拉动力法则：为何客户需求比痛点更重要

今天的嘉宾是 Rob Snyder，《拉力的力量》一书的作者，他将创业失败转化为寻找真实客户需求的框架。

Today's guest is Rob Snyder, author of The Power of Pull, who turned startup failure into a framework for finding real customer demand.

我从麦肯锡和哈佛商学院出来，以为自己懂商业运作，以为自己知道自己在做什么。我们做了大量客户研究、访谈、实验，招募了设计合作伙伴。所以一切看起来都准备就绪，应该会成功。我们甚至从风投那里融到了钱。但后来它没有成功。我们试图销售，却没人愿意买，我们也不知道为什么。

I came out of McKinsey and Harvard Business School thinking I knew how business worked, thinking I knew what I was doing. We had done a bunch of customer research, interviews, run experiments, recruited design partners. And so everything seemed like it was lined up to succeed. We had even raised money from VCs. And then it didn't work. Then we tried to sell it and nobody would buy it and we couldn't figure out why.

然后偶然间，我们另一个产品从零做到了 400 万美元的年度经常性收入。我完全不知道它为什么在成功时成功，也不明白它为什么在失败时失败。

And then randomly, we went zero to 4 million ARR on a different product. I had no idea why it worked when it worked. And then I didn't understand why it didn't work when it didn't work.

与潜在用户一起做设计评审之类的。你为什么认为那也没有用？

With the potential users, doing design reviews and things like that. Why did you think that also didn't work?

我后来意识到，我们做所有这些客户研究时找错了东西。我们寻找痛点和问题，但那并不是促使人们购买的原因。在我看来，有一种东西叫做需求，那才是我们要找的。但我们误解了需求是什么。它不是对产品的渴望。需求存在于世界中，是某人试图做某事，就像他们待办事项清单上的一个项目。我们要找的是那些现在正在优先处理待办事项中某件事的人，而他们现有的选择在生活中并不好。所以这就是拉力框架的用武之地。它描述了必须存在的四个不同组成部分，如果这些条件同时具备，人们就会把产品从你手中抢走。如果他们不买你的产品，那才奇怪。如果这四个条件没有同时具备，他们买了你的产品，那才奇怪。所以这四个条件是

What I've since realized is we do all of this customer research looking for the wrong things. We look for pain points and problems, but it is not what's going to cause people to buy. In my mind is that there is a thing that's called demand, which is what we're looking for. But we misinterpret what demand is. It is not desire for a product. Demand is out there in the world. It's somebody trying to do something. It's like a project on their to-do list. What we're looking for is somebody who is prioritizing something on their to-do list right now, where their existing options for doing that thing are not good in their life. And so that's where the pull framework comes in. It describes four different components that have to exist, where if these things are present, somebody will pull the product out of your hands. It would be weird if they didn't buy your product. And if these four things aren't present simultaneously, it would be weird if they bought your product. And so the four things are

嘿，暂停一下。我做 TechLead Journal 的目标很简单：向科技界最优秀的人学习，让我们共同成长。如果这引起了你的共鸣，请点击订阅关注频道。这是支持节目、帮助我们继续为你带来优秀嘉宾和见解的最大方式。感谢你的到来，让我们回到节目。

Hey, quick pause. My goal with TechLead Journal is simple. Learn from the best in tech so we can all grow together. If this resonates with you, hit subscribe to follow the channel. It's the biggest way for you to support the show and help us keep bringing great guests and insights to you. Thanks for being here and let's get back to it.

大家好。欢迎回到 Tech Lead Journal 播客的另一期新节目。今天这期节目主要涵盖产品管理和构建成功创业公司。今天和我一起的是 Rob Snyder。他最近出版了一本书，书名是《拉力的力量：关于客户需求你需要知道的事，以构建成功创业公司》。我觉得这本书的核心论点与过去可能见过的通常的产品管理书籍有些不同。所以我想让 Rob 实际告诉我们更多关于拉力的概念。似乎是个双关语“pull”。但没错，让他多谈谈。欢迎来到节目，Rob。

Hello everyone. Welcome back to another new episode of the Tech Lead Journal podcast. Today we have an episode mostly covering about product management and building a successful startup. I have Rob Snyder today with me in the episode. He just recently published this book titled The Power of Pull. What you need to know about customer demand to build a successful startup. So I think I find this book the core thesis is something a little bit different than what the usual product management book that maybe have seen in the past. So I'd like Rob to actually tell us more about the concept of pull. Seems to be a pun intended pull. But yeah, let him speak more about it. So welcome to the show Rob.

非常感谢邀请我。

Thanks so much for having me.

Rob，也许一开始先告诉我们更多背景。你为什么写这本书？因为我知道你背后也有故事。你创业公司的失败故事促使你写了这本书。

Like Rob, maybe in the very beginning tell us a little bit more background. Why did you come up with this book? Because I know you have a story behind it as well. And your kind of like failure story with your startups and that led you to writing this book.

是的，完全正确。我从麦肯锡和哈佛商学院出来，以为自己懂商业运作，以为自己知道自己在做什么。显然，从麦肯锡和哈佛商学院出来后下一件酷事就是创建一家创业公司，你知道，成为科技亿万富翁。所以我想，酷，那显然是下一个要勾选的框，所以我要去做。我们做了所有应该做的事。我们做了大量客户研究，做了很多访谈，做了实验，招募了设计合作伙伴。我们证明了很多价值，他们每天都在使用我们的产品。所以一切看起来都准备就绪，应该会成功。我们甚至从风投那里融到了钱，展示了所有吸引力、所有研究和我们做的所有事情。但后来它没有成功。我们试图销售，却没人愿意买，我们也不知道为什么。就这样持续了我人生中最糟糕的两年。就像我每天都在用头撞墙，太糟糕了。然后偶然间，我们另一个产品从零做到了 400 万美元的年度经常性收入。一个客户打电话给我说：“嘿，我不需要你想卖给我的东西。你能做这个其他东西吗？”你可能会想：“好吧，酷，你搞定了。”不，我没有搞定。我完全不知道它为什么在成功时成功，也不明白它为什么在失败时失败。所以从那以后我一直在试图弄清楚，这本书是我试图将过去——我不知道，十年——试图弄清楚这个问题所学到的东西形式化的尝试。

Yeah, totally. So I came out of McKinsey and Harvard Business School thinking I knew how business worked, thinking I knew what I was doing. And obviously the next cool thing to do after McKinsey and HBS was to build a startup and and you know become a tech billionaire. And so I was like cool, that's that's clearly the next box to check and so I'm going to do that. And we did kind of like everything you're supposed to do. We had done a bunch of customer research. We'd done cut bunch of interviews. We had done run experiments. We had recruited design partners. We had proved a ton of value. They were using our product every day. And so everything seemed like it was lined up to succeed. We had even raised money from VCs showing all the kind of like traction and all the research and all the stuff that we had done. And then it didn't work. Like then we tried to sell it and nobody would buy it and we couldn't figure out why. And so that went on for two of the worst years of my life. It was just like I was hitting my head against the wall every single day, and it was terrible. And then, randomly, we went from zero to four million ARR like on a different product. A customer called me and was like, "Hey, I don't need what you're trying to sell me. Like, could you could you do this other thing?" And you would think, "Okay, cool. Well, you figured it out." No, I didn't figure it out. I had no idea why it worked when it worked, and then I didn't understand why it didn't work when it didn't work. And so, I have just been kind of trying to figure that out since, and the book is a is my attempt to formalize what I have learned in the past, I don't know, decade trying to figure this out.

是的，所以当我读你书的引言时，你也讲了那个故事。我觉得非常吸引人，因为我确信许多创业公司创始人都有共鸣，而且我认为许多创业公司创始人，即使最终成功了，对吧？他们最初的想法并不是最终让他们成功的那个，对吧？所以，我认为这是一个非常经典的问题。但是，你提到你也遵循了所有经过验证的方法论，现在许多创业公司也在做，对吧？你知道，做研究、询问、也许访谈潜在用户，对吧？做设计评审之类的。你为什么认为那也没有用，即使我们现在已经大量使用这种方法论？

Yeah, so when I read the intro of your book, you tell told the story as well. I find it quite fascinating because I'm sure it is quite relatable for many startups founder, and I think many startup founders, even though they succeeded in the end, right? They didn't start with the same idea. With what made them successful in the end, right? So, I think this is quite classic a problem. But, you mentioned that you also kind of like followed all the true tested methodology of how many startups these days are also doing, right? You know, doing research, asking, maybe interviewing potential users, right? Doing design review and things like that. Why did you think that also didn't work, even though we have kind of like prayed, you know, using this methodology a lot in these days?

