$RDDT

$RDDT · The Intrinsic Value Podcast · 2026-11-14

AI 摘要 · The Intrinsic Value Podcast

RDDT 的毛利率约为91%,资本回报率超过100%,但Google流量如今是最大风险:搜索引荐流量正在下滑,下一季度日活用户可能两位数缩水。公允价值低于一年前,结果区间也更宽,但说话者乐于在100美元以下买入,对应约20倍出头的所有者收益倍数,而指引增长约为40%;这对长期股东并非生存性威胁。

看涨他说的是几年内
《Reddit (RDDT): Record Cash Flow, Collapsing Share Price》
The Intrinsic Value Podcast发布 2026-11-14 · 146 段内容

播放时,YouTube 可能接收设备与播放信息并使用 Cookie。 隐私说明 · Google 隐私政策

00:22/ —打开原视频

逐段跳播

146 段内容
0:2261:36

而本周,Reddit 警告称其日活用户下个季度可能减少,随后股价在一天内下跌了30%(截至录制时),尽管上一季度营收同比增长超过50%。

And this week, Reddit warned that its daily users could shrink next quarter, and then the stock fell 30% in a day at the time of recording, even with revenue growing more than 50% year over year in the last quarter.

那么,Reddit 是一个无人能真正取代的网络,还是只是一个租客,而它的房东刚刚在隔壁开了一家竞争门店?

So is Reddit a network nobody can truly replace, or just a tenant whose landlord just opened a competing store next door?

我从高中起就断断续续地用 Reddit,但多年来我从没把它当作一项投资。它的股票当时还是非公开的,而且说实话,我都不记得自己在这个平台上见过广告。所以我真的不认为它有什么值得投资的商业逻辑。

但从那以后,管理层的成就让我彻底惊叹。增长快得惊人,他们从每个用户身上赚到的钱也比过去多得多。现在正是重新审视 Reddit 的好时机,因为截至录制时,Reddit 股价在发布第三季度财报后已下跌30%,使该股今年累计下跌60%,这对我们2025年表现最好的标的之一来说,真是彻底的180度大转弯。

I've used Reddit on and off since high school, but for years I never thought of it as an investment. The shares were private anyways and honestly, I could not remember ever seeing an ad on the platform.

So, I really didn't think that there was much of a business case to be made for it, but ever since then I've been just absolutely blown away by what management has accomplished.

Growth is off the charts and they're earning much more per user than they ever did in the past and it's a good time to revisit Reddit because at the time of recording here, Reddit shares have dropped 30% following their Q3 earnings report, dragging the stock down 60% for the year, which is really quite the 180 for one of our best performers from 2025.

是啊,而实际上最疯狂的是,这已经是过去3个月里第二次单日超过20%的回撤了。7月份时,公司公布了一个营收增长超过60%、利润几乎翻了三倍的季度,第二天股价就跌了约21%。而昨天,Reddit 的营收增长甚至更高,营收和利润都超出华尔街预期,利润率也扩大了,股价却依然被狠狠打压。

通常这样的下跌都伴随着一个灾难性的季度,而这两次都不是这样。

Yeah, and actually the craziest part is that this is actually the second 20% plus drawdown in a single day in the last 3 months. So, back in July the stock fell about 21% the day after the company reported a quarter where revenue grew more than 60% and profits nearly tripled.

So, yesterday Reddit grew revenues even more than beat Wall Street on both sales and profits, expanded its margins, and the stock still got crushed. So, normally a drop like that comes with a disastrous quarter and in neither case was that true.

作为 Reddit 的股东,这样的下跌已经不再让我震惊了。我的意思是,从我持有它的第一天起,这只股票就完全是过山车,而我确实赶上了第一趟。

And drops like that don't shock me anymore as a Reddit shareholder. I mean, this stock has been an absolute roller coaster ride ever since the first day I owned it and I was really there for the first ride.

如果有人错过了我最初对 Reddit 的推荐,我在节目笔记里放了链接。不过我最终参与了 IPO,而且我对 Reddit 相当熟悉,因为正如你提到的,我是 TIPs 子版块的版主,因此我获得了以34美元 IPO 价格认购的配额。

作为背景,Reddit 想让最活跃的用户有机会以与机构投资者相同的价格买入,从而在他们喜爱的平台上拥有一份所有权。所以,因为担任版主,我有机会参与,基本上 IPO 当天股价就涨了约50%,我试图尽快锁定这部分收益,因为我知道自己并不愿意持有一家当时还不盈利的社交媒体公司,尤其是看到此前 Twitter 上毁灭了多少价值之后。

但我也知道 Reddit 是一个热门 IPO,所以我算是赌了一把。押注股价在 IPO 当天上涨,这有点像一个有趣的赌博,而它确实涨了。但我非常清楚,我不想持有超过那一天。我不是长期持有者。

说实话,我至今仍认为那大概率是正确的思路,尽管在我卖出之后股价确实涨了很多。我的想法是,如果你无法安心持有某样东西,卖掉也没关系。

For anyone who missed my original pitch on Reddit, I do have it linked in the show notes, but I actually ended up participating in the IPO and I was quite familiar with Reddit since as you had mentioned I was a moderator of TIPs subreddit and for doing so I got an allocation at the $34 IPO price, which for context, Reddit wanted to give its most active users the chance to buy in at the same price institutional investors could and and have a stake of ownership in the platform that they love.

