跳到主内容
@wquguru
精选88SaaStr 博客(RSS)产品与增长

AI销售初创的两大陷阱与六大自动化GTM动作

Sam Blond Met 200+ AI Sales Startups at Founders Fund, Then Built Monaco: Why Most Get It Wrong, Where AI Beats Your Reps, and the $15,000 Plane

原文
发到 X
推荐理由

直接给出了AI时代GTM的具体分工判据:6个可全量自动化的动作清单与2个必须保留人类的场景,附带避坑指南,创业者今天就能对照优化流程。

Sam Blond scaled sales at Brex, then joined Founders Fund as a partner. Because his background is sales, a large share of the AI GTM startups pitching Founders Fund ended up in his meetings. He met with hundreds of them. Then he co-founded Monaco, an AI-native revenue platform for startups, and raised a Series A and a Series B inside about four months.

Sam Blond 在 Brex 负责扩大销售规模,随后加入 Founders Fund 担任合伙人。由于他的背景是销售,大量向 Founders Fund 推介的 AI GTM(面向客户增长)初创公司最终都安排与他会面。他接触了数百家公司。之后,他联合创立了 Monaco——一个专为初创企业打造的 AI 原生营收平台,并在大约四个月内完成了 A 轮和 B 轮融资。

At SaaStr AI, Sam covered what he learned from both sides of the table: the two traps most of those 200+ startups fell into, the GTM work where AI already beats humans, the smaller set where it doesn’t, and the monthly process Monaco uses to produce brand moments like the plane towing a Monaco banner over the event.

在 SaaStr AI 大会上,Sam 分享了他从双方视角学到的经验:那 200 多家初创公司中最常陷入的两个陷阱、AI 已经超越人类的 GTM 工作领域、AI 尚未超越人类的那部分较小领域,以及 Monaco 每月用于打造品牌时刻的流程,例如在活动上空用飞机拖曳印有 Monaco 标志的横幅。

1. Most AI GTM Startups He Saw Were Either a Feature or a Solution in Search of a Problem

1. 他看到的绝大多数 AI GTM 初创公司要么是某个功能,要么是为寻找问题而存在的解决方案

Almost every AI sales startup Sam met at Founders Fund fit into one of two buckets.

Sam 在 Founders Fund 遇到的几乎所有 AI 销售初创公司都可以归入两类之一。

Bucket one: this should be a feature, not a company. His hypothetical example was a startup that takes the content from a call recording and writes it into a field or set of fields in your CRM. The workflow makes sense, because the rep doesn’t have to do it. The company doesn’t. That capability belongs inside the call recorder or the CRM. Sam thinks the late B2B era and the start of the AI era produced far too many point solutions this narrow.

第一类:这应该是一个功能,而不是一家公司。他举了一个假设的例子:一家初创公司提取通话录音中的内容,并将其写入你 CRM(客户关系管理)系统中的某个字段或一组字段。这个工作流程看似合理,因为销售代表无需亲自操作。但作为一家独立公司则不合理。这种能力本应内置于通话录音工具或 CRM 系统中。Sam 认为,晚期 B2B 时代与 AI 时代的开端催生了太多如此狭窄的点解决方案。

Bucket two: a solution in search of a problem. This was a common phrase at Founders Fund. His example: a startup that sends every rep an AI-generated video at the end of each week recapping what they did well, what they did poorly, and how to improve. It sounds interesting, but nobody he knew was asking for it. The usual founder response is the Henry Ford line: customers don’t ask for it because it doesn’t exist yet. To Sam, that usually means the founder skipped the customer research.

第二类:为寻找问题而存在的解决方案。这是 Founders Fund 中常见的说法。他的例子是:一家初创公司在每周结束时向每位销售代表发送一段由 AI 生成的视频,回顾他们做得好的地方、做得不好的地方以及如何改进。听起来很有趣,但他认识的人中没有人要求这样的产品。创始人通常的回应是引用亨利·福特的名言:“客户不会要求它,因为它还不存在。”但在 Sam 看来,这通常意味着创始人跳过了客户调研环节。

Most of the companies he saw landed in one of those two buckets. Before Monaco, he and his co-founders ran hundreds of customer conversations to make sure they weren’t building a third version of either.

