market motion, Joe Schmidt maps prospective sales environments across two axes: Buyer Exposure (the operational, regulatory, or customer-facing failure risk) and Proof Traveling (how strongly the market values premier reference customers). A Lighthouse strategy suits the upper-right quadrant of high risk and high proof portability, typical in tightly knit, regulated industries where buyers require marquee validation to adopt unproven technology. Schmidt cites legal AI platform Harvey as a prime example, where winning top-tier law firms was necessary to dispel risk concerns across the wider sector. Conversely, a Land Grab strategy occupies the lower-left quadrant of lower exposure and low reliance on social proof, targeting buyers with existing budgets that prioritize mathematical ROI. Schmidt highlights Stuut, an AI accounts receivable platform, as an example winning mid-market deals through straightforward cost savings and working-capital efficiency. For Land Grab motions, founders should avoid fixating solely on Silicon Valley logos and aggressively sell into regional mid-market hubs like Ohio, Chicago, or St. Louis. Seizing External Urgency and Mid-Market Velocity (6:57) Andy McCall explains that structural market shifts create fleeting windows for aggressive Land Grab campaigns. Drawing from his time at Samsara, McCall notes that the federal Electronic Logging Device (ELD) mandate enacted between 2016 and 2019 forced commercial fleets to adopt automated digital tracking, creating an immediate, mandatory spending surge. Instead of chasing elite enterprise fleets that rejected an early-stage startup, Samsara captured mid-market buyers who decided faster and provided vital product feedback loops against legacy incumbents like AT&T and Verizon. Schmidt and McCall view current executive and board pressure around generative AI as a modern equivalent to the ELD mandate. McCall references Meraki's early enterprise networking pivot around 2009 as another Land Grab playbook, which bypassed entrenched rivals like Cisco by targeting mid-market IT teams with webinars offering free cloud-managed access points. For early-stage companies crossing $10M to $50M in ARR, McCall advises that once an Annual Contract Value (such as $15,000) clears cost-to-serve hurdles, founders should execute repeatedly at that tier rather than prematurely chasing complex enterprise accounts. Controlling AI Pilots and Platform Replacements (22:04) Because AI software can be highly configurable and subject to rapid model changes, McCall warns early-stage sales teams against letting customer evaluations devolve into indefinite 'science projects.' Early sales leaders must enforce rigid guardrails by setting fixed evaluation windows (typically 30 to 60 days) and agreeing upfront on objective, quantitative success metrics. Companies like Decagon and Further AI succeed by establishing clear performance benchmarks and structured onboarding before trials begin. Schmidt adds that these operational constraints coincide with a broader macro shift in enterprise procurement. Over the preceding 15 years, foundational cloud monoliths forced startups to enter enterprises via lightweight 'wedge' products. In contrast, generative AI and autonomous agents are driving enterprise buyers to reconsider core workflows from scratch. Buyers are now open to replacing foundational platforms like CRM and ITSM entirely, creating rare openings for startups to pitch large-scale, high-ACV enterprise systems. Strategic Transitions, GTM Infrastructure, and Quotas (26:15) Startups typically cycle through both strategies across their growth lifecycle. A company may execute a Land Grab in mid-market accounts and later verticalize into an enterprise Lighthouse model, or secure marquee Lighthouse accounts first before sweeping down-market. When Meraki scaled into public-sector school districts, it adapted from high-velocity sales to a Lighthouse motion governed by long procurement cycles, elected boards, and seasoned enterprise reps. McCall advises founders to spend 1% of their time on strategy and 99% on field execution, warning against selecting a Lighthouse model merely for the vanity of closing prestigious enterprise brands. To support execution, McCall recommends hiring a dedicated Sales Operations lead earlier than most founders expect, ensuring that territory alignments, account lists, and commission rules are established before administrative friction impedes growth. Furthermore, McCall urges early sales organizations to target 100% quota attainment to sustain sales team momentum, arguing that early market capture matters far more to long-term valuation than short-term cost-of-sales optimizations. Early sales success depends on aligning go-to-market motions with buyer risk tolerance rather than prestige. McCall and Schmidt emphasize that public markets reward market dominance and repeatable execution, leaving no premium for revenue won through unnecessarily difficult sales cycles.

a16z合伙人Joe Schmidt与Andy McCall指出，人工智能初创公司在制定市场进入（GTM）策略时，应基于买方风险敞口与行业背书流动性，在“灯塔策略”与“跑马圈地策略”之间作出明确抉择，切忌陷入过度分析而拖累执行。 双轴矩阵与两大销售路径的判定 (0:00) 针对硅谷众多软件公司扎堆争夺旧金山湾区少数科技标杆客户的同质化竞争，Joe Schmidt提出了一个基于两项核心指标的2x2评估矩阵：买方风险敞口（软件故障、合规处罚或对外暴露带来的业务风险）以及证明的流动性（行业对头部标杆客户背书的重视程度）。在强监管或顶级客户高度集中的行业中，买方承担极高业务风险，交易极度依赖信任与行业名牌的背书传导，这属于典型的“灯塔策略”（Lighthouse Strategy）；而在买方风险较低、缺乏全行业通用背书效应的场景中，客户已有现成预算编制，决策高度依赖可衡量的即时投资回报率（ROI），属于“跑马圈地策略”（Land Grab Strategy）。Schmidt建议，采取跑马圈地策略的公司不应局限于湾区，而应积极拓展俄亥俄州、芝加哥或圣路易斯等更广阔的区域市场。 跑马圈地的催化契机与算术逻辑 (6:57) Andy McCall结合其在Samsara与Meraki的早期销售经验指出，外部政策或技术变革往往能为跑马圈地创造宝贵的时间窗口。例如，美国在2016至2019年间分阶段实施的联邦电子记录仪（ELD）合规法令，迫使整个交通运输行业放弃纸质日志转用数字追踪系统，从而激发了全行业的采购紧迫感。Samsara当时绕开了拒绝接纳初创企业的头部大型车队，转而集中攻坚中端市场，凭借更快的成单速度与产品迭代完成了早期起量。McCall和Schmidt认为，当前企业董事会与高管层对AI的焦虑感，正在扮演与ELD法令类似的催化角色。在此类策略中，销售比拼的是清晰的数学账本而非长期的“传教式”教育。例如利用AI优化应收账款与催收业务的Stuut，以及取代传统客服系统的Pylon，都是在中端市场通过证明直接节省开支和提升资金周转率来推进销售。McCall建议，在年经常性收入（ARR）达到1000万至5000万美元的早期阶段，只要年度合同价值（ACV）能够覆盖服务成本并展现可复制性，团队就应全力铺开，而非过早追求复杂的头部企业大单。 灯塔客户突破与POC的严格管控 (13:11) 当软件采购需要对客户进行深度的概念普及与内部立项论证时，企业必须采取灯塔策略。法律AI领域的Harvey便是一个典型案例：在法律行业最初将AI自动化视为高风险任务的背景下，Harvey通过拿下少数顶级精英律所，迅速建立了可跨机构传播的行业信任，从而撬动了原本持审慎态度的市场。对于保险等极度厌恶风险的传统垂直领域，初创企业如Further AI和Decagon通过将安全与合规控件内置于平台，并配备前线部署工程师协同实施，以赢取数十亿美元级核心承保商的背书。由于AI产品配置复杂且模型演进迅速，McCall警告概念验证（POC）极易失控沦为无休止的“科学实验项目”。早期销售团队必须设定不可推迟的严格截止日期（如30天、45天或最长60天），并在上线前明确书面约定可量化的成功标准，确保一旦达标即执行购买；Schmidt同时提醒，必须将底层软件性能与客户自身的操作适配风险严格区分，防止因客户配置不当而承担不应有的产品指责。 策略动态演进与企业软件形态回归 (26:15) McCall指出，绝大多数成功的企业在全生命周期内都会综合运用这两种模式：初创公司既可以从中端市场的跑马圈地起步积累现金流，随后通过团队垂直化转向攻关行业前五大巨头（例如Meraki通过搞定全州最大的K-12学区树立标杆，进而带动周边学区）；亦可先攻克超级标杆客户，再向中下游市场横扫。不过，对于像Applied Intuition这类面向高度集中车企（OEM）市场的公司，由于目标客户基数极其有限且客单价极高，则会维持纯粹的灯塔模式。此外，Schmidt认为当前行业正在经历重大分水岭：在过去15年的传统云时代，底层平台（如CRM、HR）被巨头固化，初创企业只能通过产品驱动增长（PLG）或轻量级“契子”切入；而当前的生成式AI时代，企业高管的技术认知已大幅提升，普遍寻求利用自主智能体彻底重构业务流，这为初创企业重回数百万美元级、全面替换传统核心平台的大型企业软件销售打开了历史性窗口。 销售组织建设、配额机制与执行纪律 (38:49) 在组织落地层面，McCall提出创始人常犯的致命错误是在销售策略上陷入“分析瘫痪”，实际上应将1%的精力用于确定初始路线，99%的时间投入一线销售，因为资本市场并不会给耗时更久、更艰难的收入赋予额外估值溢价。在人员与组织搭建上，跑马圈地依赖进取心强、能深入现场（如卡车停放场或工业厂房）快速累积成单的年轻销售，而灯塔模式则需要深谙漫长政企采购与内部政治的资深销售。初创团队往往过晚引入销售运营（Sales Ops）职能，即使仅设立一人岗位，对于厘清业务区域、维护目标客户名录、裁决佣金纠纷及制定销售准则也至关重要。最后，针对早期团队的业绩考核，McCall主张应制定合理的指标以力求实现100%的配额达成率（Quota Attainment）；他认为早期销售主要依靠士气驱动，40%到50%的达标率暴露出配额制定过高或人员招聘不匹配的问题，在扩张初期抢占市场势头远比过度苛求销售成本更具战略意义。 无论是先跑马圈地后垂直化，还是先立灯塔再向下渗透，创始团队必须尊重客观的采购动力学，根据自身产品的风险敞口与行业背书效应选择阻力最小的初始路径。在企业级AI平台迎来整体替代潮的当下，快速获取市场真实验证并维持组织势头，远比追求纸面上的完美战略更具决定性。

## Background 该视频由著名风险投资机构 a16z 于 2026 年 8 月 20 日发布在官方 YouTube 频道上。节目由主持人 Elena Burger 对话 a16z 投资人 Joe Schmidt 与拥有 Meraki 及 Samsara 资深操盘经验的销售高管 Andy McCall。内容聚焦于企业级 AI 初创公司的销售路径抉择，深入探讨是该集中火力攻下少数几家具有极强背书效应的“灯塔客户”，还是凭借明确的 ROI 算账逻辑迅速在广泛的中端市场展开“跑马圈地”。 ## TL;DR 节目系统性地提出了基于“买家风险敞口”与“证据流动性”的 2x2 销售策略决策框架，明确划分了依靠权威背书的“灯塔模式”与依赖纯算术 ROI 替代既有工作流的“圈地模式”。结合 Samsara 借势 ELD 强制法案在中端市场快速起量以及 Meraki 免费赠送 AP 攻坚的历史经验，嘉宾们深入拆解了 AI 时代如何避免概念验证（POC）沦为无休止科研项目的实战纪律。对谈最后强调，几乎所有伟大企业最终都会走向双轨合流，早期创始人切忌因虚荣心迷恋顶级大客户或陷入分析瘫痪，而应坚持“1% 战略、99% 执行”，在阻力最小的战场迅速建立获胜势能。 ## Atmosphere & style 整期对谈节奏紧凑、直击本质，洋溢着一种来自真实商业战场前线的粗粝感与务实作风。Andy McCall 带着历经多次技术周期沉淀后的老练与幽默，用打靶场、海钓船甚至垃圾处理场签单的生动经历，彻底打破了硅谷精英对高大上销售的幻想；Joe Schmidt 的系统化模型则与之形成了极佳的互补。交流中充满对创业者虚荣心态的辛辣剖析与对一线执行力的由衷推崇，既有战略层面的拨云见日，又带有一种催促创始人立刻“冲出办公室去给客户打电话”的强烈紧迫感与行动召唤。 ## Bulletin - [00:00](https://www.youtube.com/watch?v=aakZLqxRQfo&t=0s) 企业级 AI 软件销售正在进入重新销售底层大平台（Big Software）的黄金窗口，企业不再满足于微小的功能点修补，而是系统性重构工作流。 - [02:37](https://www.youtube.com/watch?v=aakZLqxRQfo&t=157s) a16z 提出 2x2 评估矩阵：以买家风险敞口和证据流动性为维度，决定采用重背书的“灯塔模式（Lighthouse）”还是重 ROI 算账的“抢地盘模式（Landgrab）”。 - [21:39](https://www.youtube.com/watch?v=aakZLqxRQfo&t=1299s) 早期 AI 公司必须对概念验证（POC）设立硬性截止日期与预先量化的成功验收标准，避免试用沦为无休止的需求蔓延与“科研项目”。 - [37:31](https://www.youtube.com/watch?v=aakZLqxRQfo&t=2251s) 创业初期销售团队应追求接近 100% 的配额达标率以建立获胜势能，创始人应将 1% 精力用于战略制定，99% 精力扑向一线执行与客户沟通。 ## Key takeaways - [31:42](https://www.youtube.com/watch?v=aakZLqxRQfo&t=1902s) 当前软件采购正经历底层重塑，生成式 AI 和 Agent 的出现让企业预算重新向核心平台型大软件汇聚，打破了过去十年依赖微小功能点切入的 PLG 局限。 - [02:37](https://www.youtube.com/watch?v=aakZLqxRQfo&t=157s) 决定销售路线的核心取决于买家风险敞口（Buyer's Exposure）与行业证据流动性（Proof Travels）两个关键维度。 - [05:06](https://www.youtube.com/watch?v=aakZLqxRQfo&t=306s) “灯塔策略”适用于高合规风险、需要品类开拓的新兴市场，依靠攻克顶尖标杆企业产生的强示范效应来打破全行业的采购顾虑（如 Harvey 和 Decagon）。 - [05:06](https://www.youtube.com/watch?v=aakZLqxRQfo&t=306s) “抢地盘策略”适用于已有成熟预算的工作流替换赛道，销售无需漫长背书，仅凭清晰的投资回报率（ROI）算术题即可快速在中端市场规模化铺开（如 Samsara、Meraki 与 Further AI）。 - [07:50](https://www.youtube.com/watch?v=aakZLqxRQfo&t=470s) Samsara 在早期抓住联邦 ELD 法案强制淘汰纸质日志的政策红利，避开大企业漫长决策链，坚决在中端市场走量完成快速迭代验证。 - [17:24](https://www.youtube.com/watch?v=aakZLqxRQfo&t=1044s) 客单价（ACV）纪律要求在前期严格核算单位经济模型以迈过盈利门槛，一旦模型成立就必须专注复制与放量，而非为了凑数承接亏损订单。 - [18:15](https://www.youtube.com/watch?v=aakZLqxRQfo&t=1095s) Meraki 早期通过向企业免费赠送云管理 AP 的激进试用策略打透中端市场，证明了优秀体验带来的极高转化率能够覆盖硬件前期成本。 - [21:39](https://www.youtube.com/watch?v=aakZLqxRQfo&t=1299s) AI 销售中必须严格管控 POC 流程，通过设置 30 至 60 天的不可动摇截止期以及白纸黑字的量化验收标准，防止试用期无限蔓延。 - [23:36](https://www.youtube.com/watch?v=aakZLqxRQfo&t=1416s) 初创企业切忌替客户内部的执行或管理不力兜底，必须在交付首日与客户明确界定软件交付指标与客户自身使用责任的边界。 - [27:15](https://www.youtube.com/watch?v=aakZLqxRQfo&t=1635s) 几乎所有成功的成熟企业最终都会走向双轨合流：圈地起步的公司在向上渗透时会按垂直领域打灯塔，而灯塔起步的公司在确立壁垒后会横向收割长尾品类。 - [28:20](https://www.youtube.com/watch?v=aakZLqxRQfo&t=1700s) 灯塔策略需要深谙复杂招投标与长周期的资深企业级销售，而圈地策略更依赖高执行力、进取心强的早期年轻销售团队。 - [37:31](https://www.youtube.com/watch?v=aakZLqxRQfo&t=2251s) 早期收入不存在所谓的“艰难加分项”，盲目执念于拿下摩根大通等知名虚荣客户往往会拖垮团队，创始人应花 1% 时间定战略、99% 时间抓执行。 - [40:39](https://www.youtube.com/watch?v=aakZLqxRQfo&t=2439s) 初创团队应当尽早引入哪怕只有一人的销售运营（Sales Ops），提前梳理区域划分、客户名单与佣金规则，避免规模化扩张时遭遇机制减速带。 - [41:45](https://www.youtube.com/watch?v=aakZLqxRQfo&t=2505s) 销售团队在业务初期高度依赖赢球势能，早期应当以 100% 配额达标率为目标设计指标，过低的达标率通常反映了配额脱离实际或招聘画像错配。 ## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=aakZLqxRQfo&t=0s) 重新销售“大软件”的时代窗口与两种销售打法 播客在开篇引入了一个极具冲击力的行业论断：当前正处于一个“重新去销售大软件（big software）”的关键时间窗口，整个商业运作模式都在发生根本性的重构。针对这一转变，AI 创业公司究竟该如何选择进入市场的路线？是该聚焦于拿下少数几家举足轻重的标杆客户，还是该在大众市场迅速铺开？ 在讨论这一核心命题时，嘉宾们提到了创业者常陷入的心理陷阱。以 Meraki 在 2009 年的早期经历为例，当时企业网络市场几乎完全被 Cisco 和 HP 瓜分，外界认为 Meraki 绝无可能打进全球最大的龙头企业，但团队通过强调“配置更快、部署更快、使用更简单”的价值主张，成功走出了一条典型的抢地盘（Landgrab）路线。反观当下，许多 AI 创始人宁愿削尖脑袋去争取摩根大通（JP Morgan Chase）这样听起来无比性感的标志性大客户，也不愿意拉下身段去给名不见经传的普通企业打电话、坐飞机去拜访潜在客户，而这种对“虚荣客户”的盲目执念往往是早期创业者最容易犯下的致命错误。 ### [00:58](https://www.youtube.com/watch?v=aakZLqxRQfo&t=58s) 101 高速公路上的广告牌大战与销售框架的诞生 主持人 Elena Burger 正式开场，指出如何在早期做销售是 AI 创始人面临的“单笔成本最高”的战略抉择。a16z 的投资人 Joe Schmidt 刚刚发表了一篇题为《灯塔还是抢地盘》（*Lighthouse or Landgrab*）的文章，梳理了企业级 AI 初创公司中最主流的两种销售策略。Joe 分享了激发他撰写这篇文章的灵感来源： 几个月前，Joe 开车行驶在旧金山湾区的 101 高速公路上，眼前的一幕让他深感荒谬：高速公路两旁立满了巨幅广告牌，两家生产一模一样软件的竞对公司隔路对峙。不仅如此，整条街的公交车被广告贴纸完全包裹，甚至天上还有小型飞机拉着横幅在空中巡游宣传。似乎所有 AI 初创公司都心照不宣地认定，世界上唯一值得销售的对象就是旧金山本地的企业，以及每天行驶在 101 高速上的那一小撮科技从业者。 这种同质化严重的营销动作催生了他的思考：创始人真的必须把所有精力都耗在旧金山、纽约这种一线大都市的名牌客户身上吗？显然不是。创始人完全可以去俄亥俄州、芝加哥、圣路易斯寻找那些真正对产品有迫切需求的传统客户。为了帮助创业者不再盲目扎堆于抢夺“知名 Logo”，Joe 提炼出了一套系统化的决策模型。 ### [02:37](https://www.youtube.com/watch?v=aakZLqxRQfo&t=157s) 评估框架：买家风险敞口与证据流动性的 2x2 矩阵 为了清晰量化销售场景，Joe 提出了一个咨询风格的 2x2 评估矩阵，主要围绕两个核心维度展开： - **Y 轴：买家敞口（Buyer's Exposure）**：指采购该方案时买家所承担的整体风险与暴露程度。这不仅包括“万一选错软件会不会背锅导致离职”的个人职业风险，还包含该软件在企业业务链路中的暴露位置——它是完全处于内部后线，还是会直接暴露在企业的终端客户面前；一旦产品出现偏差，是否会引发监管处罚甚至违法违规。 - **X 轴：证据流动性（Proof Travels）**：指在一个特定市场中，“同行背书”或“社会认同”能否高效传导。即其他行业买家在做采购决策时，是否极度看重行业内领头羊的示范效应。 根据这两个维度的组合，市场被划分为截然不同的两种类型： ```mermaid flowchart TD accTitle: 销售策略 2x2 评估矩阵 accDescr: 展示以买家敞口和证据流动性为坐标轴的灯塔与抢地盘销售策略分类 subgraph Matrix ["市场分类矩阵"] Lighthouse["【灯塔模式 Lighthouse】<br/>· 象限：右上角<br/>· 驱动因素：社会背书 (Proof)<br/>· 特征：高买家敞口、高监管风险、证据高效流动<br/>· 适用：高度监管行业、品类开拓型产品"] Landgrab["【抢地盘模式 Landgrab】<br/>· 象限：左下角<br/>· 驱动因素：投资回报算术题 (Math)<br/>· 特征：低买家敞口、已有现成预算、证据流动性低<br/>· 适用：工作流替代、效率改进型产品"] end ``` 在右上角的高敞口、强证据流动市场中，属于标准的“灯塔模式（Lighthouse）”。