多邻国数十亿产品战略背后的秘密：如何扩展至一亿用户 - 塞姆·坎苏

我说，我觉得我学不会韩语。

I said, I I don't think I'm gonna learn Korean.

你最终放弃了。

You gave up eventually.

是的，我放弃了。

I did.

我们刚结婚，我当时在努力学习。

We had just gotten married, and I was trying.

我当时想，我真的不行。

And I was like, I can't.

我真的做不到。

I can't do it.

我们后来聊了商业模式。

Well, we talked about business models.

我觉得这个商业模式存在挑战，因为你已经买了光盘。

I think this is a challenge of that business model because you already had bought the CDs.

所以，他们在教授韩语的方式上并没有太多持续改进的动力。

So there's not a lot of incentive for continuous improvement on how they teach Korean in a way.

是的。

So Yeah.

因为以前我们用CD的时候是那样的。

Because it was like a well, back in the day when we had CDs.

你知道的？

You know?

对。

Yes.

这暴露了我的年龄。

That dates me.

好吧，谢谢你。

Well, thank you.

好的。

Okay.

所以我们在Netflix时其实也非常严格，我认为我们的目标是娱乐全世界。

So so we were actually pretty ruthless at Netflix too, and I think, you know, we our goal was to entertain the world.

对吧？

Right?

为了娱乐全世界，你必须拥有最优质的内容。

And so in order to entertain the world, you have to have the best content.

但内容并不一定是你认为有趣的东西。

And and the content isn't necessarily what you think is, you know, entertaining.

而是世界认为有趣的东西。

It's what the world thinks, you know, is entertaining.

然后你还得能轻松找到让自己娱乐的内容。

And and then you have to you have to be able to find things easily to entertain yourself.

上传视频不能耗时太久。

The uploading of the video can't take forever.

就像，它必须能被看到，你知道的，所有这些方面。

Like, it's gotta be seen you know, all all those things.

所以我喜欢这一点，但我觉得很多公司都没有这种专注力。

And so I I like that, but I think that a lot of companies don't have that focus.

对吧？

Right?

他们确实没有。

They don't.

我认为这对我们来说显然非常好。

And I think it it obviously is great for us.

显然，这对你也有帮助。

Obviously, it's it's helping you.

但为什么团队和组织在专注上会这么困难呢？

But but why do teams, organizations have such a hard time focusing?

这是个很好的问题。

It's a great question.

我不确定有没有完美的答案，但我觉得可能有两个难点。

And I don't know if I have the perfect answer, but I would say maybe two things are hard.

一是，从后见之明来看，告诉我这个故事并说‘是的’总是很容易的。

One is it's always easy in in hindsight for me to even tell the story and say, oh, yeah.

我们曾有一个洞见。

We had this one insight.

我们围绕它开发了产品，效果非常好。

We built product around it, and it worked really well.

但我认为达成这个洞见的过程相当混乱。

But I think getting to that was a quite messy process.

意思是，我们一直在测试和投入，我的意思是，这并不是说我们只做了一件事，而是一个最终改变了多邻国增长轨迹的洞见。

Meaning, we were testing and invest I mean, it's it's also not like this is we did one thing, but it was one insight that ended up changing the trajectory of Duolingo's growth.

我认为获得这个洞见并不简单，你会不断追逐许多死胡同，直到最终找到它。

And I think it was it's not simple to get that insight, and you end up chasing a lot of dead ends until you get to that insight.

我认为，最终奏效的公式是快速迭代你的假设，直到找到那一个、两个或少数几个真正能推动业务的点。

And I think that I feel like the formula that ends up working is really fast iteration cycles on your hypotheses until you land on the one or two or a few that is really going to drive the business.

但我认为，达成这个过程看起来总是会很混乱。

But I think getting there is always going to look messy.

这需要一些时间。

It's gonna take some time.

但我认为，要达到最终状态——即明确什么真正会对业务产生影响——需要通过测试各种假设来实现。

But I think the way to get to the end state, which is clarity on what really is going to make a difference in the in the business happens through testing various hypotheses.

无论是产品问题、组织问题，还是业务问题，无论是什么，都必须亲自尝试。

Whether that's honestly a product challenge, an org challenge, a business challenge, whatever it might be, it really requires trying.

实际行动才能带来洞见，而不是靠白板讨论或撰写战略文档。

Like, action really gets you the insight rather than white boarding and writing strategy docs.

我觉得人们常犯的错误是做那些不是实际行动、而是讨论行动的工作。

I feel like people make the mistake of doing the work that is not the actual action, but the discussion of the action.

这真的无法推动你的思考前进。

And that really doesn't move your thinking forward.

至少我认为，这就是解决我的问题的方法，我

That's kind of what I believe at least is the solution to I a

你说得对。

think you're right.

就像我在思考的那样，我觉得很多时候大家都在讨论、分析和思考，但实际行动却不多。

Like, as I'm thinking about it, I think there's a lot of discussing and analyzing and thinking, and sometimes there's not a lot of doing.

我认为很多人，包括领导者和组织，都害怕风险和失败。

And I think a lot of humans, but leaders and organizations are afraid of risk and failure.

我们在Netflix的时候，有着非常强大的A/B测试文化，但同时我们也常说：干脆先试试看。

We were very I mean, we were a huge AB testing culture at Netflix, and and but we also were like, let's just try it.

如果行不通，我们就会知道它行不通。

And if it doesn't work, we will we'll know it doesn't work.

对吧？

Right?

如果你不去尝试，你就永远不知道结果。

And, like, there's no if you don't try, you don't know.

当你身处这些行业时——就像我以前和你现在一样——没人知道答案。

And when you're in these businesses like like I I was and you are, nobody knows the answer.

你只能边走边摸索。

You're making it up as you go along.

你不知道。

You don't know.

你得我

You gotta I

我认为，建立一种实验文化比你决定进行的具体实验更重要，因为它是复利增长的。

think that's why the building a culture of experimentation is more important than the specific experiments you decide to run because it compounds.

对吧？

Right?

比如，我们可以争论如何最好地进行订阅实验，但更重要的是组建一个能够大规模运行、提出良好假设的团队，这样更有成效。

It's like, think we can debate, for example, the best way to run a a subscription experiment, but it's way more productive to build a team that can do that at large numbers, that can run good hypotheses.