是的，所以基本上我花了很长时间才弄清楚为什么那些方法没有用。起初，我以为我只是无能，就像：“嗯，我显然很不擅长这个。”但后来我看到我的一群朋友和世界各地的一群其他创始人尝试做这些事情，对他们也没有用。所以我想：“好吧，也许我在这里不是白痴。”所以我后来意识到的是，我们做所有这些客户研究时找错了东西。我们寻找痛点和问题，我们寻找它们来说：“哦哇，这会非常有价值。”我们寻找签约设计合作伙伴，我们在所有这些客户研究中主动避免销售。所有这些都是错的。就像所有寻找痛点和问题听起来是对的事情，但它并不是促使人们购买的原因。所以你最终做了所有这些工作，严谨的工作。我见过数百次客户访谈寻找错误的东西，然后你找到了它，你基于你找到了正确东西的理论建立了一个业务或产品。

Yeah, so it's it's basically it's taken me a long time to figure out why those things didn't work. At first, I thought I was just incompetent, and it was just it was like, "Well, I'm clearly very bad at this." Um, but then I've seen a bunch of my friends and a bunch of other founders from all across the world try to do these things, and they also don't work for them. And so, I was like, "Okay, maybe I'm not the idiot here." And so, what I what I've since realized is that we do all of this customer research looking for the wrong things. We look for pain points and problems and we look for them to say, "Oh wow, this would be super valuable." And we look to sign on design partners and we actively avoid sales in all of this customer research. All of that is wrong. Like all of that looking for pain points and problems sounds like the right thing, but it is not what's going to cause people to buy. And so you wind up doing all of this work, rigorous work. I've seen hundreds of customer interviews looking for the wrong thing and then you find it and you build a business or a product based on the theory that you found the right thing.

所以所有这些都指向一点：如果你按照我们都被教过的这种学术式研究技巧去做，成功了反而会显得奇怪。

And so all of this leads to it would be weird if you succeeded following these kind of like academic style research techniques that we've all been taught.

是的，当你提到那个时，我脑海中把它和比如“待完成的工作”联系起来了，对吧？我觉得那是个挺有名的概念：你找到某个特定用户的画像，弄清楚他们要完成的工作是什么，他们有哪些痛点，然后你试着用你的产品去解决。或者另一件事是，如果你也知道亨利·福特的那句名言，对吧？如果你问用户想要什么，他们只想要一匹更快的马，而不是汽车，对吧？所以你对这些怎么看？因为这个概念似乎已经深深植根于产品管理领域了。

Yeah, so when you mentioned about that, I think in my mind I associated this with like for example jobs to be done, right? So I think it was quite kind of like a quite famous like you find the persona of a particular user, find out what the what are the jobs to be done, what are the pain points that they have and you try to solve that with your product. Or the other thing is like if you if you also know about Henry Ford's quote, right? So if you ask the users what they want, they just want a faster horse and not a car, right? So what is your take about all these because it seems like it is kind of ingrained in the product management world about this concept.

是的，所以这个有点难展开。我们得从头开始：我们必须有一个关于是什么导致人们购买产品的理论。具体来说，这个理论必须说明，当市场上已有大量产品时，是什么会导致人们购买一个新产品。所以这就有点难了，也就是说外面有很多噪音。比如你提到的亨利·福特那句话。在我看来，归根结底是有一个叫做“需求”的东西，那才是我们要找的。如果某人有某种需求，他们就会购买创业公司的产品。但我们对需求的误解是，我们认为需求意味着想要一个产品，或者有痛点导致你需要一个产品之类的。如果你认为需求意味着想要一个产品，那么你就会说：“好吧，我在做一个新东西，人们怎么会知道他们想要它呢？”于是你就会去寻找痛点之类的东西，说：“好吧，他们有痛点，所以也许他们会买，也许他们会传播。也许我能说服他们。也许我能通过创造品类来取得成功。”对吧？所以这是一种默认心态，或者另一种说法是，如果他们有一个问题，而且问题足够大，他们就会想买我们的产品。但事实证明，人们有很多问题却什么都不做。你和我，对吧？我确信我们的生活就是一系列痛点和问题，而对于大多数问题，我们什么都不做，也不会去做。所以，问题、痛点、想要产品之间是有区别的。那并不完全是需求。需求更像是试图做某事，采取行动，就像我待办清单上正在优先处理的一个项目。这确实和“待完成的工作”有点押韵，但并不完全相同。这里面有一些细微差别。

Yeah, so so this one this one's kind of hard to unpack. So where all start is we have to have a theory about what causes somebody to buy a product. And specifically that theory has to state what's going to cause somebody to buy a new product when there are plenty of existing products on the market. And so, this gets kind of hard, which is to say there's a lot of noise out there. There's the Henry Ford quote like you mentioned. What it comes back to in my mind is that there is a thing that's that's called demand, which is what we're looking for. If somebody has some kind of demand, then they will buy a startup's product. But we misinterpret what demand is. We think that demand means wanting a product. Or having pain points for a that lead you to a product or some something like that. And if you think that demand means wanting a product, well, then you're going to say, "Well, I'm building something new. How could people know they want it?" And so, you're going to kind of look for pain points and stuff like that to say, "Well, they have pain points and so maybe therefore they will buy it and maybe they'll evangelize. Maybe I'll convince them. Maybe I'll, you know, category create my way to success here." Right? And so, that's kind of the the default mindset or or it's on the other side of it of if they have a problem and that problem is big enough, they will want to buy our product. It turns out there are plenty of problems that people have that they do nothing about. You and I, right? I'm sure our lives are a series of pain points and problems and for most of them we don't do anything and we're not going to do anything. It does it's not So, there's a difference there between problems, pain points, wanting a product. It's that's not quite what demand is. Demand looks more like trying to do something, taking action like a project that I am prioritizing on my to-do list. And that does kind of rhyme a little bit with jobs to be done, but it's not exactly the same. It's a there's there's some nuances there.

是的。我觉得你提到了很多关于需求的内容。那么，也许让我们试着展开你的方法，对吧？拉力的力量，对吧？所以，我让你来解释什么是拉力，这样我们就能更多地探讨需求这件事。

Yeah. So, I think you I mentioned a lot about demand. So, maybe let's try to unpack your method, right? The pull the power of pull, right? So, I'll let you explain what is pull so that we can cover a lot more about this demand thing.

是的，所以核心是，需求是一个独立的东西。它不是对产品的渴望。需求存在于世界中。它是某人试图做某事。就像他们待办清单上的一个项目。现在，某人可能有一个正在优先处理的待办项目，但并不购买你的产品，因为所有现有产品都足够好，对吧？他们有现有产品。他们有目前正在使用的东西，而且足够好。所以，我们要找的是那些正在优先处理待办清单上某事的人，而他们现有的选项在他们看来不够好。这就是拉力框架的用武之地。它描述了必须存在的四个不同部分，四个必须存在的组件，如果这些条件都具备，人们就会把产品从你手中抢走。如果他们不买你的产品，那才奇怪。如果这四个条件没有同时具备，他们买了你的产品，那才奇怪。所以，这四个条件是：他们待办清单上有一个项目。它是不可避免的。U 是第二个。P 是项目。U 是不可避免的。第一个 L 是他们有一系列完成该项目的选项。第四个，也就是第二个 L，是这些选项有局限性，阻止他们完成该项目。所以，所有这四个条件必须同时存在。项目、不可避免、选项列表有局限性阻止他们。他们试图做某事，但用现有工具、可用工具、方法、选项做不到。如果是这样，他们就会从你手中抢走能让他们突破障碍的东西。

Yeah, so so the core thing is that demand is a standalone thing. It is not desire for a product. Demand is out there in the world. It's somebody trying to do something. It's like a project on their to-do list. Now, somebody can have a project on their to-do list that they're prioritizing right now and not buy your product because all the existing products are good enough, right? They have existing products. They have something they currently use that's good enough. So, what we're looking for is somebody who is prioritizing something on their to-do list right now where their existing options for doing that thing are not good enough in their mind. And so, that's where the pull framework comes in. It describes four different pieces of that have to exist four different components that have to exist where if these things are present, somebody will pull the product out of your hands. It would be weird if they didn't buy your product. And if these four things aren't present simultaneously, it would be weird if they bought your product. And so, the four things are there's a project on their to-do list. It's unavoidable. U This is the second one. P is project. U is unavoidable. First L is they have a list of options for getting that project done. And the the fourth or the second L is those those options have limitations that prevent them from getting that project done. And so, all four of those things must be present simultaneously. Project, unavoidable, list of options will have limitation preventing them. They're trying to do something they can't do with their existing tools, available tools, methods, options. If that's the case, they will rip something out of your hands that just unblocks them.

是的，我喜欢你对“拉力”这个词的玩法，突然就像是有人在拉你的产品，对吧？可以说那就是需求，对吧？所以，嗯，有了这些解释，肯定说得通，对吧？就像潜在客户待办清单上有一个当前高度优先的项目，对吧？嗯，它是不可避免的，他们有一些如何解决的选项列表，但呃，他们认为不够好。那些选项有局限性，这就是为什么他们在寻找你也许正在创造的新产品。那么，也许在这四个条件中，在如今典型的创业世界里，当人们构建产品时，你认为最缺失的是哪一块？

Yeah, I like the way you play with this word pull and suddenly you it's like someone pulling your products, right? That's the demand, so to speak, right? So, um definitely makes sense with all this explanation, right? That's like the potential customer is having a project in their to-do list that is a high priority at the moment, right? Um they is unavoidable and they have some list of options how to solve it, but uh they think it's not good enough. They have limitations to those options and that's why they are looking for this new product that maybe you're creating. So, maybe out of these four in a typical startup world these days when people build product, what do you think are the most missing piece?