这类市场多为高度受监管的行业，全行业的买家高度关注行业龙头的选择。一旦买错系统产生失误，后果不堪设想；而一旦某家顶尖龙头公司公开采用并给予背书，这种示范效应将瞬间传导至整个行业。 而在左下角的低敞口、弱证据流动市场中，则属于典型的“抢地盘模式（Landgrab）”。在这里，企业通常拥有成熟的既有预算，习惯于为某类既定服务付费。销售的核心逻辑不再是“谁在用它”，而是纯粹的“数学算术题（Math）”——创始人只需要向采购方清晰论证：AI 驱动的方案在效率和成本上直接优于他们现有的纯人工或旧软件方案，就能快速促成交易。 ### [05:06](https://www.youtube.com/watch?v=aakZLqxRQfo&t=306s) 现实中的战略映射：品类开创 vs 工作流替换 曾在 Meraki 和 Samsara 操盘顶尖销售团队的 Andy McCall 针对这一框架进行了行业映射。他指出，选择灯塔策略的公司往往在进行“品类开创（Category Creation）”。由于这种产品所定义的品类在市场上尚不存在，客户没有参考依据，创始人必须倾尽全力拿下行业内最顶级的名字来证明该路线的可行性与合规性。 相反，当前市场上大量的 AI 初创公司在本质上属于工作流的改造者或替换者（Workflow Replacement）。企业早已有专门的人力和软件预算用于解决该项业务，初创团队只需直接切入既有预算池，告知对方“我们借助 AI 打造了一种更先进的达成路径”。这类企业无需等待漫长的大客户背书周期，完全可以采取自下而上、全面铺开的抢地盘策略。 ### [06:30](https://www.youtube.com/watch?v=aakZLqxRQfo&t=390s) 时代拐点与 Samsara 抓住 ELD 法案的经典战役 谈到抢地盘策略的极致应用，Joe 特别引述了 Andy McCall 在 Samsara 的亲身操盘经历。当时美国联邦出台了电子记录装置强制令（Electronic Logging Device, 简称 ELD Mandate），要求所有商用卡车必须在规定期限内安装电子记录仪替代纸质日志。这一政策红利瞬间强制催生了一个全国性的大爆发市场，谁能以最快的速度覆盖最多的车队，谁就能赢得整场战役。 Andy 回顾这段经历时感慨道，当一个人在科技行业摸爬滚打足够久（他笑称自己的头发已经“从金色变成了花白”），就会亲历多次巨大的技术浪潮转换——从最早的互联网兴起，到后来的移动化（Mobile）、云计算（Cloud），再到如今的 AI。每一次底层技术的大换代都会引发深远的范式转移，逼迫企业寻找全新的 Go-To-Market（GTM）打法。然而，底层销售逻辑并未发生颠覆性的全盘翻转，而是获得了更强大的辅助工具。Andy 坦言，任何一家最终取得极高成就的伟大企业，其背后无一例外都伴随着关键的时间窗口（Timing）与不可或缺的运气成分。 ### [07:50](https://www.youtube.com/watch?v=aakZLqxRQfo&t=470s) Samsara 的早期历程与 ELD 法规红利带来的“圈地”机遇 Andy McCall 回顾了 Samsara 创立初期的探索过程。Samsara 于 2015 年成立，Andy 于 2017 年加入。起初团队的核心愿景是围绕物联网与联网传感器：他们相信未来所有价值链都会实现传感器化，核心命题是如何把传感器布置出去、采集数据，并以可理解、可操作的方式回馈给企业管理者。 真正让 Samsara 迅速起飞的第一个突破口是车联网远程信息处理（Telematics）设备。在 2016 年之前，美国长途货运行业普遍采用极其原始的人工纸质日志。卡车司机每开四小时车、停下来休息二十分钟，或是开两小时车停下吃午饭，都必须亲手在日志本上手写记录；一旦在公路上被巡警拦截，警察会现场抽查日志本，确认司机没有超时疲劳驾驶。这一切在 2016 年前后迎来了根本性转折——美国政府正式出台 ELD（Electronic Logging Devices，电子记录仪）法规，强制要求通过技术手段自动追踪车辆移动与司机的休息状态，彻底消除人工记录环节的篡改和漏洞。 ```mermaid flowchart TD accTitle: ELD Mandate Market Shift accDescr: 政策强制推动货运行业预算重构并为新进入者打开市场切口的流程图 subgraph Regulation ["政策驱动阶段 (2016-2019)"] A["美国出台 ELD 法规"] --> B["淘汰纸质日志，强制采用电子记录仪"] end subgraph MarketDynamics ["市场预算与竞争格局"] B --> C["全行业被迫设立专项硬件及软件采购预算"] C --> D["传统巨头占据先机 (AT&T, Verizon 等)"] C --> E["新型初创企业迎来破局窗口 (如 Samsara)"] end subgraph GrowthPlaybook ["圈地打法落地"] E --> F["切入中端市场 (Mid-Market) 快速走量"] end ``` 这项法规在 2016 至 2019 年间分阶段强制推行，为所有电子记录仪厂商带来了巨大的政策顺风。市场上原本已有年营收数亿甚至五亿美元的传统巨头（如 AT&T 和 Verizon 旗下方案），但政策红利“水涨船高”，让整个货运行业在一夜之间必须拿出真金白银的预算去采购合规设备。其中相当一部分企业开始主动调研市场上的新兴技术方案，这为刚刚入局的 Samsara 提供了极为关键的突围跳板。 - Samsara 创立于 2015 年，Andy McCall 于 2017 年加入。 - ELD 法规落地跨度为 2016 至 2019 年，属于典型的由合规驱动的行业刚需爆发。 - 既有市场中存在 AT&T、Verizon 等年收入数亿美元的成熟玩家，但强制合规迫使海量中小企业寻找全新替代方案。 --- ### [09:58](https://www.youtube.com/watch?v=aakZLqxRQfo&t=598s) 放弃大企业冷板凳：用中端市场换取快速迭代 当 Elena Burger 问及既然长途货运是严监管行业、客户理应极其在乎背书，Samsara 是否必须先拿下高速公路上最知名的大型车队来建立社会认同（Social Proof）时，Andy 坦率地打破了这种“完美战略”的执念。他坦言，当时并没有经过宏大的理论论证去纠结“选灯塔还是选圈地”，团队纯粹是在倾听客户反馈并接受现实的敲打。 作为一个成立仅 18 个月、毫无知名度的新公司，去给全美最大的物流巨头打冷电话，收到的答复往往是冷冰冰的一句“我们绝不可能买你们的产品”，紧接着便是挂断电话的嘟嘟声。当时的产品功能极度初级，核心诉求就是先验证“到底有没有人愿意掏钱”。于是团队果断转向中端市场（Mid-Market）。 选择中端市场带来了两大核心战术优势：其一，中端企业不需要深厚的社会认同与品牌背书，他们更关注产品能否满足最基本的合规与远程信息处理需求；其二，中端客户的销售周期极短，能够快速签约部署，为产品研发提供高速反馈闭环。相比之下，越大的企业部署周期越长、反馈周期越滞后。通过在中端市场迅速拿下大量交易，Samsara 得以在极短时间内完成了产品的快速试错与功能迭代。 --- ### [11:34](https://www.youtube.com/watch?v=aakZLqxRQfo&t=694s) 识别当下 AI 浪潮的销售信号：预算替换与教育成本的权衡 Joe Schmidt 指出，当前的 AI 创业环境虽然没有美国政府直接下达的“强制令”，但全美几乎所有大型企业的 CEO 与董事会都在向管理层施压，要求必须在规定期限内完成 AI 采购与落地，这在大型组织内部激发出前所未有的“狂暴动能”。然而，判断自己究竟处于“圈地（Landgrab）”还是“灯塔（Lighthouse）”市场的核心分水岭，在于客户当下的实际反应：客户是否愿意立刻接听你的电话？是否愿意启动 POC（概念验证）并真正把产品用起来？ 如果客户愿意见面并快速采购，创业者就应当全力进入中端市场快速“圈地”，而不是死磕顶级金融机构去碰壁；但如果连电话都打不进去，那就必须采取深度驻场、前置部署的“灯塔战略”。Andy 进一步总结了判断标准： | 评估维度 | 圈地模式（Landgrab） | 灯塔模式（Lighthouse） | |:--- |:--- |:--- | | **预算来源** | 存在明确的现有预算，属于既有供应商替换 | 没有现成预算，需要跨部门立项甚至创造新类目 | | **市场认知** | 痛点明确，客户清楚知道自己要买什么 | 属于全新概念，需要耗费大量精力进行市场教育 | | **销售逻辑** | 纯靠投入产出比（ROI）和数据算账驱动 | 传道士式销售（Missionary Work），引导认知旅程 | | **切入策略** | 迅速扑向中端市场，依靠高频成交流水化作业 | 绑定行业标杆深度共建，用顶级背书撬动整个生态 | Andy 强调，如果产品是对既有软件或流程的替代，创始人就应该果断执行圈地策略，告诉客户“把你付给厂商 A 的钱转付给我们”；但如果是在创造一个前所未有的新品类，客户在掏钱前必须经历漫长的认知教育以完成内部合理化论证，这就必须走灯塔路线。 --- ### [13:26](https://www.youtube.com/watch?v=aakZLqxRQfo&t=806s) 圈地标杆案例：Further AI 与应收账款自动化的大幅提效 话题随后深入到具体企业的实战案例。在探讨 Joe 文章中提到的企业时，Joe 详细拆解了专注于企业应收账款（Accounts Receivable, AR）自动化赛道的 Further AI（由 Tareq 与 Ben 创立）。 应收账款催收在企业级市场向来是一项极其繁琐且不够光鲜的工作。传统模式下，企业需要雇佣庞大的催收团队，配合笨重的传统财务软件与客户往来沟通，不仅流程高度依赖人工，效率也十分低下。Further AI 的切入点在于利用 AI 擅长语言交互与内部数据处理的能力，实现催收流程的端到端自动化，随后进一步覆盖“从订单到现金”（Order-to-Cash）的整个应收账款产品矩阵。 Further AI 展现出了教科书般的圈地战术：他们带着无可辩驳的财务测算模型直接走向中端市场客户，用极其明确的 ROI 证明自己的效率远超现有人工催收团队——不仅能大幅改善企业的营运资金周转、节省人力成本，还能从坏账中追回更多现金。销售对话变得极其直接：“我们有明确的数学算账逻辑证明能帮你省钱多赚，这个方案你买还是不买？”凭借极强的地推执行力，Further AI 在中端市场迅速席卷客户，成为了当下 AI 圈地打法的典型代表。随后，话题转向了另一端作为灯塔战略典型代表的法律科技公司 Harvey。 ### [15:45](https://www.youtube.com/watch?v=aakZLqxRQfo&t=945s) Harvey 的灯塔示范与 Pylon 的圈地突围 对于高度敏感或规避风险的行业而言，“灯塔客户”（Lighthouse）所带来的背书效应往往具有决定性意义。在法律科技领域，初创公司 Harvey 展现了一场教科书级别的灯塔战略落地：法律行业原本对让 AI 介入日常工作、接管初级律师（Junior Lawyers）的日常事务抱有极大的不确定性和风险顾虑，但 Harvey 顶着这种高风险，在早期精准签下了几家声誉卓著的顶级标杆律所。当这几家行业灯塔被攻克后，其信用背书迅速向全行业扩散，让原本承担巨大合规与业务暴露风险的买方彻底放下了防备，意识到“购买这套方案实际上是非常安全的”。 与 Harvey 依靠顶尖标杆建立信任的路径相反，被投公司 Pylon 则代表了典型的“圈地模式”（Land Grab）。作为一家 AI 原生的客户支持（Customer Support）解决方案商，Pylon 的打法简单而凶狠：直接走向市场告诉客户“我们有一种更优越的解决方式”，通过高效的 GTM 销售团队以替代现有方案为目标全力执行。在客单价（ACV）上，Pylon 最初从相当适中的单价起步，随着产品迭代和攻城略地，一步步向上攀爬 ACV 阶梯，展现了圈地模式在广阔市场中快节奏铺开的强悍爆发力。 - **Harvey 的攻坚核心**：在法律等高风险垂直领域，灯塔客户的示范效应能打破认知阻碍，把“使用激进新技术”扭转为“经权威验证的安全选择”。 - **Pylon 的执行策略**：瞄准客户支持赛道进行存量替换，依靠执行力极强的销售团队高频触达，逐步抬高客单价。 --- ### [17:24](https://www.youtube.com/watch?v=aakZLqxRQfo&t=1044s) ACV 门槛纪律：算清商业模型，然后只管全速开跑 在创业公司从 1000 万到 5000 万美元 ARR 的成长过程中，究竟该如何平衡客单价（ACV）与业务拓展节奏？Andy McCall 分享了一条看似矛盾却极其精辟的心法：“你必须在前期对它深思熟虑，而一旦确认后，就彻底别再去想它。”所谓深思熟虑，是指创始人必须严格审视单位经济模型（Unit Economics），确保所追求的 ACV 能够迈过盈利与健康的基准门槛，绝不能为了单纯跑量去接那些单位经济为负的“亏本买卖”。 然而，一旦客单价越过了及格线，团队就不该再在单笔数字上患得患失，而是要立刻聚焦于销售动作的标准化与规模化。如果一家公司能够建立一套依靠 1.5 万美元 ACV 就能健康运转的 GTM 引擎，那就绝不要为了凑数去碰 8000 美元的订单，而是全力以赴拿到尽可能多的 1.5 万美元单子，向这个已经验证的飞轮疯狂“添柴浇油”。随着你在中低客单价市场积累了足够多的胜场，品牌和口碑自然会向上渗透，吸引规模更大的企业客户，顺理成章地沿着阶梯一步步攀升 ACV。 ```mermaid flowchart TD accTitle: ACV Discipline and Expansion Engine accDescr: A flowchart depicting the unit economics validation leading to repeatable execution and gradual enterprise ascension. subgraph Evaluation ["第一阶段：单价与模型校验"] A["评估单位经济模型"] --> B{"是否迈过利润门槛？"} B -- "否 (如 $8K 负收益单)" --> C["坚决放弃 / 拒绝接单"] B -- "是 (如 $15K 健康单)" --> D["锁定目标 ACV 基准"] end subgraph Scale ["第二阶段：全速圈地放量"] D --> E["打造标准化销售引擎"] E --> F["高频拓客，快速复制标准单"] end subgraph Ascend ["第三阶段：向上渗透"] F --> G["累积大量成功案例与口碑"] G --> H["大客户主动涌入，抬升 ACV 梯级"] end ``` --- ### [18:15](https://www.youtube.com/watch?v=aakZLqxRQfo&t=1095s) 从 MIT 屋顶到云端网络：Meraki 的免费 AP 圈地战术 回顾企业级网络巨头 Meraki 的崛起，其早期同样经历过极为经典的圈地战役。Meraki 的技术雏形源于 2006 年两位 MIT 博士生的研究项目“Roofnet”，最初旨在麻省剑桥市的屋顶上部署大规模无线网状网（Mesh Wi-Fi），为市政和公共公园提供无线覆盖。虽然技术惊艳，但团队在头两年就撞上了商业壁垒——市政 Wi-Fi 根本没有成熟的变现模式，公司不得不掉头转向企业市场。 到了 2009 至 2010 年间，Meraki 试图进军企业网络时，外界几乎一致认为他们“疯了”，因为当时 Cisco 和 HP 早已在十年前彻底瓜分了企业网络天下。然而，恰逢云计算浪潮初起，Meraki 团队研发出了通过云端集中配置和管理网络硬件的革命性方案。在巨头林立的格局下，Meraki 根本不可能拿下大型灯塔企业，因为那些财富 500 强的 IT 部门早已被精通 Cisco 命令行代码的工程师锁死。于是 Meraki 坚决推行中端市场（Mid-market）圈地战略：对于缺乏专业命令行运维人员的中型企业，Meraki “即插即用、极速部署、极简运维”的特性直击痛点。为了打消客户疑虑，Meraki 推出了一项极其大胆的获客手段： > “我们当时举办在线研讨会（Webinar），直接对客户承诺：‘只要你来参加 Webinar，我们就免费送你一个企业级 AP（无线接入点）。’你拿回去插在插座上试一下，只要他们一试，脑子里的灯泡瞬间就亮了，他们会惊叹：‘天啊，这比我原来用的东西简单太多了，我为什么还要买以前那套？’” 这种激进的试用与评估（Trial & Eval）策略不仅击穿了中端市场的防御，甚至在硬件包含生产成本的商业考量下，依然通过极高的试用转化率证明了其经济性，为 Meraki 后续在 2012 年被 Cisco 高价收购奠定了核心基本盘。 --- ### [21:39](https://www.youtube.com/watch?v=aakZLqxRQfo&t=1299s) AI 时代的 POC 陷阱：别让概念验证沦为无休止的“科研项目” 当话题回到今天的 AI 浪潮，Joe Schmidt 提出了一个现实难题：Meraki 当年送出硬件即可让用户立刻体验优劣，但现在的 AI 软件往往高度复杂、具备极强的可配置性，“就像直接把一辆法拉利交到客户手里，他们甚至连怎么点火都不知道”。面对这种技术落地断层，初创公司在推进概念验证（POC）或试用期时往往走入误区。 Andy McCall 警告称，当前 AI 初创公司在销售端面临的最大威胁，是把 POC 做成了无限延期的“科研项目”（Science Projects）。由于基础模型和 AI 工具以天为单位飞速迭代，销售人员一旦缺乏定力，面对客户不断提出的“你们能做这个吗？能再展示那个吗？”，往往下意识地回答“能”。随之而来的便是漫无边际的功能测试与需求蔓延，导致试用期永远无法收口，商务合同迟迟无法落笔。 要在当下的 AI 销售中建立严密的 POC 纪律，必须牢牢把握两大核心原则： - **设定不可动摇的硬性截止日期**：根据产品配置难度合理划定试用周期（通常为 30 天、45 天或 60 天），如果前置安装需要 2 周，就绝不能设置 2 周的过短期限，但一旦时间敲定，必须到期即止。 - **前期双方共同确立验收标准**：在交付产品的第一天，买卖双方必须明确界定“满足何种指标即算成功”。通过合同化约定，只要在约定期限内达成了既定目标，客户就必须推进正式采购，彻底斩断无休止的需求拉扯。 ### [23:36](https://www.youtube.com/watch?v=aakZLqxRQfo&t=1416s) POC 的范围边界与落地风险 在企业级销售与落地过程中，界定概念验证（Proof of Concept，POC）的范围极为微妙。Andy McCall 与 Joe Schmidt 指出，在很多受高度监管或风险敏感的行业中，客户甚至根本不会允许初创公司做 POC；但只要能够推进 POC，就必须在前端白纸黑字明确两件事：固定的截止日期（end date）以及清晰量化的成功指标（success criteria）。 Joe Schmidt 进一步剖析了当前 AI 落地中最棘手的难题：当你在自动化某些前所未有的业务流程时，往往伴随着极高成本的复杂系统配置。产品本身可能完全没有问题，但一旦客户部署方式不当，或者成效需要耗费 30 到 45 天以上才能显现，责任边界就会变得模糊不清。以销售赋能工具为例，软件功能运转正常是一回事，客户的销售人员是否选对了目标客群、是否以正确的方式使用则是另一回事。如果初创公司盲目为"产品好不好用"以及"客户员工是否用得正确"同时背负兜底风险，项目极易陷入泥潭。因此，早期创始人和营收负责人必须在销售早期就向客户做好预期教育，明确产品所承诺交付的核心指标，切忌让软件厂商为客户内部的执行与管理不力埋单。 ### [25:09](https://www.youtube.com/watch?v=aakZLqxRQfo&t=1509s) 行业布道与灯塔标杆：Decagon 与 Further AI 当被问及哪些公司在向客户传教、引导入职部署方面做得最为出色时，Joe Schmidt 举了两个具有代表性的案例： - **Decagon（AI 智能客服）**：客服市场暴露风险极高，任何故障都会直接对准终端消费者，绝非轻而易举的事。Decagon 并不回避这种复杂性，而是强力向市场传达"我们能将客户支持做到极致"的理念。他们会在销售与部署期设立极其严格的考核基准，并在规定时间内精准兑现指标，从而在看似拥挤的赛道中建立起强大的威信。 - **Further AI（保险垂直领域 AI）**：Joe 担任董事的 Further AI 将大模型引入了历来对新技术最为保守的保险行业。他们打动全球顶级大型保险机构的关键，在于打造了一套治理优先（governance-first）、高安全性的 AI 架构，并派遣前置部署团队（Forward-Deployed Teams）深入一线协同客户落地。 这正是典型的"灯塔策略"（Lighthouse Strategy）：通过啃下行业中最顶尖的巨头，利用其强大的示范效应与背书，进而通吃整个保险长尾市场。Joe 纠正了一个常见的认知误区——很多人误以为只有行业老兵才能做灯塔策略，但现实是：你完全不需要出身于该行业，核心在于走出去直接与客户建立信任关系，带着你所掌握的"认知秘密"（earned secret），向他们证明 AI 与新技术能给他们的业务带来质的跃升。 ### [27:15](https://www.youtube.com/watch?v=aakZLqxRQfo&t=1635s) 终局融合：大公司最终都会双轨并行 针对创始人经常纠结的"灯塔"与"圈地"（Landgrab）二选一问题，Andy McCall 指出，几乎所有从微小成长为庞然大物的成功企业，最终都会同时运用这两种策略： - 某些公司以广撒网的圈地模式起步，随着体量变大、进入成熟期后，再转向灯塔策略攻克特定行业标杆； - 另一些公司则依靠灯塔客户建立不可撼动的口碑，再将产品横向泛化，进行大规模的市场圈地。 对于早期创始人，Andy 给出最中肯的建议："在当前阶段，做对你业务最合理、阻力最小的事情。"走出去与客户对话，找到最快、最容易开单的切入口快速突进，这绝不代表永久放弃另一种模式，只是时机未到。 ```mermaid flowchart TD accTitle: 销售策略演进流程 accDescr: 展示初创企业从单一模式发展为双轨并行的成长路径 subgraph EarlyStage ["早期阶段：寻找阻力最小的路径"] E1["与客户密集沟通"] --> E2{"发掘最初的破局点"} E2 -- "ROI清晰 / 替代旧工作流" --> L1["圈地策略 (Landgrab)<br/>快速铺量打知名度"] E2 -- "行业门槛高 / 依赖背书" --> H1["灯塔策略 (Lighthouse)<br/>集中火力攻克头部"] end subgraph MatureStage ["成熟阶段：双轨合流"] L1 -- "切入企业级大客户 / 行业垂直化" --> M1["转向大客户攻坚<br/>（锁定各细分前五名）"] H1 -- "标杆背书确立 / 体系化沉淀" --> M2["横向辐射长尾市场<br/>（大规模下沉铺开）"] M1 --> Both["大企业最终形态：<br/>双轨并行策略"] M2 --> Both end ``` 以 Meraki 和 Samsara 的演进历程为例，两家公司起初都采用狂飙突进的圈地策略。但随着公司体量扩大、向高端企业级（Enterprise）迈进时，团队开始按垂直行业细分（Verticalize）——比如列出运输业前五强、仓储业前五强、公共部门前五强，此时销售逻辑便自然转变为逐个击破的灯塔策略。 ### [28:20](https://www.youtube.com/watch?v=aakZLqxRQfo&t=1700s) 垂直化场景下的销售模式转换与团队画像 Andy McCall 回顾了 Meraki 与 Samsara 开拓公共部门（Public Sector）的具体经历。在 Meraki，团队初期抓取的低垂果实是公立学区（School Districts）。学区间的信息极度透明、互通有无，只要拿下各州规模最大的标杆学区，其他学区便会顺理成章跟进采购。而在 Samsara，切入城市、郡县与州政府车队管理时，也采用了同样的打法。 之所以在这些场景下必须由圈地转变为灯塔策略，核心在于**销售动作（sales motion）发生了根本性突变**： - 销售周期大幅拉长； - 采购与招投标流程（Procurement）极为特殊且受法规约束； - 核心决策链条与中小型企业截然不同。 要求同一拨擅长快节奏打中端市场（Mid-market）的销售去啃复杂的政府招投标项目是不切实际的。这也决定了两种模式对销售人员画像的要求大相径庭： | 维度 | 灯塔策略销售画像（Lighthouse Sellers） | 圈地策略销售画像（Landgrab Sellers） | |:--- |:--- |:--- | | **经验要求** | 资深、成熟的企业级销售（Seasoned Enterprise Sellers） | 可以更早期、年轻化（Earlier in career） | | **核心能力** | 深谙复杂的组织架构渗透、漫长销售周期与严格采购流程 | 极强的进取心与执行力，招聘看重"态度与悟性"（Attitude & Aptitude） | | **工作方式** | 针对特定名单（如金融业前 15 大客户）逐一精细化攻坚 | 快速触达海量客户，以极高频次迅速堆叠胜场（Stack wins） | 探讨尾声，Andy 和 Joe 均强调，当前的 AI 浪潮为处于职业早期的销售人员创造了绝佳的历史机遇。随后话题引向了第三种截然不同的软件扩散形态——由开发者主导的自下而上（Bottoms-up）草根式采纳，这是否正在从根本上重塑销售人员的定义。 ### [31:42](https://www.youtube.com/watch?v=aakZLqxRQfo&t=1902s) 时代周期的转变：从 PLG 细分切入到重新销售“大软件” 在讨论产品主导增长（PLG）是否随着 AI 浪潮逐渐式微时，Joe Schmidt 明确表示 PLG 依然活跃，消费者式的采购体验和新型购买行为仍层出不穷，正如 a16z 投资的 Coda 以及其他优秀公司的表现。然而，站在更宏观的软件创新周期来看，过去的 PLG 浪潮其实是特定历史时期的产物。Joe 提到他在《Trading Margin for Moat》一文中的分析：大约在 2000 年至 2010 年间，第一批云平台巨头在 CRM、HR、ITSM 及安全等核心领域拔地而起，牢牢占据了底层平台的位置。在此之后的十余年里，初创公司若想在大型企业和中端市场打开缺口，唯一的办法就是打造极其锋利的“楔子产品”（wedge product），以单点切入并逐步“着陆与扩张”（land and expand）。当时从本地部署转向云端是一次巨大的范式转移，但若只是在云端之间互相替代，“无论你的按钮做成绿色还是蓝色，哪怕多了一两个小功能，客户也绝不会因为这点细微差别而大费周章地更换 CRM”。 如今，企业内部正涌动着前所未有的“巨大动力”（kinetic energy）。生成式 AI 和 Agent 的出现，彻底打破了过去那种拟物化的一对一功能替换模式。企业开始从底层重新审视 CRM、HR 和 ITSM 的运作逻辑：人类员工将不再从事繁重机械的琐碎劳动，转而执行高价值决策，将大量日常操作全权交由 AI Agent 接管。这种根本性的重塑让创业市场迎来了一个千载难逢的窗口期——重新向企业销售大型软件与核心底层平台（sell big software again）。对于今天的创始人而言，不仅要理解 PLG，更需要深入研究十五年前初代云平台巨头如何构建软件大厦及其周边生态的经典打法。 ```mermaid flowchart TD accTitle: 软件创新周期与企业采购逻辑变迁 accDescr: 对比传统云平台巩固期与当前生成式 AI 时代的软件销售模式与企业诉求差异 subgraph Past["过去 15 年：云巨头巩固期 (2010-2023)"] A["底层平台锁定<br/>(CRM / HRIS / ITSM 巨头成型)"] --> B["切入难度极高<br/>(从云到云替换意愿低)"] B --> C["PLG 与楔子策略<br/>(单点切入，Land & Expand)"] end subgraph Present["当前周期：AI 范式重塑期 (2024+)"] D["业务流程底层重构<br/>(从被动工具转向 AI Agent)"] --> E["企业重塑预算涌现<br/>(告别微小的界面修补)"] E --> F["重回平台型大单销售<br/>(重新销售 Big Software)"] end ``` - **认知门槛的大幅降低**：Andy McCall 补充指出，现在的买家成熟度远胜五年前、十年前甚至十五年前。他们对自己的真实诉求心知肚明，销售团队不再需要花费数周去教育客户“为什么需要某类软件”或“它底层如何运作”，只需聚焦于证明“为什么自家公司是解决这一诉求的最佳方案”。 --- ### [35:32](https://www.youtube.com/watch?v=aakZLqxRQfo&t=2132s) 灯塔策略的演进与终局：通吃品类还是坚守灯塔？ 针对“一家公司是否能永远执行灯塔策略（Lighthouse Strategy）”的疑问，Andy McCall 解释道，除非企业走多产品线路线、不断推出全新产品去攻克一个个全新的垂直领域，否则大多数公司在拿下核心标杆后都会发生战略演变。灯塔模式的本质是建立社会认同（social proof）：当你拿下全球排名前五的金融机构后，下一阶段的重心必然是向下通吃整个细分品类（run the category underneath it）。此时的销售动线会变得更标准、更轻量，团队无需再耗费几个月去为单个巨头进行深度定制和公关式攻坚，而是转向“这是顶级客户的背书，如果不放心可以调取参考案例，接下来请看产品为何适合你”的高效转化模式。 然而，Joe Schmidt 也指出，在极少数极高客单价、目标客户总量高度受限的垂直深水区，确实存在必须自始至终贯彻灯塔模式的企业。最典型的代表便是 a16z 投资组合中的 Applied Intuition。由于其面向自动驾驶与复杂汽车工程软件领域，全球范围内具备采购实力的主机厂和大型制造商数量极为有限。在这样封闭且 ACV 极高的市场中，团队必须将每一个潜在客户都视作战略级灯塔，“以极大的精细度和耐心去呵护每一笔交易”。 - **策略转换的标志**：从最初依赖社会认同的攻坚式慢销，转变为手握标杆案例后的全品类快速收割。 - **终生灯塔模式的适用场景**：潜在客户总体基数极小（如全球核心车企）、客单价极高（ACV 达到数百万至数千万美元量级）、替换成本极高的深科技垂直市场。 --- ### [37:31](https://www.youtube.com/watch?v=aakZLqxRQfo&t=2251s) 早期创始人最容易踩的坑：1% 战略与 99% 执行 当被问及创始人为什么容易误判销售模式时，Andy 直言不讳地指出，很多时候是因为虚荣心作祟——“卖给摩根大通（JP Morgan Chase）听起来可比卖给一家不知名的小公司性感多了”。但他观察到早期团队最普遍、最致命的毛病是**过度分析（analysis paralysis）**： > “创始人最容易犯的错误就是在战略上耗费了太多时间。战略固然重要，但你应当只花 1% 的精力制定战略，剩下 99% 的精力全部用来执行。” Andy 强调，早期创业没有任何“艰难收入加分项”（no bonus points for hard-earned revenue）。投资人和市场不会因为你的收入来自难啃的大公司就给予更高的估值乘数，每一块钱的收入在算账时都是等价的。与其关在房间里纠结究竟该走 Lighthouse 还是 Landgrab，不如立刻冲出门去与真实客户对话，找出当下愿意为现有功能买单的群体并全力拿下。在完成第一年的营收里程碑后，团队随时拥有回头复盘和调整策略的机会，切莫在分析中消耗宝贵的起步窗口。 --- ### [38:49](https://www.youtube.com/watch?v=aakZLqxRQfo&t=2329s) 快问快答：无处不在的 Landgrab 销售战场 在访谈尾声的快问快答环节，Joe Schmidt 询问 Andy 在销售生涯中签下单子的“最诡异地点”。Andy 回忆，做企业销售尤其是面对实体行业的 Landgrab 扩张期，成败往往取决于销售团队是否愿意走到客户真实的工作场景中，而不是局限于光鲜亮丽的高端写字楼。 - **非传统签约地点**：出海海钓的渔船上、射击打靶场、棒球场看台，以及各地的卡车转运货场（truck yards）与大型环卫垃圾处理场（sanitation sites）。 - **实地销售的韧性**：Andy 和 Joe 一致认为，在 Landgrab 这类快速跑马圈地的销售动态中，最突出的核心精神就是“愿意到任何能促成交易的地方去与客户面对面交流”，这种下沉到泥土里的执行力正是早期销售团队冲破阻力的关键武器。 ### [39:50](https://www.youtube.com/watch?v=aakZLqxRQfo&t=2390s) 销售新人职业生涯的核心法则：寻找能同频成长的伟大公司 当被问及如果能穿越回职业生涯早期会给年轻时的自己什么建议时，Andy McCall 直言这是他常常分享给刚入行新人的忠告，因为这也是他早年踩过的坑。在销售职业起步阶段，唯一应当聚焦的重点就是：尽一切可能去找到一家你能加入的最优秀的公司。 早期从业者极易被短期的虚荣指标迷惑，陷入“去哪里能拿最高提成”、“哪项技术最火热”、“去哪里能拿到最大的头衔”的误区。Andy 强调，到头来这些表面光鲜的东西毫无意义。真正重要的是找到一家有潜力爆发式增长的伟大企业。一旦选对赛道并登上一艘火箭般的公司，它就如同一部“职业升降机”（career elevator），带领个人的能力与回报随企业水涨船高。因此，新人绝不能被自尊心（ego）牵着鼻子走，不要过分纠结于“我要总监头衔”、“我要多高的底薪”或“我能谈下多少提成点数”，放下执念去选最好的公司才是正道。 ### [40:39](https://www.youtube.com/watch?v=aakZLqxRQfo&t=2439s) 销售运营（Sales Ops）的提前布局：避免高速扩张中的减速带 关于初创企业普遍过晚招募的销售岗位，Andy 指出具体情况往往取决于创始人的背景与风格。如果创始人本身非常擅长销售，他们可能在销售负责人（Sales Lead）的招聘上可以拖得久一些。但在普遍意义上，绝大多数公司都拖延了对**销售运营（Sales Operations / Revenue Operations）**的招聘。 Andy 特别澄清，过早建立庞大的 RevOps 团队并无必要，早期哪怕只有一个人也至关重要。公司必须有一位专职人员每天系统性地梳理核心规则：划分销售区域（territory alignment）、整理目标客户名单（named account lists）、调解佣金与成单归属纠纷（commission skirmishes），并确立团队的“销售基本法”（sales constitution）。当业务全面进入规模化扩张模式（scale mode）时，这些底层机制必须已运转成熟，否则它们会变成致命的“减速带”（speed bumps）。这类基础建设不能指望销售主管去兼顾，因为负责人的精力几乎全部被招兵买马和推进具体大单所占据。 ```mermaid flowchart TD accTitle: Sales Operations Infrastructure Before Scale accDescr: 销售运营在团队进入规模扩张前所需建立的核心机制与职责划分 subgraph SalesOps ["销售运营专员职责 (早期仅需1人)"] A["销售区域划定 (Territory Alignment)"] B["客户名单管理 (Named Lists)"] C["业绩与提成仲裁 (Commission Skirmishes)"] D["制定销售宪章 (Sales Constitution)"] end subgraph Value ["核心业务价值"] E["消除规模化阻碍 (Remove Speed Bumps)"] F["解放销售总监精力 (Focus on Deals & Hiring)"] end SalesOps --> Value ``` ### [41:45](https://www.youtube.com/watch?v=aakZLqxRQfo&t=2505s) 配额达标率与造势：为什么早期阶段应该追求 100% 达成率 在探讨初创期销售团队应当有多少人完成业绩配额（quota）时，Andy 毫不犹豫地给出答案：“100%”。许多人倾向于将销售团队维持在一种饥饿甚至近乎狂热的内卷状态，但 Andy 坚信，在业务发展初期，销售团队完全是依靠势能（momentum）运转的。管理层要做的是招聘天生具备赢家心态的人，并为他们创造真正能够赢下比赛的舞台。 虽然指标设计必须兼顾单体经济模型（unit economics）并保证公司盈利，但初创公司若想用优秀的软件去抢占市场，吸引顶尖销售人才的唯一王道就是让他们切实看到拿满提成并达成配额的机会。目前不少早期公司仅有 40% 到 50% 的销售能完成配额，Andy 认为这不仅是对团队的消耗，也是在自毁前程——这说明要么指标脱离现实定得过高，要么招聘画像完全跑偏。 在创业早期，销售成本率（cost of sales）并非决定生死的关键指标。Andy 调侃道，当公司经历披荆斩棘最终敲钟上市并统治市场时，绝不会有任何投资人或分析师跳出来抱怨：“哎呀，六年前你们初创期的销售成本真是太差劲了。”大家真正在意的是公司当初是否踩准了能够持续走向成功的增长轨道。 - **职业发展金律**：摒弃追求浮夸头衔与提成比例的短期心态，首选高增长潜力的平台作为职业跳板。 - **Sales Ops 关键抓手**：负责区域切分、客户排他名单梳理与佣金争议裁决，为规模化扫除隐形摩擦。 - **达标率指标（Quota Attainment）**：早期追求 100% 达标以建立团队获胜势能，40%-50% 的低达标率通常暗示配额或画像失误。 - **战略视角**：在软件发展的各个周期与时代，这种回归基本面的商业常识始终恒久不变。 ## Notable quotes > [18:15](https://www.youtube.com/watch?v=aakZLqxRQfo&t=1095s) 我们当时举办在线研讨会（Webinar），直接对客户承诺：‘只要你来参加 Webinar，我们就免费送你一个企业级 AP（无线接入点）。’你拿回去插在插座上试一下，只要他们一试，脑子里的灯泡瞬间就亮了，他们会惊叹：‘天啊，这比我原来用的东西简单太多了，我为什么还要买以前那套？’ — Andy McCall > [27:15](https://www.youtube.com/watch?v=aakZLqxRQfo&t=1635s) 在当前阶段，做对你业务最合理、阻力最小的事情。 — Andy McCall > [31:42](https://www.youtube.com/watch?v=aakZLqxRQfo&t=1902s) 无论你的按钮做成绿色还是蓝色，哪怕多了一两个小功能，客户也绝不会因为这点细微差别而大费周章地更换 CRM — Joe Schmidt > [37:31](https://www.youtube.com/watch?v=aakZLqxRQfo&t=2251s) 卖给摩根大通（JP Morgan Chase）听起来可比卖给一家不知名的小公司性感多了 — Andy McCall > [37:31](https://www.youtube.com/watch?v=aakZLqxRQfo&t=2251s) 创始人最容易犯的错误就是在战略上耗费了太多时间。战略固然重要，但你应当只花 1% 的精力制定战略，剩下 99% 的精力全部用来执行。 — Andy McCall > [39:50](https://www.youtube.com/watch?v=aakZLqxRQfo&t=2390s) 尽一切可能去找到一家你能加入的最优秀的公司。 — Andy McCall > [41:45](https://www.youtube.com/watch?v=aakZLqxRQfo&t=2505s) 哎呀，六年前你们初创期的销售成本真是太差劲了。 — Andy McCall > [41:45](https://www.youtube.com/watch?v=aakZLqxRQfo&t=2505s) 100% — Andy McCall ## Mentioned Elena Burger, Andy McCall, Joe Schmidt, a16z, Samsara, Meraki, Cisco, HP, JP Morgan Chase, Harvey, Decagon, Further AI, Pylon, Coda, Applied Intuition, AT&T, Verizon, Tareq, Ben, Lighthouse or Landgrab, Trading Margin for Moat, ELD Mandate, Roofnet ## Video description Elena Burger is joined by a16z's Andy McCall and Joe Schmidt to break down two very different ways AI startups can go to market: the lighthouse and the landgrab. Should founders win a handful of marquee customers whose credibility unlocks an entire industry, or move quickly across a broad market where the ROI already speaks for itself? Drawing on Joe's Lighthouse or Landgrab framework and Andy's experience building sales organizations at Samsara and Meraki, they explore how founders can determine which strategy fits their market, when social proof matters more than math, and why the current rush to adopt AI has created a rare window for startups to sell big software again. They also get tactical on POCs, pricing and ACV, hiring early sales teams, moving from mid-market to enterprise, and why founders shouldn't spend too much time perfecting their GTM strategy before talking to customers. As Andy puts it: spend 1% of your time on strategy and 99% executing. Timestamps: 00:00 - Intro 00:58 - The Lighthouse vs Land Grab Framework 06:57 - Samsara's ELD Mandate: The Perfect Land Grab Moment 13:26 - Lighthouse in Practice: Harvey, Decagon & Further AI 17:00 - ACV Discipline: Clear the Hurdle, Then Just Go 27:14 - Every Big Company Eventually Deploys Both Strategies 33:54 - Why Now Is the Moment to Sell Big Software Again 37:49 - The Biggest Mistake Founders Make Resources: Read Joe Schmidt's "Lighthouse or Landgrab": https://a16z.com/lighthouse-or-landgrab-how-to-pick-your-ai-sales-strategy/ Follow Andy McCall on LinkedIn: https://www.linkedin.com/in/amccall/ Follow Joe Schmidt on X: https://x.com/joeschmidtiv Follow Elena Burger on X: https://x.com/VirtualElena Stay Updated: If you enjoyed this episode, be sure to like, subscribe, and share with your friends! Find a16z on X: https://twitter.com/a16z Find a16z on LinkedIn: https://www.linkedin.com/company/a16z Listen to the a16z Show on Spotify: https://open.spotify.com/show/5bC65RDvs3oxnLyqqvkUYX Listen to the a16z Show on Apple Podcasts: https://podcasts.apple.com/us/podcast/a16z-podcast/id842818711 Follow our host: https://x.com/eriktorenberg Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see http://a16z.com/disclosures. [Source](https://www.youtube.com/watch?v=aakZLqxRQfo)

[18:15] Andy McCall: 我们当时举办在线研讨会（Webinar），直接对客户承诺：‘只要你来参加 Webinar，我们就免费送你一个企业级 AP（无线接入点）。’你拿回去插在插座上试一下，只要他们一试，脑子里的灯泡瞬间就亮了，他们会惊叹：‘天啊，这比我原来用的东西简单太多了，我为什么还要买以前那套？’ [27:15] Andy McCall: 在当前阶段，做对你业务最合理、阻力最小的事情。 [31:42] Joe Schmidt: 无论你的按钮做成绿色还是蓝色，哪怕多了一两个小功能，客户也绝不会因为这点细微差别而大费周章地更换 CRM [37:31] Andy McCall: 卖给摩根大通（JP Morgan Chase）听起来可比卖给一家不知名的小公司性感多了 [37:31] Andy McCall: 创始人最容易犯的错误就是在战略上耗费了太多时间。战略固然重要，但你应当只花 1% 的精力制定战略，剩下 99% 的精力全部用来执行。 [39:50] Andy McCall: 尽一切可能去找到一家你能加入的最优秀的公司。 [41:45] Andy McCall: 哎呀，六年前你们初创期的销售成本真是太差劲了。 [41:45] Andy McCall: 100%

[0:00] Andy McCall: spend 1% of your time on strategy and 99% executing. [13:11] Joe Schmidt: hit the pavement better than anyone. [16:28] Andy McCall: pour fuel on the fire [35:31] Andy McCall: analysis paralysis [39:51] Andy McCall: career elevator [40:41] Andy McCall: sales constitution

[0:00] There's a moment right now to go sell [0:01] big software again. We're now looking at [0:03] a different way of doing business [0:04] entirely. [0:04] >> What are the lighthouse and land grab [0:06] sales playbooks? [0:07] >> Here's the framework for evaluating [0:09] which playbook should you be following. [0:10] There's this very obvious one, go after [0:12] the very obvious companies. You're in [0:13] San Francisco that probably have some [0:15] sort of proof or social value associated [0:16] with them. Or like go out and sell in [0:19] Ohio. Find people who need your [0:20] solution. [0:21] >> If you think about sort of the [0:22] enterprise networking world in 2009, [0:25] people thought we were crazy. Like we [0:26] had no chance of getting into the [0:28] largest corporations in the world [0:30] >> lighthouse because Cisco and HP had them [0:31] all tied up. But what we could do is we [0:34] could say, "Listen, we can configure, we [0:36] can deploy faster, we're simpler to [0:38] use." And that was very much a land grab [0:39] strategy. [0:40] >> Too few people are willing to pick up [0:41] the phone and willing to get on the [0:42] plane and willing to get, you know, in [0:43] front of those customers right now [0:45] because they feel like it sounds way [0:46] more sexy to sell the [music] JP Morgan [0:49] Chase than the shmish shmish Morgan. [0:51] >> [laughter] [0:52] >> I think the biggest mistake that I see [0:54] [music] founders make at an early stage, [0:56] honestly, is just [0:58] >> Welcome back to the A16Z podcast. Today [1:01] we're getting into the single most [1:02] expensive question an AI founder makes, [1:05] how you sell. Joe Schmidt just wrote a [1:07] piece called lighthouse or land grab, [1:10] which gets into the two dominant [1:11] playbooks he's observed among enterprise [1:13] AI startups. Joe, tell us about the [1:15] piece in your own words. What are the [1:18] lighthouse and land grab sales [1:20] playbooks? [1:21] >> Yeah, and I and this this piece actually [1:22] all stemmed back from an observation [1:25] uh that I had actually driving up the [1:26] 101 freeway. Uh maybe this is like [1:28] three, four, five months ago, I don't [1:30] remember. And you just realize that you [1:32] have the same kind of like two competing [1:33] companies. Like one's on one side of the [1:35] freeway and the other's on the other [1:36] side of the freeway and they're selling [1:37] the exact same piece of software. And [1:39] for some reason they all have decided [1:41] that the only relevant companies for [1:43] this piece of software are in San [1:44] Francisco and driving on the 101 [1:46] freeway. And then this has gotten even [1:47] more ridiculous. It's obviously every [1:49] bus has been wrapped and every, you [1:51] know, it's like planes are flying up [1:52] above and they're now towing, you know, [1:54] towing startups. And so I think it's [1:55] it's all very clever, but it's all [1:57] targeting the same kind of sales motion. [1:58] The reality is [2:00] you don't always have to do that. Like, [2:02] you don't always have to sell to the [2:02] same companies in San Francisco. And so, [2:04] what I wanted to try to do is to sell [2:06] tell founders like, "Hey, here's the [2:07] framework for evaluating which playbook [2:09] should you be following? There's this [2:11] very obvious one, which is like go after [2:13] the very obvious companies here in San [2:14] Francisco, in New York City, in a major [2:16] metro that probably have some sort of [2:19] proof or social value associated with [2:20] them. Or like go out and sell in Ohio. [2:24] Go out and sell in Chicago. Go out and [2:25] sell in St. Louis. Find people who who [2:27] who need your solution. Um and so, that [2:29] was the whole point of the piece is [2:30] like, you don't always have to go sell [2:31] to these notable logos. And [2:33] um [2:34] we'll see how that plays out, but that's [2:35] why. [2:36] >> And and just to get a little bit deeper, [2:38] like when does it make sense for a [2:41] founder to, you know, go and buy a giant [2:44] billboard that you see when you're [2:46] driving from SFO into the city? When [2:49] does it make sense for you to kind of do [2:51] a more targeted sales activity or like [2:54] motion elsewhere? [2:55] >> Yeah, and I think the way that we tried [2:57] to make this make sense, of course, we [2:58] did very [3:00] uh consulting style with a 2 by 2 [3:01] matrix. Um I did never work at a [3:03] consulting firm, but um I'll do my best. [3:06] Uh and so, we really were thinking [3:07] about, "Okay, what are the the axes that [3:08] we should be kind of mapping [3:09] opportunities against?" And so, we [3:11] decided upon like the Y axis is really [3:13] what we called like a the buyer's [3:15] exposure. Um and it's it's intentionally [3:17] called exposure because there's the [3:18] exposure of, you know, making a mistake [3:20] with the solution that you buy. There's [3:22] also the the exposure of the solution [3:24] inside of your company. Like, does this [3:26] does this product that I'm selling to my [3:27] customer end up you know, being is it [3:29] shown to their end customers, right? [3:31] Which is actually an important [3:32] distinction. Um and really just kind of [3:34] the overall risk associated with buying [3:36] this piece of software. So, that was [3:37] like kind of the Y axis, and it goes [3:39] from like high to low. Um and then the [3:41] the X axis is whether or not proof [3:43] travels in any given market. And so, you [3:46] know, if you think about like the top [3:47] right would be proof travels in this [3:48] market. Um and it's it's high buyer [3:51] exposure and high you know, high buyer [3:52] risk. Um and the bottom left would and [3:55] and that's a lighthouse market, right? [3:56] This is to be clear. [3:57] Um and the bottom left would be, you [3:59] know, low low uh you know, low proof [4:01] traveling, but also low buyer exposure. [4:03] Uh and that would be a land grab market. [4:05] Um and so I think those are actually [4:06] quite different and if you think about [4:08] the the, you know, the standard markets [4:10] that fit into the lighthouse model, [4:11] it's, you know, regulated industries. [4:14] Often times there's a a more constricted [4:16] number of logos. Um if you're wrong in [4:19] the industry, like if the buyer buys the [4:20] wrong piece of software and it ends up [4:22] doing the wrong thing, uh it can lead to [4:24] a you know, very bad things happening [4:26] for your firm, uh including, you know, [4:27] potentially getting in trouble with the [4:28] regulator, you know, even going uh and [4:31] and doing something illegal. That's very [4:32] bad, uh of course. Um and then uh you [4:35] know, on the on the flip side, when you [4:36] when you're looking at the more of a [4:37] land grab market, there's an established [4:38] budget. People have very, you know, used [4:41] to and accustomed to paying for a type [4:43] of service, and you can kind of come in [4:45] there and show like the end buyer the [4:47] math of like, "Hey, my solution is [4:49] better than whatever solution you're [4:50] you're using today, whether that's, you [4:52] know, a software driven solution or [4:53] human driven solution." So, you know, [4:55] the the the kind of distinction that we [4:57] drew was between like proof on the the [4:59] top right of the of the of the quadrant [5:01] and math on the bottom left of the of [5:03] the quadrant. Um and that's how we kind [5:04] of thought about the framework. [5:05] >> Yeah, and [5:06] >> Andy, we'll get into your background uh [5:08] in a little bit, but I think first maybe [5:10] it's it's good to kind of categorize [5:12] some of these modern day AI startups [5:15] within these two frameworks. And I know [5:18] you both work with a lot of these [5:19] companies. [5:20] >> Yeah. So, so Joe, I'm not sure if you [5:21] want to call out specific examples or [5:23] Andy, if you want to call out specific [5:25] examples and really just kind of like [5:26] talk through the the sales strategies [5:29] that you're seeing. [5:29] >> Yeah, I mean, well, you're like, you [5:30] know, the the the land grab uh you know, [5:33] mastermind. So, maybe you want to talk [5:34] about some of the stuff that you've [5:35] seen. [5:35] >> Yeah, well, I mean, yeah, I think you [5:37] you framed it up really well in in your [5:38] in your article, but um yeah, I think [5:40] that concept of like the the lighthouse [5:42] being more in industries that require, [5:45] you know, regulation, a lot of social [5:47] proof, and so forth. And those tend to [5:49] be companies that are going after like a [5:51] category creation, right? It doesn't [5:53] exist today, so therefore you got to go [5:55] out and you got to kind of prove [5:55] yourself with the big the big names. And [5:57] obviously we have a bunch of portfolio [5:58] companies out there. There's a there's a [6:00] bunch of the companies that I think you [6:01] mentioned a couple in your article that [6:02] are doing that. And then on the other [6:03] side the the the sort of land grab, you [6:06] know, less social proof, but those tend [6:08] to be especially in the AI world today, [6:10] those are companies that are maybe [6:12] replacing or improving workflows that [6:14] already exist. And there's existing [6:16] budget and again we have a whole bunch [6:17] of portfolio companies you named a [6:19] couple in the article [6:20] that are doing that today. They're [6:21] inserting themselves in saying hey we [6:23] built a better way to go after this [6:24] through AI [6:26] and and I think those are the great you [6:28] know companies going after land grab [6:29] strategy. [6:29] >> Yeah, and it's interesting. I mean for [6:31] for those that aren't familiar with with [6:33] Andy's background, you know, he's he's [6:35] built some of the the best sales [6:37] organizations that I've ever heard of at [6:39] Samsara and Meraki. [6:41] And I've always found like these these [6:43] stories like really fascinating and [6:44] illuminating like the the Samsara, you [6:46] know, story around kind of the ELD [6:48] mandate as the way I understand it kind [6:50] of basically forcing a category to [6:52] happen everywhere all at one point and [6:54] it's just who can go out there and do it [6:56] fastest. Like maybe just if you if you [6:59] could for for the audiences edification [7:01] like share a little bit about what that [7:02] like big why now moment was because [7:04] we're kind of having one right now [7:06] and and how you guys kind of went and [7:07] captured it. [7:08] >> Yeah. Yeah, I think if you if you're in [7:10] the industry for long enough and I've [7:12] got the gray hair to to prove that I [7:13] have been [7:14] >> Blonde. [7:15] >> Blonde. [7:15] >> [laughter] [7:15] >> Blonde. Blonde. [7:16] >> Yeah, bordering on gray. [7:18] >> Sure. [7:18] >> Um you know, you you tend to see these [7:20] big you know, transitions and and I'm [7:22] old enough to have seen you know, the [7:24] internet come about and [7:25] you know, certainly mobile and cloud and [7:27] obviously now AI and each each of those [7:29] causes this this transition and it and [7:31] it causes new ways to think about how [7:33] you go to market. [7:35] Not wholesale changes, but always tools [7:37] to help improve upon them. [7:40] And and I think uh know, the other thing [7:42] I would say [7:43] um I I I can't name too many companies [7:46] that have become hugely successful [7:48] without some element of timing and luck, [7:50] right? And so you talk about Samsara and [7:52] um you know, the the company was uh [7:54] founded back in 2015. I joined in 2017. [7:57] Um and you know, the idea back then was [7:59] you know, at the very beginning it was [8:00] like internet connected sensors, like [8:02] all the value chains are going to come [8:03] sensored up. How do we you know, get [8:05] sensors out there and ingest this data, [8:06] give it back to business owners in in [8:09] digestible and usable ways. Um [8:11] and one of the first products that [8:13] started gaining traction were these [8:15] telematics units. And you know, taking a [8:18] step back if you think about you know, [8:19] the the world of transportation, [8:20] long-haul trucking, right? Um you know, [8:23] prior to like 2016, [8:25] uh they had these manual logbooks. If [8:27] you were driving a truck and you stopped [8:29] to take a break, you would write down in [8:31] your logbook, I had just drove for 4 [8:33] hours, now I'm taking a 20-minute break, [8:34] and then then you drove for 2 more hours [8:36] and you stopped for lunch. And if you [8:38] got pulled over by highway patrol, [8:40] they would ask to see your logbook and [8:41] they'd audit and make sure you weren't [8:43] driving too long. It was a it was a [8:44] safety regulation. [8:46] And in the US, roundabout 2016, they [8:49] implemented this ELD mandate, right? And [8:51] stands for electronic logging devices. [8:53] And the idea there was, hey, we can use [8:55] technology to actually track when the [8:57] vehicle's moving and when it isn't and [8:59] are they taking enough breaks and so [9:00] forth, right? Take the human [9:02] uh and and and data element out of it, [9:04] um [9:05] the human element rather. And and so you [9:07] know, over a 2-year period, that was [9:08] basically implemented between 2016 and [9:10] 2019 with various phases of of sort of [9:13] compliance. But what it did is it it [9:15] provided this huge tailwind for anybody [9:18] making these electronic logging devices. [9:20] And we just happened to be one of the [9:22] newer companies doing it. And there were [9:24] some very very established players, [9:25] right? AT&T had a solution, Verizon had [9:27] a solution. There were a number of [9:28] companies that were in sort of a you [9:29] know, hundreds of millions, half a [9:31] billion in revenue already doing this. [9:33] And you know, rising tide floats all [9:35] boats. It helped everybody. But if you [9:37] were a new entrant into the market, like [9:39] we were at Samsara, it really helped [9:41] because basically the entire industry [9:44] all of a sudden had to find budget to go [9:47] out and and buy these things. And a [9:49] certain percentage of them, you know, [9:51] clearly would say, "Hey, let's check out [9:52] what's new out there. Are there any new [9:54] entrants into the field?" And so it [9:55] really helped kind of, you know, give us [9:57] a a boost. [9:58] >> How did you navigate like the the social [10:00] proof side of that? Cuz like, you know, [10:02] I think the casual observer might think [10:04] about that and say, "Okay, wow, this is [10:05] regulated. Like we can't screw this up." [10:07] So you probably have to go win, I don't [10:09] know what the largest long-haul trucking [10:11] company is. I'm trying to think of the [10:12] ones I see on the freeway, but in any [10:13] event, you probably have to go win that [10:14] one. But it doesn't sound like that's [10:15] what you did from what I understand. [10:16] It's not what you had to do. So how did [10:18] you navigate that social proof element? [10:20] >> Yeah. Well, I think [10:22] you know, [10:23] I would maybe I'm doing a little bit of [10:25] disservice to to [10:26] the amount of strategy that went into [10:28] this, but there wasn't a lot of strategy [10:30] that went into like do we chase [10:31] lighthouse accounts or do we go after [10:33] land grab? [10:33] >> Who's willing to pay us? [10:34] >> We kind of listened to our customers, [10:36] right? And and it doesn't take too many [10:37] cold calls into the largest trucking and [10:39] transportation firms when you're you [10:41] know, 18-month-old