在六个月的时间里，与其我们争论哪个最好，不如让系统自行实验，这样能探索更广阔的领域。

And in a span of six months, rather than us debating what's best, we will cover way more ground by letting the system kind of experiment.

是的。

Yeah.

那么，作为实验体系的领导者，你如何让你的团队认同这一点？

So how as a leader of a system of experimentation, how do you get your team to to be on board with that?

你如何培养这种文化呢？

How do how do you foster it?

我喜欢这个问题。

I love this question.

我觉得我们至少在内部经常讨论这个话题。

And I I feel like we love talking about this a lot internally at least.

其中之一是让它变得简单。

One of them is make it easy.

比如，尽可能消除公司内任何人面临的障碍，至少就产品实验而言，我现在最感兴趣的是组织实验，这是我最近很喜欢的事情。

Like, if remove any friction to for anyone in the company as much as possible, at least let's take product experimentation, which I I also now my new favorite thing is org experiments, which is which is something I enjoy now.

让组织变革不像混凝土和石头那样固化，而更像一种实验。

Making org changes not like concrete and stone, but more like an experiment.

但无论如何，

But anyway,

让我来谈谈这个。

let I me get to wanna talk about this.

我喜欢这个。

I like that.

这太对我的胃口了。

This is right up my alley.

我猜到了。

I assumed.

但我觉得一个是减少摩擦。

But for I think one is reduced friction.

对我们来说，我们为自己使用场景开发了完整的A/B测试工具，并让任何团队都能轻松地进行实验。

For us, we built our entire AB testing tools ourselves for our use cases, and we made it quite easy for any team to go run an experiment.

这至少让起步变得容易。

That at least makes it easy to get started.

第二是鼓励团队保持实验精神，只要我们在展示给用户的设计或要做的变更上达成一致。

Two is encouraging teams to be experimental as long as we align on kind of what designs we're showing the user or what change we're going to make.

只要我们达成一致

As long as we're aligned

你不能更改标志。

You can't change the logo.

没错。

Exactly.

别改标志，或者我们有一系列设计原则，比如Duo必须保持某种特定外观。

Let don't change the logo Or we have like a bunch of design principles, like Duo has to look a certain way.

他必须是特立独行的。

He needs to be unhinged.

不管是什么情况。

Whatever it is.

我们了解到，这类情况有很多。

We've read there's many of these.

只要我们在设计规范和我们要做的改动上保持一致，团队就应该进行大量实验。

As long as we're aligned on on the design guidelines and what we're gonna do, teams should be running many experiments.

而且说实话，在文化上，做实验本身带有一点社交资本。

And honestly, it's a bit of a social currency culturally to run.

运行100个实验的团队比运行70个实验的团队更酷，这在某种程度上也是一种文化，因为这是好事。

A team that runs a 100 experiments is kinda cooler than the team that runs 70 is is is also the culture in a way because it's it's it's a good thing.

我认为这可能正是文化的一部分，就是鼓励团队进行实验。

And I think that's maybe maybe kind of the culture thing, is encouraging teams to be experimental.

是的。

Yeah.

你们在面对失败时有没有做什么特别的措施？

Is there something that you also do around failure?

因为很多实验都会失败。

Because a lot of the experiments fail.

对吧？

Right?

比如，你们会

Like, do

平均来说，它们都会失败。

of them fail on average.

比如你做了100个，你知道，有一半其实并不怎么样。

Like, you do a 100 of them and, you know, half of them aren't great.

还有一些只是还行。

And some of them are just okay.

就像，什么？

Like, it's like, what?

你知道的。

You know?

有一点，但还不够。

A little bit, but not not enough.

我觉得实验有一半失败是完全正常的。

I think it is completely expected that half of experiments will fail.

而且你提到了失败。

So and you mentioned failures.

我觉得当实验失败时，我们都不会在意。

I feel like we don't blink an eye when an experiment fails.

我认为，至少对于产品实验而言，我们共同形成的视角是：如果实验失败了，唯一要做的就是分析原因，然后看看是否能设计一个新的实验来解决失败的根本原因。

I think that the at least for for product experiments, the lens we have collectively created is if the experiment failed, the thing the only thing to do is look at why, and then if there's a new experiment we can run that would fix the why on why it failed.

如果不能，那就直接放弃。

If not, throw it out.

这里没什么可做的了。

There's nothing to do here.

我们继续下一个假设吧。

Let's move on to the next hypothesis.

我觉得与其说只是——我的意思是，庆祝失败当然重要，因为试错很重要，但我们的视角并不是关于庆祝与否，而是能从中学到什么？

I feel like that's rather than just I mean, celebrating failures is certainly important because trial is important, but I feel like our lens is rather than celebration or not, it's what can we learn from this?

一旦我们从中提炼出经验，就继续推进到下一个。

And once we've extracted that, let's move on to the next.

所以，听起来可能像是那种永远在拼命工作的态度，但我觉得我们就是这样看待的。

So to sound like maybe like a completely always on the grind mentality, I feel like that's how we view it.

就是说，好吧。

It's like, okay.

我们学到了什么？

What did we learn?

我们运行了这个实验。

We ran this thing.

洞察是颜色应该是红色，而不是绿色。

The insight was maybe the color should be red, not green.

好的。

Okay.

我们来试试三种不同的红色色调。

Let's try three different shades of red.

这就是接下来要进行的新实验。

That's now the new experiment to run.

对。

Right.

对。

Right.

是的

Yeah.

我非常喜欢这一点。

I I love that.

好的。

Okay.

那我们来谈谈这些组织实验吧。

So let's talk about these org experiments.

我很好奇。

I am curious.

告诉我吧。

Tell me.

嗯，我意识到，我认为进行组织调整总是很痛苦的，尤其是对于负责组织健康的人而言。

Well, I've realized, I think, making org changes is always painful, especially for someone that is responsible for the health of an organization.

进行组织重组，没人喜欢做。

Doing reorgs, nobody loves to do.

然而，这在某种程度上是一种心理技巧：如果你称某件事为试点，说我们先试一段时间，或者称之为组织实验，大家首先会更加开放地接受。

However, this is this is a kind of a mental trick in a way, which is if you call something a pilot that we're gonna try for a while, or if you call it an org experiment, everyone is, number one, way more open to it.

