本集简介
双语字幕
仅展示文本字幕，不包含中文音频；想边听边看，请使用 Bayt 播客 App。
卡罗琳，我太兴奋了。
Carolyn, I'm so excited.
今天，我们请到了彼得·莱因哈特做客节目。
Today, we have Peter Reinhardt on the on the show.
他创办了Segment公司，2011年参加了Y Combinator，现在又创立了Charm Industrial，一家做碳去除的公司。
He was the founder of Segment, which went through Y Combinator in 2011, and now founder of Charm Industrial, which is a carbon removal company.
嗨，彼得。
Hi, Peter.
欢迎你，彼得。
Welcome, Peter.
是的。
Yeah.
见到你们俩真好。
Good to see you both.
谢谢你能来参加节目。
For coming on the show.
咱们可有一阵子没见了。
It's been it's been a while.
是啊。
Yeah.
确实好久没见了。
It has been a while.
见到你们真好。
It's good to see you.
我超想听听你关于Charm的事。
I cannot wait to hear more about Charm.
不过在那之前，我想先聊聊Segment，你2011年夏天参加Y Combinator的时候，还不是做Segment的。
But first, I wanna start with Segment because you went through Y Combinator in 2011 in the summer, but not with Segment.
Segment啊，这简直就是一次彻底的转型。
I mean Segment, it's a story of a pivot.
所以我想从Classmetric和你当初的申请说起。
And so I wanna start with Classmetric and your application.
你能跟我们说说那时候的事吗？
Can you tell us a little bit about about that time?
是的。
Yeah.
挺有意思的。
Fun.
对啊。
Yeah.
那时候我们是MIT和罗德岛设计学院的学生。
I mean, we were students at MIT at the time, or MIT and Rhode Island School Design.
我们想给学生一个按钮，一按就能说‘我糊涂了’。
And the idea was to give students this button to push to say, I'm confused.
教授们开着电脑，就能看到学生困惑程度随时间变化的图表。
And the professor, they'd all have their laptops open or whatever, and the professor would see this graph over time of how confused their students were.
我们觉得，教授们其实很想改进讲课方式，至少也想实时回应学生问题，而学生呢，更愿意匿名提问。
I mean, we had this notion that professors really wanted to improve their lectures or that they at least wanted to be able to answer questions in kind of a real time and that students would prefer anonymous questions and so on.
所以呢，这主意听起来挺不错的。
So, you know, it was a nice idea.
你可能会觉得他们确实会想要这个。
You would think they'd want to.
是的。
Yeah.
对。
Yeah.
我觉得这正好说明了，有时候你做的东西理论上看着挺对，但人家压根儿不想要。
I think it's a good example of, like, you know, building something that's seems right in theory, but people may not actually want.
那你们当时怎么就决定要创业，还去申请Y Combinator了呢？
And what what made you want to move forward with a startup and apply to Y Combinator anyway?
你们那时候还是大三学生。
You were juniors at MIT.
对啊。
We were.
是的。
Yeah.
其实，我们大概在那之前一两年就开始读保罗·格雷厄姆的博客了。
Actually, I think we started reading PG's essays maybe, like, a year or two before that.
所以，作为室友一起创业这个想法让我们特别兴奋。
And so we were pretty amped at the idea of of building a company together as roommates.
我们花了很多时间讨论这个事。
We spent a lot of time talking about it.
那会儿正好赶上MIT的独立活动期，我们就天天捣鼓各种点子。
We spent like MIT's independent activity period kinda hacking on ideas there.
说白了，那时候创业这事儿在校园里已经挺流行的了，不过我猜最初是被Hacker News和保罗的一些文章给忽略了。
So, you know, was I guess it was in the water at that point, but I think it was initially kind of omitted by Hacker News and some of PG's essays.
就你们俩吗？还是说人更多？
Was it just two of you, or was it there's more of you.
哦，有
Oh, there
我们一共四个人。
were four of us.
对。
Yeah.
一直都是这样，从头到尾
The whole time for for
我们三个都是麻省理工的，还有一个是YC刚开始时从罗德岛加入的。
Three of us three of us from MIT, and then one who joined right at the beginning of of y combinator from Rhode
网站。
Island's website.
明白了。
Okay.
所以你们收到了面试邀请，我记得……
So you got invited to interview, and I remember
天哪。
Oh, bless.
我记得大家都特别喜欢你。
I remember everyone really liked you.
就是那个团队，你知道的。
Got, you know, the group.
你们都是顶尖的程序员，我知道你本来是搞航天工程的。
And you were hotshot programmers, and I know you were an aerospace engineer, technically.
你也写代码吗？
Did you program too?
我写。
I did.
对。
Yeah.
明白了。
Okay.
所以你们全都编程很厉害。
So you're all good programmers.
但没人真觉得这主意有多棒。
You but no one really loved the idea.
说来有趣，我翻了下当时的笔记。
And actually, I got a kick out of this because I looked at the notes.
里面保罗写的笔记是这么写的：‘提议重启Etherpad’，那是YC以前投过的一个没做起来的项目。
And in it, Paul's note was, quote, proposed resuscitating Etherpad, which was an old y c startup that that didn't really take off.
所以很明显，保罗在面试时还在推新点子，这本身就说明了这主意的潜力。
And so it was clear if Paul's pitching new ideas during the interview, that says something about the idea.
你还记得那次面试吗？
Do you remember the interview?
当然记得。
Oh, yes.
不，我不记得了。
No.
不过面试的事我还真有印象。
I do remember the interview.
我记得特别清楚，PG当时说，是啊。
And I I remember very specifically one part of it where PG was like, yeah.
你知道吗，我当学生的时候就一直想要这样的东西。
You know, I always wanted this as a student.
我当时就想，太好了。
And I was like, thank you.
对。
Yeah.
然后他转向了，是罗布吗？
And then he turned to, is is it Rob?
我导师是他吗？
Is that my supervisor?
他转向了罗布。
He turned to Rob.
然后
And
他跟Rob说：‘你会用这个吗？’
he's like, Rob, would you use this?
我们在做产品调研的时候，早就专门找过Rob了。
And we had specifically gone through Rob ahead of time when we were doing, like, product discovery.
当时我们就问过Rob：‘这东西你会用吗？’
And we'd asked, like, Rob, is this a thing you'd use?
后来在MIT他的办公室里，他亲口说会用。
And in person, when we were in his office at MIT, he'd said yes.
可是在访谈时，他直视着PG的眼睛，却说不会用。
And then in the interview, he looked, p g in the eye, and he said no.
我当时就愣了。
I was like
这简直是背叛。
The betrayal.
等等。
Wait.
我正想说呢，这背叛得太狠了。
I was just gonna say the betrayal.
所以你们真的去MIT见了罗伯特？
So you actually met with Robert at MIT.
我的意思是，他还专门抽时间见了你们。
Like, he took the meeting with you.
是啊。
Yep.
这简直太离谱了。
That that that's like
就是
a Just
作为教授，你面对的可是潜在客户之类的啊。
as a professor, you could be with a potential customer and so on and so forth.
不管怎样，我觉得他当时也没多热情，但态度确实明显变了。
Anyways, I don't think he was that amped on it, but he definitely he definitely had a change of tune.
这太搞笑了。
That's hysterical.
真逗。
That is funny.
所以那时候你完蛋了？
So it toast at that moment.
没有。
No.
你没完蛋。
You weren't toast.
我也会这么想。
I would think that too.
因为我们认识的每个人都知道。
Because everyone real we knew.
到那时候，大家都知道了，毕竟都2011年了。
We knew at this point, because it's 2011.
我们渐渐意识到，创业不一定非得靠点子本身。
We'd sort of realized it's not always about the idea.
如果你有一群关系融洽、直觉好、编程能力强的创始人，那就投他们。
If you have a great group of founders who get along well and have good instincts and are good programmers, fund them.
投的是他们调整方向的能力。
Fund for the pivot.
但咱们录取你们的时候，你们不是还坚持要做课堂那个点子吗？
So but you guys must have insisted you're still gonna do the classroom idea when we accepted you.
是的。
Yeah.
我们就是比较固执。
We're more stubborn.
没错。
Yep.
所以整个夏天我们都在埋头开发那个应用，还试着在斯坦福的一些暑期课程里测试，结果不太理想。
And so we spent the whole summer basically building out some like, building out the application, and we just tried testing it in some summer classes at Stanford and so on, which didn't go great.
但我们决定，好吧。
But we decided, okay.
我们打算认真试一把，先在一些课堂里启动，然后搬回波士顿，毕竟那儿大学多，觉得是推这个产品的好地方。
We're gonna give this a proper trial, like, you know, kickoff with some classes, and we're gonna move back to Boston because Boston has so many universities that we actually figured it was a pretty good place to to launch it.