哦，天哪。呃，所有都缺失。所以这不是人们使用的框架，也不是人们倾向于思考的方式。我们倾向于从我们的产品或想法出发，去思考潜在客户，然后说：“既然我的想法是制造业的智能体式工作流之类的，对吧？制造业中有哪些痛点或问题我们可以关联、可以谈论、可以传播？有哪些用例？”这些都是从我们的产品想法或产品出发，推理到市场的方式。而拉力是一种完全不同的思维方式，因为你必须完全忘记你的产品或想法存在，你只是在观察自然栖息地中的购买者，观察他们的生活，并意识到大多数时候他们过得很好。他们不需要帮助。他们没有卡住。你在寻找那些点，再次假装你没有产品，寻找他们卡住的地方，他们试图做某事但被阻碍的地方。所以，这种思维方式需要创始人做最难的事，那就是完全假装你的产品想法和你自己不存在，在冷光下看清人们在哪里卡住，然后围绕那个进行设计。完全反直觉，它改变了你处理产品开发的方式、处理销售的方式，以及从那个视角处理一切的方式。

Oh, gosh. Uh all of them. So, so this is not a framework that people use and it's not a way that people tend to think. We tend to think from our product or from our idea to the potential customer and say, "So, given that my idea is agentic workflows for manufacturing or whatever it is, right? What are the pain points or problems in manufacturing that we can tie to, that we can talk about, that we can evangelize? What are the use cases?" Like, these are all ways to reason from our product idea or our product to the market. And pull is a totally different way of thinking because you have to entirely forget that your product or idea exists and you're just observing buyers in their natural habitat, going about their lives, and realizing that most of the time they're just fine. They don't need help. They're not stuck. And you're looking for the points, again, just pretending you don't have a product where they are stuck, where they're trying to do something and blocked. And so, this way of thinking requires the hardest thing a founder can possibly do, which is to totally pretend your product idea and you don't exist and see in the cold light of day the the where people are getting stuck and then designing around that. Totally counterintuitive, it changes how you approach product development, how you approach sales, how you approach kind of everything from that point of view.

但没错，我看到的普遍情况是，他们并不——鉴于我们看待世界的方式，这甚至不是一种我们可用的思考方法。

But yeah, the general thing I see is that they don't—this is not even a method of thinking that is available to us given how we see the world.

是的，我能理解为什么这不常见，因为现在很多人，尤其是当他们上产品管理课、创业课、学院之类的课程时，他们被教导要执行自己的想法、自己的愿景，对吧？不一定是去发现——观察人们如何做事以至于陷入困境。所以这意味着，什么样的研究、什么样的方法（如果存在的话），能真正帮助你观察到人们在日常中遇到的这些局限？

Yeah, I can see why it's not something that is usual because many people these days, especially when they go through product management class, startup class, academy or something like that, they're taught to execute their idea, their vision, right? Not necessarily finding—observe what people do in such a way that they are stuck. So which means that what kind of research, what kind of methodology, if it exists, to actually help you observe these limitations that people are having in their day-to-day?

是的，我会说有两种宽泛的方法论是我推荐的。第一种——其实我强烈不推荐客户访谈，尤其是——我主要关注从零到一的事情。所以你可以做访谈，但我发现它太容易出错，而且最后没有一种寻求真相的机制，你可以学到一堆不同的东西，但如果没有人被迫认真对待并购买或投入资金，你实际上不知道你是否学到了任何东西。所以我真正推荐的第一种方法就是销售。销售是一种非常非常有用的学习方式。它感觉像研究，因为在销售中，如果你做得对，你是在提问并得到答案。然后在最后，你通过他们是否尝试购买来看到你是否学到了东西。好的，酷。但如果你实际上没有假设——好的，某人在尝试做什么并卡住了？那么我推荐一种我称之为“去现场”的方法。这来自日本制造业。就是去工作发生的地方。去车间。去观察。做那份工作。亲自体验你试图销售的东西。所以一个例子是哈佛商学院的布拉德告诉我他的——他刚写了一份他可能有兴趣为其构建产品的行业列表，然后他花了一周时间跟随每个行业的小企业主。就像，“好的，酷。我先看看我是否喜欢这个行业。我要试着对这些人的直觉建立起来。然后我要试着找到他们尝试做某事并卡住的地方。”所以第一手经验远比二手研究、访谈和调查之类的东西好得多，保真度高得多。

Yeah, I'd say there are kind of like two broad methodologies that I recommend. The first—so I actually strongly do not recommend customer interviews or especially—I mostly focus on taking things zero to one. And so there are ways you could do interviews, but I just find it's way too error-prone and there's no kind of truth-seeking mechanism at the end where you can learn a bunch of different things, but if somebody isn't forced to get real and buy or put money down, you don't actually know if you learned anything or not. So the first methodology that I actually recommend is just selling. Sales is a really, really useful way to learn. And it can kind of feel like research because in sales, if you're doing it right, you're asking questions and you're getting answers. And then at the end of it, you see if you learned anything by whether they try to buy or not. Okay, cool. But if you don't actually have a hypothesis of—okay, what is somebody—where is somebody trying to do something and stuck? Then I recommend a method that I call go to the gemba. This is from Japanese manufacturing. It's like go to the place where the work is happening. Go to the shop floor. Go observe. Do the job. Get a first-hand experience yourself doing the thing that you're trying to sell into. So an example is HBS—Brad told me about his—he just wrote out a list of industries that he was interested in maybe building a product for, and he went and spent a week just riding along with a small business owner in each of the industries. And it was like, "Okay, cool. I'm just going to first see if I like this industry. I'm going to try to build an intuition for these people. And then I'm going to try to find the places where they're trying to do something and stuck." So first-hand is way, way better, way higher fidelity than second-hand research and interviews and surveys and all that nonsense.

是的。我喜欢你所说的销售本身就是研究的这个概念，对吧？所以我认为很多人可能避免把销售作为第一件事。他们想先构建一些东西。尽管现在他们考虑 MVP 或者不，比如构建一个 MVP 并尝试让客户或用户使用它。所以我认为销售显然是真正知道你是被拒绝还是被接受的最佳方式，对吧？所以我认为这绝对有道理。我有的一个反论点是，对于那些创新的或愿景性的产品呢？我的意思是想想太空行动，比如建造火箭去火星之类的。

Yeah. I like the concept when you say sales is a research by itself, right? So I think many people maybe avoid doing sales as a first thing. They want to build something first. Although these days they think in MVP or no, like build an MVP and try to have a customer or user using it. So I think selling is obviously the best way to actually know whether you get rejected or get accepted, right? So I think that definitely makes sense. One counter argument that I have is what about for products that are innovative or a visionary thing. I mean think of it like the space act things like building a rocket to Mars or something like that.

是的，是的。所以那些我实际上——我在一些深度科技公司帮了很多忙，是的，我们在构建一些创新的东西。我认为这里的误解是，因为一个产品需要很长时间开发，或者技术需要很长时间开发，或者它是一个疯狂的想法，因此我们现在无法找到需求。而这实际上恰恰相反，真相是：想法越疯狂，构建所需的时间越长，现在找到需求就越重要，因为它成功的可能性要小得多。所以我看到的是，这里有一种误解，就像“哦，因为我们在创造一个新类别，因为我们在构建一个完全创新的产品，人们不知道他们是否想要它。他们直到看到它才会知道他们是否想要它。”这不是真的。这完全不是真的，因为记住，那是将需求与想要供给混为一谈。那是亨利·福特关于更快的马的名言。就像，好吧，他们想要更快的交通，当一个概念摆在他们面前，一种更快的交通方式，他们甚至在它建成之前就从卖家手中抢走了它，这就是我在这些创新的深度——尤其是深度科技和未来事物中倾向于看到的，SpaceX 是一个很好的例子。你认为他们尝试过预售吗？是的，他们尝试过预售。他们发现政府将是他们的头号客户，他们一直与政府交谈。他们试图预售这个，政府给他们拨款并说酷。这里——他们在三个月内从成功的火箭发射到政府合同，就像三个月内数亿美元的政府合同。除非你进行了销售对话，就像“好的，这个政府合同的最后一个复选框是你必须将火箭送入轨道”，否则这不会发生。所以这些都是某种误称或误解。

Yeah, yeah. So those I actually have—I've helped out a lot at fellow deep tech places where yes, we're building some kind of innovative thing. I think the misconception here is that because a product is going to take a lot of time to develop or technology is going to take a lot of time to develop or it's a crazy idea, that because of that we can't find demand right now. And that's actually exactly the opposite of the truth, which is the crazier the idea, the longer it's going to take to build, the more important it is to find demand right now for it and because it's way less likely to work. And so what I've seen is that there's a kind of misconception here that's like oh because we're creating a new category, because we're building a totally innovative product, people don't know if they want it. They won't know if they want it until they see it. It's not true. That's not true at all because remember that's conflating demand with wanting supply. That's the Henry Ford quote about faster horses. Like okay, they wanted faster transportation and when a concept was put in front of them, a means of faster transportation, they pulled it out of the seller's hands even before it was built and that's what I tend to see in these innovative deep—especially deep tech and future things—and SpaceX is a really good example. Do you think they tried to pre-sell this? Yes, they tried to pre-sell. They found the government was going to be their number one customer and they were talking with the government all the time. They tried to pre-sell this and the government was giving them grants and saying cool. Here—they went from successful rocket launch to government contract in like three months, like hundreds of millions of dollar government contract in three months. That doesn't happen unless you had the sales conversation that was like okay the last check box on this government contract is that you have to get a rocket into orbit. So these are all kind of misnomers or misconceptions.