其次，如果它没有成功——因为组织实验有一半的时间都不会成功——你也很容易 revert 回去，因为你从未说过‘我要把某人固定安排在这里领导B部门’。

Two, it gives you the chance that if it doesn't work out, because org experiments also half the time don't work out, it's a very easy reversion because you never said, I am putting, you know, a person here to lead b.

你从来没有真正承诺过它。

You you never committed to it.

所以这使得尝试新事物变得更加容易。

So it makes it easier to try something.

因此，我现在倾向于将很多组织调整作为实验来提出，当然不是所有事情，但很多情况都是如此。

And that's why I I now I I propose a lot of org changes as experiments, not everything, of course, but a lot of things.

是的。

Yeah.

而且说实话，这让你更愿意尝试更多事情，从这些实验中学习更多，而不是像我刚接手这份工作时那样，每次组织调整都觉得：‘好吧，我得说服18个人。'

And it's it honestly gets you to try more things, to learn more from those experiments rather than I think when I first honestly stepped into my job, every org change felt like, okay, I gotta convince 18 people.

每个人都必须同意，然后我们才正式投入。

They all have to say yes, and then we are gonna commit.

我们下个季度就做。

We're gonna do it next quarter.

这变成了一种常态，大多数事情都可以先试试。

It became a whole thing where most things can be just tried.

如果成功了，我们就让它成为永久的。

And if it works, we make it permanent.

如果不成功，我们就恢复原状。

If it doesn't, we revert.

是的。

Yeah.

不。

No.

我非常喜欢。

I love it.

这实际上也是我在奈飞处理事情的方式。

And it it's actually how I approach things at at Netflix.

我记得，以前当我们在做这些事的时候，帕蒂也和我在一起。

And I remember, you know, back in the day and, you know, Patty Patty was with me too when we were doing these things.

但当帕蒂、团队和我试图取消假期时，人们都说：你不可能取消假期。

But, you know, when when Patty and and the team and I were, like, getting rid of vacation, right, it was like, you can't get rid of vacation.

没人会去休假的。

Nobody's gonna take vacations.

当时大家都很焦虑，这还是我们早期阶段的事情。

Like, you know, and everybody was sort of freaking out, and so this was the sort of early on in our thing.

但我们觉得，这就像一次产品测试。

But we were like, it's it's just like a product test.

我们要测试一下。

We're gonna test it.

我们来看看会发生什么。

Let's test what happens.

如果我们错了，我们就把假期重新加上。

And if we're wrong, like, we'll we'll put vacation back.

不是我们能不能把它恢复，而是我们需要尝试一下。

It's not we can put it back, but, you know, we need to try.

我们就试试吧。

And let's just try it.

如果有效，那就太好了。

And if it works, great.

如果无效，也没关系。

If it doesn't work, that's okay.

我们学到了东西。

We learned.

我觉得情况一样，因为当时公司里大部分都是工程师和产品专家，他们不喜欢组织结构上太多变动。

It's the same I think we tried to because at the time, most of the company were were engineers and product professionals, and they don't like a lot of change on the org stuff.

他们就是不喜欢。

They just don't.

他们就是不乐意。

They don't like it.

但我们试图让他们明白，这只不过是他们的本职工作。

But we tried to get them to see that it was just, like, their job.

就像我们不得不这么做，是的。

Like, we had to Yeah.

情况是一样的。

It's the same.

我们必须像你们一样进行创新。

We have to innovate just like you guys are.

你们就是我们的产品，不管你们愿不愿意

You're our product, be it what you want or

都不行。

not.

不管你喜不喜欢。

Whether you like it or not.

你们就是产品，所以我们一定会尝试。

You're the product, so we're gonna try.

我们要进行测试。

We're gonna test.

但我很喜欢这一点。

But I love that.

所以好吧。

So okay.

所以跟我们说说吧，也许有很多领导者在听这个。

So tell us about well, maybe we have a lot of leaders that listen to this.

所以跟我们说说最近你做过的一个成功的组织实验吧。

So tell us about maybe an org experiment that you've done recently that did work.

让我想想。

Let me see.

这可能不是一个确切的组织变革，但确实是我们尝试过的一个流程变更，而且最终成功了：我们和其他大多数公司一样，有定期召开的名为产品评审的会议，团队带着产品提案前来，一群包括我在内的高级人员——我们的CEO、设计负责人——会参与讨论，判断是否应该推进这个项目。

This is maybe not an exact org change, but it is a process change that we tried, it it ended up working, which is we have these, I guess, like most companies, these meetings called product review where they are scheduled, where teams show up with the product proposal and a bunch of senior people that includes me, our CEO, our head of design.

我们会讨论这个提案，决定是否要进行这个项目。

We discussed the proposal to say, okay, should we do this or not?

我们想尝试的试点或实验是，这些会议有时会偏离主题，走向非常奇怪的方向。

The pilot or experiment we wanted to try was these meetings some in some percent of the time would go in very wild directions.

比如，我们的CEO会抛出一个想法，我们会跟着追，然后团队在二十到三十分钟的会议结束后，却感到不确定我们到底达成了什么共识。

Like, would discuss, our CEO would throw something, we would chase that, and then the team would walk out out of a twenty to thirty minute meeting being like, I'm not sure what we aligned on.

所以这需要改变，但由于我们都关心产品，都有自己的观点，讨论有时会陷入循环。

So this needed to change, but it was since we all cared about the product, we all had opinions and it was sometimes circular discussion.

因此，我们决定严格规范会议格式，将时间缩短到十分钟。

So what we decided to do is really tighten the format, make it ten minutes.

会议必须遵循严格的发言顺序，谁先说、谁后说，CEO必须最后一个发言。

It needs to follow a strict order of who speaks first and the CEO has to go last.

很多人对此不太认同，觉得十分钟能讲清楚所有细节吗？

A lot of people were kinda averse to it because it was like, okay, ten minutes, what are we can we actually cover all the all the details?

当时支持的人不多，但我们决定先试三个月。

Not a lot of people were on board, but we said, we're just gonna try it for three months.

我们就先试三个月吧。

Let's just try it for three months.

你知道吗，我甚至有时会站在自己提案的对立面，说：‘我不确定这个方案会不会成功’，然后我们就在下一个季度把它改回来。

You know, I would I would even sometimes take the opposing view of my own proposal to be like, you know, I'm not even sure if it'll work, and then we'll revert it by by the next quarter.