So, for being a moderator I had the chance to participate and basically on IPO day the stock popped about 50% and I tried to lock that gain in as quickly as possible because I knew I wasn't comfortable owning what was at the time an unprofitable social media business after seeing especially how much value had been destroyed at Twitter previously.

And yet I knew Reddit was a hot IPO. So I sort of rolled the dice. It was a bit of a fun gamble betting that the shares would pop on IPO day, which they did. But I definitely knew I didn't want to hold it for any longer than that.

I wasn't a long-term holder. And honestly, I still think that was probably the right thought process even though the stock did actually go much higher after I sold. And the way I think of it is if you can't get comfortable owning something, it is fine to sell.

所以在我最初的推荐中,我对 Reddit 的估值是每股120到135美元。当时它的交易价格其实高于这个水平。但我说过,如果它哪天回落到那个价格区间,我会很乐意持有。后来它确实回落了。

我们最终在组合中以约105美元的平均成本建立了仓位。所以我的 Reddit 投资经历包括:一笔我个人赚了钱、但作为价值投资者大概本不该参与的短线交易;以及一笔投资,它在峰值时翻了一倍多,然后截至录制时又回到了原点。

So in my original pitch, I valued Reddit at between $120 and $135 a share. And at the time, it was actually trading above that price. But I said that I would love to own it if it ever came back down to that price range.

And it did. And we ended up building a position in the portfolio at an average cost of about $105. So my Reddit investing experience includes one trade I made money on personally, but probably shouldn't have been in as a value investor.

And then one investment that more than doubled at its peak and then came full circle as of the time of recording.

是啊,当一个仓位翻倍时,人很容易变得懒惰,不再重新审视它。但当它在几个月内把涨幅全部吐回去时,你知道,人又很容易恐慌。所以我认为这两种都不是做决策的最佳方式。在目前这轮下跌之后,它大约占组合的3.5%。

所以如果我们仍然相信这项业务,这可能是一个加仓的机会。

Yeah, so when a position doubles, it's pretty easy to get lazy about re-underwriting it. When it gives it all back in a few months though, you know, it's very easy to panic. So I think neither is probably the optimal way to make decisions.

So after this current drop, it's close to being about 3 and 1/2% of the portfolio. So if we still believe in the business, this could be a chance to add.

你最初那期节目给了我们一个很好的时间胶囊。你录制那期时,Reddit 刚刚实现第一个盈利年度,日活用户大约1亿。但最大的问题是:一个以限制用户数据为核心搭建的平台,能否像 Meta 那样卖定向广告,还是它的变现会像 Snapchat 那样停滞?

现在快进到今天,Reddit 单季度营收约9亿美元,增长超过50%,每1美元销售额中约有三分之一转化为利润。所以在过去12个月里,它产生了远超10亿美元的自由现金流。所以我认为 Reddit 显然是一家有能力印钞、而且印很多钞的公司。

但更难的问题是接下来会怎样。这是近两年来第一个增长低于60%的季度,管理层指引下个季度增长大约40%,略低于华尔街的期望。而市场,你知道,他们讨厌增长减速,即使增速仍然非常非常高。

And your original episode gave us a really nice time capsule. So when you recorded it, Reddit had its first profitable year and daily users were around 100 million. But the big question was whether a platform built around limiting user data could sell targeted ads like Meta or whether its monetization would stall out like Snapchat.

Now if we fast forward to today, Reddit just did about $900 million of revenue in a single quarter, up more than 50% with about a third of every dollar sales turning into profit.

So over the last 12 months, it's generated well over a billion dollars of free cash flow. So I think Reddit is really a business that is clearly capable of printing cash and a lot of it.

But the tougher question is just what comes next. This was the first quarter in nearly two years that growth came in under 60% and management guided to roughly about 40% growth in the next quarter, which is a little bit below what Wall Street wanted it to be.

And the market just, you know, they hate decelerating growth even when it's still very very high.

没错。当一家企业能以几乎超过市场上任何公司的速度增长,却在投资者认定下一季度不会那么神奇的那一刻被狠狠打压,尤其是当它的估值相对市场处于极高的倍数时。这就像一个一直拿全A的孩子,却因为,你知道,一个A减就被禁足。

所以我觉得我不会做那种因为这点事就禁足儿子的爸爸,不过距离这真的可能发生,我还有很多年。但为了判断这里的放缓,我们需要再往引擎盖下面多看一点。Reddit 不只是一台引擎,而是好几台引擎。

有些运转得相当火热,有些可能在噼啪作响地出故障,还有几台可能根本还没被启动。所以如果我们理解每一台引擎,就能更好地判断市场当时的恐慌是否正确。我得承认,在做这期节目之前,我并没有真正认真想过 Reddit 的商业模式。

我只想知道怎么种生菜。直到我多研究了一点,我才意识到这门生意到底有多好。

Exactly. And when a business can grow faster than almost anything in the market and still get absolutely hammered the moment investors decide the next quarter won't be as magical, particularly when it's priced at a huge multiple to the market.