他看到的大多数公司都落入了这两类之中。在创立 Monaco 之前,他和联合创始人进行了数百次客户对话,以确保自己不是在构建上述两种情况的第三种变体。

2. Pick the One Category You’re Qualified to Found In

2. 选择你有资格创立的那一类公司

Sam is a non-technical founder who spent his career in sales and GTM. By his own account, that qualifies him to found exactly one type of technology company: a sales or GTM technology company.

Sam 是一位非技术背景的创始人,职业生涯一直从事销售和 GTM 工作。据他自己所述,这使他具备创立唯一一种类型科技公司的资质:即销售或 GTM 技术公司。

He uses Brex and Ramp as examples:

他以 Brex 和 Ramp 为例:

  • Brex: Henrique Dubugras and Pedro Franceschi had already built Pagar.me in Brazil, essentially a Brazilian Stripe. They came to the US, got into YC, and started a virtual reality company with no background in VR. While in the batch, they noticed none of their batchmates could get corporate cards. They pivoted into fintech, their second company in the category.
  • Ramp: The founders had previously sold a fintech company to Capital One.
  • Brex:Henrique Dubugras 和 Pedro Franceschi 此前在巴西创建了 Pagar.me,本质上是一家巴西版的 Stripe。他们来到美国,加入 YC,并在毫无 VR 背景的情况下创办了一家虚拟现实公司。在 YC 期间,他们注意到同期的创业者都无法获得企业信用卡。于是他们转型进入金融科技领域,这是他们在该类别中创办的第二家公司。
  • Ramp:这两位创始人此前曾将一家金融科技公司出售给 Capital One。

The largest outcomes Sam has seen came from founders with deep experience in the category before they started. His three company-building rules:

Sam 所见到的最大成果,来自于那些在创业前就在该领域拥有深厚经验的创始人。他总结了三条公司建设法则:

  • Build in a space you understand better than anyone.
  • Find the problem through hundreds of customer conversations.
  • Time it to a platform shift.
  • 在你比任何人都更了解的领域进行构建。
  • 通过数百次客户对话来发现问题。
  • 把握平台变革的时机。

3. The Customer Pain at 12 People Was Tool Overload

3. 12 人规模时的客户痛点是工具过载

In those hundreds of conversations, the recurring complaint was too many tools.

在那数百次对话中,反复出现的抱怨是工具太多。

A typical early-stage founder Sam talked to had 12 people and one or two reps, and had to pick:

Sam 接触过的典型早期阶段创始人拥有 12 名员工和一到两名销售代表,并且必须做出选择:

  • HubSpot, Attio, or Salesforce for the database
  • Apollo or ZoomInfo for the data
  • Outreach or Salesloft for sequencing
  • A call recorder on top of all of it
  • 使用 HubSpot、Attio 还是 Salesforce 作为数据库
  • 使用 Apollo 还是 ZoomInfo 获取数据
  • 使用 Outreach 还是 Salesloft 进行序列营销
  • 以及在其之上叠加一个通话录音工具

They didn’t know what worked or how to use any of it. Sam had dealt with the same setup building GTM teams at startups.

他们不知道什么有效,也不知道如何使用任何这些工具。Sam 在初创公司组建 GTM(Go-To-Market)团队时也曾经历过同样的配置。

Point solutions existed because building everything took too long. One company couldn’t afford to build a CRM, a data provider, a sequencer, and a call recorder, so best-of-breed won the last cycle.

之所以存在单点解决方案,是因为构建一切所需的时间太长。一家公司无法负担同时构建 CRM、数据提供商、序列营销工具和通话录音工具的成本,因此“最佳单品”在上一个周期胜出。

AI changed the cost of building. Monaco has built a call recorder, an email sequencer, a database provider, and more in under 18 months. A small team can now ship what used to take large teams years, which makes the all-in-one platform viable again.