嗯，谢谢。我觉得这绝对有道理。对吧？所以我想现在我对你的世界很好奇。你说的这个需求和供给是什么意思？对吧？因为我认为你对这个需求和供给与典型的经济学定义有一些略微不同的视角。

Well, thanks. I'm thinking that definitely makes sense. Right? So I think now I'm quite curious in your world. What do you mean by this demand and supply? Right? Because I think you have some slightly different perspective about this demand and supply versus the typical economic definition of it.

是的，所以我们上过经济学入门课，对吧？有需求曲线、供给曲线，所以我实际上从之前在 Basecamp 的 Ryan Singer 那里得到了这种思维方式，他实际上用产品术语描述了这一点。他有一个非常棒的视频叫做“功能请求不是需求”，这就像它只是打破了我的大脑。所以这是一个非常有用的视频。我想它有大约 15,000 次观看，而我有大约 14,000 次。所以但是当我们——正如我一直在说的，对吧？我们倾向于将需求视为想要供给或愿意为供给付费，这意味着它是一个只存在于供给参照中的概念。如果它只存在于供给参照中，那么唯一的答案就是，好吧，让我们构建供给并为其创造需求。那行不通。而那些腾飞的产品根本不是那样表现的。

Yeah, so we took Econ 101 right? There's a demand curve, supply curve and so I actually got this way of thinking from Ryan Singer who was at Basecamp before and he was describing this actually in product terms. He's a really great video about this is called "feature requests are not demand" and this is like it just like broke my brain. So it's such a useful video. I think it has like 15,000 views and I have like 14,000 of those. So but when we—as I've been talking about right? We tend to think of demand as like wanting supply or willing to pay for supply which means it's a concept that only exists in reference to supply. And if it only exists in reference to supply, well then the only answer is well, let's build supply and create demand for it. That doesn't work. And the products that take off do not behave anything like that.

我们必须认识到，需求和供给是两回事。供给是我们的产品、服务，是我们提供的一切。需求是存在于外部世界的一个独立事物。它并不是想要供给；它可能导致购买供给，或者表现出想要供给的行为，但它是独立存在的。它就像我待办事项清单上的项目。我把需求想象成我大脑里的 Trello 看板或 Asana 看板。它是我试图去做的所有事情，与产品或功能无关。我正试图弄清楚如何从第一批客户那里收钱。我正试图弄清楚如何阻止 bug 进入生产环境。这些都是我试图去做的事情，然后我可能会跳到供给端说：‘好吧，有什么工具能帮我做到这一点？’或者‘我用什么方法来做这件事？’所以你必须把这两者分开，并认识到需求是供给的上游。它独立存在，但它决定了供给是否相关。一旦你明白了这一点——我称之为加入需求教——当你开始看到这一点时，一旦你开始看到需求就是存在于外部的东西，你就会去寻找它，去人们正在尝试并受阻的地方，然后先找到它，再围绕它设计产品。

What we have to realize is that demand and supply are separate things. Supply is our product, our service, all the things we offer. Demand is a separate thing out there in the world. It's not wanting supply; it might result in the purchase of supply, or behavior like wanting supply, but it's a separate thing. It's like projects on my to-do list. I think of demand as the Trello board or Asana board in my brain. It's all the stuff I'm trying to do that doesn't have anything to do with products or features. I'm trying to figure out how to collect money from my first customers. I'm trying to figure out how to stop bugs from getting to production. It's all these things I'm trying to do that I might then jump over to the supply side to say, 'Okay, what's the tool that can help me do that?' Or what's the method by which I'm going to do that? So you have to separate those two and realize that demand is upstream of supply. It exists independently, but it causes supply to be relevant or not. Once you have that—I call it joining the cult of demand when you start seeing this—once you start seeing that demand is just a thing out there, you go looking for it, where people are trying and stuck, and then you find that first and then design a product around that.

是的。所以我认为这是对我们通常理解的需求和供给概念的一个小变体，对吧？感谢你提到这一点。假设这是有道理的。那么你在书中强调的下一步是找到那个单一的‘绝对肯定’型客户，对吧？那么为什么找到这样一个客户如此重要，哪怕只是一个肯定的答复，而不是‘嗯，还行’？事实上，人们在谈论的是那些会真正喜欢你的产品的累积潜在用户。只有那样他们才认为值得去做。

Yeah. So I think that's a slight variant of what we are used to in terms of the concept of demand and supply, right? So I think thanks for mentioning that. So let's say this makes sense. So the next step for you in your book you emphasize a lot about finding the single 'hell yes' kind of customer, right? So why is it so important to find this even just a single yes, not like yeah, in fact, people are talking about cumulative potential users that will really love your product. Only then they think it's worth to do.

是的，我的意思是，我倾向于发现，我们，尤其是产品人，倾向于用抽象思维思考。以群体、人物角色、细分市场来思考。但商业只关乎人，个人购买和使用东西。从来没有——没有哪个细分市场、哪个人物角色会使用东西。事实证明，当你从抽象层面看时，很难弄清楚该做什么。所以你总是必须从一次一个客户开始。你总是一次卖一个客户。客户一次一个地孤立体验你的产品。所以我发现，当你开始思考——当我开始用这些抽象思维思考时，我总是创造我想要存在的抽象。我创造我想要存在的人物角色，而这个人恰好喜欢我的产品。所以对此的解药，因为显然那个人不存在。没人在乎我的产品。没人早上醒来说：‘你知道吗？Rob 的 B2B 软件今天早上对我来说真的很有趣。’不，那太疯狂了。所以我发现唯一合理的解药是一次一个客户。这就是为什么我喜欢从一对一销售开始，无论你是在构建消费产品还是 B2B 企业产品，你都是一次一个。这不仅仅是‘我能让某人购买吗？我能说服某人购买吗？我能强迫某人吗？’不，因为我想要的是一个绝对肯定的客户，一个把产品从我手中拉走的人，而不是我必须施加努力让他们购买。然后他们购买后，使用它就像上瘾一样。如果他们流失了，那会很奇怪。我的工作是先找出一个这样的客户，因为我无法找出 10 个。你必须先找出一个，然后弄清楚如何达到 10 个。所以我倾向于在初创公司发现——我已经做过很多次了——比如你得到前 10 个客户。其中一半你不得不推。另一半有点拉。有些人可能真的拉。然后有些人购买后恨你，绝对恨你。他们说：‘哦，这是最糟糕的东西。’有些人相当高兴。然后有些人说：‘哦天哪，这是最伟大的东西。’所以我发现你必须比较那些绝对肯定的客户和其他人，以弄清楚到底是什么创造了一个绝对肯定的客户。这里的需求是什么？我实际上应该针对谁？然后你对整个业务进行刷新，只专注于你的绝对肯定客户。

Yeah, I mean, so what I tend to find is that we, especially product people, tend to think in abstractions. In groups, in personas, in niches. Business only ever like people, individuals buy and use things. There is never—no niche, no persona ever uses a thing. And it turns out when you look at an abstract level, it's very hard to figure out what to do. And so you always have to start—you go one customer at a time. You always sell one customer at a time. Customers experience your product one at a time in isolation. And so what I've found is that when you start thinking—when I start thinking in these abstractions, I always just create the abstraction that I want to exist. I create the persona that I want to exist who happens to love my product. And so the antidote to that, because obviously that person doesn't exist. Nobody cares about my product. Nobody's waking up in the morning saying, 'You know what? Rob's B2B software is really interesting to me this morning.' Like no, that's insane. And so what I find the only reasonable antidote to that is going one customer at a time. And that's why I like starting with one-to-one sales, whether you're building a consumer product or a B2B enterprise product, you're one at a time. And it's not just can I get somebody to buy? Can I convince somebody to buy? Can I twist somebody's arm? No, because what I want is I want somebody who is a hell yes customer, who pulls the product out of my hands more than I have to exert effort to get them to buy. And then after they buy, they use it as if they are addicted to it. It would be weird if they churned. My job is to first figure out one of those because I can't figure out 10 of those. You first have to figure out one to then figure out how to get to 10. And so what I tend to find at startups—and I've done this a bunch of times—is like you get your first 10 customers. Half of them you had to push. The other half kind of pulled. Some people maybe really pulled. And then some people after they buy hate you, just absolutely hate you. They're like, 'Oh, this is the worst thing ever.' Some people are quite happy. And then some people are like, 'Oh my gosh, this is the greatest thing ever.' And so I found that you have to compare the people who are hell yes customers versus the others to figure out what actually creates a hell yes customer. What is the demand here? Who should I actually be targeting? And then you do a refresh of the entire business to just focus on your hell yes customers.

是的。专注于绝对肯定的客户。我认为这绝对是一个非常——我该怎么说？非常务实的方法，对吧？所以你真正找到的是你产品的狂热粉丝，而不是推动、哄骗人们实际购买你的产品，最终这不是一个绝对肯定的客户，对吧？因为这可能像是你对他们影响太大了。所以对于许多目前不以这种方式运作的初创公司创始人来说，他们仍然有点在传统产品管理的世界里，你称他们为生活在痛苦洞穴中。如何意识到他们实际上如此努力，但实际上他们的产品没有需求？

Yeah. Focus on the hell yes customer. I think definitely it's a very—how should I say? Very pragmatic, right? So you find really the raving fans of your product instead of pushing, cajoling people to actually buy your product, which in the end is not a hell yes customer, right? Because it's probably like you influenced them too much about it. So for many startup founders who are currently not operating in this way, they are still kind of in the traditional world of doing product management, you refer to them as living in this pain cave. How to realize that they're actually trying so hard, but actually there's no demand for their product?