It's like the kid who consistently gets straight A's then gets grounded for, you know, one A minus. So I don't think I'll be the dad grounding my son for that, but I have many more years until that potentially becomes a reality.

But, you know, just to judge the slowdown here, we need to kind of look a little more under the hood. So Reddit isn't just one engine, it's a handful of engines. So some are running quite hot, some are probably sputtering, and a couple of have probably not even been switched on yet.

So if we understand each of them, we'll have a much better sense of whether the market was right to panic. And I will admit before this episode, I hadn't really given Reddit's business model that much thought.

I just wanted to know how to grow lettuce. I didn't appreciate how good the business really was until I dug in a little bit more.

大多数 Reddit 用户从没想过自己其实身处一家企业之中。而当这个话题被提起时,它在 Reddit 上其实是一个非常禁忌且敏感的话题。Reddit 常常让人觉得只是一个大家去争论哪款吸尘器吸力最强的地方。

但这些其实可以成为非常肥沃的广告空间。而在过去几年里,Reddit 在销售和变现这些空间的能力上已经有了大幅提升。

Most Reddit users never think about the fact that they're inside a business at all. And when it does come up, it's actually a very taboo and sensitive topic on Reddit. Reddit often feels like just a place where you go to argue over which vacuum has the best suction power.

But those can actually make for very fertile ad space. And over the last few years, Reddit has drastically improved at its ability to sell and and monetize those spaces.

对。那场吸尘器之争很好地说明了这整套东西到底是怎么运作的。Reddit 大约有1.3亿日活用户,以及超过270亿,是的,270亿条帖子和评论。它们都被整齐地组织进所谓的 subreddit,也就是围绕单一话题建立的讨论论坛。

有种生菜的版块,有养高能量犬的人的版块,还有成千上万个你根本想不到地球上会有单个人在意的冷门话题版块。用户发帖,其他用户评论,然后大家投票。最好的回答浮到顶部,最差的回答沉到底部。

Reddit 并不创作内容,内容全部来自用户,有点像 YouTube。对 Reddit 来说更好的是,他们有一支志愿版主大军,负责运营每个社区,做设定规则、清理垃圾信息之类的事。

而他们做这些全都是零报酬。所以 Reddit 的工作基本上就是当房东。他们让大楼立着,让灯亮着,然后把所有墙面空间租出去。而那里的墙面空间才是真正赚钱的地方。广告约占收入的95%。

其余收入绝对是个热门话题。Reddit 把它标为其他收入。这部分主要来自向 Google 或 OpenAI 这类 AI 公司授权数据。此外还包含 Reddit 高级订阅的收入。

Right. And that vacuum argument is a pretty good illustration of how this whole thing really works. So Reddit has about 130 million daily users and over 27 billion yes, 27 billion posts and comments.

They're all neatly organized into something called subreddits, which are these discussion forums that are built around a single topic. There are ones for growing lettuce. There's one for people with high energy dogs and then tens of thousands more for topics that you wouldn't even think that a single person on earth would care about.

Users make a post, other users add comments, and then people vote. The best answers float to the top and the worst answers sink to the bottom. Now, Reddit doesn't create the content.

It's all user based, kind of like YouTube. To make things even better for Reddit, they have an army of volunteer moderators who run each community and do things like set the rules and clean up any spam.

And they do this all for zero dollars. So Reddit's job is basically just to be the landlord. They keep the building standing, keep the lights on, and rent out all of the wall space.

But the wall space there is where the real money is. So advertising is about 95% of the revenue. The rest of the revenue is definitely a hot topic. So Reddit labels it as other revenue.

And this ends up being mostly from data licensing to AI companies such as Google or OpenAI. Also lumped in is its revenue from Reddit premium subscriptions.

是的,高级订阅那部分相当微不足道,就是个四舍五入的误差。但确实有很多人谈论 Reddit 在那一行其他收入里算得上是一家 AI 公司。正如你提到的,那部分只占销售额的5%,而且上季度增长比广告还慢。

所以我肯定会反驳市场上把 Reddit 说成 AI 公司或 AI 受益者的叙事,不管你怎么看这件事。我一直更愿意关注核心业务,而不是数据授权收入,因为后者在我看来从来都不够持久;而且作为这款产品的长期用户,我更感兴趣的是他们如何变现像我这样的长期用户。

更有意思的 AI 故事其实在广告这一侧。AI 让 Reddit 的广告定位更精准、投放更容易,而真正的钱已经证明在那里。但 Reddit 的广告业务和 Facebook 或 Instagram 之类的并不一样。

Meta 对用户了解很多,它知道你的年龄、住哪里、朋友是谁,还知道过去几周你一直在看圣诞树。而在 Reddit 上,大多数用户是匿名的,很多用户根本不登录。所以它不是针对单个用户定向,而是针对人们在 subreddit 里进行的对话定向。