company that they've [10:43] never heard of to hear, "We're not [10:46] buying you." Click. To figure out, hey, [10:48] you know, who can we sell to? And you [10:50] know, in in 2017, 2018, you know, we had [10:54] minimal features. We were just looking [10:56] for do we have something that somebody [10:59] wants to buy? And so for us at that [11:01] point in time, the mid-market was the [11:03] place to go. For a couple reasons. [11:04] Number one, it didn't require as much [11:06] social proof, right? It was it was more [11:09] about hey, are you satisfying my my need [11:11] for for telematics? Do you do you fit [11:13] the requirement? And then, you know, the [11:16] other reason was we could get really [11:18] fast feedback product, right? Cuz the [11:20] sales cycles were short. We could get it [11:22] implemented quickly. They would deploy [11:23] quickly, right? The bigger the account, [11:25] the longer these deployments, the longer [11:27] the feedback loops. And so it really [11:28] helped us with the the sort of product [11:30] innovation side as well just to get as [11:32] many deals out there and get as many [11:33] wins as we could. [11:34] >> Yeah, and I think this is actually [11:35] really important though like as like you [11:37] know early stage founders evaluate this [11:39] moment in time like what do we don't [11:41] exactly have like you know the mandate [11:43] from the you know US government saying [11:45] you have to adopt AI but like you know [11:47] CEOs everywhere are saying you do have [11:49] to adopt AI company. There are AI boards [11:52] at every enterprise right now saying [11:53] like here's what we need to buy and we [11:55] need to do it by X period of time and [11:56] and that surely will go away but there [11:58] is this moment of like crazy you know [12:00] kinetic energy inside of big companies [12:02] and and so I think that like what he [12:04] just said actually what Andy just said [12:06] is is actually a good a good kind of [12:07] bellwether of whether or not you're in a [12:08] land grab or versus a lighthouse market. [12:10] Are people willing to actually buy from [12:12] you is part of the land grab you know [12:15] math here is like are they willing to [12:16] get on the phone with you? Are they [12:17] willing to buy your product? Are you [12:18] going through POCs and are you figuring [12:20] out like how to get someone to use it? [12:22] You know and so if you can't get that [12:24] done then it's all about going and like [12:26] doing like very deep you know forward [12:28] deployed [12:29] lighthouse like a you know arrangements [12:31] and figuring out how to then get to your [12:33] next customers but [12:34] I just think too like too few people are [12:36] willing to pick up the phone and willing [12:37] to get on the plane and and willing to [12:39] get you know in front of those customers [12:40] right now because they feel like oh this [12:42] is like this new category moment I have [12:44] to go I have to go to JP Morgan Chase to [12:46] to sell my deal. [12:47] >> Yeah, and you said a really good thing [12:49] there it's it's you know do they have [12:51] existing budget? Do they have you know [12:53] is it is it a replacement product? If it [12:55] is then you're probably going to lean [12:57] more towards a land grab. [12:58] >> Yeah. [12:58] >> If this is a brand new product and and [13:01] there's so many of those today right? We [13:02] talked to founders every day companies [13:03] are being born with brand new products [13:05] and they're going after brand new [13:06] markets. If you have to do a lot of [13:08] education to your market if they don't [13:09] have existing budget if you're going to [13:11] have to take them this educational [13:12] journey before they can go out and [13:14] justify the purchase internally that's [13:16] probably more of a lighthouse strategy. [13:18] You're taking them on this educational [13:20] journey and it's a lot more missionary [13:22] work than it is hey take that money [13:24] you're spending with vendor A and move [13:25] it over to us. [13:26] >> Yeah, yeah. [13:26] >> So maybe it's worth getting specific [13:28] about some of these companies. I know in [13:30] the piece you talk about Heavy A and [13:32] Harvie as as classic lighthouse [13:34] examples. And then Stut and Deckagon as [13:38] as land grabs. So, maybe do you guys [13:40] want to talk about some of those those [13:41] playbooks that you've seen or maybe like [13:43] give other examples? [13:44] >> Yeah. Sure, I mean I I I can go I mean [13:46] you jump in. Um you know so for example [13:49] I highlight Stut in the article as like [13:51] the the prototypical example of of a [13:54] land grab company that we're seeing in [13:55] this new age. And so [13:57] Stut this amazing business founded by [13:59] you know two incredible entrepreneurs [14:01] Tarc and Ben. And what they are going [14:03] after is the accounts receivable market. [14:05] And for listeners who maybe have never [14:08] thought about like AR accounts [14:09] receivable basically you know this is [14:11] when someone owes you money you know in [14:13] an enterprise context and you have to go [14:14] collect the money from them. This is not [14:15] glamorous. Um but however there has been [14:18] you know a mechanism to do this [14:19] historically. Right? There are [14:20] collections teams and there's [14:22] big pieces of software like [14:24] you know I won't say their names because [14:25] compliance will probably bleed them out [14:26] anyway. So, there are big companies that [14:28] do lots of AR that that sell in this [14:30] market. But it's been very manual. [14:31] Right? These these human teams have to [14:33] interact with this piece of software and [14:34] they have to go out there and they have [14:35] to sell [14:36] or have to go they have to go collect. [14:37] And so what Stut said was hey AI is [14:39] actually quite good at you know [14:41] basically having conversations with with [14:42] people. It's very good at looking at [14:43] information internally and basically [14:45] doing this process end to end. And so we [14:47] could reimagine this historic you know [14:49] way of doing collections and instead of [14:50] having humans do it we can have humans [14:52] plus AI do this and and do it even more [14:54] effectively. And that what that they [14:56] then saw that open up was like the rest [14:58] of kind of like you know order to cash [14:59] and basically the entire kind of [15:01] accounts receivable suite. [15:03] And so what they basically went out and [15:04] showed all of the early stage buyers was [15:07] in doing this we have the math to prove [15:09] it. We'll be more effective than your [15:11] current solution and your current human [15:12] teams at collecting. And this will do [15:14] XYZ for you know improve working capital [15:17] by a tremendous amount. It'll save you [15:19] money. It'll actually make you more [15:20] money. [15:21] So, they were able to go out to the [15:22] mid-market and just basically show the [15:23] math and be like, "Would you like to, [15:25] you know, have this solution, yes or [15:27] no?" Um, so that's a really good example [15:28] of a of a of a land grab market. Um, and [15:31] those entrepreneurs are just like [15:32] unbelievable sellers. They hit the [15:34] pavement better than anyone, um, just as [15:36] good as anyone as anyone I've ever seen. [15:38] Um, and uh and and they're doing a great [15:40] job. You know, another example on the on [15:42] the lighthouse side would be, you know, [15:44] we had we had like Harvey in in our [15:45] article, um, and they just did a [15:47] fantastic job of winning the right law [15:49] firms for this very new, very, you know, [15:51] theoretically high-risk initiative where [15:53] you're augmenting your human workforce [15:55] with, you know, AI capabilities and [15:57] really automating what, you know, um, [16:00] you know, junior lawyers would be doing, [16:02] um, on a day-to-day basis. And so when [16:04] they won the first kind of, you know, [16:05] few critical lighthouse accounts, um, [16:09] inside of their market, like that proof [16:11] traveled like big time. Um, and uh and [16:14] then, you know, the buyers that had like [16:15] this tremendous amount of exposure [16:16] realized, "Hey, it's actually safe for [16:18] me to buy this this solution." [16:19] >> So those are the two kind of counter [16:21] example or, you know, two examples in [16:22] this market. I don't know if there's [16:23] anything you you'd highlight from other [16:25] companies you're working with or [16:26] >> yeah. Things you've seen. [16:27] >> Yeah, I mean, those are two great [16:28] examples. Um, yeah, I'm doing uh a [16:30] decent amount of work with a company we [16:32] invested in called Pylon. Uh, they're [16:33] basically the AI native um customer [16:36] support and they're a great example of [16:38] of uh of land grab. They're doing a [16:40] fantastic job right now of just going [16:42] out and saying, "Hey, we've got a better [16:43] way of doing this." Uh, and, you know, [16:46] they have uh you know, they they've been [16:48] climbing up the ACV ladder, but you [16:49] know, they started at pretty modest ACVs [16:52] uh and have been working their way up [16:54] just by going out and replacing. And [16:55] they have a fantastic go-to-market team [16:57] that's just out executing. [16:58] >> This this ACV question is actually kind [17:00] of an interesting one and I'd be curious [17:01] how you thought about it at at Meraki or [17:03] Samsara. Um, [17:05] you know, there's there's so much demand [17:06] out there and there's so many different [17:07] ways of like kind of building your [17:08] go-to-market engine. Like how much did [17:10] you actually even think about what you [17:12] were landing at with these like, you [17:14] know, it maybe go back to when you were [17:16] at like, I don't know, 10 or 50 of ARR [17:18] at one of these businesses. Were you [17:20] optimizing for that or was it just like, [17:21] let's basically figure out how to get [17:23] enough reps in and [17:24] >> Yeah, it's a good question. I the answer [17:26] is you think about it a lot and then you [17:28] try and not think about it at all. And [17:29] what I mean by that is you want to make [17:31] sure that the ACV that you're going [17:33] after [17:34] it it has to it has to top the hurdle, [17:37] right? In in other words, you look at [17:39] your unit economics and is it healthy or [17:41] not? You don't want to be taking deals [17:42] that are, you know, negative to your [17:44] unit economics. [17:45] >> Yeah. [17:45] >> But if it passes the threshold, then the [17:47] answer is you don't think about it, you [17:48] just go and you get as many of those as [17:50] you can. So, if you can build a [17:51] go-to-market engine just theoretically [17:53] that could live off of 15K ACV deals, [17:55] >> Yeah. [17:56] >> fantastic. Like, don't take 8K ACV [17:59] deals, but go get as many 15K AC deals [18:01] as you can. And and you want to just [18:02] build a repeatable engine and pour fuel [18:05] on the fire and get as many of those as [18:06] you can. And then what happens over time [18:08] is you start inching up, right? Bigger [18:09] and bigger companies like what you're [18:11] doing then and and you start stacking [18:13] up, you know, the wins and [18:14] >> Yeah. [18:14] >> going up the ACV ladder. [18:15] >> And and can I one follow up on that [18:17] because this is interesting and [18:19] I'm so much enjoying getting you talking [18:21] about this stuff, but like Meraki, you [18:23] guys had and maybe tell the the audience [18:25] what Meraki does, but you know, [18:27] basically access points, right, for for [18:28] internet connectivity [18:30] and you guys had this really clever [18:31] program where you would kind of give it [18:32] an access point away for free is the way [18:34] I understood it almost, but [18:36] and then of course like you do that, [18:37] that impacts gross margin, but also [18:40] you then have a hardware element as a [18:41] part of your gross margin calculus like [18:43] we don't have as many companies out [18:45] there that are doing, you know, maybe AI [18:46] applications with with with a hardware [18:48] element, but there is an inference [18:49] element which impacts gross margins and [18:51] so I don't know, how did you think about [18:52] like these kind of tradeoffs in the [18:53] early days of Meraki and and of course [18:55] you had the same thing at Sentara with [18:56] with hardware, too. [18:57] >> Yeah, so so just just for for the [18:58] audience's education, so Meraki was a [19:01] basically a cloud networking company [19:04] at this point 20 years ago. Still part [19:06] of still still is a very healthy [19:08] business within Cisco. Was was by Cisco [19:10] back in 2012, Um, the the company was [19:13] actually founded back in 2006. Uh Uh, [19:15] the co-founders were working on a a [19:16] research project as PhD students at MIT [19:20] and started the business. And and what's [19:22] interesting about that, the the research [19:23] project was called Roofnet. The [19:25] technology they built was basically [19:27] large-scale like mesh Wi-Fi and they'd [19:29] install it on the roofs in like [19:31] Cambridge and the idea was you could, [19:33] you know, outfit these municipalities [19:35] and you know, parks and you know, public [19:37] areas with Wi-Fi. Um, it was fantastic [19:39] technology. Uh, within the first couple [19:41] years they figured out it wasn't a great [19:42] business model. There wasn't a lot of [19:43] revenue in municipal Wi-Fi and so they [19:46] made this pivot into enterprise. [19:47] >> Mhm. [19:48] >> Um, and the reason I say that is, [19:50] uh, you know, if if you think about sort [19:52] of the enterprise networking world in [19:54] you know, 2009, 2010, like people [19:57] thought we were crazy. Like, why would [19:59] you be trying to build an enterprise [20:00] networking company in 2009? You don't [20:02] you know that market was won 10 years [20:04] ago by Cisco and HP. Um, but the reality [20:07] was and again back to sort of the the [20:08] timing, that was right when cloud was [20:10] coming about. And and the the big [20:12] innovation around Meraki was the the [20:15] product and engineering folks, fantastic [20:18] team there, figured out how to configure [20:20] and manage this networking equipment [20:21] through the cloud, which sounds like, [20:23] yeah, total, of course. Back then it was [20:25] it was a little bit innovative. Anyway, [20:27] um, [20:28] our our problem and that was very much a [20:30] land grab strategy, right? We had no [20:32] chance of getting into the largest [20:33] corporations in the world, the [20:34] lighthouse, because Cisco and HP had [20:36] them all tied up. But what we could do [20:39] is we could say, "Listen, we can [20:41] configure, we can deploy faster, we're [20:43] simpler to use." Well, who cares about [20:45] that? The mid-market, right? Where they [20:46] don't have substantial IT teams that [20:49] have been trained in command line code [20:51] and and this kind of stuff. Um, and so, [20:54] you know, our kind of firm belief at [20:55] that point in time was, well, what's the [20:57] best way to get them to understand that [20:59] our networking equipment is simpler to [21:01] use than their Cisco that they're about [21:04] to buy? [21:04] >> Mhm. [21:05] >> And the answer is get them to try it. [21:07] >> Yeah. [21:07] >> And so what we do is we'd run these [21:09] webinars and we'd say, "Hey, you attend [21:10] the webinar, we'll send you a free [21:12] access point. [21:12] >> Mhm. [21:13] >> You plug it in, try it out." And the [21:15] idea was if they if they try it, the [21:17] lightbulb goes off and they say, "Wow, [21:19] this is just so much easier than what [21:21] I'm using. [21:22] >> Yeah. [21:22] >> Why don't I why don't I use that?" And [21:23] it was it was very very successful for a [21:25] long period of time and [21:26] even as the company matured, you know, [21:28] we were very very liberal in our trial [21:30] and eval because you just fundamentally [21:32] want customers to experience the [21:35] technology. [21:36] >> Yeah. [21:36] >> Um and realize that it's better than the [21:38] alternative. [21:39] >> And do you think like there's there's an [21:40] element of um like that people can kind [21:43] of learn from that right now? It's it's [21:45] hard though because there is a [21:46] an aspect of configurability with a lot [21:49] of the new AI stuff, right? Like if you [21:50] just give somebody this like, you know, [21:53] Ferrari, they they might not know [21:54] exactly how to even turn it on and um [21:56] and so I don't I I don't know how how [21:58] you even think about like the delivery [21:59] mechanism for some of these like trial [22:01] periods, POCs with some of the AI [22:03] companies you're working with right now. [22:04] >> Yeah. I I think it's become more [22:06] challenging, right, in the world of AI [22:07] because number one, things are moving so [22:10] fast. Like things are changing daily. [22:12] And if you think about like a a proof of [22:14] of concept or a trial, you know, the [22:16] whole idea, if you're on the sales side, [22:17] is I want the customer to experience [22:19] this, I want to prove that it works for [22:21] them, but I want to do it in a in a [22:23] period of time that doesn't go on [22:24] forever, right? And so what you what you [22:26] have to stay away from, and I think one [22:28] of the real dangers today is these [22:30] things turn into like science projects, [22:31] right? Like I'm going to deploy this, [22:32] well, can it do this? Can it do this? [22:33] Can you show me this? Can you show me [22:34] this? And of course things are advancing [22:36] every day, so the answer is probably [22:38] yes, I could, but then you run the risk [22:40] of these trials or proof of concepts [22:41] going on forever. [22:42] >> Yeah. [22:43] >> And so it just takes a lot of [22:44] discipline, I think, in today's day and [22:46] age to to really box that in and say, [22:48] "Listen, here's what our solution does, [22:51] and we're going to, you know, define it [22:52] this way, and we're both going to agree [22:54] that if it, you know, has done this [22:55] after 45 days, it's success and you're [22:57] going to move forward with the [22:58] purchase." [22:59] >> so there's there is a lot of that today. [23:01] >> And did you um like I guess would you [23:03] recommend having like kind of auto [23:04] converts on these as much as you can or [23:07] Yeah, like I don't know how like other [23:09] like kind of learnings and lessons from [23:11] this type of like 30-45 day [23:14] And then we can talk about like the time [23:16] you know the the right amount of time [23:17] that you're giving people with the [23:18] product. [23:18] >> I think I think the timing depends a [23:20] little bit on the complexity of your [23:21] product, right? If it's going to take, [23:22] you know, 2 weeks to set up, then you [23:24] can't make it a 2-week trial type thing. [23:25] But I think the two biggest things are [23:27] make sure you have an end date, right? [23:29] >> Yeah. [23:30] >> It's a 30-day trial, it's a 45-day [23:31] trial, it's a 60-day trial, period, end [23:32] of story. And then the second one is you [23:34] have to define the success criteria up [23:36] front. Here is what we are proving that [23:38] we can do for you, right? And in some of [23:40] these companies, [23:42] you know, you can't do a proof of [23:43] concept because, you know, maybe it is [23:46] regulatory, maybe there's, you know, too [23:47] much risk in there, and they're not [23:48] going to let you do it. But where you [23:50] can, I think you want to make sure you [23:52] have both an end date and you have the [23:53] the success criteria clearly defined. [23:55] >> Yeah. I think this is so tricky right [23:57] now where, you know, so many, you know, [23:59] like you have to basically define the [24:02] scope that you're going after, and the [24:04] success criteria, but often times if [24:06] you're like automating something that [24:07] has never been automated before, there's [24:09] like significant amount of [24:10] configuration, and that like costs [24:12] money. And then your product could could [24:14] work, but it might be deployed [24:17] improperly, and and and or the results [24:20] take longer than 30 or 45 days. And so [24:22] how do you actually kind of There's like [24:24] a There's like a difference of hey, the [24:26] product is working, and you just and we [24:27] and basically you need to work with the [24:28] customer then optimize whatever they're [24:30] doing with the product. [24:31] Right? Like imagine your [24:33] I mean I don't I always go back to sales [24:34] cuz it's easier to think about sales, [24:35] but like imagine you're a [24:36] you know, a sales tool. Like the product [24:38] needs to work, and then you need to use [24:40] it and target the right customers for [24:42] whatever you're you're selling. [24:44] And and they're kind of like two [24:45] different parts of the equation. And so [24:47] if all of a sudden you're taking risk on [24:48] for whether or not your product works [24:49] and then whether or not it's being used [24:50] correctly, very tricky. And so I think [24:52] this is actually a really important [24:54] topic for founders and for early-stage [24:56] revenue leaders today to figure out how [24:58] you're educating your customer on like [25:00] here is the thing we are signing up for. [25:02] We are not signing up for, you know, you [25:04] know, maybe you are, maybe you aren't, [25:05] but um you know, whether or not your [25:06] employees are using our tool the right [25:08] way. [25:09] >> Who who have you seen that does the best [25:11] of evangelizing and and kind of like [25:13] taking people through the onboarding [25:14] process? [25:15] >> You know, I think uh [25:17] who would I put in that bucket? Um [25:21] I mean, I think like Decagon has done an [25:23] amazing job um at this. Like I think [25:25] they go in, they basically evangelize [25:27] that they're doing customer support [25:29] better than anyone else. Um and then [25:31] they're very good about saying here are [25:32] the benchmarks that we are signing up to [25:35] hit and then they hit them um you know, [25:37] in in their time period, right? And so I [25:38] think this is [25:39] you know, I think some people might [25:41] trivialize, you know, this. It's very [25:43] hard to do customer support effectively. [25:45] Like this is a high-risk exposed market [25:48] um you know, where where like you you [25:49] you don't want to screw this up. And so [25:51] I think they've done a really good job [25:52] of of evangelizing. Um I think the guys [25:54] at Stude back to like the the example I [25:56] just I just talked about. And then [25:58] another example would be excuse [25:59] [clears throat] me [26:00] a company that I'm on the board of [26:01] called Further, Further AI, and they [26:03] were they sell into the insurance space [26:05] and they're kind of they're basically [26:06] evangelizing the idea of bringing AI to [26:08] insurance. Um you know, [26:11] people haven't been using AI in [26:12] insurance. Like this is not, you know, [26:13] generally a first adopter of of of of [26:15] technology. But a lot of their customers [26:17] are some of the biggest insurance [26:18] companies in the world and it's because [26:20] they're very comfortable with like, [26:21] "Okay, hey, here's like an AI solution [26:22] that's built in a govern it uh you know, [26:25] a a a uh secure, you know, [26:27] governance-first, you know, form and [26:29] fashion." Um and then they work with [26:31] with forward-deploy teams, you know, [26:33] their customers to get it up and [26:34] running. So those are a couple examples. [26:35] I don't know if you have any [26:35] >> And and that's that that last one's a [26:37] good example of like a lighthouse [26:38] strategy, right? Like they've they've [26:39] gone after the bigger insurance [26:41] companies and you then you get that [26:42] social proof and you know, on down the [26:44] the long tail of insurance. [26:46] >> And I and I I think like sometimes [26:47] people think that you have to be from a [26:50] given market to do a lighthouse strategy [26:52] and maybe there's someone