而且我女朋友当时在波士顿。
And and also my girlfriend was in Boston.
现在是我老婆了。
Wife now.
所以这步棋走对了。
So it was it was the right move.
真的吗？
And Really?
不管怎样，我们后来在好多课堂里都上线了。
I anyway, so we've launched it in a whole bunch of classroom.
我们还办了演示日。
We did demo day.
我们筹了大概六十万美元。
We raised, like, 600 k.
那真是一段特别熬人的日子。
It was a bit of a grind.
那时候我睡在Codecademy那帮人的地板上，他们倒是坐着私人飞机去见投资人。
I was sleeping on the floor of the CO2cademy guys who were out, like, you know, meeting VCs on jets.
我整天忙着约见面，连饭都顾不上吃。
And I was, like, scrambling for meetings and forgetting to eat.
那日子简直糟透了。
It was terrible.
真的，惨得不行。
It was, like, it was awful.
你当时在等什么？
Were you wait.
你搬回波士顿后，是睡在Codecademy那帮人地板上的吗？
Were you sleeping on the Codecademy guys floor because you had moved back to Boston?
那你是
And were
是的
Yeah.
其他人都搬回波士顿了。
The other guys had moved back to Boston.
对啊
Mhmm.
扎克·西姆斯坐私人飞机？
And Zach Sims is on a private jet?
对。
Yeah.
没错。
Exactly.
扎克·西姆斯就是这样的。
Zach Sims is like yeah.
我不清楚。
I don't know.
在停机坪上跟海岸法律的人碰面之类的。
Meeting Coast Law on the tarmac or whatever.
他那边term sheet多得都快溢出来了，而我呢，连60万美金都凑得特别吃力。
And, you know, he's got term sheets coming out of his ears, and I was, like, you know, struggling to pull together 600 k.
那你有没有跟风投说过，其实客户是大学，他们会为这个买单？
But did you tell VCs, like, universities are gonna pay for this?
你的客户根本就是大学，他们才是要付钱买你软件的人。
Like, your customer was actually the university, and they were gonna pay for your software.
你本来就是靠这个
That's how you were
没错。
That's right.
对。
Yeah.
我们当时主要跟Blackboard竞争。
We were competing with Blackboard, really.
我们想做成一个学习管理系统。
We wanted to become, a learning management system.
哦，明白了。
Oh, got it.
这正是我们撬动Blackboard的突破口。
And this was, an entry point to dislodge Blackboard.
嗯，懂了。
Okay.
那为啥没人用你们的软件呢？
Why didn't anyone adopt your software?
其实那年秋天，有六七个教室的教授用了我们的系统。
We actually we got, like, I don't know, six or so classrooms, professors to adopt it in in the fall.
也不算太差吧。
Like, it wasn't it wasn't terrible.
其实报名的人更多些，比如麻省理工有几个，波士顿大学也有几个，之类的。
We had maybe more signed up, like, you know, couple at MIT and a couple at Boston University and so on.
我们在市场推广上下了挺大功夫，但真正的问题是，进了教室之后才暴露出来。
We'd worked pretty hard on the go to market side, and then the problem actually was that started putting in the classroom.
我们站在教室后面，心想：学生们上课到底在干嘛？
We, like, stood in the back of the class, and we were like, what are students actually doing during class?
他们用这个系统了吗？
Like, are they using it?
结果发现，根本没人用。
And what we found is that they they weren't.
简直是一团糟。
It was, a total disaster.
上课刚开始，60%的学生都在刷Flickr、收Gmail、刷Twitter，或者在eBay上购物。
Like, the you know, at the beginning of class, 60% of students would be on, like, Flickr and Gmail and Twitter and whatever else, shopping on eBay.
等下课的时候，这个比例已经涨到80%了。
And then by the end of the class, it'd be 80%.
所以当时我就想，哎哟。
So it's like Oh.
我的结论就是：天哪。
My my conclusion was just like, oh my god.
赶紧把笔记本电脑都从教室里拿走。
Like, get the laptops out of the classroom immediately.
这简直太糟了。
Like, this is terrible.
谁都不该把笔记本电脑打开。
No one should have their laptop open.
学生们根本就没在听讲。
The students weren't even paying attention to the lecture.
当然了，他们一脸懵。
Of course, they were confused.
他们根本没在听讲。
They weren't listening to the lecture.
大概那时候，《纽约时报》登了一篇文章，说咱们得确保每个学生上课都能用上笔记本电脑。
And around this time, an article came out in New York Times that was like, you know, maybe we should make sure that every student can have a laptop in in lecture.
我当时就惊了，天哪。
And I was like, oh my god.
绝对不行。
No.
咱们得赶紧停掉这个项目。
Like, we gotta shut down the product.
得把所有设备都从教室里清出去。
We gotta, like, get all the devices out of their classrooms.
结果短短一周半，我们彻底反转了对科技在课堂中作用的看法。
So we we did, like, you know, complete one eighty on the role of technology in classrooms over the course of about a week and a half.
然后你不是还和保罗开过办公时间吗？
And then you had office hours with Paul, didn't you?
因为他记得，你当时边走边说……
Because he was saying he remembered that office hours when you were walking
那是一年半以后的事了。
That was a year and a half later.
天哪。
Woah.
好吧。
Okay.
天哪。
Woah.
天哪。
Woah.
天哪。
Woah.
那这一年半里到底发生了什么？
So what happened in this year and a half?
嗯，我们当时挺尴尬的。
Well, we were pretty embarrassed.
所以我们把演示日的所有投资人都叫回来了，跟他们说：嘿。
So but we called back all of our investors from demo day, and we told them, hey.
我们要转型了。
We're gonna pivot.
我们要做个数据分析工具。
We're gonna build an analytics tool.
那是我们当时想到的最好的点子。
That was the best idea we had.
典型的陷阱项目。
Classic tar pit idea.
反正呢，我们打算跟谷歌分析这些玩意儿正面刚。
And, anyways, we're gonna, like, compete with Google Analytics and and and so on.
然后我们就躲起来埋头苦干，开始做一款分析工具，整整
So then we, like, disappeared into a hole and, like, started building an analytics tool for for a
一年。
year.
天啊。
Wow.
结果发现，要让别人能任意方式查询数据，技术上难度还挺大的。
Which turns out to be a reasonably hard technical problem to, like, allow people to query data any which way.
所以花了不少时间才搞出来，这其实是个糟糕的决定，而且我们也没好好测试市场推广。
And so it took a long time to build, which was a bad idea, and we didn't really test the go to market that well.
我们根本不懂怎么搞B2B销售，根本没主动去拉客户。
We didn't know how to do b to b sales, so we weren't, like, out trying to really get customers that way.
是啊。
Yeah.
大概有两个投资者在我们转型时把钱撤了。
And I guess there were there were two investors that took their money back at the pivot.
那次转型挺艰难的，之后我们就又埋头苦干了一年。
It was it was kind of a rough pivot, and then we, like, disappeared into a hole for a year.
对。
Yeah.
你们四个人整整一年都在坚持做这件事。
And all four of you kept working on this for for the entire year.
你们就一直埋头干着。
You just were you just logged it out.
谁都没去找别的工作。
No one got other jobs.
就是一直这么熬着。
Like, you just kept working.
对。
Mhmm.
天哪。
Wow.
每人不到六百美元，要是有人更少的话。
With the $600 or less if some of them
当时说的是四块五美金。
It was saying it was $4.50.
不过，要是省着点花，你们也能撑一年
Still, you can you could last for
一年
a year
只要日子过得紧巴点。
if you're living cheaply.
那你们是什么时候觉得这个数据分析搞不下去的？
So at what point did you say this analytics thing isn't working?
接下来咱们怎么办？
What are we gonna do?
我觉得，等我们做出一个自己都觉得挺有意思的东西，开始真正推出来的时候，才终于说服自己：别再纠结了，赶紧发布吧。
I think by the time we built something that we thought was pretty interesting, and then we we actually started, like, launching it, and we finally convinced ourselves, like, we gotta just get over this and start putting it out.
都一年了。
Like, it's a year.
时间够长了。
It's been long enough.
我们真得从边上站出来，动手干点啥了。
Like, we have to just get off the the bench here and do something.
后来我们发现，那些之前说‘这挺有意思的’的人。
And what we started to find is all these people who had told us, like, oh, that's really interesting.
他们都说想多了解了解啥的，但其实都是嘴上说说。
I'd, you know, I'd love to learn more, etcetera.
根本没真打算花钱买，这简直太打击人了。
It was really, like, idle interest, and there was not any real interest behind it in terms of, like, a business actually purchasing anything, which was brutal.
就是那种要命的、假惺惺的客户反馈。
It's like, you know, the sort of ruinous empathy feedback from from customers.