我们试了试，当然，这种格式还是有些奇怪的地方。

And we tried it, and of course, there was some bizarreness in the format.

十分钟对于某些讨论来说实在太短了，但最终效果出奇地好。

Ten minutes is really short for some some discussions, but eventually, it worked out beautifully.

我们甚至都没讨论过要不要改回去，因为大家都觉得这好太多了。

And we didn't even discuss reverting it because everyone's like, this is way better.

没错。

Right.

大家都看在眼里。

Everybody saw it.

是的。

Yeah.

然后我们就不用再花几个月争论这个格式是不是对的了。

And then we didn't have to debate for months on, like, is this the right format?

这些是正确的与会者吗？

Is this the right attendees?

我们只是试了一下，结果就成功了。

It was just like, we just tried it and it worked.

这并没有直接回答你的组织问题，但它确实是一项流程变更，显著影响了产品经理如何进行产品评审。

So it wasn't exactly answering your org question, but it it was a process change that quite affected how, for example, product managers approach product review.

我非常喜欢这一点。

I love that.

你知道，我觉得这是一个我经常思考的问题。

You know, I think this is something that I think about a lot.

我们在工作中花费时间的方式，还没有得到足够的实验。

The the way we spend our time at work isn't experimented on enough.

对吧？

Right?

确实如此。

So true.

这简直就像单调乏味。

It's almost like monotony.

你知道，我们都知道有些会议很糟糕，但没人会取消它们。

Like, you know, we know there are terrible meetings, but nobody will cancel them.

是的。

Yes.

我们知道有些会议已经开了好几年，其实根本不需要再开了，但每个人还是继续参加。

We know that there are meetings that we've had for years that we don't need to have anymore, but everybody keeps going to them.

惯性是真实存在的。

Inertia is is real.

惯性真的非常真实。

Inertia is very real.

对。

Yeah.

我曾在播客上采访过一个人，他说了点挺有意思的事。

And I I had a guy on on the podcast, and he's like and it was interesting.

他说这跟你那十分钟的说法有关。

He was like and this goes back to your ten minute thing.

他说，如果会议已经排在日历上，标了一个小时，会持续多久？

He's like, if a meeting's on the calendar, do you know if for an hour, how long will it last?

我当时说，一个小时。

And I was like, an hour.

他问，如果会议在日历上标了二十分钟，会持续多久？

He's like, if the meeting's on the calendar for twenty minutes, how long will it last?

没错。

Exactly.

二十

Twenty

分钟。

minutes.

所以，如果它只有十分钟

So so it's like, if if it's ten minutes

就是十分钟。

It's ten minutes.

就是这样。

That's it.

分钟。

Minutes.

对吧？

Right?

所以我觉得，你们这么做我很喜欢，我们在Netflix也做了很多类似的实验。

Like so I think, you know, I love that you guys did that, and we we did a lot of that experimentation at at Netflix too.

因为again，我觉得如果你能实现运营卓越，就能取得很多成就，但我就是不明白，为什么公司都不愿意这么做。

Because, again, like, I think if you can have operational excellence, like, you can achieve a lot, but I but I just I'm so perplexed why companies are unwilling to go there.

有些时候，你懂我的意思吧？

Some you know what I mean?

也许是因为这样比较舒服。

Like, maybe it's comfortable.

我不知道。

I don't know.

但大家都明白这很糟糕，也都明白它可以变得更好。

But it's like everybody knows it sucks, and everybody knows it could be better.

但我们还是继续这么做。

But yet we just keep doing it.

就是，我不知道。

It's just I don't know.

这太有趣了。

It's so interesting.

我认为惯性是真实存在的。

I think inertia is real.

与其承认并可能进行一场尴尬的对话，说‘我想取消这个会议’，不如忍受持续的轻微痛苦来得容易。

It is easier to sit through consistent low pain than to acknowledge and maybe have an awkward conversation and say, I would like to kill this meeting.

所以，忍受持续的痛苦要容易得多。

So it's way easier to sit through consistent pain.

即使从累积的角度来看，情况也糟糕得多。

Even though if you think about it cumulatively, it's way worse.

很难举手提出一场令人尴尬的直接对话。

It is hard to raise your hand and have an acute awkward conversation.

大多数人会避免这种情况。

Think most people avoid that.

而为了避免这种尴尬，你最终会浪费大量会议时间，例如那些没有充分利用的时间。

And in avoiding that, you end up with a lot of meeting time, for example, that isn't well used.

是的。

Yeah.

在那些根本不重要的事情上，浪费了太多金钱和时间。

There's so much wasted money and time on, like, things that just don't matter.

我只是希望，如果公司能真正做到一点，人们会意识到，提出问题确实存在障碍。

And it's like, I just wish if companies could actually one, people would I do think that there is a problem with people raising their hand.

对吧？

Right?

而且，是的。

And and Yeah.

比如，有时候是恐惧，有时候有些人就是，我不知道。

Like and I don't sometimes it's fear or sometimes there are people like, I don't know.

我不在乎。

I don't care.

我还不够在意去，你知道的，举手发言。

I don't care enough to, like, you know, raise my hand.

我就只是坐着听完整个会议。

I'll just sit through the meeting.

但我确实认为人们应该这样做。

But but I do think people need to do that.

但我认为领导者，然后路易斯在这方面做得很好，比如里德在这方面就非常出色。

But I think leaders, and then probably Louis is pretty good at this, like fostering Reed was really good at this.

他会看到组织中出现的这些问题，比如沟通不畅、目标不一致、效率低下。

Like, he he would see these, like, things bubbling up in the organization, like miscommunication, misalignment, slowness.

你知道，我们并不想要任何官僚主义，对吧？

You know, we we did not want any bureaucracy or you know?

而且也不是说我们没有流程，但我们不想搞那些愚蠢的流程。

And and it's not like we didn't have process, but we didn't want stupid process.

对吧？

Right?

但当我们开始看到这些东西悄然滋生时，这让他感到焦虑，于是他会介入系统，进行调整、实验，试图消灭并去除它们。

But the moment we started to see that stuff creeping, like, it was anxiety for him, and he would, like, get in the system and tinker and experiment and try and kill and get rid of.

我的意思是，‘消灭’听起来很糟糕，但有时候你确实必须消灭某些东西，因为它们会害死你。

And, I mean, kill sounds so terrible, but sometimes you you have to kill things because they'll kill you.