如果你在一个家用意式咖啡机 subreddit 里,你可能会看到咖啡磨豆机的广告,这种情况叫情境广告,而真正的魔力在于这些广告出现的时机。Instagram 可能在你正看小狗视频或者你在 Instagram 上干的别的什么时给你看跑鞋广告,而 Reddit 会在你阅读“适合扁平足人群的最佳鞋款”subreddit 时给你看鞋广告。

所以可以这样理解:Meta 就像一个销售员,一直跟着你在商场里到处走,你走到哪就给你建议;而 Reddit 则像一个销售员,站在鞋类货架旁,而你已经在找鞋、也准备好促成转化了。这大概就是我能想到的最简单的比较方式。

Yeah, the premium subscriptions is pretty negligible. It's a rounding error. But there has definitely been a lot of talk about Reddit being something of an AI company within that other revenue line.

That's only 5% of sales, as you mentioned. And it grew slower than ads last quarter. So I definitely would push back on the market narrative of Reddit being an AI company or an AI beneficiary, however you want to think about it.

I've always preferred to focus on the core business than the licensing money for data because it just never seemed that durable to me and as a long time user of the product I was always more interested in how they could monetize other long-term users like myself and the more interesting AI story is actually on the advertising side.

AI is what's making Reddit's ads better targeted and easier to buy and that's where the real money has proven to be but Reddit has a different ads business compared to something like Facebook or or Instagram.

Meta knows a lot of things about its users. It knows your age, where you live, who your friends are and that you've been looking at Christmas trees over the past few weeks and on Reddit most users are synonymous and many users never log in at all so instead of targeting individual users it targets the conversations that people are having on subreddits and if you're on a home espresso subreddit you might see an ad for a coffee grinder and it's called contextual advertising when this happens and the magic really is in the timing of these ads.

So where Instagram might show you a running shoe ad while you were just watching videos of puppies or whatever whatever else you do on Instagram Reddit will show you a shoe ad while you're reading a subreddit on the best shoes for people with flat feet and so a way to think of it is Meta is like a salesperson who's just following you around the mall at all times making suggestions wherever you go whereas Reddit is like this salesperson that's standing in the shoe aisle when you're already looking for shoes and prepared to help drive that conversion and that is sort of the simplest way I would compare the differences.

不,我觉得这个比较很好。所以我认为向广告主推销的说法是:用你的比喻,那个站在鞋类货架旁的销售员大概率能成交更多。而我认为数字实际上也支持这一点。今年夏天 Reddit 说其活跃广告主数量增长了约70%,转化量翻了一倍多;本季度广告收入增长超过50%,既受更高价格推动,也受更多广告展示推动。

但我觉得真正有意思的部分,是想想平台上的广告主都是谁。随着 Reddit 规模扩大,这一点已经向好的方向改变。一开始,Reddit 的广告业务主要只是那些通过巨型广告代理公司投放的全国性品牌。

但今天,这部分仍然在业务中,同时他们也在发展中小型企业这一块。来自这一群体的收入实际上同比增长了一倍,其中很大一部分来自 Reddit Max,也就是它的 AI 广告活动工具,这个工具在过去这个季度刚刚向所有广告主开放。

No, I think that's a great comparison. So the pitch to advertisers I think is that the shoe aisle salesperson using your metaphor is probably going to close more deals and the numbers I think actually back it up.

So over the summer Reddit said that its number of active advertisers grew about 70% and conversions more than doubled and this quarter ad revenue grew by more than 50% driven by both higher prices and more ads shown but I I the really interesting part is when you think about who the advertisers are on the platform.

It's changed for the better as Reddit has scaled up. So in the beginning, Reddit's ad business was mostly just, you know, national brands who are buying through these gigantic ad agencies.

But today, they still have that as part of their business, but they're also growing the small and mid-sized business segment as well. So revenue from that group actually doubled year over year and a lot of that is coming from Reddit Max, which is its AI campaign tool which just became available for every advertiser this past quarter.

Reddit Max 对小型广告主来说是件大事。大多数小企业没有广告团队,而有了 Reddit Max,企业只需告诉 Reddit 自己想要什么,无论是促销还是应用下载,然后 Reddit 的算法就会判断谁应该看到这条广告以及如何触达他们。

这帮助 Reddit 从只有大型广告代理公司才能利用的业务,变成连一个经营园艺中心这种平凡生意的本地店主也能在一个下午内设置好并使用的东西。我知道你喜欢你的,呃,你的生菜牛。而且增长未必来自某一个热门品类,对吧?

今年夏天,管理层说其15个行业垂直领域中有11个增长超过50%,这种广度本季度再次保持,由旅游和零售等板块领跑。但他们在多元化和发力购物等方面做得非常好。他们有个叫动态产品广告的东西,可以帮助把商店的产品目录变成广告。

目前已有500家商店接入。所以广告故事在这里变得复杂,也正是我们可以讨论到底是什么开始让市场紧张的地方,对吧?并非每个访客对广告主的价值都一样。而这是其股价目前处于低谷的一个很大原因。

Reddit Max is a big deal for smaller advertisers. Most small businesses don't have an ad team and with Reddit Max, a business just tells Reddit what it wants, whether that's a sale or an app download and Reddit's algorithm then figures out who should see the ad and how to get it to them.