那会儿大概是2012年十月、十一月吧。
And so that was, like, October, November of of twenty twelve.
我刚结婚，度完蜜月回来，咱们就坐下来商量：
I had just gotten married and went on a honeymoon, came back, and we're like, okay.
这事儿，怕是真不行了。
But this is probably not working.
所以我们当时就想，这下该怎么办？
So was like, what do we do?
天哪。
Oh gosh.
那时候你的职业士气肯定低到谷底了吧。
Is it like, your professional morale must have been pretty low at this point.
低得不能再低了。
Extremely low.
我差不多每周都要经历一次恐慌发作。
I mean, I was getting panic attacks about once a week.
我都住进医院了。
Like, I was in the hospital.
啊，对。
I Oh.
是啊。
Yeah.
那会儿可糟了。
It was bad.
真的特别糟。
It was really bad.
什么？
What?
天哪。
Gosh.
我觉得我失去的是。
What I think I lost.
我当时已经挺脆弱的了。
I was already pretty lightweight.
我当时就是觉得。
I was like
你那时候一个用户都没有。
You didn't have any users.
对吧？
Right?
还是你其实有？
Or or did you?
对，就是这样。
Well, that's that's yeah.
所以我才那样，就是因为这个。
That's why I was that's why I
然后呢？
was And
所以就靠各种套路了。
so hacks.
这根本行不通。
This is not working.
是啊。
Yeah.
但彼得，你那时候 probably 就才25岁吧。
But, Peter, you were probably only, like, 25 years old.
对吧？
Right?
我那时候23。
I was 23.
23岁啊。
23.
所以23岁根本不算失败。
So it's like, you're not a failure at 23.
你那时候才刚起步，你不是已经拿到学位了吗？
That is like, you could start like, you have a you you did get your degree.
对吧？
Right?
不对。
Nope.
还是你中途退学了？
Or did you oh, you dropped out.
明白了。
Okay.
哦，原来如此。
Okay.
行吧。
Alright.
我本该……
I should have
回学校继续念的。
gone back.
你现在经常恐慌发作吧？
So you're having panic attacks.
而且还是因为这个住过院。
And in the hospital for them, no less.
因为我当时以为是心脏病发作之类的，结果发现根本不是。
Well, because I thought they were heart attacks or whatever, and then it turns out it's not.
但也没办法。
But whatever.
你根本不知道该怎么办。
You, like, you don't know what to do.
对吧？
Right?
所以，确实挺糟的。
So, yeah, it wasn't it wasn't good.
那段时间特别难受。
It was pretty miserable.
嗯。
Okay.
那后来怎么样了？
So what then happened?
接下来最重要的转折点是什么？
What was the next big turning point?
对。
Yeah.
下一个重大转折点是我们联合创始人艾恩写的那份《Analytics JS宣言》，他就是来自罗德岛设计学院的那个。
The next big turning point was when our co founder Ian, the one from Rhode Island School of Design, wrote this Analytics JS manifesto.
就一页纸，用Markdown写的，放在GitHub上。
It was like a one page, markdown file on GitHub.
他 basically 就说：瞧好了。
And he was basically like, look.
我们一年多前搞了个叫Analytics JS的小库。
We built this little library called Analytics JS a little over a year ago.
当时就是为Classmetric内部用的，因为我们想把所有数据分析数据都发到Mixpanel、Google Analytics和Kissmetrics这三个平台。
We built it to use just inside of Classmetric because we wanted to send all of our analytics data to Mixpanel, to Google Analytics, and to Kissmetrics.
所以这个库一开始就能把数据分发到这三个服务上。
So it had those three kind of services that it could route data to.
他说，这背后肯定有个大生意。
And he's like, think there's a big business behind that.
你瞧。
Look.
这是个开源库。
It's an open source library.
我们过去一年半根本没怎么管它，就只是扔到网上，结果GitHub上居然有五十多个星了。
We've given it no love over a year and a half aside from just putting it up on on the web, and it has, like, 50 stars or something on GitHub.
很明显，有人在用它。
So, like, clearly, is doing something with it.
这真是我们做过最棒的东西。
It's the best thing we've done.
咱们应该全力把它做成一个生意。
Like, we should double down on it as a business.
我当时就说，老兄，这不就是个开源库嘛，才两百行代码。
I was like, dude, it's an open source library, and it's 200 lines of code.
我真的搞不懂，你怎么可能靠这个做生意。
Like, I just do not understand how you can possibly build a business around this.
这根本啥都没有。
Like, it is there's nothing there.
是，也许有人想用，但我真不明白怎么把它变成生意——挺讽刺的是，这问题后来成了我接下来十年的困扰。
Like, yeah, maybe people wanna use it, but I don't understand how to turn it into a business, which is funny because that became my problem for the next decade.
不过，这也就是我当时反对的理由。
But anyway, so that was my argument against it.
我怎么都甩不掉这个想法。
And I just I couldn't I couldn't shake him off of it.
我就是甩不掉卡尔文的这个想法。
I couldn't shake Calvin off of it.
埃利奥特嘛，倒没明确表态，但意思大概是：这大概是我们唯一能选的路了。
I think Elliot was, like, neither here nor there on it, but was kinda like, it's probably the best option we have.
于是我就回家了。
And so I went home.
我和我妻子艾丽卡聊了这事，我说：你帮我想想办法，怎么才能彻底放弃这个点子。
My wife, Erica, and I were talking about it, I was like, help me figure out how to kill this idea.
这事儿就像滚雪球，越滚越停不下来，太糟了。
It's like getting momentum, it's a terrible one.
所以我们俩熬夜 brainstorm，琢磨怎么才能把这玩意儿彻底干掉。
And so we I think we stayed up late brainstorming, like, how we could freaking kill this thing.
哦，天哪，我好像那晚根本没睡，或者只睡了几个小时。
And Oh, that's came I think I stayed up all night or something.
我也不太记得了。
I don't know.
我累得不行。
I was exhausted.
第二天一早我就来了。
Came in the next day.
我说：行吧，咱们先给 Analytics JS 做个落地页，发到 Hacker News 上看看。
Was like, alright, Let's build a landing page for Analytics JS, and let's put it up on Hacker News.
咱们别一头扎进去，埋头搞一年。
Let's, like let's not disappear into a hole and build something for a year.
先发出去看看会发生啥。
Let's just launch it and see what happens.
大家也都同意这个做法。
And so we could all get on board with that.
行，就这么办。
We're like, okay.
对。
Yeah.
咱可不想再重演去年那出。
We don't wanna repeat the last year.
就这么干。
Let's do that.
然后我们就把这东西发到 Hacker News 上，直接冲上了榜首。
So then we we put it up on Hacker News, and it went straight to the top.
一整天都稳坐第一。
Like, sat at number one all day.
GitHub 上一下子拿到了三千多个星，差不多吧。
Got, like, 3,000 stars on GitHub or something like that.
Hacker News 上几百个赞，几千人注册了邮箱。
A few 100 upvotes on Hacker News, few thousand email sign ups.
所以在 Hacker News 上直接炸了。
So it was just like just exploded on Hacker News.
这样一来，决定就很简单了。
So that made the decision easy.
等等。
Hold on.
你和埃里卡当初计划的时候，是不是在琢磨怎么阻止这事？
Was your and Erica's plan that you spent thinking, how do we thwart this?
就是说，行吧。
Was to say, okay.
咱们现在就动手吧，做完好继续下一步。
Let's just make it happen now so that we can move on.
肯定会失败的。
It will fail.
失败了咱们也能赶紧撤，不耽误事儿。
We can move on quickly.
没错。
Exactly.
所以咱们在Hacker News上根本没引起任何关注，然后就直接跳过了。
So we get no attention on Hacker News, and we'll move on.
结果完全相反。
It was totally the opposite.
是啊。
Yeah.
完全相反。
Totally the opposite.
展开剩余字幕（还有 480 条）
这太搞笑了。
That's hysterical.
好吧。
Okay.
所以你就说，好吧。
So then you said, okay.
这儿有东西，好像挺受人喜欢的。
There's some something here that people seem to like.
于是你就带着这个宣言往前推进了。
So you sort of moved forward with this with this manifesto.
我觉得那时候差不多就是我们跟PG开办公时间前后吧。
Well, I think then somewhere in right around this time, might have been right before or right after we had office hours with PG.
可能就是那天之后，或者差不多那时候，我们就在想：要不要做这个事儿？
I think it might have been, like, the day after that or something like that, where we're like, should we do this thing?
或者你知道的，埃利奥特又想了个点子，怎么找朋友一起干点啥。
Or, you know, Elliot has another idea for, like, how to find things to do with your friends.
这个我知道。
Which which I know.
我知道保罗怎么说的。
I know what Paul said.
对。
Yeah.