如果你不消灭它们，它们就会害死你。

They'll kill you if you don't kill them.

我的意思是，作为领导组织的人，最重要的资源就是领导力的精力。

I mean, the the biggest the most important thing as people who lead orgs is the most important resource you have is leadership bandwidth.

任何消耗这种精力的事情，哪怕只消耗了2%，你也必须极其审慎地质疑：我是否把这2%用在了正确的地方？

Anything that chips away at that, even if it chips away 2%, you have to be exceedingly questioning of am I spending my 2% on this correctly?

我觉得有时候人们只是忘记了这就是现实。

And I think sometimes people just forget that that is that is reality.

就只是另一个二十分钟的会议而已。

It's just like, it's another twenty minute meeting.

我就去参加一下吧。

I'll just attend it.

但这可是每周都有。

Well, it's every week.

所以一年下来，你在这上面花了不少时间。

So in the course of a year, you spent quite a bit of your time on this thing.

你得到了什么回报？

What return did you get?

我觉得你也会觉得，Luis 在这方面也非常出色。

And I think you are like, Luis is is also excellent at this.

如果某件事根本没达到它原本的目的，首先该讨论的是：我们今天要不要取消它？

If something is just not doing the job it was intended to, the first thing to discuss is, should we kill this today?

事情就是这样。

Is how it goes.

我也坚持这个观点。

And I I I stand by that too.

我学到的一个方法——不是技巧，而是一种健康的习惯——就是我的大多数会议都不会设置为永久重复。

I I do this the one trick I've learned, not trick, but I think a healthy habit with meetings is most of my meetings are not scheduled to recur forever.

它们都有结束日期。

They always have an end date.

因此，我发现这是一个非常好的习惯：定期回顾，判断是否应该继续所有会议，或者因为它们非常有价值而重新安排。

So that is a very good habit I've found to revisit if everything should continue, or should we reschedule them because it was so valuable.

重新安排是更好的主动参与方式，因为你是在重新选择加入，而不是任由它在日历上自动重复。

Rescheduling is a way better way to opt in because you're re opting in than just letting it recur on your calendar.

我喜欢这个做法。

I like that.

这真是个很好的技巧。

That is a very good trick.

听好了，各位领导者。

Listen up, leaders.

给这个会议设置一个结束日期。

Put an end date on that meeting.

好的。

Okay.

你知道吗，也许这次通话中的一些人会想了解这一点。

You know, maybe this is something that some of the folks on the call might wanna know.

但作为产品领域的领导者，有没有过这样的时刻——你真的有一个想法，也许它违背了所谓的最佳实践，或者你想怎么称呼都行。

But is there a time, like, when as a leader for product where you've you really had an idea, maybe it was against best practices or whatever you wanna call it.

我不喜欢这个词。

I hate that term.

我想知道，是谁说这些就是最佳的？

Who said they they were best is what I wanna know.

谁决定的？

Who decided?

谁决定的？

Who decided?

对吧？

Right?

我想知道是谁定的，说这些做法都是最好的。

I wanna know this person who decides that all these practices are the best.

但你有没有过这样的时候，你是那个孤身一人认为这个主意很棒，却依然坚持推动它的人？

But was there a time where you really, like, you were the kind of lone person out there thinking that this was a great idea and you pushed for it anyways?

最近几年里，可能有几个项目是我大力推动的，当时并不明显。

Maybe there's a couple initiatives in recent history that I've certainly been part of pushing really hard on that wasn't obvious at the time.

我选一个现在看来相当有名的例子，就是Duolingo的Unhinted品牌，我不知道你们有没有用过我们为应用提供的小部件，它本质上是一个四图标大小的小部件，可以添加到你的iOS或Android设备上。

I'll I'll pick one which I think is is now quite famous as part of Duolingo's Unhinted brand, which is I don't know if you use our widget for for your app, which is basically this four app icon size widget you can add either on your iOS device or Android.

它主要用在天气类应用中，比如你可以以更大的视图查看天气，或者历史上人们会在这里放一些信息，比如股票价格之类的。

And it was it's mainly used with, like, weather apps where you kinda look at your weather in a bigger view or it historically is like you put some information there like your stock price, whatever it is.

但它从未被用作一种有趣的方式来传达你的品牌或应用，而这在当时根本不是一种被认可的做法。

But it's it was never quite used as a fun way to communicate your brand or your app, and that was not a thing that was done.

我当时有个假设，我们的品牌非常有趣，而且有很多用户对他们的连续打卡有着忠诚的习惯。

I had this hypothesis that we have such a fun brand and and a lot of people with loyal habits to their streak.

连续打卡是我们用户群体中最重要的元素之一。

The streak is one of the most important things to our user base.

它是一种社交货币，用户用它来向其他用户展示。

It it is it is the social currency you communicate it to other users with.

你会说：我有100天的连续打卡记录，而别人则用他们的1000天记录超越你，无论具体是多少天。

You say, I have a hundred day streak, and someone else trumps you with their thousand day streak, whatever it might be.

但这是一个非常重要的点，我真的很希望我们的用户安装我们的小部件，里面会显示他们的打卡天数，Duo会呈现这个数字。

But it is an important thing, and I really wanted our user base to install our widget, which would have your streak number in there, and Duo would would present that number.

这其实是个相当简单的东西。

It's a fairly simple thing.

当时并没有人完全相信这对Duolingo会是一个巨大的突破，但现在几乎有50%的日活跃用户在使用我们的小部件。

It was not fully believed that this would be a big big thing for Duolingo, but now almost 50% of our DAUs use our widget.

所以它的实际采用率相当高。

So it's actually a quite high adoption.

我不确定我有没有。

I don't know if I have it.

我强烈建议你安装它。

I would highly encourage you to install it.

这非常有趣。

It's very fun.

我们在里面加入了很多有趣的Duo时刻，如果你愿意，也可以把你的小部件展示在主页上。

We put a lot of funny Duo moments in there, and you can showcase your widget on your homepage if that's also what you would like to do.

如果你还没有连续打卡，那你得先建立一个连续打卡记录，才能展示出来。

And if you don't have a streak, you gotta get a streak so that you can showcase it.

是的。

Yeah.

嗯，我孩子用我的手机。

Well, my kid uses my phone.

好的。

Okay.