And it helps turn Reddit from a business only large ad agencies would be able to utilize into something that a local owner of something as mundane as a gardening center could be able to set up and use in one afternoon.

I know you like your uh your Lettuce Cow. And the growth isn't necessarily coming from one hot category, right? Over the summer, management said 11 of its 15 industry verticals grew more than 50% and that breadth held up again this quarter led by things like travel and retail.

But they're doing a really good job diversifying and leaning into things like shopping. They have something called dynamic product ads which can help turn a store's product catalog into ads.

And so they have 500 stores that are already connected. And so this is where the ad story gets complicated and also where we can have a conversation about what is really starting to make the market nervous, right?

Not every visitor is worth the same to an advertiser. And that is a big reason that its share price is is currently in the dumps.

是的,而且我觉得我花了一点时间才真正想明白这件事。那1.3亿日活用户实际上分成两个完全不同的群体。让我来拆解一下。第一组大概就是Sean所属的那一组。他们登录、有账号、使用这个应用、订阅不同的社区,并且相当规律地回来。

第二组大概就是我属于的那一组。这些人通常是从谷歌搜索过来的。他们可能只读一个帖子然后就离开了。但一个登录用户的价值大约是一个未登录用户的三倍,原因很简单:他们停留的时间更长,因此能看到更多广告。

事实上,管理层表示,登录用户和应用用户贡献了Reddit超过80%的收入。因为我住在落基山脉附近,让我用滑雪场来举个例子。假设你有一批季票持有者,他们倾向于在滑雪季尽可能多的周末去山上。

他们不仅使用这座山,而且每次去还会花10美元买价格虚高的热巧克力,再加上他们最终从山间小屋买的食物。然后你还有一日游游客。这些人是心血来潮就开车过来的。他们可能滑几趟就回家了。

滑雪场肯定两者都需要,但很明显哪一类人群会给滑雪场带来最多的收入。

Yeah, and I think this took me a little while to really get my head around it. So those 130 million daily users are really in kind of two completely different groups. So let me break this down.

So the first group is probably the group that Sean belongs to. You know, they're people who are logging in, they have accounts, they use the app, they subscribe to different communities and come back on a pretty regular basis.

The second group is probably where I'd belong. These are people who arrive usually from a Google search. They might just read one thread and then they're just going to leave. But a logged-in user is worth roughly three times as much as a logged-out one simply because they just stick around longer and can therefore see more ads.

In fact, management says logged-in and app users drive more than 80% of Reddit's revenue. Since I live by the Rocky Mountains, let me kind of give you a ski hill example here. So let say you've got these kind of season pass ticket holders who tend to go to the mountain for as many weekends as possible during the ski season.

They're not only using the mountain, but they're also paying $10 for overpriced hot chocolate every time they go, plus whatever food they end up getting from the lodge. Then you've got the day trippers.

These are the ones that just drive up on a whim. They might ski a few runs and then just head home. The hills definitely need both of them, but it's obvious which cohort is going to generate the most revenue for the ski hill.

上次我去单板滑雪时,我绝对是个一日游游客,但我最后还是买了价格虚高的热巧克力。>> 稍微延伸一下你的滑雪场比喻,季票持有者住在哪里也很重要,这在现实生活中其实没那么重要,但你知道,上个季度,举例来说,Reddit从每个美国日活用户身上赚了将近14美元,而从每个国际用户身上赚的不到2.50美元。

Now the last time I went snowboarding, I was definitely a day tripper, but I still ended up buying the overpriced hot chocolate. >> Just uh to stretch your ski hill metaphor a little bit, it also matters where the season pass holder lives, which wouldn't actually matter so much in real life, but you know, last quarter, for example, Reddit earned almost $14 from each daily American user versus less than $2.50 from each international user.

稍微延伸一下你的滑雪场比喻,季票持有者住在哪里也很重要,这在现实生活中其实没那么重要,但上个季度,举例来说,Reddit从每个美国日活用户身上赚了将近14美元,而从每个国际用户身上赚的不到2.50美元。

所以一个美国用户的价值是国际用户的五倍多。这并不意味着美国人更是更好的Reddit用户。这只是意味着美国广告商每条广告的付费远高于其他国家的广告商。这就是为什么Reddit大约80%的收入来自美国,尽管国际用户实际上占了其用户基数的大多数。

所以Reddit最有价值的用户是登录的美国用户。而这个群体本季度又只增长了1%。然后真正的重击来了,CEO Steve Huffman说,来自谷歌搜索的流量在本季度大幅下降,而且实际上到10月份还在继续下降,每周减少超过一百万次访问。

如果这种情况持续下去,Reddit的日活用户数下个季度可能会以两位数的百分比缩水,这非常非常可怕。而且他们与谷歌也还没有续签数据协议。所以目前Reddit和谷歌之间肯定存在紧张关系。

另一件让人侧目的事情发生在今年早些时候,当时管理层表示将逐步停止报告登录用户与未登录用户的拆分数据,而恰在此时登录用户的增长正在停滞。当一家公司突然停止报告某项指标时,通常不是因为该指标在改善,对吧?