没错。
Yeah.
保罗说：天哪。
Paul was like, oh my god.
不是的。
No.
所以我觉得他的建议 basically 就是：我也说不准。
So he I I think his advice is basically like, I don't know.
这事儿看着真有点门道。
This seems it seems like it's there's something here.
你干脆直接去追这个吧。
You should just go after this.
我对这事的记忆有点模糊了。
It's my vague memory of it.
我不确定你面前有没有笔记。
I don't know if you have notes in front of you.
他不是说了吗？他记得自己跟你说过，你都花了一半百万美元了，可啥成果都没出来。
Didn't he well, he told he remembered that he said to you something like, you've spent, you know, half $1,000,000, and you have nothing to show for
对啊。
it.
是的。
Yep.
那时候你觉得这话刺耳吗？
Did that feel harsh at the time?
感觉特别真实。
It felt real.
我的意思是，确实挺伤人的，但感觉特别真实。
I mean, it hurt, but it felt real.
对。
Yeah.
没错。
Yeah.
我们当时真的特别为此感到羞愧。
We definitely were carrying a lot of shame about that.
是啊。
Yeah.
但我们还没真正把这事彻底了结，你知道的，就是没下定论。
But we we didn't we hadn't quite, like, you know, put a pin on it.
明白了。
Okay.
所以他最终给这事定了性。
So he put the he put the pin on that one.
然后你们在 Hacker News 上发布了，有点起色，开始有人注册，就这么启动了。
And and so you guys launch it on Hacker News, get some traction, you get some sign ups, and you're you're off.
算是吧，但这是个开源库。
Sort of, except it's this open source library.
那你们到底拿它干嘛呢？
And so, like, what do you what do you do with it?
后来这就成了一个永恒的难题：好吧，
And then it became the challenge forever of, okay.
到底大家真正愿意为这种客户数据路由的概念，付钱买什么样的服务？
Well, you know, what what do people actually want that's a paying service around this concept of routing customer data?
结果发现，第一点，人们根本不想用那个开源库。
First thing it turns out is people don't really wanna use the open source library.
他们想要的是一个托管版本，能点几下按钮，就指定数据发到哪儿。
They wanna use a hosted version where they can push buttons to say where they wanna send the data.
所以我们就这么不断打磨，最终做出了这个产品、这项服务，也成了这家公司。
So that was what we built and built, and that became the product and the service and the company.
然后我们花了大概六个月的时间，专心搞增长。
And then we kinda focused on growth for, like, six months.
下一次我们和PG开办公时间的时候，他盯着增长曲线说：这我见过最平、最直的增长曲线了。
And I think the next office hours we had with with PG, he was looking at the growth graph, and he was like, this is the most linear, straightest growth graph I've ever seen.
他接着说：通常曲线都会往上或往下拐，可这个居然一点弯都没有——你们是不是在搞付费推广啊？
He's like, they always bend one way or the other, but this is just he's like, what are you are you guys doing paid acquisition or something?
我真想不通，这怎么可能这么直？
Like, I don't understand how it could possibly be this straight.
不是，根本不是。
It's like, nope.
其实就是个超级简单的病毒循环，就一个人传一个人，就这么简单。
That was just like it was just a virality loop or whatever of one, exactly one.
我能再问一下那个 analytics.js 宣言的事吗？
Can I just go back to this analytics JS manifesto thing?
Ian 这个想法，是受你之前在做的东西启发才产生的吗？
Did Ian have the idea sort of in response to something you had been working on before?
还是说这本来就是他独立想出来的点子？
Or was this just a separate standalone idea that he had?
这跟我们当时在做的分析工具是相关的。
It was adjacent to the analytics tool that we were building.
对吧？
Right?
可能是因为我想，要是能让用户同时把数据发给我们和Mixpanel，大家比一比，说不定更容易让人接受我们。
And maybe I maybe it came from the idea of, like, oh, maybe we could make it easier for people to adopt us if they could send data to both us and Mixpanel, and we could kinda get a bake off going.
这是我记得的版本。
That's my memory.
不过呢。
But yeah.
这想法其实酝酿了很久，挺复杂的，慢慢成型的。
It was it was sort of like a complicated and marin marinating, I think, for a long time.
明白了。
Okay.
但我想说的是，你那一年在那家公司花的时间其实没白费，至少为Segment埋下了种子。
But it just it would've what I'm sort of getting at is that it wasn't a complete and utter waste of time that year that you spent working on the other company, because it sort of planted the seed for Segment.
没错。
Sure.
嗯，明白了。
Okay.
彼得对这个说法并不完全认同。
Peter's not totally on board with that.
那之后你们就一路顺风了？
And so then was it all smooth sailing for you guys?
我记不太清了。
I don't remember.
我觉得，对初创公司来说，已经算相当顺利了。
I mean, I maybe it's close to smooth as startups get.
你们后来又融过资吗？
Did you raise more money?
是啊。
Oh, yeah.
对。
Yeah.
哦，这样啊。
Oh, okay.
我们那时候快没钱了。
So we I mean, we were starting to run out of money.
大概只剩一个月的现金流时，我就给投资人发了封邮件。
So, when you we had, like, a month of cash left or something, I kinda emailed our investors.
我说：嘿。
And I was like, hey.
我们这儿已经有起色了。
We have something that's working.
他们倒问：你谁啊？
And they were like, who the hell are you again?
因为我都没怎么发过更新，啥都没管。
Because I hadn't been sending updates or anything, whatever.
这真是个大错。
That was a huge mistake.
你突然就一点动静都没了。
Oh, you go you went you went really quiet.
对啊。
Yeah.
公司要是不发消息了，那基本就凉了。
You know, a company's dead when they stop sending updates.
对吧？
Right?
没错。
Oh, yeah.
尤其是卡罗琳，我们都知道，一没消息准没好事。
Car Caroline, especially, what we know, when you don't hear, it's never good news.
可不是坏消息嘛。
Not good news.
对吧？
Right?
倒也不是完全没有例外。
Really one exception to that, maybe.
是啊。
Yeah.
那就会是个惊喜了。
That then it's a happy surprise.
你能给咱们的听众讲讲，Segment 的产品到底是干什么的？它解决了什么问题？
So can you explain for our listeners what Segment's product did and, you know, the problem it was solving?
嗯。
Yeah.
那时候，一大批 SaaS 公司突然就冒出来了。
I mean, there were all these both software as a service companies that were exploding onto the market at the time.
那时候，所有这些新推出的软件服务都跑在云上，通过浏览器访问，对吧？
All these new software offerings hosted in the cloud were online in the browser, right?
还有各种新推出的手机App，也是接二连三地冒出来，对吧？
And all these new mobile apps getting launched, right?
这都是2012年的事儿了。
This is like 2012.
Facebook的手机App火了，手机也一下子普及到哪儿都是。
Facebook's mobile app had gotten big and mobile phones were becoming just everywhere.
于是乎，各种手机App就跟雨后春笋一样涌进市场。
And so all the mobile applications were, like, flooding into the market.
所以当时市面上突然冒出一大堆新软件和服务，还有这些App。
And so you had all these new software services and apps out there.
它们都想知道，用户到底怎么用它们的App或者网站的，对吧？
And they all wanted to understand how people were using their apps, right, or or their websites, web apps.
为了搞清楚这个，它们都想收集数据——比如用户点了啥、买了啥、开了哪些邮件，所有这些跟用户互动的行为数据。
And, you know, in order to understand that, they wanted to get data about people clicking things and buying things and, opening emails and all that all that data about how people are interacting with their apps.
他们想把这些数据传到分析工具里，好看看用户都在干嘛。
And they wanted to send that data into an analytics tool where they could look at charts of what people were doing.
他们还想发到邮件系统里，好根据用户买没买过产品，发更精准的邮件。
They wanted to send it into their email tool so that they could send smarter emails to people based on products they had bought or not bought.
他们也想传到广告平台，看看哪些广告真能带来实际收入。
They wanted to send that data into their advertising tools so that they could see which ads were actually converting into into closed won revenue.
他们还想把这些数据扔进数据仓库。
They wanted to send it into their data warehouse.
所以他们有好多地方都想把数据发过去。
And so they had all these different places where they wanted to put their data.
就像你所说的，每次想加个新工具，都得去找工程团队：‘哎，我想把数据推到这个新系统里，能帮我接一下吗？’
And to your point, they would have to go and ask their engineering team every time, like, oh, I have a new tool that I wanna put my data in.
你能帮我们集成一下这个吗？
Can you integrate this thing?
时间一长，工程团队就烦了：‘别来烦我了。’
And then after a little while, the engineering teams are like, screw off.
我都已经搞过十次了。
Like, I already did this, like, 10 times.
是的。
Yeah.
太折腾人了。
Laborious.
对。
Yeah.