用Duolingo。

With Duolingo.

太棒了。

Excellent.

那其实是他的连续打卡记录，所以我不能把功劳归于自己。

That solves So it's really his streak, so I can't take credit for it.

但我还是要打开小部件，因为我觉得我不仅仅是在App里使用。

But I'm gonna turn on the widget because I don't think I'm just on the app.

我觉得我没有这个功能。

I don't think I have it.

所以。

So

好的。

Okay.

不过，不深入讨论产品功能的细节，过去通常不会把小部件用于娱乐用途。

Well, without going into more of the details of of the product feature, it was not common practice, and it was widgets were not used for fun stuff.

意思是，你的猫头鹰吉祥物现在不会跑出来展示它的屁股了，因为这符合我们的评分标准

Meaning, your owl mascot wouldn't come and show his butt now that, you know, this fits our rating

太好了。

Great.

在小部件上。

On the widget.

我们的小部件确实如此。

Our widget does.

因此，我们为娱乐性小部件采取了完全不同的立场，这并不符合常规做法。

So we took a completely different stance for an entertaining widget, which was not conventional.

这肯定不是任何应用的最佳实践，但我认为它确实引起了我们用户的共鸣，因为他们喜欢Duo，也喜欢他们的连续打卡记录，这正是小部件的核心所在。

It certainly wasn't best practice for any app to do, but I think it really ended up resonating with our audience because they love Duo and they love their streak, that's what the widget is all about.

正如我们之前讨论测试时提到的，我们了解了哪些方法有效。

And as as similar to what we talked about with testing, we learned what works.

我们在小部件上测试了很多东西，最终，这些正是让用户感到兴奋的领域。

We tested a bunch of things in our widget, and eventually, this is this is the realm of things that was really exciting for our users.

那是我真正坚定支持的一件事。

That was one of those things I really stood behind.

我认为公司里大多数人并不相信这会有什么效果。

I think most of the company did not believe this would do anything.

是的。

Yes.

但它确实有效。

But it did.

但它确实有效。

But did.

你一直坚持它。

You you stood for it.

那么，对于处在类似情况的人，你有什么建议？

So what's the advice for someone in a similar situation?

是仅仅靠决心，还是说，你是如何顶住反对声音的？

Was it just determination or, you know, how'd you stay strong against the naysayers?

我觉得可能有两个方面。

I think maybe two things.

其中之一是，对于某些事情，如果你已经高度确信那是正确的做法，那么即使大多数其他有判断力的人认为这可能不对，你也应该尽力坚持下去。

One of them is for some things, if you have reached high conviction that that is the right thing to do, you should obviously do your best to stand by it even if most other high judgment people believe it might not be the right thing to do.

所以这是第一点。

So that's one.

我认为，如果你有高度的信念，就不要保持沉默。

I think if you have high conviction, don't stay silent.

每当我这么说时，大家都会问一个显而易见的问题：我该如何对一个尚未被证明的事情产生高度确信呢？

Now there's always when I say this, the obvious question that I get is, well, how do I get high conviction on an unproven thing?

我认为，这取决于具体是什么事情，可能还有更具体的办法。

And I think there's depending on what that thing is, there's probably more specifics.

但对于产品来说，最好的方法是制作原型并可视化它，然后自己先去使用测试。

But with product, the best way is to prototype and visualize it, and you test it on yourself.

如果你自己很喜欢，并且看到身边的人也喜欢，这就非常有力地表明，世界上还有一群人会喜欢你正在打造的东西。

If you love it and you can see other people around you loving it, that is an extremely good sign that there's another group of people in the world that will love the thing you're building.

所以对于产品，你要先自己测试。

So with product, you test it on yourself.

可能对于任何事情，你都该先自己测试。

Probably with anything, test it on yourself.

如果你已经说服自己这是正确的事，那就是一个很好的高信念起点。

If you've convinced yourself that that is the right thing to do, that is a great high conviction starting point.

对于组织变革也是如此，对吧？如果你想为别人安排一个十分钟的会议，先自己试一次十分钟的会议，看看是否可行。

And with org org changes too, right, if you wanna do a ten minute meeting for others, do a ten minute meeting on yourself first to see if it works.

所以我会从这里开始。

So that that's where I would start.

第一步，培养信念。

Grow conviction, number step one.

第二步，如果你已经做到了，那就坚定地坚持下去。

Step two, if you have done that, obviously, stand by it.

第三步，要承认，任何实验中有一半都会失败。

And step three, acknowledge that, again, half of any experiment will not work.

所以，即使你觉得自己信念坚定并坚持了下去，最终仍有可能失败，但这是正常的。

So even if you think you have high conviction and you stood by it, there's a chance of failure at the end, but that's normal.

意思是，你会坚持十件事，其中五件可能不会成功。

Meaning, you will stand by 10 things and five of them will not work out.

而我的记录到现在可能更糟了。

And that's certainly, my track record might even be worse by now.

如果我坚持了十件事，大概只有三四件成功了，其余六件都失败了。

If I stood by 10 things, probably four of them, three of them worked, and six of them did not.

所以是的。

So Yeah.

这实际上就是事情的运作方式。

And that's honestly just how it works.

我总是觉得，真正出色的产品领导者就像优秀的棒球击球手。

I always thought about, like, like, really great product leaders were, like, really great baseball hitters.

因为那些顶尖击球手的失误次数比安打次数还要多。

Like, the because, like, the big hitters have more misses than they have big hits.

但那些重磅安打可以将留存率提升百分之五，或者你打出一记本垒打，就能赢得比赛。

But the big hits can increase retention by, you know, 5% or, you know, it's or you, you know, you get a home run and you win the game.

我的意思是，这确实很对。

I mean, that's very true.

击球次数才是关键，我认为这在商业或产品中也是如此。

At bats is what matters is is is I think how think how it goes for business or product.

如果你有更多击球机会，第一，你打出本垒打的能力会提升，从而获得技能成长。

If you get more at bats, one, your your muscle of hitting these home runs get better, so you get the skill development.

第二，你获得击出本垒打的机会也更多。

But two, you get more chances at hitting home runs.

所以我认为，是的，棒球这个类比我们其实也经常用。

So I think, yeah, the baseball analogy we use a lot too, actually.

真的吗？

Do you?

好的。

Okay.

很好。

Good.

对。

Yeah.