是的，所以有一些非常常见的症状，我基本上每一个都经历过。所以如果你觉得被评判了，相信我，我也做过。最常见的症状是，如果你与潜在客户交谈，听到很多热情。然后他们不购买，而且不清楚为什么。这就是经典信号。或者如果他们问：‘哦，我们能免费试用吗？哦，我们能——也许我们应该试试。’然后就开始延长、延长、延长。这是最常见的事情。我要告诉你的是，它不会变得更好。并不是说如果他们不基本上立即尝试购买——我知道一家初创公司刚做了分析，他们在增长，他们意识到他们的试点——任何要求试点的人，这类客户的转化率是 0%。真的是零。而他们在增长。每个真正有需求的人都不需要试点就购买。所以你要寻找的是那些真正会从你手中拉走产品的人，那些会很快尝试购买的人。第一次通话，也许第二次通话，他们实际上在尝试购买。如果你没有这种情况，如果你听到的是‘哇，这太棒了。这正是——这是一个重要的产品。这正是行业需要的。’或者甚至‘哦，这每年能为我们节省一百万美元。’你没有听到的是‘我现在就需要这个。’

Yeah, so there are some very common symptoms that I have experienced basically every single one of these. So if you feel judged by this, trust me, I've done it. So the most common one is if you have conversations with potential customers and you hear a lot of enthusiasm. And then they don't buy, and it's not clear why. That is just like the classic signal. Or if they ask for, 'Oh, could we get a free pilot? Oh, could we get—maybe we should try.' And then it just kind of starts extending and extending and extending. That is the most common thing. And what I'll tell you is it doesn't get better. It's not like if they don't try to buy it basically right away, there's a startup I know that just did an analysis and they're growing and they realized that their pilot—anybody who asked for a pilot, their conversion rate on that kind of customer is 0%. Literally zero. And they're growing. Everybody who actually has demand buys without a pilot. And so what you're looking for are people who are actually going to pull this out of your hands, who are going to try to buy pretty quickly. First call, maybe second call, they actually are trying to buy. And if you don't have that, if you're hearing things that are like, 'Wow, this is amazing. This is exactly what—this is an important product to be building. This is exactly what the industry needs.' Or even things like, 'Oh, this would save us a million dollars a year.' What you're not hearing is 'I need this right now.'

“我该怎么买？”如果你听不到这句话，那你几乎肯定身处痛苦洞穴。你会不断地跟进潜在客户，而他们会改期，再改期。你会想：“我到底在干什么？我在推他们做免费试点。到底怎么回事？”然后当你终于让某人进入免费试点或付费试点——谁在乎呢——然后你又在推他们使用。感觉你所有时间都花在推动和说服上。如果这听起来像你，那你几乎肯定在痛苦洞穴里。这需要你做到我们之前谈过的那件事：完全停止思考你的产品——把它从你的大脑中消除——首先关注谁有拉力。哪里真的有人正在尝试但被卡住了？这会解决问题，但这需要你——这非常难做到。

How do I buy? If you're not hearing that, then you are almost certainly in the pain cave. You're going to continually follow up with a potential customer, and they're going to reschedule, and they're going to reschedule. And you're going to be like, "What the heck am I doing? I'm pushing them to do a free pilot. What the heck is going on?" And then when you finally get somebody into a free pilot or a paid pilot—who cares—and then you're pushing them to use it. It just feels like all you're doing is spending all of your time pushing and convincing. If that sounds like you, then you are almost certainly in the pain cave. It requires you to have this thing that we talked about earlier, where you stop thinking about your product at all—just eliminate it from your brain—and focus first on who has pull. Where is there actually somebody who is trying and blocked? That will solve it, but it will require you to—it's very hard to do that.

是的，我记得在我之前的创业经历中——创业世界，对吧？听起来和我一样，对吧？所以总是当你向一些用户介绍你的产品时，你知道，随机的人，他们会想：“哦，哇，非常非常有趣，很酷。”等等。但然后他们从不拿出信用卡真正购买。所以我认为你刚才说的绝对有道理，尤其是对很多人来说很有共鸣。甚至我认为人们假设这就是过程，尤其是在 B2B 领域，当你试图向大企业销售时。他们倾向于做试点等等。你认为这种方式仍然合理吗，因为传统上——不？好吧。

Yeah, I remember in my previous startup experience—startup world, right? It sounded the same like me, right? So always when you introduce your product to some users, you know, random people, they think, "Oh, wow, it's very, very interesting, it's cool." And all that. But then they never put out their credit card to actually buy. So I think what you said just now definitely makes sense, and especially I think it's quite relatable for many people. And even like I think people assume this is the process, you know, especially in the B2B world when you're trying to sell to enterprise big corporates. They will tend to do pilots and all that. Do you think it's still kind of like makes sense in this way because traditionally— No? Okay.

不，不，不，不，不。是的，是的，就像这——有很多——所以我给你举个例子。我认识一家初创公司，花了两年时间只是撞墙。麻省理工超级聪明的博士们，就像撞墙一样试图把一些 AI 产品销售到人力资源之类的领域。但没成功。他们得到了所有这些免费试点，就是我们谈到的那些东西，对吧？免费试点，人们使用它，人们喜欢它，但没有人真正购买。然后他们偶然获得了一个进入企业制药领域的机会。他们在企业制药领域毫无经验。他们的产品完全不相关。所以他们既没有产品，也没有想法要卖什么。他们基本上只是说：“你们愿意付我们多少钱来解决你们的问题？你们愿意付我们多少钱来做？”那个人说：“哦天哪，非常感谢你问。我们正在尝试做 X。”那种需求如此强烈，以至于在没有产品的情况下，他们能够在两个月内销售到企业制药领域，没有试点。我再次看到这种情况——比如一家大型财富 100 强媒体公司，非常类似的事情发生了，一家企业购买了——我们倾向于认为，哦，因为是大企业，所以必须花很长时间。是的，也许如果它是 500 个优先级中的第四或第五。但如果真的有人正在尝试做某事但被卡住了，无法做他们试图做的这件非常重要的事情，他们会从你手中抢走东西，即使产品还没建好，即使你没有 SOC 2，即使所有这些不同的东西都不存在。这就是你要寻找的。

No, no, no, no, no. Yeah, yeah, like that's this is like there's this there's all this— So, I'll give you an example. There's a startup I know that spent 2 years just hitting their head against the wall. MIT super smart PhDs, like hitting their head against the wall trying to sell some AI product into like HR or something. And it didn't work. They got all these free pilots, just all these things that we talked about, right? Free pilots, people using it, people enjoying it, and nobody actually buying. And then they randomly got an intro into enterprise pharma. They had no experience in enterprise pharma. Their product was totally not related to it. And so they didn't have a product, they didn't have an idea of what they would sell. And they basically just said, "What would you pay us to do to solve for you? What would you pay us to do?" And the person was like, "Oh my gosh, thank you so much for asking. Here, we're trying to do X." And that demand was so intense that without a product they were able to sell into enterprise pharma in like two months with no pilot. And I've seen this again—like a big Fortune 100 media company, a very similar thing happened there where an enterprise bought—like we tend to think that it's oh because they're big enterprises it has to take forever. It's like, yeah, maybe if it's priority four or five out of 500. But if there's actually somebody who is trying to do something and stuck and can't do this really important thing they're trying to do, they will pull something out of your hands even if the product isn't built yet, even if you don't have your SOC 2, even if all of these different things are not present. And that's what you're looking for.

是的，所以嗯是的，这是一个类似的故事，就像你在书中提到的，在你的引言中，同样的经历，就像你试图卖东西，但最后，你知道，有人要求你做别的事情，对吧？所以我认为这是一个非常相似的故事。所以你提到了 AI 产品。我认为现在是 AI 炒作的时代等等。人们认为构建产品、构建软件实际上相当容易，非常便宜。可以在我不知道，几天内完成。呃，你对此有什么看法，因为它给人的印象是你可以只是构建、构建、构建，你知道，这么多不同的产品。这实际上是一个好策略吗？

Yeah, so um yeah, this is similar story just like what you mentioned in your book right in your intro like same experience like you were trying to sell something but in the end, you know, somebody asked you to do something else right? So I think it's quite similar story. So you mentioned about AI product. I think now it's like the time of you know, AI hype and all that. People think building product building software is actually quite easy, very cheap. It can be done within I don't know, days probably. Uh what's your take about this because it gives the impression that you can just build build and build, you know, so many different products. Is it actually a good strategy?