你就是一站式解决。
So you are the one stop shop.
把数据丢给Segment就行，剩下的我们来处理。
Just give it to give your data to Segment, and then we'll handle it.
没错。
That's right.
对。
Yeah.
所以不用工程团队去搞集成，市场人员或者任何人只要点个按钮，就能把数据发到Salesforce、Adobe、Google，或者任何地方。
And so instead of the engineering team doing the integration, a marketer or whoever could push a button to say, send my data to Salesforce, send my data to Adobe, send my data to Google, send my data to wherever.
明白了。
Okay.
这属于那种消费者根本不知道、但企业却急得不行的幕后功能。
One of these sort of unsexy things that consumers don't really know about, but the companies are desperate for.
没错。
That's right.
这个
This
说起来有点突然，你们是怎么定价格的？
is kind of random, but how did you decide what to charge the customers?
你们明明在做别人特别需要的东西，可到底怎么决定收多少钱呢？
Because you were obviously building something that they really wanted, but like, how did you decide what they were gonna pay you?
哎呀，我们根本没数。
Oh, we had no idea.
所以我们一开始定价是每月10美元。
So we started at $10 a month.
哇哦。
Wow.
那简直太离谱了。
Which was very clowny.
但我们当时来自开源和MIT的环境，真的一点都不懂该怎么定价，就像你说的那样。
But we were coming from this, like, open source MIT world, so we had no idea what to charge, to your point.
我们觉得既然是开源库，理应免费，结果却硬着头皮跟人说：‘麻烦每月付10美元吧。’
It seemed like it should be free because it was an open source library, so we very awkwardly asked people to pay $10 a month.
我永远记得，后来收到一封来自Eduardo的邮件，他是巴西一家叫Maxi Store之类的公司的人，我记不太清了。
I'll never forget we got this email back from Eduardo, Maxi Store or something like that, some Brazilian company, I think.
他回信说：‘好，我付每月10美元。’
And he emailed back, he was like, I'll take $10 a month.
但他接着说：‘不过你们要是长期不涨价，我可能就换别的了——你们这价格根本撑不起这个服务。’
But if you don't raise your pricing in general, like, I may decide to migrate off because there's no way that you can support this service at $10 a month.
因为我根本不尊重你。
Because I don't respect you.
是的。
Yeah.
其实就是因为
Well, it's just like because
你觉得自己迟早要倒闭。
you thought you'd go out of business.
对。
Yeah.
你要是还收十块钱一个月，那百分百是要倒闭的。
It's like a 100% you're going out of business if you're charging $10 a month.
所以他意思是特别希望你们能成功，但十块钱一个月真不行。
So he's like, really want you guys to succeed, but, like, $10 a month ain't it.
所以我挺感激他的。
So I was very kind of him.
后来我们雇了一位销售顾问，米奇·米兰达。
And then we hired a sales consultant, Mitch Miranda.
我都不记得是怎么跟他搭上线的了，但他直接跟我说：‘彼得，你该收的是每年12万美元，不是120美元。’
I forget how we got connected with him, but and he he was like, Peter, you should be charging a $120,000 a year, not a $120 a year.
我当时就觉得，这简直是我听过最离谱的事了。
And I was like, that's the most bananas thing I've heard.
是啊。
Yeah.
那是2013年吗？
And this was 2013?
大概是那会儿吧？
'20 around then?
2013年。
2013.
对。
Yeah.
2013年。
2013.
对。
Yeah.
2013年。
2013.
你真收了那么贵的费用？
Did you charge that?
我们第一次去见客户，就是跟马特·弗里德曼，那时候他还在管Xamarin这家公司。
Well, so we went to our first sales meeting, which was with Matt Friedman when he was running his company, Xamarin.
他们当时用我们的产品，好像是按每年120美元的套餐来的。
And they were they were using us, I think, on the $120 a year plan.
他就说：行吧。
And he's like, okay.
在去见他之前，我们在旧金山市中心，他就跟我们说：你们必须跟马特开口要每年12万美元。
Think before we went in, and this is in Downtown San Francisco, he's like, you have to ask Matt for a $120,000 a year.
我当时就说：这我可干不了。
And I was like, I can't do that.
这也太离谱了吧。
That's crazy.
价格直接涨了一千倍。
It's a thousand x increase in price.
他说：行吧，那我就不当你的销售顾问了。
And he's like, well, then I quit as your sales advisor.
说完他就转过身，真走了。
He, like, turned around and started walking away.
我当时就说：行吧。
I was like, okay.
行吧。
Okay.
行吧。
Okay.
我就说，我要提12万的薪资要求。
Like, I'll I'll I'll ask for a 120 k.
这真是硬核关爱啊。
What a tough love.
是啊。
Yeah.
于是我们就进去提了12万的薪资。
So we went in and asked for a 120 k.
我说完，感觉这事儿算完了，还挺兴奋的，反正就是那样。
And I turned you know, this is the end of the conversation, and I was excited or whatever.
然后我一下子脸就通红通红的。
And I just turned, like, just deep red.
我特别尴尬。
I was very embarrassed.
我记得他先要了1万2，我接着要了1万8，最后就定在1万8了。
And I think when he asked for 12 k, and then I asked for 18 k, and we settled at 18 k.
我妈，人特别好。
Mom Which was very nice.
销售顾问。
Sales consultant.
那销售顾问对18k这事啥反应？
How did the sales consultant feel about the 18 k?
他说：‘下次你肯定能谈得更好。’
He was like, well, you'll do better next time.
但说实话，我心里根本不在乎，因为在我眼里，这已经涨了150%了。
But, you know, in my mind, it didn't really matter because in my mind, it was already up a 150 x.
对吧？
Right?
我当时整个人都震惊了。
And so I was like my mind was blown.
你懂的。
You know?
我当时真没想到，哪怕再过一千年，这事儿也能成。
I was like, never in a thousand years that I think that it would be.
所以这个基于价值定价的想法，说实话，我一开始根本没搞明白。
And and and so the the whole idea of, like, value based pricing where I don't know.
我们其实每年能省下五百块左右，都是些他工程团队根本不想碰的垃圾集成活儿。
We were probably saving that, like, 500 a year or something on stupid integration stuff that his engineering team didn't even wanna do.
对他来说，这投资回报率简直爆表。
And so for him, it a great ROI.
但要真正理解‘按价值定价’而不是‘按成本定价’——毕竟那是个开源库——我们花了好长时间才想通。
And but this concept of, like, value based as opposed to, like, the cost based side of, like, it's an open source library took a long time for us to really learn.
然后你就慢慢把价格抬上去了。
And you slowly raised your prices.
下次我直接更有底气地开口要了十二万。
Well, I just more confidently asked for a 120 k the next time.
结果你猜怎么着？真给到二十四万了。
And lo and behold, we got 24 k.
然后我就想，行吧。
And then I was like, okay.
现在这事儿已经能复制了。
Well, like, now it's a repeatable thing.
接着我们拿到了36千，然后54千，再后来直接干到108千。
And then we got 36 k, and then we got 54 k, and then we got a 108 k.
后来Angie's List是第一个超标的。
And then Angie's List was the first over.
大概有24万左右吧。
Was like 240 k or something like that.
天啊。
Wow.
我就这么一路继续往上走。
So I kept going.
哇哦。
Oh, wow.
还行吧。
Not too bad.
彼得，这个……
Peter, this
这也太突然了吧。
is so random.
但我真想听听你这个故事。
But I wanna ask you this story.
你之前说过，你雇了个新销售，本来你根本没想到会招他，因为他太油嘴滑舌了。
You talked about how you hired this, like, new sales guy who was not someone you thought you'd hire because he was very slick.
但你从他那儿学到了不少，他业绩直接炸了。
But you learned a lot from him, and he crushed his sales numbers.
能讲讲这个故事吗？
Can you tell that story?
因为这事儿挺反直觉的。
Because it's sort of counterintuitive.
是的。
Yeah.
这绝对是创业过程中最让人谦卑的经历之一，真的算得上是其中最扎心的一次。
It's definitely one of the more humbling I mean, of the many humbling experiences in building the company, it's definitely one of the more humbling ones.
我们第一个销售是拉斐尔·帕克，那家伙简直太牛了，你懂的，啥都能卖出去，谁都买账。
So our our first sales guy, Rafael Parker, who was incredible, like, you know, could sell, I don't know, could sell anything to anybody.
他特别成功。
And and he was really successful.
我们做到了差不多一百万到两百万的年经常性收入。
We got to, I don't know, like, million or 2,000,000 ARR.
然后他说：行了，该招第二个销售了。
And then he was like, okay.
我说：太好了，拉斐尔。
It's time for us to hire our second sales rep.
太好了，拉斐尔。
And was like, great, Raf.