sitting at [26:53] home and they're like oh like if only [26:55] like I worked at this company I could go [26:57] do this and it's like no the reality is [26:59] like go build a relationship. [27:01] >> Yeah. [27:01] >> You know I see you laughing cuz I know [27:03] you're like go build a relationship with [27:04] your customer. Go find someone and show [27:06] them like here's you know I have an [27:08] earned secret. This earned secret is [27:09] that AI can help your business or [27:10] technology can help your business in [27:12] this way and you know work with us on [27:14] that front. [27:15] >> I I I I'll also say that you know every [27:16] every small company wants to become a [27:17] big company. I don't know too many very [27:20] large successful companies that at some [27:22] point in time haven't deployed both [27:23] strategies. You might start off with [27:25] land grab but then you mature and you [27:28] have a lighthouse strategy or you start [27:29] with lighthouse and then you get big [27:30] enough that you can go broad you know [27:33] into land grab. So I think for founders [27:35] when I get this question early on do [27:37] what makes the most sense for your [27:39] business right now. What does that mean? [27:41] Go out and talk to customers. Find out [27:42] where the early earliest and easiest [27:44] sales are and pursue that strategy. It [27:47] doesn't mean you're completely you know [27:49] punting on the other one. It just means [27:50] come back to it. And in in both of the [27:52] last companies that I worked for both [27:54] Meraki and Samsara we started with land [27:57] grab but as soon as we matured and [27:58] started getting up in enterprise then [28:00] what do you do? Well you verticalize and [28:02] all of the sudden like great who are the [28:04] top five you know transportation [28:05] companies who are the top five [28:07] warehousing companies who are the top [28:08] five you know public sector and then you [28:11] want to go take those down and so you [28:12] can morph into a land grab strategy. You [28:15] just want to do what's most efficient [28:17] and most effective for the stage of the [28:19] company you're at. [28:20] >> Talk Can you talk a little bit about [28:21] like one of those like key markets that [28:23] you unlocked with like you went from [28:24] land grab early to you know lighthouse [28:26] in a given market? Like what did you how [28:28] did you set up that team um or like was [28:31] it just you were the founders that was [28:33] going in this new market and then like [28:35] you know maybe a little bit about some [28:36] of those deals that you closed. I'm just [28:38] curious how this played out with the [28:39] sequencing. [28:40] >> Well I I think I you know at both Meraki [28:43] and um Samsara [28:45] you know the the sort of earliest [28:47] example of of uh lighthouse was when we [28:51] verticalized. And in both those [28:53] instances, it was basically in like [28:55] public sector. At Meraki, we were going [28:56] after like school districts because that [28:58] was sort of low-hanging fruit. And you [29:00] know, if you're if you if anybody [29:01] anybody's ever sold to school districts, [29:03] like they all talk to each other. They [29:04] all know each other. You want to find [29:05] the biggest school districts in each [29:07] state. And if you can take that down, [29:09] every school district underneath them [29:11] what did that one buy? Great. All of a [29:13] sudden the social proof is there. Now, [29:15] why why use a lighthouse strategy in say [29:19] school districts or in Sam's case when [29:21] we went into public sector when you [29:23] start selling to cities and counties and [29:24] states, you know, why shift? Because the [29:27] sales motion is fundamentally different. [29:30] Like the sales cycles are different, the [29:31] way you sell, the decision-makers are [29:33] different, the way they procure is [29:35] different. And so, asking the same sales [29:38] team to shift from, you know, selling to [29:40] a mid-market customer and enterprise [29:41] customer over to selling to a city or a [29:43] county or a school district, it's just [29:45] it's just different. And so, you know, [29:47] that's a a point where you might want to [29:49] shift and say, "Okay, great. You know, [29:50] once we're verticalized, we want to [29:51] shift to a [29:53] a lighthouse strategy." [29:54] >> What what would you say there the [29:55] differences are between great sellers in [29:58] lighthouse models versus great sellers [30:00] in land grab models? Are there any [30:01] differences like from what you saw maybe [30:03] when you're opening new markets? Is [30:04] there a profile [30:07] that that was most effective? [30:08] >> It's always difficult to generalize. I I [30:10] think, you know, in a true lighthouse [30:11] strategy, and and when I envision [30:13] lighthouse, it's like, "Hey, we're going [30:15] to go after whatever the financial [30:17] sector. You know, here are the top 15 [30:19] accounts in finance. And you know, here [30:21] are their logos. Like how many can we [30:22] get into this quarter, next quarter, [30:24] next quarter?" Right? Like that's a [30:25] lighthouse strategy. [30:26] You know, generally you want more [30:28] seasoned, [30:30] you know, enterprise sellers that, you [30:32] know, [30:33] know how to work within those accounts. [30:35] They understand the sales cycles. They [30:37] understand the the sort of procurement [30:39] cycles. In in more of a land grab [30:41] strategy where you're just saying, "Hey, [30:43] we have the best technology. We're [30:44] replacing this workflow. We're replacing [30:46] this product." You just want, you know, [30:49] very aggressive, like hire for attitude [30:51] and aptitude. [30:52] >> Right? You can go earlier in career. You [30:54] know, you just want those people to get [30:55] out there and and hit as many of those [30:57] customers as you can because, you know, [31:00] at that point it's like you're hitting a [31:02] big market and you just want to stack [31:04] wins as fast as you can. [31:06] >> Yeah, and I I, you know, my my [31:08] unsolicited advice to any [31:10] early-in-career seller or potential [31:12] seller right now is there's never been a [31:13] better time to work at some of these [31:14] companies in our portfolio. [31:16] >> That's absolutely true. It's it's a [31:17] super fun time to be in the market. [31:19] Yeah. [31:19] >> I I wanted to ask [31:22] about maybe a third kind of of selling [31:25] or or or product diffusion that we that [31:28] we haven't talked about, which is like [31:30] the developer kind of bottoms-up, [31:33] like more grassroots adoption. And is [31:36] that is that something that you guys are [31:38] seeing? Like is is the idea of the [31:40] seller sort of becoming, I wouldn't say [31:42] obsolete, but like for a particular kind [31:45] of product just just less relevant now? [31:47] Like a developer saying to, you know, [31:50] their CTO or their, you know, CIO or [31:53] whoever, like, "Hey, this is great. Like [31:54] let's just get this." And like there's [31:57] less of a sales motion needed? [31:59] >> You know, [32:00] I I guess definitely not. Like there's a [32:02] bunch of PLG that's still happening [32:03] today. I think people are buying things [32:05] in a consumer fashion all the time and I [32:07] think that there's new buyer behaviors [32:09] being, you know, kind of discovered. And [32:11] of course, like, you know, we were big [32:12] investors in Coda and everyone saw that [32:14] played out and there's a bunch of other [32:15] examples of this. [32:16] However, like if I take a giant step [32:18] back and and talk about where we are in [32:20] this like current cycle, like I wrote [32:22] this long piece called "Trading Margin [32:24] for Mode" about a year or year and a [32:25] half ago or whatever it was, [32:27] talking about like the cycle. [32:29] And if you think about like why, you [32:31] know, basically the last 12 years before [32:33] 2024 or 10 or 15 years before 2024, you [32:36] saw so much PLG is just kind of where we [32:38] were in like the software, you know, [32:40] innovation cycle. You know, and a lot of [32:42] the big cloud platform businesses, if [32:44] you think about like CRM, HR, ITSM, [32:47] security, like a lot of those big [32:49] platform businesses were founded in the, [32:51] you know, call it 2000 to 2008 [32:54] to 2010 period. Um, and those businesses [32:57] went out and solved like the big [32:58] platform opportunities. And so then the [33:01] only way to really break in at the [33:02] enterprise, right? And we're talking [33:04] more about like enterprise enterprise [33:05] sales or mid-market enterprise sales, [33:07] was to build like some sort of wedge [33:08] product, wedge in with this product, and [33:10] say, "Hey, I'm going to solve this part [33:12] of your of your, you know, suite for [33:13] you, and then try to expand over time." [33:15] And so this like land and expand model [33:17] became super, you know, obviously super [33:19] in vogue, but it was really based on [33:21] where we were in this like adoption [33:22] cycle. And there were other people, of [33:24] course, that said, "Hey, I want to go [33:25] build a new CRM. I want to go build a [33:26] new HRIS. I want to go build a new [33:28] ITSM." But like the reality, you know, [33:31] was going from, you know, on prem to [33:33] cloud was a big enough shift for people [33:34] to like switch to something new. But [33:36] going from like cloud to cloud for CRM, [33:38] like I don't care if the button is green [33:41] or blue. I don't care if there's like [33:42] one little feature difference. I'm not [33:44] going to switch. And so you can you [33:47] compare that cycle where we were looking [33:49] at everything for the last 15 years, [33:50] which was all PLG and all the time, this [33:52] is incredible, to where we are right [33:53] now, there's this crazy kinetic energy [33:56] inside of companies where they're [33:57] saying, "Hey, it could be even something [33:59] as, you know, fundamental as CRM, right? [34:02] It could be as fundamental as HR, ITSM. [34:05] We're now looking at a different way of [34:06] doing business entirely. This is not a [34:07] skeuomorphic one-to-one replacement, [34:09] green to blue. We're now thinking about [34:11] like humans are going to be doing [34:12] something completely different, way more [34:13] high value. We're going to be doing way [34:15] less of the the same kind of, you know, [34:16] mundane wrote work. Um, and instead [34:19] agents are going to be doing that doing [34:20] that. And so that's I think the [34:21] opportunity right now, and it's why [34:23] instead of talking about like of course [34:24] we got to talk about PLG and like all [34:26] these other sales models, there's a [34:28] moment right now to go sell, you know, [34:30] big software again, right? And to [34:32] And to go sell platforms. And And it's [34:35] because of this moment. Um and so I [34:37] think people need to be studying kind of [34:39] the 15 years ago models of like how [34:41] people build this and the and the [34:42] ecosystems around it um to go and have [34:45] success. [34:45] >> Yeah, I think that's right. I I will say [34:47] I think the consistent trend as long as [34:49] I've been doing this is that, you know, [34:50] every year and with every technology [34:52] transition, buyers become more and more [34:55] educated, right? So the buyer today is [34:57] just fundamentally has a better idea of [34:59] what they want than they did 5 years [35:00] ago, 10 years ago, 15 years ago. [35:02] >> Yeah. [35:02] >> And so, you know, that does lend itself [35:04] more towards, you know, if you can hit [35:07] that buyer when they're in their [35:09] decision mode, you're going to have a [35:11] better chance. It doesn't have to [35:12] necessarily be PLG, but it can be more [35:14] self-serve. It can be it should be an [35:16] easier sell than, you know, 10, 15, 20 [35:18] years ago where you had to take them on [35:20] this entire education journey, why you [35:22] need this, how it works, here's how [35:23] you're going to write. Like they're just [35:24] so much more educated