它有——有好处也有坏处。所以好消息是，你不再有借口说你必须埋头 6 个月来构建正确的东西。这意味着你被迫面对冷酷的现实：嘿，你刚花了一周时间构建的东西没人在乎。而且哦等等，实际上如果我们找不到需求，我们构建什么并不重要。所以从这个意义上说，这是一件好事。一件非常非常好的事。你知道，AI 的缺点是现在每个人都认为他们可以构建，或者很多人认为他们可以构建任何东西。所以现在更难找到需求，因为现在用 Claude Code 自己动手在很多不同情况下是一个可行的选择。而且在很多情况下是一个非常好的选择。所以，你必须在世界上找到那种“自己动手用 Claude Code”不可行的拉力需求。

It has—there's upsides and downsides to it. So the good news is that you no longer have the excuse that you need to put your head down for 6 months to build the right thing. Which means you are forced to deal with the cold reality that hey, this thing you just spent a week on building nobody cares about. And oh, wait, actually it doesn't matter what we build if there's no demand, if we can't find demand. And so, in that sense, it's a great thing. It's a really, really good thing. You know, the downside of AI is that now everybody thinks they can build or many people think they can build anything. And so, it's even harder now to find demand when now do-it-myself with Claude Code is a viable option in a lot of different cases. And it's a very good option in a lot of cases. And so, you have to find pull demand out there in the world where do-it-myself with Claude Code isn't a viable option.

是的，所以我认为很多人都在考虑氛围编程，尤其是在 SaaS 世界，对吧？很多人——有些人认为 SaaS 已死，仅仅因为他们认为，你知道，我可以构建自己的 SaaS，你知道，定制任何我想要的东西，对吧？因此，可能拉力现在更难找到，对吧？所以，你对于那些构建 SaaS 之类的人有什么建议，比如为了保护他们的产品不被别人氛围编程？

Yeah, so I think many people are thinking about vibe coding and especially in the SAS world, right? Many people—some people think SAS is dead simply because they think, you know, I can just build my own SAS, you know, to customize whatever I want, right? And hence probably like the pull is probably something that is more difficult these days, right? So, what is your probably advice for people who are building SAS something like this like in order to protect, you know, their product from being vibe coded by someone else?

我在这里看到两种方法都有效，可能还有更多。我目前只看到两种，第一种是专注于那些“自己氛围编程”不是选项的人。所以这就是为什么——我不知道你是否关注 YC，但你经常看到的是，人们进入 YC 时构建开发工具，离开 YC 时销售到保险领域。所以，保险，对吧？或者物流之类的。就像，很多这些主街企业，他们不想——就像，是的，他们可以学习氛围编程，但他们是普通人，不在乎。不想——就像，我不想学习这种东西。所以，第一条路是找到那些 DIY 不是选项的人。第二条是找到那些尝试过 DIY 但撞上了 DIY 无法实现的墙的人。所以，一个例子是我朋友的公司。他们发现很多人试图 DIY 自己的公司品牌。所以你制作自己的个人大脑，你使用 Karpathy 的东西，你做所有那些有趣的事情。然后，当你试图将其扩展到你的团队时，你就想，酷。

I've seen two methods here that have both worked and there's probably more. I've just only seen two so far, which is the first one is to focus on people for whom vibe coding it myself is not an option. So, this is why—I don't know if you pay attention to YC, but you see what often happens is people enter YC building for dev tools and leave YC selling into insurance. And so, insurance, right? Or logistics or whatever it is. Like, a lot of these Main Street businesses, they don't want— Like, yeah, they could learn to vibe code, but they're like normal human beings who don't care. Don't want— Like, I don't want to learn this kind of stuff. And so, that's path one is find people for whom DIY is not an option. The second is find people who have tried DIY and hit some wall that DIY does not make possible. So, an example of this is a friend's company. What they've found is that a lot of people are trying to DIY their own company brands. So, you make your own kind of personal brain, you use the Karpathy stuff, you do all that fun stuff. And then, the second you try to expand this to your team, you're like, cool.

让我们把所有邮件都连接起来。把所有 Slack 都连接起来。把所有会议记录工具都连接起来，然后全部汇入这个公司大脑。那会很棒。然后有人说：“等等，我的一对一会议记录在记录工具里。我的 HR 对话也在里面。我们该怎么办？”这就成了一个难题，你会觉得：“哦，这是一堵墙。”于是，这时就产生了一种需求，导致 DIY……就像：“哦，我不……那看起来真的很难。”DIY 不再是一个选项。所以，这些就是我见过的两种情形。

Let's all connect our emails. Let's all connect our Slacks. Let's all connect our meeting note-takers, and we'll pull it all into this company brain. It's going to be awesome. And then, someone says, "But wait a second, I have my one-on-ones in our note-taker. Like, my HR conversations are in there. Like, what do we do?" And that becomes a hard place where you're like, "Oh, that's a wall." And so, there becomes a need at that point that leads to where DIY... It's like, "Oh, I don't... That actually seems really hard." Where DIY becomes no longer an option. And so, those are the two permutations that I've seen.

嗯。是的。我觉得我肯定……是的，我觉得对于尝试过这种 vibe coding 方法的人来说，对吧？显然，有时他们会成功，对吧？你知道，构建东西。最初，他们可能觉得：“哦，我能构建一切。”但是，是的，肯定有些复杂的产品，尤其是企业需求，肯定需要更多，你知道，能力。所以，也许说……

Mhm. Yeah. I think I definitely... Yeah, I think for people who have tried this method of vibe coding, right? Obviously, sometimes they succeed, right? You know, building stuff. Initially, they probably feel, "Oh, I can build everything." But yeah, definitely some complex product, especially enterprise needs, definitely requires much more, you know, capability. So, maybe to say...

关于这一点，抱歉打断一下，但我还发现，人们在尝试过 DIY 之后，会成为极好的客户。因为他们知道自己需要什么。他们懂得欣赏产品。

On that, just sorry to interrupt, but like what I found too is that after people have tried to DIY it, they become excellent customers. Because they know what they need. They know they can appreciate the product.

是的。有道理。那么如今 AI 模型越来越聪明，对吧？我认为这也是一个方面……我不知道，就像在世界上，对吧？我们认为 AI 将取代一切，许多许多不同的东西。从构建初创公司、构建技术产品，甚至构建 AI 产品的角度，你对此有什么感觉？

Yeah. Makes sense. So these days AI models are getting smarter and smarter, right? And I think this is also one aspect of... I don't know, like in the world, right? We think that AI is going to replace everything, many many different things. What's your feel about this in the perspective of, you know, building startups, building technological products, right? Or even building AI products, right?

是的，我的意思是，大家一直在预测 Harvey 和 Lovable 的末日，好像永远在预测，但它们似乎仍在增长。我不知道。我不在内部。我完全不知道。但我想说的是，我只是尽量不预测未来，因为事实证明未来很难预测，而我没那么聪明。所以我看到的是，是的，模型不断变得更好。随着模型变得更好，更多事情可以 DIY，100% DIY 成为越来越多人的选择。很好。这可能会继续。但似乎同样真实的是，产品的最后一英里非常重要。所以对于 Lovable 来说，有很多人不想学 quad code。有很多人想要一个网站，但不想学 quad code。对于 Harvey 来说，好吧，这里有很多……最后一英里似乎非常……你可以在最后一英里保持领先，甚至可以进一步定制，我不知道。而且这就像……这不是静态的。所以你们……我猜 Harvey 团队和 Lovable 团队一直对此感到焦虑和抓狂，那种焦虑和抓狂，以及看到新模型做出的神奇事情，同时关注他们的客户需要什么，他们特定客户需要什么，他们特定潜在客户需要什么。这些力量促使他们不断想出新的东西，那就是……我不知道，那就是生活。

Yeah, I mean everybody's been predicting Harvey and Lovable's doom for like ever, and they still seem to be growing. I don't know. I'm not on the inside. I have no idea. But so what I'd say is I just try not to predict the future because it turns out future is quite hard to predict and I'm not that smart. And so what I've seen is that like yeah, the models keep getting better. 100% DIY becomes an increasing option for a lot of different people as the models get better and more things can be DIY. Great. That is probably going to continue. But what seems to also be the case is that the very last mile of a product matters a lot. And so for Lovable, like there are a lot of people who don't want to have to learn quad code. There's a lot of people who want a website don't want to have to learn quad code. For Harvey it's like okay there's a lot there are there's like the last mile here that seems to be very... and you can kind of stay ahead and you can kind of tailor it even more at the last mile and like I don't know. And you just like it's not a static thing. So you all... I assume the Harvey team and I assume the Lovable team are like constantly anxious and freaking out about this and that anxiety and freaking out and like seeing all the magical things new models do and then also paying attention to what their customers need, what their specific customers need, what their specific prospects need. Like those forces cause them to keep figuring new stuff out and that's... I don't know, that's what life is.

是的，似乎迭代现在越来越快，你可能也有竞争对手轻易地克隆同样的东西，对吧？所以我想让你谈谈另一个所谓的反模式，对吧？因为我看到许多创始人在开始创业时也认为我想解决自己的问题，对吧？有点像我是第一个解决的人，第一个客户，对吧？他们有点活在这个世界里，贯穿公司的余生，对吧？在你看来，这也是一种可行的方法吗？

Yeah, it seems that iteration now getting faster and faster and you probably have competitors also easily, you know, cloning the same thing, right? So I'd like to for you to cover another maybe so called anti-pattern, right? Because I see many founders when they started their startup they also think I want to solve my own problems, right? And kind of like I'm the first person, first customer, first person to solve, right? And they kind of like live in this world for the rest of the company, right? Is this also a viable approach in your view?