你去招人吧。
You hire.
我对销售一窍不通。
Like, I don't I don't know anything about sales.
你就直接雇这个人吧。
You should just hire this person.
他来了之后，带了个叫克里斯·萨尔迪的人来面试。
So he came in, and he and he brought Chris Surdy in and for an interview.
这人出生长大都在新泽西，头发梳得油光水滑，一副典型的兄弟派做派，满身纹身，却穿着西装。
And, yeah, new like, born in born and raised in New Jersey, slicked back hair, like, real bro y attitude, tatted up, but in, like, a suit.
你懂的。
You know?
那种感觉特别到位。
It was like a vibe.
真的很有范儿。
Real vibe.
我当时就想，这哥们儿真能搞定开发者吗，拉夫？
And and I was like, I don't think this guy can sell to developers, Raf.
我真的看不出来他有这本事。
Like, I don't I'm not seeing it.
他却说，你信我一次就对了。
And he's like, just trust me on it.
我就回他：行吧。
And I was like, okay.
那就干吧。
Like, let's do it.
我懂什么销售啊？
What do I know about sales?
于是我们雇了克里斯，结果他每季度都超额完成业绩，后来听说连续八年里有六年都是销售冠军。
So so we we hired Chris, and, yeah, he he beat the number every I think he was the top sales rep for, like, six of or six of the next eight years or something like that.
这事直到一年后，我跟着他去纽约跑销售，才慢慢明白过来——当时真觉得有点不可思议。
And and and the whole thing did really didn't make that much sense to me even for, like, a year afterwards until I started going on sales trips with him, like, to New York in particular.
我后来注意到，他有些对话听起来挺别扭的，但他特别主动去摸清客户到底有什么痛点。
And what I started to notice is that he some of the conversations were a little awkward, but he actually just was more aggressive about understanding what the customer's pain point was.
他就会硬着头皮往前冲，哪怕方式特别尴尬，也要把这事儿问清楚。
And he would kinda push through the awkwardness of trying to get to that, sometimes in very awkward ways.
比如在纽约有一次，他发了冷邮件，才让对方同意见一面。
Like, you know, there was one conversation in New York where it he had done some cold outbound to get this person to take a meeting.
我们到了之后，
And we then got there.
就我、他，还有个售前工程师，坐下来了。
It was, like, me and him and a sales engineer, and we sat down.
那客户就说：行啊，
And the guy was like, cool.
欢迎。
Welcome.
感觉全场的气氛都压在你身上，得赶紧开始推销自己了。
You know, like, energy is on you to, like, start pitching yourself.
克里斯·特里就问了：‘那我们到底为啥在这儿？’
And Chris thirty's like, so why are we here?
我一听，心都凉了。
And and, like, my heart stopped.
我当时就想：天哪。
I was like, oh my god.
接下来会发生啥啊？
Like, what is gonna happen next?
来自Segment的每个人都尴尬得不行。
And, like, everyone from Segment was just, like, like, mortified.
现场足足沉默了十秒，大家都在拼命缓和气氛。
And, literally, it was very awkward, like, ten seconds of, like, recalibration happening in the room.
然后那个人说：‘集成软件本来就不容易。’
And then the person that said was like, well, it's not easy to integrate software.
克里斯·特里就问：‘你这话啥意思？’
And Chris Terry's like, what do you mean?
他开始深入挖掘，做了大量调研。
And he started digging in and doing all this discovery.
结果他提前花了二十分钟，先问了一堆关于他问题的问题，而不是直接上手。
So he ended up doing twenty minutes before discovery, asking questions about his problems before he was trying, like, okay.
这挺棒的。
Well, cool.
我来给你看看Segment是干嘛的。
Let me show you what Segment is.
我本来尴尬得要死，但十分钟之后，我心想：卧槽。
And I was for like, went from mortified to then, like, ten minutes in, I was like, oh, shit.
你真得看看这事儿怎么发展的。
Like I mean see what's happening.
外联。
Outreach.
对。
Yeah.
他根本啥都不懂，但特别聪明地问出了所有需要的信息，好精准地做推介。
Like, like, he he went in knowing nothing, and he, like, very cleverly got all the info he needed to, like, pitch the right thing.
而且这事儿挺公平的。
That's And it's, like, it's fair.
他们确实答应了见面。
They took the meeting.
那他们为啥愿意抽出时间来谈这个呢？
So, like, why did they take the meeting?
为啥要专门腾出一小时来聊这个？
Why did they set aside an hour to talk about this?
他其实特别生气，心想：你凭什么这么浪费我的时间？
But he was almost, like, incensed of, like, how why are you using my time this way?
整个情况完全颠倒过来了。
It's like, the whole thing was just, like, totally reversed.
但这特别有意思，因为我根本不是个会卖东西的人。
But that's so interesting because I'm not a very good salesperson at all.
但有意思的是，到最后，他通过不断提问，就能顺理成章地说：‘这个能解决你的问题，还有这个、这个’，把整个推销话术量身定制出来。
But it's interesting because by the end of that, like, his his questioning, he could then say, well, this solves your problem and this and this and, like, tailor the pitch.
没错。
Yes.
反而那些看起来像推销员的人，一进来就给你来一套硬邦邦的推销话术。
And in fact, someone who looks like they're doing sales, who comes in and, like, gives this really hard pitch.
我完全搞错了客户的期待。
So, I totally missed the mark with the customer.
客户根本感觉不到自己被听见、被理解。
The customer is not going to feel heard or listened to.
这根本算不上什么好销售。
It's not actually good sales.
不知为啥，他就有股劲儿，能扛住探索阶段的那种尴尬。
And so for whatever reason, he's got the energy to push through the awkwardness of discovery.
客户反而特别喜欢，因为他讲的全是他们自己说出来的需求。
Customers And actually then love it because then he's speaking he's speaking to their needs because they've told them what their needs are.
克里斯·塞迪在Segment公司全程都在吗？他有没有离开过？
Was Chris Cerdy at Segment the whole did he ever leave?
我觉得他一直都在。
I think he was there the whole time.
他本人其实也是企业文化的重要传承者。
I he and turned out to be a huge culture carrier too.
什么‘团队合作让梦想成真’啊，简直像纹身一样刻在他身上，成了Segment内部的口头禅。
Like, teamwork makes the dream work was, like, you know, tattooed on his leg, and was like a it was like a a thing inside Segment.
是的。
Yeah.
天哪。
Oh my gosh.
这太棒了。
I love that.
我想问问，你们当时考虑被收购的时候，是怎么想的？
I wanna ask about, like, when you thought about getting acquired.
还是说你们自己考虑过被收购，还是说只是Twilio主动找上门的？
Or or did you think about it, or did you just get approached by Twilio?
是的。
Yeah.
在被收购前的几年里，杰夫每隔一个季度或者半年就会联系一次。
Jeff, for a couple years prior to the acquisition, had been kinda reaching out once a quarter or once every six months.
老是说：咱们合在一起会更好。
Being like, You know, we'd do better together.
你们有数据层，我们有通信层。
We've got the communication layer and you've got the data layer.
其实通信加数据，才能真正驱动数据行动，而数据也让通信变得更强大。
It turns out communication plus data is how you action the data, and data is what makes the communication better.
这两块合在一起，效果确实更好。
And it's like these two layers are better together.
但很长一段时间，我都没怎么信这个说法。
And for a long time, I didn't really buy it.
然后到了2019、2020年左右，我开始发现客户真的在主动提这个需求。
And then, in, like, 2019 2020, I started to see customers actually asking for that.
那时候我才真正意识到，关键转折点来了。
And so that was the that was the key shift for me.
客户们开始觉得这两块东西其实是紧密相关的。
It was, like, customers actually were starting to view these things as being adjacent to each other.
他们想把这两者以各种有意思的方式整合起来。
They were wanting to integrate them in interesting ways.
这下子，我的想法就彻底变了。
So that that shifted things in my mind.
不好意思。
Sorry.
咱们先回退一下。
Just to go back.
咱们刚才直接跳了六年，有点快了。
You bet we sort of skipped ahead, like, six years.
对吧？
Right?
不好意思。
Sorry.
呃，也不是那么回事。
Like, it well, no.
不是。
No.
我只是在想，你当初起步可真够艰难的。
Just more like I was thinking to myself, you had this really rough start.
你结婚了。
You got married.
那时候压力大得不行，各种事儿都堆一块儿了。
You were super stressed out, all the stuff.
这六年里，你肯定经历了创业和扩张带来的各种压力。
Those six years, you're obviously there's all these stresses that come from running and growing a company.
但总的来说，那几年还挺开心的，因为事情进展顺利，你也不用再往医院跑了，啥的。
But for the most part, were were these, like, happy years because things were going well and you weren't going to the hospital anymore and all this
我的意思是，确实比那个月好多了。
I mean, it's definitely It's still better than that month.