我一直以来都是这么想的。

I always thought about it that way.

好吧。

So okay.

我们时间到了，但我还想和你聊聊人工智能，因为大家都在谈AI。

We're ready out of time, but I wanna talk to you a little bit about AI because everybody's talking about AI.

但我认为，AI在学习方面有一个非常棒的机会。

But I think that there is a very cool opportunity with learning with AI.

我知道你一定在思考这个问题。

And I'm I know you're you have to be thinking of this.

你很可能已经在做了。

You probably are doing it.

但你能否与我们分享一些信息，或者你觉得听众在这一方面应该关注些什么？

But is there anything that you can kind of share with us or any things you think listeners should just be paying attention to on this front?

当然。

Definitely.

我想说的是，多邻国大量使用了人工智能。

I would say we have Duolingo uses AI heavily.

让我从这里开始说起。

Let me start there.

如果我要归纳人工智能对多邻国带来的三种变革性影响，首先是内容创作。

And if I was to categorize the three ways it's been transformational to to Duolingo is one is content creation.

多邻国，如果你想想我们教授的所有内容，我们教授的语言超过40种。

So Duolingo, if you think about all the the things we teach, we teach more than 40 languages.

我们还教授数学、音乐和国际象棋。

We teach math, music, and chess.

创建既有效又有趣的内容是一项极其艰巨的任务。

Creating the the content that is effective and fun is an extreme extreme task.

这是一项繁重的工作。

It's it's a lot of work.

现在，我们的许多内容创作都借助了AI，因为我们能够以更高的规模和更好的质量完成。

And now a lot of our content creation is helped with AI because we can do it at much higher scale and at higher quality even.

这是第一点。

So that's one.

因此，内容创作通过使用AI得到了极大的变革。

So content creation is quite transformed by using AI.

顺便说一下，Duolingo 有时会被报道说，AI正在取代Duolingo的员工队伍。

And this by the way, Duolingo sometimes has been in the news of like, oh, like, AI is replacing Duolingo's employee base.

这完全是误解，我想澄清这个谬误。

That is a complete I just wanna debunk that myth.

AI真正帮助我们的员工提高了效率，这基本上就是我们想说的全部。

AI has really helped our employees be more productive, and that's honestly the end of our story there.

是的。

Yeah.

不。

No.

我的意思是，这太有趣了，因为很多公司，比如亚马逊最近的裁员，大家都说这是AI造成的。

I mean, it's so it's so funny because, you know, a lot of companies, like, the the Amazon layoff recently, and everybody's like, oh, that's AI.

拿走吧。

Take it.

我觉得。

I'm like,

不。

no.

不。

No.

实际上并不是这么回事。

That's not the story, actually.

并没有发生。

It's not it's not happening.

人工智能并不是在取代工作。

Like, AI isn't taking the jobs.

你知道的。

You know?

是的。

Yes.

它只是从生产力角度提升了工作效率。

It's just it's it's it's enhancing things from a productivity perspective.

没错。

Exactly.

我觉得我们正在见证这一点。

And I think we're seeing that.

比如内容创作，我们用同样数量的员工能产出十倍的内容，我们非常喜欢这样。

Like, content creation, we can do 10 x more, but with the same number of employees, and we love that.

我们做的另一件事是功能开发。

The other thing we're doing is feature creation.

对于语言学习来说，能够练习口语是一种非常棒的体验。

So with language specifically, it is such a good experience to be able to practice speaking.

在大语言模型达到如今这种能力之前，我们没有很好的技术来提供可以自由练习对话的口语功能。

And before LLMs were this capable, we didn't have great technology to provide speaking features where you can freely practice conversation.

现在我们可以了。

Now we can.

因此，我们利用大语言模型来创造大量的对话练习。

So we use LLMs to create a lot of conversation practice.

我们有一个功能叫做视频通话，你可以拨打我们其中一个角色的电话，她的名字叫Lily，你可以和她交谈，她会根据你所学语言的确切水平进行调整。

It's a feature that we call video call where you call one of our characters, Lily is her name, and you can talk to her, and she adjusts to your exact proficiency level of the language that you're learning.

我们正在解决以前用A无法解决的问题

So what we're solving problems we couldn't do before with A

这太棒了。

Which is amazing.

是的。

Yeah.

是的。

Yeah.

当我的儿子用他六岁的小孩子声音对着电话说话时，有时候他并不明白自己在说什么。

When my son is, like, speaking into the phone in his little six year old voice, sometimes he doesn't understand what he's saying.

但这更多是因为他无法，你知道的，他发音不准。

But it's it's more because he can't you know, he's not pronouncing things.

公平地说，我们必须改进语音识别技术，因为这不仅仅是你儿子的问题——当你使用一门不熟练的语言时，有时会说些胡言乱语，这是一个很难解决的语音识别问题。

We to be fair, we have to get better at speech recognition because it's it's not only your son because it's a language you're not proficient in, so you say gibberish some of the time, and that's a hard speech recognition problem to solve.

是的。

Yeah.

我相信人工智能会帮助解决这个问题。

Well, I'm sure AI will help with that.

人工智能肯定会帮助解决这个问题。

AI will certainly help with that.

所以，这属于我的第二类问题：我们能够解决以前无法解决的新技术难题。

So this is kind of my second category, which is we're able to solve new technical problems we couldn't solve before.

很好的对话练习和语音识别，我相信未来还会有其他应用。

Great conversation practice, speech recognition, and I'm sure there'll be others in the future.

第三点我想说的是通用生产力。

And then the third I would say is general productivity.

我觉得我们的写作现在更好了，因为你可以让AI帮你校对任何内容。

I think our writing is better now because you can have AI proofread whatever you're doing.

我每天都把它当作写作助手使用。

I use it as a writing assistant every day.

比如，我们的工程师也在用它来辅助编程。

And and, for example, our our engineers are using it for helping their coding.

我们能够更容易地做出原型。

We are able to do prototypes easier.

所以，无论如何，每个功能使用它的方法都不同，但生产力确实有所提升。

So, anyway, every function uses it differently, but really productivity has has has gone up a bit as well.

这三点大概总结了我们目前使用AI的大部分内容。

Those three really probably summarize most of the stuff we're doing with AI.

我们将其视为推动我们所有工作的全面加速器。

We see it as a complete accelerant to everything we're doing.