now on what they [35:27] want to buy that your job is just [35:28] convince them that your company is the [35:30] right solution for that. [35:31] >> Amen. [35:31] >> Yeah. [35:32] >> Yeah. I guess there the lighthouse kind [35:35] of like definition just gets pushed ever [35:37] outwards or you kind of like, yeah, just [35:40] have to keep on conquering new [35:42] territory. I'm curious I I assume a [35:45] company just can't stay a lighthouse [35:47] forever and you you even alluded to [35:48] this, Andy. Like what what is kind of [35:51] the average amount of time that a [35:53] company kind of can just, you know, chew [35:57] off those those bigger logos and then, [36:00] you know, like like what is that [36:02] transition moment like, I guess? [36:03] >> Yeah. I mean, theoretically, you could, [36:06] right? If if you were a company that was [36:07] dedicated or committed to a, you know, [36:10] multi-product strategy, you could just [36:11] keep coming out with new products and [36:13] keep going after, you know, more [36:14] verticals, right? Like you could [36:16] theoretically do that and I think [36:17] there's probably some examples we could [36:18] come up with of companies that did that. [36:20] But But yeah, in general, you know, if [36:22] you start with a lighthouse strategy, [36:24] you've built the social proof, you've [36:25] gotten these big names, then what you [36:28] want to do is you want to run the [36:29] category underneath it, right? If I've [36:31] gotten the top five financial companies [36:33] in the world, I want to run down, you [36:35] know, the list after that. And so, that [36:37] fundamentally becomes a little bit [36:39] different of a sales motion, right? [36:41] You're spending less time with, you [36:43] know, the huge organization, the big [36:45] logo, you know, selling and you're [36:47] spending more time, [36:49] you know, less time in the social proof [36:50] and more time on, "Hey, here's a [36:52] reference if you need it. Otherwise, [36:54] this is why my product Best Buy it. [36:55] >> Yeah. And I mean, I I think And to to [36:57] Danny's point, there are certain [36:58] companies I think that actually do just [37:00] stay lighthouse the entire time. And [37:01] like, that would be like I'm trying to [37:02] think of a the best example might be [37:03] like Applied Intuition in our portfolio. [37:05] Like, there's there is a very, you know, [37:07] set number of people who are, you know, [37:09] buying that kind of autonomous software. [37:12] Um and a certain number of, you know, [37:13] car manufacturers and so on and so forth [37:15] in the world. Um not to say that that's [37:17] the only people they can sell to, but [37:18] like of course, these are these are good [37:19] trade markets. And so, you then have to [37:21] treat every one of these with immense [37:22] care cuz they are huge um ACB [37:24] opportunities. And And they've obviously [37:26] I think they're they're arguably the [37:27] best in the world at doing that. Um so, [37:30] yeah. That's a good example of that. [37:31] >> Why why would a founder sort of mis- [37:34] misjudge what game they're playing? Or [37:37] how how have you seen founders kind of [37:39] misjudge whether they're doing [37:41] lighthouse or land grab? [37:42] >> I mean, it sounds way more sexy to sell [37:44] to JP Morgan Chase than to, you know, [37:46] Schmitty Schmitty Mormon. [37:48] >> [laughter] [37:49] >> I think I think the biggest mistake that [37:51] I see founders make in an early stage, [37:54] honestly, is just spending too much time [37:56] trying to figure it out, right? It's [37:58] like too much time on the strategy. And [38:00] strategy's important, but you should [38:02] spend like 1% of your time on the [38:03] strategy. Pick it and then spend 99% of [38:05] your time trying to execute. So, you [38:07] know, rather than sit back and say, [38:09] "Well, you know, should we do lighthouse [38:10] or should we do land grab?" Get out, [38:12] talk to your customers. Figure out which [38:14] ones, you know, are are willing to, you [38:16] know, buy your product, the features and [38:18] the services that it delivers today, and [38:20] then chase that path, right? There's no [38:23] bonus points for hard-earned revenue. [38:25] You know, you don't get extra, you know, [38:27] multipliers on your revenue if you get [38:28] the big logo or something. Like go after [38:30] the customers you can, you know, and [38:32] then, you know, constantly be improving [38:34] your product and then, you know, you can [38:35] always reassess your strategy. You You [38:37] after the first year, if you get all [38:38] your your revenue milestones, you can [38:40] look at it and say, "Well, could we be [38:41] more effective doing this?" Maybe. But [38:43] just don't spend too much time in [38:45] analysis paralysis mode. [38:47] >> Yeah, totally agree with that. I totally [38:48] agree with that. [38:49] >> Joe, Joe, you had some you had some fun [38:51] questions for Andy. I It was a lightning [38:53] round question. [38:54] >> Yeah, lightning round, okay. [38:55] >> Lightning round. I'll see what other [38:56] ones I can come up off the top of my [38:57] head, but [38:58] the first one is maybe like your [39:01] favorite place you ever weirdest place [39:03] you ever closed a deal. [39:04] >> Yeah, I mean, we've had some customers [39:06] that, you know, have have taken taken us [39:09] to some some funny places, right? I [39:10] mean, I've closed deals on fishing [39:12] trips. I've closed deals, you know, out [39:14] at shooting ranges. I've closed deals at [39:17] ballparks. So, yeah, I think those are [39:19] all, you know, kind of unique, like [39:20] anywhere that wasn't a meeting room. [39:22] >> Um airport Chili's? [39:24] >> [laughter] [39:24] >> Uh probably. Probably over probably over [39:27] the years. [39:28] Yeah, certainly at Samsara where we're [39:30] selling to, you know, logistics and and [39:32] transportation companies, you know, a [39:33] lot of truck yards and and that kind of [39:35] stuff. Um you know, sanitation sites. [39:37] >> Yeah, yeah. [39:38] >> There's some There's some pretty [39:40] >> When you're doing land grab motions, I [39:41] think this is like the the permeating [39:43] theme is just like people willing to go [39:44] and and sell wherever. Maybe Andy, if [39:47] you could go back and tell yourself one [39:48] thing when you're building out these [39:50] these teams or like early in your [39:51] career, what would you tell yourself? [39:52] >> Um early in my career. Well, I mean, [39:54] this is the advice I give a lot of uh [39:57] early in career folks. [39:58] Um and it was a mistake I made early on. [40:01] The only thing you should really be [40:02] focused on when you're starting your [40:03] sales career is find the best company [40:07] you can possibly find to work for. [40:09] Uh I made this mistake early in my [40:11] career. I was chasing, you know, where [40:12] can I make the most commission? Where [40:13] can I make the most money? What's the [40:14] hottest technology? What Where can I go [40:17] get the biggest title? At the end of the [40:19] day, none of that matters. What you want [40:20] to find is you want to find the the the [40:22] great company that's going to grow. And [40:25] if you do that, then it's like a career [40:26] elevator, right? Like you will grow with [40:28] the company. [40:29] Yeah. But you can't you can't let your [40:31] ego get in the way and I want a director [40:33] title or I want this big of a base [40:34] salary or I think I can get this much of [40:36] a commission rate. Just go find the best [40:38] company you can work for. [40:39] >> Yeah. Totally agree. What's one role you [40:41] think companies should hire for earlier [40:43] than they normally do? [40:44] >> In sales work? [40:46] I mean, it really depends on the company [40:48] and and how and you know, specifically [40:49] like how comfortable the founders are. [40:51] Like if you're a founder that's very [40:53] comfortable sales and you know, you can [40:55] wait longer to hire sales here and that [40:56] kind of stuff. Um so, it's a little bit [40:58] of a generalized question, but I I would [41:00] say sales operations is is probably one [41:02] that I see companies waiting a little [41:04] too long on. And I am not a proponent. I [41:06] don't think you need to to, you know, [41:08] stand up a gigantic revenue operations [41:11] organization. It can be literally like [41:12] one person, but you need somebody that [41:15] every day is thinking through territory [41:16] alignment, you know, name lists, doing [41:19] you know, [41:20] commission skirmishes, setting a sales [41:22] constitution. Like all that kind of [41:23] stuff becomes really really really [41:24] important because [41:26] when you get into like, you know, scale [41:28] mode, you want all that stuff like [41:29] largely figured out. You don't want it [41:31] to become speed bumps otherwise. So, you [41:33] want somebody thinking about that and [41:34] it's it's generally not going to be your [41:35] sales leader cuz they're thinking about [41:37] how do I hire the next person? How do I [41:39] bring the next deal on? [41:41] So, I'd say, yeah, maybe sales [41:42] operations, sales revenue operations. [41:44] Yeah. [41:45] >> Um maybe you know, one sentence or two [41:47] sentences on, you know, [41:48] how you think about how what percentage [41:51] of sales teams at the early stage [41:52] companies you're at should be hitting [41:54] quota? [41:55] And like kind of like, [laughter] you [41:56] know, sales thinking [41:57] >> 100%. [41:58] >> Yeah, 100%. See, I was going to say [42:00] Well, I think there's you have some [42:01] interesting comments on like basically, [42:02] you know, trying to keep them low, get [42:03] everyone kind of rabid and and really [42:05] >> I mean, listen, I think in the early [42:07] days, like sales teams run off of [42:09] momentum, right? You you to hire winners [42:12] and give them a chance to win. So, yes, [42:14] you want to, you know, be able to bring [42:15] people in. You want to set reasonable [42:17] goals. [42:18] Um it's got to be profitable for the [42:19] company. Your unit economics have to [42:21] work. Like you can't, you know, you [42:22] can't change the math. But at the end of [42:24] the day, like if you're an early stage [42:26] company and you've got a great product, [42:28] you want to hire the best possible sales [42:30] talent to get that product solution out [42:32] to market. And the way you do that, the [42:34] way you attract those people is give [42:35] them a chance to hit quota. [42:37] >> Totally. [42:37] >> So, yeah, I think, you know, some of [42:38] these companies today where, you know, [42:40] 40, 50% hitting quota, I think they're [42:42] probably doing themselves a disservice. [42:44] Either their quotas are too high or, you [42:46] know, their hiring profile's off. But, [42:48] yeah, I think especially in the early [42:49] stages when, you know, cost of sales [42:52] isn't as important. When you're a public [42:54] company and, you know, you've you've [42:57] you've gone out and you've you've [42:58] conquered your market, nobody looks back [43:00] and says, you know, gosh, you know, 6 [43:01] years ago your cost of sales was really [43:02] terrible. [43:03] >> Yeah. [43:04] >> [laughter] [43:04] >> Totally. [43:04] >> Nobody cares about that. What they care [43:05] about is, you know, did you get on a [43:07] path where you where you could, you [43:08] know, be successful. [43:09] >> Well, I I think something that's cool [43:11] about this conversation is just how like [43:13] timeless a lot of the wisdom is. It it [43:16] just seems it seems like, you know, [43:18] these frameworks and and ways of [43:20] thinking about the industry are are [43:22] pretty consistent just throughout the [43:24] different software cycles and, you know, [43:26] eras of that we're that we've seen. So, [43:29] anyway, Joe and Andy, thank you so much [43:32] for joining us. [43:33] This was great and excited to have you [43:35] guys back on. [43:36] >> Yeah. [43:37] >> Yeah. [43:37] >> God, the goat. Thanks for having us. [43:39] >> Thanks, Andy. [43:39] >> Thanks. It was fantastic.