是的。
Yeah.
那个月算是最糟的了，好在就那一次。
That month was about as bad as it got, thankfully.
对。
Yeah.
我的想法是，应该还有另一个阶段，让我想想。
I mean, I think there was a I think there was another period in, like, let's see.
说不准。
I don't know.
大概是2014、2015年左右吧，我记得2014年我们推了个新产品，不仅能把数据发到分析工具，还能直接同步到邮件营销平台和数据仓库。
2014, 2015, something like that, where we I think in 2014, we launched a new product where you could send the data not just to analytics tools, but you could and email marketing tools, but you could send the data to data warehouses.
那是一个新类别，云数据仓库，比如Redshift之类的。
And that was a new category, cloud data warehouse, Redshift, and so on.
这是个快速增长的新领域，我们帮客户处理的数据，正好就是他们想要的。
Was a new category that was growing really explicitly, and the type of data that we were handling for people was exactly what they wanted in there.
其实，是和之前提到的销售代表SerDi一起出差时，我们才意识到这产品真找到市场契合点了。
And actually, believe it was a trip with SerDi, the sales rep we were talking about before, where we realized that this sort of had product market fit.
那天在纽约，我们一连跑了五场会，每一场客户都在把数据导出，存成某种特定格式，就为了把原始数据扔到某个地方。
And it was like, we were going to We had like five meetings in New York over the course of the day, and every single one was getting their data out into a particular format of just storing the raw data somewhere.
我当时就想，你们为啥要这么干？
And I was like, why would you want that?
那时候我明显感觉到SerDi的语气更主动、更急切了。
And now I was, like, getting this vibe of, like, more aggressive discovery from Serdy.
于是我开始深入追问这些客户，刨根问底。
And so I just started, like, really drilling into these customers.
到底为啥？
Like, why?
你们为啥要把数据放到 S3 上？
Why are you putting the data on s three?
结果发现他们都在做同一件事：想把数据加载到云数据仓库里，特别是 Redshift。
And it turned out they were all doing the same thing, which is that they wanted to load the data into a data warehouse, cloud data warehouse, and specifically Redshift.
所以五场会全是一样的反应，我从纽约回来就决定：咱们得赶紧做 Redshift 的集成。
And so it was, like, five for five meetings, and I came back from New York, and I was like, we're building an integration into into Redshift.
说到底，塞迪不仅帮我看清了销售的本质，还彻底改变了我对做产品的思路——让我学会更主动地深挖客户真正想要什么。
And so, actually, to wrap up on on on Serti, I actually created not only with helping me understand sales, but, actually, also, like, totally changing how I think about doing product, by doing, like, much more aggressive discovery of what customers really want.
幸好我问了关于塞迪的事。
I'm glad I asked about Cerdy.
我知道他起着关键作用。
I knew that he played an important role.
没错。
For sure.
而且不管怎样，从那以后，这事就开始飞速发展了。
And I think the and, anyway, that that was then started growing explosively.
所以第二年，我们的收入从二百五十万直接涨到了一千万。
So the next year, we went from, like, two and a half million in revenue to 10,000,000 in revenue.
天啊，真的吗？
Oh, wow.
是啊。
Yeah.
在SaaS行业，这已经快得不得了了。
For SaaS, that was was great.
不清楚。
Don't know.
现在AI公司干的事儿都太疯狂了，但那时候，这速度已经算很快了。
AI companies are doing crazy things now, at the time, that was that was pretty quick.
后来突然就慢下来了，因为客户都到了续费期，我们的定价和套餐出了问题。
And then and then it, like, all suddenly decelerated because everyone was coming up on their renewals, and we had a pricing and packaging issue.
本来计划第二年从一千万冲到两千万的。
And so we were gonna go from 10 to 20 the next year was the plan.
到年中时，我们的收入已经降到1200万，而且上半年最后一个月还缩水了50万。
And by the middle of the year, we were at 12, and we had actually shrunk 500 k in the last month of the first half.
哦。
Oh.
我当时就想：糟了。
And so I was like, I was like, oh crap.
对啊。
Yeah.
这情况可不妙。
Like, this is not good.
我总觉得，大家之所以还在，纯粹是因为之前增长得太快了。
And I had this feeling that, like, everyone was there just because it had been growing quickly.
所以谁都不太愿意把问题摊开来讲。
And so you, like, didn't wanna really share the problem that much.
我那会儿简直快疯了，整天焦虑怎么解决定价和套餐的问题，感觉这担子全压在我一个人身上。
And so I was, like, working I was getting pretty frenetic and, like, stressed about how how do we solve this pricing and packaging problem, and I felt like it was really on me to figure it out.
然后我们那个兼职CFO兼HR主管直接把我拉进一间屋子，跟我说：彼得，没人知道该干嘛。
And then our part time CFO and part time chief HR officer basically shunted me into a room, and they were like, Peter, nobody knows what to do.
我当时就说：你这话从何说起？
I was like, what are you talking about?
我说：对啊，真没人知道该干嘛。
I'm like, yeah, nobody knows what to do.
没人清楚目标是什么，啥都不清楚。
Nobody knows what the goals are or whatever.
你把所有压力都憋在心里，就是不告诉大家到底该干啥。
Like you're holding it all inside and like, you're not telling people what needs to get done.
我说：这太离谱了。
And I was like, that's crazy.
大家都清楚该做什么。
Everyone knows what to do.
他们就说：那你不如去随便找十个人问问，现在最要紧的事儿到底是什么？
And they're like, well, why don't you go talk to like 10 people and ask them what the top priority is right now?
我当时就说，对啊。
And I was like, yeah.
没人对到底该做什么有一致的理解。
Nobody had a shared understanding of what to do.
那真是我当领导的一次深刻教训，他们跟我说：你得站到全员面前。
So that was, like, that was, like, a real tough leadership learning moment for me where they were like, you need to get up in front of all hands.
你得清清楚楚告诉大家问题出在哪，每个人该做什么来解决它。
You need to tell people exactly what the problem is and what you expect each of them to do to solve it.
我就照做了。
So I did that.
我几乎觉得内疚，把担子全压在公司员工身上，而不是自己扛下最难的部分。
I felt very guilty almost of, like, putting the burden on on the people in the company as opposed to, you know, shouldering the hard part myself.
尤其是想到产品团队的那些人，我心里特别过意不去。
In particular, felt very guilty that there were, you know, all these people on the product team.
我心想：我得让这个产品团队赶紧把定价和包装的更新上线，真对不起你们。
I, you know, was like, I need this particular product team to go ship a pricing and packaging update, and I'm really sorry.
但现在都火烧眉毛了，我只能拜托你们连续加班，四天内必须把东西搞定，咱们才能翻盘。
But, like, because it's now the eleventh hour, I need you basically to stay up nonstop and and, like, just deliver it in the next, like, four days so that we can turn things around.
这件事我愧疚了好几年。
I felt so guilty about that for years.
而且
And
天啊。
and Wow.
好多年啊。
Years.
好多年啊。
Years.
是啊。
Yeah.
直接说结果吧。
To jump to the end.
因为这突然间给他们个人生活带来了巨大的冲击，要求实在太高了。
Well, because it was, like it was asking a lot of, like, very sudden massive disruption in their personal life.
然后跳到几年后，收购消息公布后，我们联系了当初那批元老，打了好些电话。
And then jumping forward several years to the acquisition, after the acquisition was announced, we did, like, phone calls with a bunch of the early people.
每一个曾经在那支团队里的人，都说那是他们整个公司生涯中最难忘的时刻。
Every single person who was on that team said that that moment was their favorite moment in the whole company.
哎呀，这太棒了。
Oh, I love that.
我一开始也震惊得不行，但回头想想其实挺合理的——他们等于救了这家公司。
I know it blew my mind blew my mind, but it makes sense in retrospect because it was like they got to rescue the company.
他们成了拯救公司的英雄。
They got to be the heroes.
我感觉我快被感动哭了。
Feel like I'm gonna get a little emotional over that.
别这样。
No.
你居然一直背负着这么重的愧疚，可他们那边却觉得，
It's crazy that you carried all this guilt, and they were meanwhile, like, that's,
简直太燃了。
like Stoked.
那简直太棒了。
That was awesome.
他们人生中最棒的时刻。
Moment ever.
是啊。
Yeah.
这种认知差距也太离谱了。
Like, what a that's crazy a disconnect.
这真是个特别棒的故事，
That's a really cool story,
不过。
though.
你还记得那次会上你说的话吗？
You remember what you said at this this meeting?
你是怎么讲这个的？
How did you present this?
你还记得自己说了哪些话吗？
Do you remember any of the sentences you used?