AI 改变了构建成本。Monaco 在不到 18 个月内构建了通话录音、邮件序列、数据库提供商等功能。现在一个小团队就能交付过去需要大团队花费数年才能完成的产品,这使得一体化平台再次变得可行。

4. Salesforce Beat Siebel on a Platform Shift, and AI Is the Next One

4. Salesforce 在平台变革中击败了 Siebel,而 AI 是下一次变革

Sam’s timing argument mirrors the last big shift in GTM software.

Sam 的时机论点与 GTM 软件领域的上一次重大变革相呼应。

Salesforce became the market leader on the move from on-prem (Siebel) to cloud. Salesforce, HubSpot, and Attio were all architected to be reactive. The user clicks, types, and updates, and the system responds.

Salesforce 在从本地部署(Siebel)向云端迁移的过程中成为市场领导者。Salesforce、HubSpot 和 Attio 在设计上均为响应式架构:用户点击、输入并更新,系统随之响应。

Monaco was architected with agents doing the work first. Before a user opens the platform, agents have already acted on their behalf, and the UI shows what they did.

Monaco 的设计架构则是让智能体(agents)先执行工作。在用户打开平台之前,智能体已经代表用户采取了行动,界面展示的是它们所做的事情。

Sam doesn’t think a pre-AI architecture can pivot to AI-native. Monaco started from scratch, and he doesn’t know of another way to do it. He expects every large enterprise software category to produce new leaders in this cycle.

Sam 认为,非 AI 原生架构无法转型为 AI 原生。Monaco 是从零开始构建的,而且他不知道还有其他实现方式。他预计在这个周期中,每个大型企业软件类别都会涌现出新的领导者。

5. Six GTM Jobs Where AI Is Already Better Than Your Team

5. AI 已经在六个 GTM 岗位上表现得比你的团队更好

On these, Sam’s claim is that AI outperforms people, and teams should spend close to zero human time on them.

在这些方面,Sam 的观点是 AI 的表现优于人类,团队应在这些任务上投入接近零的人工时间。

Building your TAM. The old way was people crawling the web, running and filtering lists, exporting to a database, and prospecting on LinkedIn for companies that fit. AI does this better.

构建你的 TAM(可触达市场)。旧的方式是人手爬取网页、运行并筛选列表、导出到数据库,然后在 LinkedIn 上针对符合条件的公司进行潜在客户挖掘。AI 能做得更好。

Scoring your TAM. Monaco sells to startups, roughly 20,000 to 30,000 companies, and they aren’t all equally good fits. Location, headcount, B2B vs. B2C, and venture-backed vs. not all feed a prioritization model. A RevOps person used to build that in spreadsheets. Now you write the scoring logic in a prompt and apply it across the whole database in a day or two.

对 TAM 进行评分。Monaco 的客户群体是初创公司,大约 2 万到 3 万家企业,但它们并非都同样契合。地理位置、员工人数、B2B 与 B2C、以及是否获得风投等因素都会影响优先级模型。以前 RevOps(营收运营)人员需要在电子表格中构建这个模型。现在你只需在提示词中编写评分逻辑,即可在一天或两天内应用于整个数据库。

Enrichment. Agents enrich every account in the TAM automatically.

信息丰富化。智能体自动为 TAM 中的每个账户补充信息。

Signals. At Brex, reps read a funding newsletter every morning and manually emailed each company that had raised. Today, AI catches funding rounds, relevant job postings, new hires, and website visits in real time and routes them into workflows.

信号捕捉。在 Brex,销售代表每天早上阅读融资新闻通讯,并手动向已获融资的公司发送邮件。如今,AI 能实时捕捉融资轮次、相关职位发布、新入职员工和网站访问等信号,并将它们路由到工作流中。

Automated multi-channel outbound. Outbound should run off the scored TAM, use the signals in the messaging, and go out over email and LinkedIn at minimum, fully automated. Humans monitor it and take over when someone replies, handling objections and booking meetings.