是的，是的，但有两种做法。第一种，我也做过两种。第一种是说，嗯，我在构建我希望存在的产品，但我的生活中并没有一个具体的时刻我会真正为它付费。对吧？所以那叫做自我项目。那不是构建产品。另一方面是……我认为每个创始人都需要基本上亲身体验那种让他们的产品变得相关的拉力。所以是的，一般来说你不够有趣，不值得为你构建。就像你知道的，在很多情况下，你只是没有遇到过那种情况，你会说“天哪，我这里需要点什么。我真的会为这个领域实际上不存在的东西付费。”也许你有，酷，就为此构建。如果你没有，不要只是构建你希望人们会喜欢的产品，或者你认为……你知道，自我膨胀，觉得每个人都会崇拜我，因为我在构建这个产品？相反，去获得第一手经验，这样你就可以为自己构建。你可以成为一个值得为之构建的人，然后为自己构建。我认为这完全没问题，但真的……那里有真正的拉力。

It is, it is, but there are two ways to do this. There's the first one and I again I've done both. Way one is to say, um, I'm building the product that I want to exist but there isn't a specific moment in my life when I would have actually paid for it. Right? And so that is called an ego project. That is not building a product. The other side of it is like... I think that every founder needs to basically have first-hand experience of the pull that makes their product relevant. And so yeah, like you generally you are not... you are not interesting enough to build for. Like you know in a lot of cases like you just haven't had something where you're like oh my gosh I need something here. I actually I would pay for something that doesn't actually exist in this space. Maybe you do, cool, build it for that. If you don't have that, don't just build the product you hope people will like or you think you know going on an ego trip and like everybody's going to worship me because I'm building this product? Instead, go get first-hand experience so that you can build for yourself. You can become someone who's worth building for, and then build for yourself. I think that's totally fine, but it's really... it was there real pull there.

是的。我发现当你提到找到你真正会购买自己产品的时刻，对吧？我认为那也是真正的时刻，对吧？因为你可能有一个花哨的想法，但最终你可能不会买它，我猜那也不是真正的拉力。那么，这和 LARPing 的概念一样吗？我的意思是，你的书里提到了 LARP 这个东西，对吧？所以，真人角色扮演。这是一样的，还是不同的？

Yeah. I find that when you mention like finding the moment when you actually will buy your own product, right? I think it's that's the true moment as well, right? Because you could probably have a fancy idea, but in the end you would probably not buy it, and I guess it's not a real pull as well. So, is this the same as the concept of LARPing? I mean, your book you mentioned this LARP thing, right? So, live action role playing. Is this the same, or is this something different?

嗯，是的，所以……这属于 LARPing 的范畴。所以，对于那些不知道的人，LARPing 是……再说一次，我对 LARPing 没有意见。它只是一件有趣的事。嗯，LARPing 基本上是人们喜欢走进树林，重演或扮演中世纪战斗，拿着剑，他们打扮起来，有那么回事。所以，这完全是一种合理、可接受的爱好。我对 LARPing 没有意见。问题在于，大多数创始人在早期阶段所做的就是 LARPing，却没有意识到自己在 LARPing。所以，你本质上就像……想象你在一个正常的商务会议上，但你穿着中世纪盔甲。当你这样做的时候，真是太尴尬了……所以，LARPing 是这一类活动，它们看起来像商业活动。它们是你为了建立企业而做的事情，但它们与建立企业毫无关系。所以你只有在事后才意识到你已经在 LARPing 上花了大量时间。所以，这些包括参加路演比赛、参加创业活动、花大量时间在你的使命、愿景、价值观声明上，在你还不知道自己在做什么之前。

Um yeah, so what... This would fall into the category of LARPing. So, for those who don't know, LARPing is... Again, I have nothing against LARPing. It's just kind of a funny thing. Um LARPing is basically there people who like go into the woods and reenact or play act like battle medieval battles with like swords, and they get all dressed up, and there's like a thing. And so, it's a totally reasonable, acceptable hobby. I have nothing against LARPing. The problem is that most of what founders do in the early stage is LARPing without realizing you're LARPing. So, you're essentially just like imagine you're in a normal business meeting, but you're dressed up in like medieval armor. It's just like so embarrassing when you do... And so, LARPing is this category of activities that look like they're business things. They're things you would do to build a business, but they have nothing to do with building a business. And so, you only realize after the fact that you've been LARPing for a huge amount of time. So, these include um going to pitch competitions, going to startup events, going to like um you know, spending a bunch of time on your mission, vision, values statement before you know what you're doing.

所有这些都反映出对创业公司成功原因缺乏理解，但感觉像是企业在做的商业事情，对吧？所以我脑海中的另一个画面就像是一个穿着西装的孩子，在电脑上敲敲打打。感觉就是这样。有时你意识到自己一直在这么做。

It's all these kinds of things that reflect a lack of understanding of what causes a startup to work, but feel like business things that businesses do, right? And so the other image I have in my head is like a child wearing a business suit, just tapping away at a computer. It feels like that. And sometimes you realize you've been doing that.

是的，总是以做营销或建立品牌、获得更多知名度或认知度为借口，对吧？所以我认为这绝对是个陷阱。就像不要做这种角色扮演的事情，我认为这会完全让你偏离业务中真正需要做的事情。那么说到为你的业务做事情，在你的书里，在你找到几个“hell yes”客户之后，对吧？我认为为了扩大规模，显然你不能只针对个人，一个一个人来做，对吧？所以你在书里提到了这个过程，你知道，收集，我不知道你怎么称呼它，成功故事工厂？请告诉我们这个概念，比如人们如何扩大产品规模，使其更加成功？

Yeah, it's always, I think, in the excuse of doing marketing or building branding and getting more popularity or awareness, right? So I think definitely it's a gotcha as well. Like don't do this larping thing, which I think totally brings you away from the true things that you need to do for your business. So speaking about doing the things for your business, right? In your book, after you found several hell yes, right? So I think in order for you to scale, obviously you can't just do for individuals one individual one individual, right? So in your book you mentioned about this process of, you know, collecting, I don't know what do you call it, success story factory? Tell us about this concept, like how can people scale their products to become much more successful?

是的，我发现我们缺乏一些非常有用的概念模型来描述创业公司实际如何运作。我发现，当你找到有拉力的人，他们从你手中拉走产品并成功使用，其结果就是一个客户成功故事或案例研究，你可以在网站上找到。这就是那种“hell yes”客户，对吧？所以你作为创始人的工作就是弄清楚如何重复这一点。你想重复这一点，而不必每次都创建一个新的，因为创建一个新的需要大量工作。如果你在构建产品，你还必须为每个人构建各种功能。所以你希望达到可重复性，每次做同样的事情。你带领某人经历成为潜在客户、成为实际客户、成为“hell yes”客户的过程。所以当你这样想时，目标是产生一堆基本相同的案例研究，一堆相同的成功故事。那么创业公司就是一个生产案例研究、成功故事的工厂。这就是为什么我称之为成功故事工厂。当你有了工厂模型，你可以借鉴一百多年来关于如何有效运营工厂、产品如何流经工厂的思考。我发现最有用的的是运营工厂的约束理论，即：嘿，我们有潜在客户，我们把他们变成实际客户，再变成成功客户。这个工厂的目标是最大化生产看起来非常相似的“hell yes”客户。那么我今天需要做什么？我需要弄清楚工厂的约束在哪里，瓶颈在哪里。哦，是不是我们没有和足够多的人交谈？如果是这样，那么我只需要专注于这一点，因为如果这是工厂的约束，那么我做任何其他事情都不会导致案例研究工厂的产量增加。所以工厂模型是一种思考业务的方式，是一种看待业务的方式，看到所有部分如何连接，你可以看到它不是——工程与产品是分开的，产品与销售是分开的。它都是一个工厂模型的一部分，你可以用它来知道在任何时间点该做什么。

Yeah, so I found that we lack some very useful conceptual models for how does a startup actually work? And so what I found is that when you find somebody who has pull, who pulls the product out of your hand and is successful with it, the result of that is like a customer success story or a case study you would find on your website. And that's a kind of like a hell yes customer, right? And so your job as a founder is to just figure out how to repeat that. And you want to repeat that and not have to create a new one every single time because it's a lot of work to create a new one. And then you have to build all sorts of features that are for each individual person if you're building a product. And so you kind of want to get towards repeatability where you're doing the same thing every time. And you're kind of bringing somebody through the process of becoming a potential customer, becoming an actual customer, becoming a hell yes customer. And so when you think of it that way, the goal is to produce a bunch of basically identical case studies, a bunch of identical success stories. Then what a startup is is just a factory for producing case studies, success stories. And so that's why I call it the success story factory. And when you have that factory model in your mind, well, you can bring to bear a hundred plus years of thinking about how to effectively run a factory, how product flows through a factory. And the most useful thing I've found is the theory of constraints for running a factory, which is: Hey, we have people who are potential customers, we're turning them into actual customers, we're turning them into successful customers. The goal of this factory is to maximize the production of hell yes customers who look very similar. Well, what do I need to do today? Well, I need to figure out where in the factory is the constraint, is the bottleneck. Oh, is it that we're not talking to enough people? If that's the case, then I just need to focus on that because if that's the factory's constraint, then if I do anything other than that, it won't lead to an increase in production out of the case study factory. So the factory model is a way to think about your business, it's a way to see your business, how to see how all the pieces connect, and you can see that it's not okay, engineering is in a separate thing from product is in a separate thing from sales. It's all part of one kind of factory model that you can then use to know what to do at any one point in time.