所以，老实说，我非常期待看到我们还能用这些技术做些什么。

So we're very honestly, I'm very excited to see how what else we can do with all of these.

是的。

So yeah.

我也很兴奋。

I'm excited too.

我迫不及待想亲身体验了。

I can't wait to experience it.

好的。

Okay.

最后一件事，然后我们就进入职业坦白环节。

One last thing, and then we'll we'll get into career confessions.

所以我总是请嘉宾给听众一些建议，比如任何你想留给他们的、一点有价值的建议。

So I always ask my guests to give a little advice to the listeners, like anything that you wanna leave them with, a little nugget of something.

那么，你今天对我们听众的建议是什么？

So what is your advice to our listeners today?

我想强调一下你在其中一个问题中提到的内容，就是你如何获得洞察力？

I think I wanna underline something we said in one of the questions you asked me, which is how do you get to insight?

我认为我的回答大致是：去行动。

And I think my answer was somewhat along the realms of do action.

行动能带来你需要的洞察力。

Action will get to insight that you need.

我们有一个运营原则，我曾经多次提到过，最终它成了我们的正式原则。

And we have this operating principle that I I guess famously, said so many times that we eventually made an operating principle.

我们常说‘先发布’，意思是去行动，而不是空谈。

We we say ship it, and that is meant to basically underline the fact that go act on the thing rather than debate the thing.

老实说，这个方法在我的职业生涯中一直非常有效。

And I've honestly, it's it's a it's a device that has served me really well in my career.

正因如此，我经常重复这句话。

It's something that I repeat a lot for that reason.

就是去行动，做这件事，或者在产品领域，就是发布它，交给用户，你会学到更多。

Just do like, take action, do the thing, or in in the, I guess, the product realm of things that is just like ship it, give it to users, you'll learn more.

这是我想要给出的一个建议，因为我觉得很多领导者是你们的听众。

That is one advice I will say because I think a lot of, I mean, leaders are your audience here.

坐在房间里争论太容易了，因为人类天性就是追求意见一致，但行动起来并看看会发生什么要快得多。

It's so easy to sit and debate in rooms because that's human nature with to get consensus on opinion, but it's way faster to act on it and see what happens with it.

确实如此。

So so so true.

发布它吧，各位。

Ship it, people.

来吧。

Come on.

就发布它吧。

Just ship it.

好的。

Okay.

那我们进入职业坦白环节吧。

So we're gonna roll into career confession.

好的。

Okay.

我会先提供一点背景信息，然后我们再进入问题。

So I'll set it up with, like, a little context, and then we'll get to the question.

有时我会发现自己在关注数字、日活跃用户、留存率、曲线和变现指标之前，没有先问一句：我们是否仍在服务真实的人？所以我的问题是，在面临变现和扩张压力时，你如何带领一个以使命为导向的产品团队，同时不失去同理心、真实性和长期的用户信任？

Sometimes I catch myself optimizing for numbers, daily active users, retention, curves, monetization metrics before asking, are we still serving real people's So my question is, how do you lead a mission driven product team, especially under pressure to monetize and scale without losing empathy, authenticity, and long term user trust?

哇哦。

Woah.

我喜欢这个问题。

I love this question.

感觉这个问题非常有针对性。

It feels very targeted even.

确实如此。

It does.

确实如此。

It does.

这很适合你。

This is right for you.

我确实很喜欢这个问题。

I do I do love this question.

我想说，可能有两种方法对于引导这艘试图平衡商业目标（如收入增长、日活跃用户增长）与公司使命目标（比如多邻国提供的优质教育普及）的船至关重要。

I will say maybe there's two approaches that are very important to guide this ship that is trying to balance business objective that is like revenue growth, DAU growth, to maybe the mission objective of whatever that company might be doing for Duolingo that is providing access to great education.

第一种是，你能否为你们试图实现的使命设定一个可量化的指标？

One is can you put a metric behind the mission that you're trying to solve?

这至少能为这一目标赋予数据驱动的力量和商业推动力。

That at least puts metric power and business power behind that thing.

有时这并不可能，但如果可以为某个目标设定指标，比如，如果公司的使命是——我不知道。

Sometimes this is not possible, but if it is a thing that you can put a metric around for example, if the mission of the company is to I don't know.

随便选一个使命，比如我们希望帮助饥饿的儿童，那么一个合适的指标就是：我们帮助了多少儿童？

Pick any mission, it's like that we would like to help starving children, then okay, a metric is how many children have we helped?

你能在业务中设定一个指标，并衡量你对这一指标的影响吗？

Is there a metric you can put in the business and can you measure your impact towards that?

对我们来说，其中一个棘手的问题是，我们可以衡量用户数量和收入，但很难量化我们实际教授了多少内容。

For us, for example, one of the hard parts that we have struggled with is it is hard to put a metric on we can put number of users and revenue, but it's hard to put a metric on how much we're teaching.

教学很难用图表来累积，很难说清楚我们到底完成了多少教学。

Teaching is hard to accumulate on a graph and and put here's how much teaching we've done.

这真的很难做到。

That that's really hard to do.

我认为，这时你就得依靠自己的判断和定性判断。

I think then you lean into your your judgment and your qualitative judgment.

我觉得很有帮助的一点是，首先，领导者必须真正认同这一使命。

The thing that I think helps a lot is one, obviously, leaders must be bought in.

公司从上到下，每一位领导者都完全认同这一使命。

Every leader in the company from the top to any level is completely bought in on the mission.

其次，他们对各自产品所实现的目标保持一致。

And two, they're aligned on what their product achieves.

因为如果你只追求数字，最终得到的产品很可能并没有做正确的事。

Because if you just optimize for numbers, you're gonna end up with a product that is probably not doing the right thing.

因为数字会把你引向错误的方向。

Because numbers will stray you somewhere.

你需要良好的判断力。

You need good judgment.

我们通过努力学习一种语言来培养良好的判断力。

A good judgment we reach by trying to teach ourselves a language.

比如，我正在努力学习西班牙语。

Like, I'm trying to learn Spanish.

我正在努力提高国际象棋和音乐方面的水平。

I'm trying to get better at chess music.

这让我们很好地判断自己是否走在正确的道路上。

And this gives us a great sense of, are we headed in the right direction?

我们是否在帮助用户实现他们的目标？

Are we helping our users reach their goals?