自动化多渠道外联。外联活动应基于评分后的 TAM 展开,在消息中使用信号捕捉到的信息,并通过电子邮件和 LinkedIn 至少两个渠道全自动发送。人工负责监控,并在有人回复时接管处理异议和预约会议。

Pipeline maintenance. Opportunity creation, stage changes, CRM field updates, and follow-up tasks should all come from the calls and emails themselves. Nobody should be clicking “+ new opportunity” or moving deals from stage two to stage three by hand.

管道维护。机会创建、阶段变更、CRM 字段更新和后续任务都应直接来自通话和邮件本身。任何人都不应手动点击“新建机会”或将交易从第二阶段移动到第三阶段。

His Brex example on that last one: as a rep, he’d get off a call and struggle to remember which corporate card the prospect said they used. The agent with access to the recording has the answer. On this work, AI does a better job than the rep.

他关于最后一点的 Brex 案例:作为销售代表,他在结束通话后往往难以记住潜在客户说他们使用的是哪家公司的企业卡。而能够访问录音的智能体则能提供答案。在这项工作上,AI 比销售代表做得更好。

Still prospecting on LinkedIn by hand?

还在手动在 LinkedIn 上进行潜在客户挖掘吗?

"If you're spending time on things that AI is better than humans at, you are likely behind your competitors," says Sam Blond (@samdblond) of Monaco (@monacogtm).

"如果你把时间花在 AI 比人类更擅长的事情上,你很可能落后于竞争对手," Monaco (@monacogtm) 的 Sam Blond (@samdblond) 说道。

Founders belong on customer calls. AI does the rest. pic.twitter.com/xYHXoIx9Fc

创始人应参与客户通话。AI 负责其余工作。pic.twitter.com/xYHXoIx9Fc

— SaaStr.ai (@saastr) November 18, 2026

—— SaaStr.ai (@saastr) 2026年11月18日

6. Two GTM Jobs Where Humans Still Win

6. 人类依然胜出的两项 GTM(进入市场)工作

Meeting customers and building relationships. You could imagine buyers talking to an avatar instead of a founder or rep. For upmarket and enterprise software, buyers don’t want that. Nobody buys Monaco or NetSuite without talking through their business needs and the implementation. Buyers of complex products want to be sold to.

与客户会面并建立关系。你可以想象买家与代表或创始人的化身对话。但对于高端和企业级软件,买家并不想要这样。没有人会在不深入讨论业务需求和实施细节的情况下购买 Monaco 或 NetSuite。复杂产品的买家希望被销售。

Creative or operationally complex campaigns. People came up with the Monaco banner plane over SaaStr and booked it with the company that flies the planes. People made the video for the Series B announcement. A video is a creative campaign and a banner plane is an operationally complex one. Humans still do both better, and those campaigns matter because every competitor has the same automated outbound.

创意性或运营上复杂的营销活动。有人策划了萨夫拉斯(SaaStr)大会期间的摩纳哥横幅飞机,并向运营该飞机的公司进行了预订。还有人制作了B轮融资公告的视频。视频属于创意型活动,而横幅飞机则属于运营复杂度高的活动。人类在这两方面依然做得更好,而且这些活动至关重要,因为每个竞争对手都拥有相同的自动化外展能力。

7. The Most Common Mistake: “Outsource GTM to AI and Just Write Code”

7. 最常见的错误:“将市场进入(GTM)外包给AI,只写代码”

Sam pushed hardest on this one. Technical, product-oriented founders see platforms like Monaco and decide they can hand off GTM entirely and spend all their time on product.

山姆对此最为力推。技术导向、产品导向的创始人看到像摩纳哥这样的平台后,便决定可以完全将市场进入工作外包,并将所有时间投入到产品开发中。

Sam argues for the opposite:

山姆主张相反的观点:

更进一步:量化金融体系

看懂新闻只是起点——沿量化金融路径,把它变成能交付的工程能力

进入量化体系 →

相似阅读

关联信息,但可能不是同一事件