是的，我认为约束理论，我的许多听众可能会理解，对吧？因为在 DevOps 世界中，一旦你解决了瓶颈，就要找出下一个约束，这很常见，对吧？总是有约束存在，对吧？并不是说它永无止境，对吧？所以只是从一个部分转移到另一个部分。是的。所以 Rob，我认为我们已经涵盖了很多内容，对吧？你认为关于这个拉力框架，人们需要理解的重要事情是什么？拉力的力量，这样我们也可以从中学习。

Yeah, so I think theory of constraints many people of my listeners would probably relate to that, right? Because that is quite common in the devops world to actually figure out what's the next constraint once you solve the bottleneck, right? And there's always constraint somewhere, right? It's not like it's never ending, right? So it's just moving from one part to the others. Yeah. So Rob, I think we have covered quite a lot of things, right? Is there any things that you think is important for people to understand about this pull framework, right? The power of pull so that we can also learn from that.

你做得非常棒，抓住了所有最重要的部分。我想说的是，拉力在商业中是一切的上游。它是每个产品的上游。它是你的市场推广是否有效的上游。它是你在市场推广中应该做什么的上游。它是你应该如何描述产品的上游。它是一切的上游。它存在于世界上，独立于你或你的想法。如果你只是静下心来思考，如果你只是尝试接受这个概念，你会花更少的时间去构建人们可能不想要的东西。你会找到真正会从你手中拉走产品的人。

You did a really awesome job getting all the biggest parts out of it. What I would say is that pull is upstream of everything in business. It is upstream of every product. It's upstream of whether your go-to-market works. It's upstream of what you should do on go-to-market. It's upstream of how you should describe your product. It's upstream of everything. And it exists out there in the world independent of you or your idea. And if you just sit with that, if you just try to embrace that concept, you will spend way less time building something that people probably won't want. And you will find people who will actually pull a product out of your hands.

是的。所以它是上游。所以先找到拉力，对吧？我认为这让我想起精益概念，对吧？不要制造太多浪费。你知道，总是拉而不是推。所以我认为这很相关。是的，谢谢提醒。那么一旦你有了这本书，一旦你提到它，教给其他人。有没有资源可供人们实际了解更多关于这个拉力概念？或者可能有在线资源之类的。

Yeah. So it's an upstream. So find the pull first, right? So I think it kind of reminds me of this lean concept, right? So don't create too much waste. You know, always pull rather than push. So I think that's pretty relevant. So yeah, thanks for reminding that. So once you have this book, right? Once you kind of mention about it, teach your other people. Is there resources available out there for people to actually learn more about this pull concept, right? Or maybe there's online resources and things like that.

是的，当然。我可能发布了太多关于这个的内容，因为我仍然认为我的工作是继续弄清楚它。它还不完整。它仍在进行中。所以，在 robchsneider.org 上，你可以找到一堆链接。我有一个新闻通讯。我有一个 YouTube 频道，我试图解释这个。我甚至录制了一个 3 小时的——这很荒谬，但一个 3 小时的所有概念的演练。然后还有书。如果那还不够，我正在开发一个 AI Rob Snyder 之类的东西，你可以插入 Claude。还没完全准备好。然后你可以和我一起工作，或者我运行一些小型的从零到一的市场推广加速器。

Yeah, totally. So I probably put out too much content out there about this because I still view that my job is to keep figuring it out. It's not complete. It's still a work in progress. And so, at robchsneider.org, you can find a bunch of links to stuff. I have a newsletter. I have a YouTube channel where I try to explain this. I've even recorded a 3-hour—it's absurd, but a 3-hour walk-through of all the concepts. And then there's the book, too. And if that's not enough, I'm working on an AI Rob Snyder kind of thing that you can plug into Claude. Not quite ready for primetime yet. And then you can kind of work with me or I run little go-to-market accelerators for zero to one.

对。是的，我希望很多人从你那里拉走所有这些，对吧？希望这个概念在产品和创业世界中变得更加重要。所以，Rob，在结束对话之前，我只有一个最后一个问题。我称之为三个技术领导力智慧，但对你来说，它不必是真正的技术性。

Right. Yeah, I hope many people pull all these things from you, right? So, and hopefully the concept becomes much more, you know, important in this, you know, product management and startup world. So, Rob, I only have one last question for you before we wrap up the conversation. I call this the three technical leadership wisdom, but I think for you it doesn't have to be really technical.

那么，你想给听众的任何建议都会很受欢迎。

So any kind of advice that you want to give to the listeners would be appreciated.

好，好。那我要给出三条超级反直觉的建议。第一条是：技术放在最后。基本上，尽可能避免技术。尤其是 AI 产品，大多数情况下你应该先能把它当作服务卖出去。所以就那么做。“最后再自动化”是埃隆那套算法里的说法，是一回事。技术放最后。好，这是第一条。

Yeah, yeah. So I'm going to give three ones that are super contrarian. And so the first one is: tech comes last. Basically, avoid tech as much as humanly possible. Especially for AI products, you should be able to sell them as services first in most cases. And so do that. Automate last is the kind of thing from Elon's algorithm. It's the same thing. Tech last. Okay, that's thing one.

第二条：让你的产品显得小，而不是大。在销售对话里，当你和潜在客户交谈时，你会以为我把产品说得越大，他们就越看重、越想要。而你实际上做的，是给他们制造更多需要理解、并最终提出反对的表面积。所以一旦你理解了让人购买的“物理原理”——他们想做某件事，却被卡住了——你只需要描述和展示最小的那件事，证明他们能被解开卡点。除此之外你展示的任何东西都会增加风阻，反而说服他们不要买。

Thing two: make your product seem small, not large. In a sales conversation, when you're talking with a potential customer, you're going to think that the bigger I make the product seem, the more they're going to value it and want it. When what you're really doing is creating more surface area for them to have to understand and ultimately object to. And so once you understand the kind of physics of what causes somebody to buy — they're trying to do something and they're blocked — all you need to describe and show is the smallest possible thing that demonstrates that they will be unblocked. Anything you show other than that adds wind resistance. It convinces them not to buy.

好，最后一条是：销售和你以为的恰恰相反。我们往往以为销售就是：我在说服你买。我作为卖家，说服你作为买家去买。销售不是这样运作的，那不是让人购买的原因。当你那样想时，第一，你会做得很糟；第二，你会竭尽全力去回避或拖延做这件事。于是你就会让你的产品失败。销售恰恰相反。你会买，是因为你想做某件事，却被卡住了。如果不是这样，那不管我做什么说什么，你从我这儿买都很奇怪。所以我在销售中的工作，是往后靠，弄清楚：你是不是想做某件事却被卡住了？如果是，就写下最小的那件事，然后往后靠，看看你会怎么做。你应该会想把它从我手里抢过去。这就是为什么销售是这么好的研究方法。你需要的不是说服别人买，而是往后靠，给他们空间去证明他们自己有拉力。所以你现在就该这么做。

Okay. Final thing is that sales is the exact opposite of what you think it is. So we tend to think that sales is: I'm convincing you to buy. I, the seller, convince you, the buyer, to buy. That's not how sales works. That's not what causes somebody to buy. When you think that way, one, you're going to be awful at sales, and two, you're going to do everything in your power to avoid it or delay doing it. And so you're going to cause your product to fail. Sales is the exact opposite of that. You're going to buy because you're trying to do something and you're stuck on it. If that's not the case, it would be weird if you bought from me no matter what I do or say. And so my job in sales is to lean back and figure out: are you trying to do something and stuck? If so, just write the smallest possible thing and lean back and see what you do. You should try to pull that out of my hands. And that's why sales is such a good research method. All you need is not convincing people to buy, it's leaning back and giving them the space to demonstrate that they have pull. So you should do that right now.

哇。确实，大多数事情——如果不是全部的话——都和大多数人所相信的有些反着来，我想。谢谢你分享这些。我觉得很有洞见。尤其是最后关于销售的那条，因为我们总以为销售就是说服别人、影响别人，而且越大越好——或者越便宜越好。但其实不是那样。关键在于找到那种需求，人们想做的那个项目，而他们被现有的选项卡住了。所以我觉得这很有洞见。

Oh wow. It's true that most things, if not all, are kind of contrarian to what most people believe, I guess. And so thanks for sharing that. I find it quite insightful. Especially the last one about sales, because we all think sales is convincing people, influencing people, and the bigger it gets, the better — or the cheaper it is, the better. But it's not about that. It's all about finding the kind of needs, the project that people want to do, and they are stuck with the options that they have. So I think that's pretty insightful.

所以非常感谢你，Rob，感谢这次对话。我希望大家能从这个概念中学到很多。我觉得挺有意思的是，可以说我们在创业界和产品管理界都有了一套新方法论，我也希望它对听众同样有益。谢谢你。

So thank you so much, Rob, for this conversation. I hope people learn a lot from this concept. I find it quite fascinating that we all have a new methodology, so to speak, in the startup world and the product management world, and I hope it's going to be beneficial for people who listen as well. So thank you for that.

谢谢。

Thank you.