我的意思是,在某些情况下,是的,被模糊化的指标肯定在恶化,但也有一种说法是,也许它不再那么能代表企业的健康状况了。这就是管理层想逐步取消它的原因。我会说,这对管理层来说是一种宽容的解读。

而我认为在大多数情况下,这大概是两者兼有。某些东西在恶化,对公司不利,但也许这项指标并不像市场认为的那么关键和重要。所以对我来说,这是一个黄旗。

Just uh to stretch your ski hill metaphor a little bit, it also matters where the season pass holder lives, which wouldn't actually matter so much in real life, but you know, last quarter, for example, Reddit earned almost $14 from each daily American user versus less than $2.50 from each international user.

So one American user is worth more than five times international ones. And this doesn't mean that Americans are are better redditors. It just means that US advertisers pay a lot more per ad than advertisers in other countries.

And that's why about 80% of Reddit's revenue comes from the US, even though international users make up about the majority of its user base actually. So Reddit's most valuable users are logged-in Americans.

And that group grew just 1% again this quarter. And so then the real gut punch came when Steve Huffman, the CEO, said traffic from Google search had fallen meaningfully during the quarter and that it's actually kept falling through October by more than a million visits a week.

If that keeps up, Reddit's daily user count could shrink by a double digit percentage number next quarter, which is very very scary. And then they still don't have a renewed data deal with Google either.

So there's definitely a tension between Reddit and Google at the moment. And the other thing that raised eyebrows happened a bit earlier this year, and that was when management said it was going to phase out reporting of the split of logged in versus logged out users right as growth in logged in users was stalling.

When a company suddenly stops reporting a metric, it's not usually because that metric is improving, right? I mean, in some cases, yes, the metric being obfuscated is definitely worsening, but also there's an argument to be made that it maybe it's not as representative of the business's health anymore.

And that's why management wants to phase it out. I would say that's a generous interpretation for managers. And I think in most cases it's probably a mix of both. Something is worsening and it's not great for the company, but perhaps the metric isn't as critical and important as the market thinks it is. So for me, it's it's a yellow flag.

是的。而且你知道,我认为这通常大概是一个红旗,或者至少是一个黄旗。所以我知道,当我看企业时,我喜欢回顾它们历史上使用的KPI,也许是几年前用的,然后我喜欢看今天,看看它们是否还在用。

所以如果我认为它们5年前有一个非常好的KPI,而现在它们不报告了,我个人就必须问为什么。而这通常标志着企业发生了根本性的变化。所以

Yeah. And you know, I think usually it's probably a red flag or maybe at least a yellow flag. So I know when I look at businesses, I like to review the KPIs that they've used historically, whether that's, you know, maybe a couple years ago, and then I like to look at today and see if they're still using them.

So if I think they had a really good KPI 5 years ago and now they're missing, I personally have to ask why. And usually it signals that something has fundamentally changed in the business. So

所以这就是我看到的基准情景。Reddit最有价值的用户几乎不增长,而它过去从谷歌获得的廉价增长现在基本上在逆转。

So here's the base case as I see it. Reddit's most valuable users are barely growing, and the cheap growth that it used to get from Google is now going basically in reverse.

而Huffman基本上说,谷歌已经从流量驱动者变成了完全不同的东西。与此同时,与谷歌的数据协议仍在重新谈判中。所以,论点是Reddit仅靠其季票持有者无法榨取足够的收益。

And Huffman basically said that Google has gone from a traffic driver to something else entirely. Meanwhile, the data deal with Google is still being renegotiated. So, the argument is that Reddit can't squeeze enough out of its seasoned pass holders alone.

提高广告价格肯定能承载很多,但你知道,它不可能永远让增长保持在50%或60%。

Raising ad prices definitely can carry a lot, but you know, it's not going to keep growth going at 50 or 60% forever.

Reddit让这些用户真正能在同一个地方相聚。而且正如你提到的,Kyle,运营这些房间的人是志愿者,对吧?你很少能找到一家由志愿者驱动的公司,但Reddit正是如此。

Reddit allows those users to actually all meet in the same place. And as you mentioned, Kyle, the people running these rooms are volunteers, right? It's not often that you find a company powered by volunteers, but that's very much the case with Reddit.

而Reddit的档案库里存有20年真实人们的经历。我真的不认为你能在一夜之间重建这些。而且Reddit已经将这些数据以现在易于搜索的方式进行了定位。这就是为什么谷歌倾向于将Reddit排得那么高,也是为什么谷歌和OpenAI都向Reddit支付了相当可观的金额来授权这些数据,至少在过去是这样。

And Reddit has 20 years of real people's experiences sitting in its archives. And I don't really think you can recreate that overnight. And Reddit has positioned that data in a way that is now easily searchable.