我记得当时气氛挺压抑的， basically 就是说，咱们的增长明显慢得不行。
I remember it feeling pretty dark and basically being like, we obviously aren't growing very quickly.
我们甚至在负增长。
We're in negative growth.
这问题可不小。
This is a big problem.
就是直白地说，问题就在这儿。
Like, it was just like, here's the problem.
除非解决这个，不然咱们肯定达不成目标。
We're not gonna hit the goal unless we fix this thing.
所以所有的压力都落在你们这四五个人身上，得赶紧把这事儿搞定，那感觉真是……
And so all the pressure is on you, this team of four or five people to, like, go fix this thing, which felt
但他们真做到了。
And they did.
然后他们真搞定了。
And then they did.
结果我们就在年底那天晚上11点50分，正好冲到了两千万ARR，现场简直疯了。
And then we had we actually hit 20,000,000 in ARR at, like, 11:50PM on the last day of the year, and it was, like, a crazy crazy scene.
玛莎。
Martha.
听着挺耳熟的。
Kinda familiar.
我感觉以前听谁讲过这个故事，就是关于这个时间点什么的。
I feel like I knew this story from someone told the story about that timing and everything.
这也太离谱了。
That's crazy.
12月31日晚上11点50分。
At 11:50PM on December 31.
对。
Yep.
我记得我们亚太区有一单，技术上算他们1月1日成交，但对我们来说是12月31日。
I think we had a deal in I think we had a deal in Asia Pacific that, like, closed technically on their January 1, but on our December 31.
所有人都还在办公室里，等着，等着，等着签合同。
And every everyone was still at the office, like, waiting, waiting, waiting for the signature.
是啊。
Yeah.
跨年夜。
On New Year's Eve.
跨年那晚。
On New Year's
当晚。
Eve.
对。
Mhmm.
哦，这真是个精彩的故事。
Oh, that's a great story.
是的。
Yeah.
哇哦。
Oh, wow.
我记得我和保罗专门去你们旧金山的办公室拜访，看看情况，那时候感觉你们业务发展得特别好。
I'm so I remember Paul and I came to your guys' office in San Francisco just to visit and check-in, and things seemed really flourishing and good.
我记不清具体是哪年了。
I forget when
那应该是去年吧。
the That was probably last year.
大概是那之后的第二年。
Probably the next year after that.
好的。
Okay.
对。
Yeah.
之后我们就跟公司说，要是达成了目标，明年就带所有人去哈利波特乐园。
After that, we, like, the and we had told the company that if we hit it, we would take everybody to Harry Potter land the next year.
所以我们真的就这么做了。
So that's what we did.
我们全体飞过去了。
We all flew down to it.
大概一月份把大家都送到洛杉矶，好好庆祝一下。
Have, like, put everyone down to LA in January or something and celebrate.
那时候团队有多大？
How big is the team
到那时候？
at this point?
天哪。
Oh my god.
我们快到一百人了吗？
Are we close to a 100?
哎呀，带这么多人去游乐园可真够热闹的。
Oh, that's a that's a crowd to take to an amusement park.
大家都想去那儿吗？
Did everyone want to go there?
是不是就在那儿
Is that where
对啊。
Oh, yeah.
这是大家一致决定的吗？
That was the collective choice?
等我们赢了的时候，人人都想去那儿了。
By the time we won it, everyone wanted to go there.
是的
Yeah.
当时是
It was
这真有趣。
That's funny.
哦，我喜欢这个故事。
Oh, I love that story.
我猜对你来说，这也是个成长的时刻吧。
And I bet it was, like, sort of a growing moment for you too.
因为很多创始人往往会把所有问题都扛在自己身上。
Because a lot of founders, I think, do sort of internalize all the problems.
有时候确实得这样，但有时候你也得跟团队说：听好了。
And sometimes you need to, but sometimes you do need to say, look, team.
咱们得这么干。
We need to do this.
我的意思是，大多数人加入公司，都是因为认同你正在做的事。
I mean, for the most part, sign up good people sign up to be at the company because they believe in the thing that you're trying to do.
所以他们真心想参与其中。
And so they want they wanna be part of it.
是的。
Yeah.
对。
Yeah.
没错。
That's right.
咱们还是回头说说Twilio吧。
We we should go back to Twilio.
就像那样。
It's like
该说说。
We should.
因为我们得赶紧推进Charm的事了。
Because we have to get going on Charm.
对。
Yeah.
那在收购过程中，有什么特别值得关注的亮点吗？
So any any highlights that would be interesting in the acquisition process?
我还想知道，你们当初是打算在什么时候做Charm的？
And I also wanna know what point you were thinking about doing Charm.
我知道，2015年在Segment的时候，你们买了一些替代方案，结果发现这事儿真够糟心的。
Because I do know that it was at Segment in 2015 that you bought know, you're buying offsets and realizing this all sucks.
这事儿太不透明了，感觉特别不舒服。
This is very opaque, and I've this doesn't feel good.
你能给我们讲讲吗？我知道这等于一口气问了二十个问题，但看你能不能捋清楚了。
Can you tell us I know that's about 20 questions in one, but see what you can do with that prompt.
对。
Yeah.
2015年的时候，我开始买这些碳抵消额度。
In 2015, I started buying these offsets.
到了2016年，我回头看看前一年干了啥，那时候正好是地球日，我就想，哎？
And then 2016, as I started to inspect what happened a year earlier, you know, it's like Earth Day twenty sixteen, I was like, hey.
去年地球日我们买的那些东西，到底咋样？
We bought that stuff at Earth Day last year.
它们加起来到底有啥效果？
What did what did it add up to?
我一看，坏了。
I was like, oh, crap.
根本没啥好结果。
Nothing good.
我当时就琢磨，哎，
And specifically, was like, hey.
亚马逊和印尼那边火势这么大，可我们之前不是还专门留了森林吗？
There's these, you know, fires raging in the Amazon and in Indonesia where we had normally set aside forests.
这些项目现在还靠谱吗？还是早就被烧光了？
Like, are they still any good, or did it, like, burn down?
我就想知道个明白。
I just wanna know.
根本查不到任何信息。
It was impossible to find out.
那时候我就觉得，唉，我对这些事的真实性真没底。
And so that was like, oh, I don't really trust what happened here.
所以我们后来换了个做法。
So then we did something slightly different.
我们开始买一些不太容易规模化的东西，比如制冷剂减排，但至少它的供应链比较透明，能审计，这让我从气候角度感觉踏实点了——不过说实话，制冷剂减排只占问题的0.1%左右。
We started buying things that aren't really scalable, like refrigerant disruption, but we're a bit had a bit more of, like, a tangible supply chain that you could audit, which I've made me feel better from a climate perspective, but it's like refrigerant disruption is only, like, I don't it's like 01% of the problem.
这点事儿根本解决不了大问题。
Like, it's not gonna get you very far.
所以到了2017年，我每个周六都泡在Segment，想看看有没有办法，把现有的工业流程改一改，让它从排放二氧化碳变成吸收二氧化碳。
And so 2017 then, I started spending my Saturdays around Segment trying to figure out if there were, like, pathways where you could take an existing industrial process and turn it from being an emitter of carbon dioxide into a sink of carbon dioxide.
后来这就演变成了Charm Industrial这个想法。
And that morphed into the idea for Charm Industrial.
名字叫Charm Farm Industrial Indenced，有点拗口。
And the name is Char Farm Industrial Indenced.
我们的想法是，从农场收集生物质，留下炭渣，再提炼出生物油。
And the idea is that we would take biomass off of farms, leave behind the char, and then you would have this bio oil.
说白了就是液态的生物质，能当工业原料，替代天然气、煤炭这些。
It's like liquid biomass, if you will, that could be used as an industrial feedstock to replace natural gas and coal and so on.
经济模型看着挺有希望，依然很靠谱。
And the economic models looked promising, still looked good.
你那时候
Were you
是自己一个人在周末做这些吗？
doing this on your own on Saturdays?
是啊。
Yeah.
对。
Yeah.
还有一位朋友，后来也成了Charm的联合创始人。
One other one other friend who became a a Charm cofounder.
然后，好吧。
And then Okay.
我们雇了两个工程师，主要由我出资——大概投了一半，另外一半是其他天使投资人出的，他们开始捣鼓这个项目，后来也成了联合创始人。
You know, we hired two engineers to kind of that I funded mostly, or I guess I've put in about half the money, another angel investors to start hacking on it who became cofounders as well.
所以，那就是当时的核心团队。
So that was the that was the kind of group.
周六的主力小分队。
The core Saturday crew.
是的。
Mhmm.
他们那时候还有别的工作吗？
And they were they had did they have jobs too?
关于 Bayt 播客
Bayt 提供中文+原文双语音频和字幕，帮助你打破语言障碍，轻松听懂全球优质播客。