This is why Google tends to rank Reddit so highly, and it's also why Google and OpenAI have both paid a pretty handsome sum to Reddit to license the data, at least in the past.

而棘手的是,与此同时,Reddit近期用户增长的很大一部分来自谷歌选择在其搜索结果中展示Reddit帖子。

And the tricky thing is that at the same time, a lot of Reddit's recent user growth came from Google choosing to feature Reddit Threads in its search results.

所以最大的区别在于,BuzzFeed的整个业务都来自那些路过访客,而Reddit幸运的是,登录用户和应用用户贡献了超过80%的收入。所以失去谷歌流量对用户数的伤害远大于对利润表的伤害,至少目前是这样。

And so the big difference is that BuzzFeed's whole business was from those drive-by visitors, whereas fortunately for Reddit, logged in and app users drive more than 80% of revenue.

So losing Google traffic hurts the user count figures a lot more than it hurts the income statement, at least for the time being.

有趣的是,Reddit即将把你提到的那个酒吧实验,Kyle,某种程度上反过来做。他们正在转向仅限登录,这有点像在门口安排一个保安,进酒吧时问你的名字。一些一日游游客会注册并成为常客,而另一些人则只会转向下一家酒吧。

And funny enough, Reddit is about to run that bar experiment that you mentioned, Kyle, sort of in reverse. They're going logged in only, and that's sort of like putting a bouncer at the door who asks for your name on the way into the bar.

Some day trippers are going to sign up and become regulars, and others are just going to move on to the next bar.

不,这是个好问题。而且我认为这是我用自己那套护城河清单来评估Reddit时感到纠结的部分原因,我喜欢用这套清单来评估我所有的企业。

No, that's a good one. And I think it's part of the reason why I was torn when I ran Reddit through my kind of moat checklist that I like to run all my businesses through.

但即使谷歌和OpenAI继续为数据付费,这并不能真正解决让人们实际点击进入Reddit的问题。

But even if Google and OpenAI keep paying for the data, that doesn't really solve the problem of getting people to actually click through to Reddit.

这就是为什么我实际上喜欢仅限登录的举措。这是一个防御性举措,即使短期内会损失一些用户。基本上,如果你愿意这样想的话,它相当于在前门上了一把锁。如果谷歌和OpenAI或其他任何人想要Reddit的对话和内容,他们就必须付费,而不是Reddit把那个产品免费交给那家实际上正在切断其应用和网站流量来源的公司。

And that 's why I actually like the logged in only move. It's a defensive move, even if it costs some users in the short run. Basically, it puts a lock on the front door if you want to think of it that way.

If Google and OpenAI or anybody else wants Reddit's conversations and content, they'll have to pay for them instead of Reddit handing that product out for free to the same company that's actually cutting the flow of traffic to their app and website.

而你说得对,Kyle,授权费并不能替代登录的习惯。所以你能从数据授权中得到的只有这么多。我的意思是,这是一笔不错的钱,但它的现值远小于你从一个新的终身用户那里获得的未来广告收入。

And where you're right, Kyle, is that a licensing check doesn't replace the habit of logging in. So there's only so much you can get from licensing out your data. I mean, it's nice money to have, but the present value of it is considerably smaller than the future ad revenue that you would get from from a new lifetime user.

对。我认为这里的另一个优势,你刚才也提到了,就是忠诚度。Reddit 的核心用户会为了他们的社区、人际关系、在网站上的身份认同而留下来,而且说实话,还有纯粹的习惯。这就是为什么登录用户数如此重要。

如果因为习惯而回访的那部分用户停止增长,那么护城河不会消失,但它就不再是增长引擎了。

Right. And I think the other advantage here, you kind of just alluded to was just in loyalty, right? So Reddit's core users stick around for their communities, their relationships, their identity on the site, and let's be honest, just plain old habit.

That's why the logged in numbers matter so much. If the group that comes back out of habit stops growing, well, then the moat doesn't disappear, but it stops being a growth engine.

好消息是,本季度登录用户和 App 用户实际上比上一季度有所增长,而且新 App 用户的留存情况好多了。所以首周留存率比一年前提高了超过50%。

Now the good news is that this quarter, the logged in and app users actually did grow from the prior quarter, and new app users are sticking around much better. So the first week retention was up more than 50% versus a year ago.

有趣的是,Reddit 出了名难懂的界面其实是一把双刃剑。老用户已经爬过了学习曲线,这很可能有助于保护 Reddit 免受潜在模仿者的冲击。但它也会吓跑像我这样的新来者,因为那些一日游访客很可能最终根本不会注册,原因就是他们不想去学这一切是怎么运作的。

And funny enough, Reddit's famously confusing interface actually cuts both ways. So long time users have already climbed the learning curve, which probably helps protect Reddit from potential copycats.

But it also serves to scare off newcomers like myself because, you know, these day pass visitors who probably never end up signing up simply because they just don't want to learn how it all works.

但在疫情期间,我花了几个小时试图破解围绕 Wall Street Bets 形成的某种暗语。而且我当时已经是 Reddit 的常客了,我发现这非常困难。

But during the pandemic, I spent hours trying to decode sort of the secret language that had formed around Wall Street Bets. And I was already a regular Reddit user at the time too, and I found it very difficult.

但我认为学习曲线既是缺陷也是特色,对吧?每个 subreddit 都有自己的俚语、自己的内部笑话,以及自己的成文和不成文规则。所以一旦你融入其中,那才真正让它感觉像一个社区。竞争对手很难复制这种文化。

But I think the learning curve is as much a feature as it is a bug, right? Every subreddit has its own slang, its own inside jokes, and its own written and unwritten rules. And so once you're in, that is really what makes it feel like a community.

And it's very difficult for a competitor to copy that culture.

而另一面是,这也是为什么那么多一日游访客从来留不住。Reddit 在某些方面会让人感觉有排他性,这也是 Reddit 为什么没有 Facebook 那么大、而且永远也不会那么大的重要原因,尽管它们差不多同时出现。

And the flip side is that it's also why so many day trippers never stick around. Reddit can feel exclusionary in some ways, and it's a big part of why Reddit isn't as big as Facebook and never will be despite them coming out around the same time.

在 Reddit IPO 前后,前10大广告主占了其收入的四分之一。所以这对 Reddit 来说是很大的集中度风险。而且 Reddit 的业务和股票也真的没有经历过广告衰退的考验,至少没有作为一家盈利公司经历过。

所以如果我们再遭遇一次2008年或2020年那样的冲击,广告预算绝对会被削减。我们就是不知道 Reddit 会受到多严重的打击。

And back around Reddit's IPO days, the top 10 advertisers made up a quarter of their revenue. And so that was a lot of concentration risk for Reddit. And Reddit's business and stock really hasn't been tested through an advertising recession either, at least not as a profitable company.

And so if we got hit with another 2008 or 2020, ad budgets would absolutely get cut. And we just don't know how hard Reddit would be hit.

但我认为 Reddit 在多元化方面做得相当不错,它正在向这些小企业倾斜。我猜客户集中度风险也相应下降了,而且很可能降得相当明显。我这么说是因为活跃广告主在过去一年增长了约70%。而中小型企业是其中增长最快的群体。

But I think Reddit has done a pretty good job on the diversification front with these small businesses that it's leaning into. My guess is that customer concentration risk has correspondingly fallen and probably pretty meaningfully.

And I say that because active advertisers grew about 70% over the past year. And small and mid-sized businesses were the fastest growing group in there.

而我们提到的像 Reddit Max 这样的工具,让长尾中的众多小广告主非常容易地出现并参与进来。所以 Reddit 没有给我们确切的数字,我无法证明。但我绝对认为,现在的结构很可能比 IPO 时健康得多。

And those tools like Reddit Max that we mentioned make it just so easy for really a long tail of smaller advertisers to show up and participate in. And so Reddit hasn't given us that exact number, so I can't prove it.

But I definitely think the mix is probably a lot healthier than it was at the time of the IPO.

是的,我非常赞赏他们实现了客户群的多元化,这绝对是我非常乐意看到的。所以在这里简单总结一下,以下是我对护城河的结论。在用户方面,我认为它非常强大。几乎没有任何其他平台能以接近 Reddit 的规模做这种基于话题、匿名、可搜索的讨论。

Yeah, I give them a lot of credit for diversifying their customer base, which is something I definitely definitely like to see. So just to kind of give a summary here, here's where I land on the mode.

So on the user side, I think it's very formidable. Pretty much nobody else does this topic-based, pseudonymous, searchable discussion at any scale that comes close to Reddit.

如果你在类似 Facebook 群组这样的地方,它和你的真实姓名绑定,这会改变人们愿意讨论的内容。当他们的家人、朋友或同事能读到的时候,人们不会去发“这一切是不是很奇怪?”这种帖子。

If you're in something like a Facebook group, you know, it's tied to your real name, which can change what people are willing to discuss. You aren't going to have people posting, you know, is this all weird?

When their family, friends, or colleagues can read it.

Discord 还不错,但社区完全是私密的,除非你是那个特定社区的一员,否则你无法真正搜索它们。X 或 Twitter,或者现在你想怎么叫它都行,往往只会演变成互相叫骂,而 TikTok 更纯粹是娱乐。

Discord, it's decent, but the communities are completely private and you can't really search them unless you are part of that specific community. X or Twitter or whatever you want to call it these days tends to just break down into shouting matches and TikTok is more purely for entertainment.

所以我确实认为 Reddit 是一个非常好的产品,我很可能会继续在更具体的搜索查询后面加上 Reddit。而在广告主方面,Reddit 必须每个季度都持续赢得预算,如果美国用户停止增长,广告主会注意到。

So I do think Reddit is a very good product and chances are high I'll continue appending Reddit to the back of my more specific search queries. Now on the advertiser side, Reddit has to keep earning its budget every single quarter and advertisers will notice if American users do stop growing.

我也不太喜欢 Google 掌控 Reddit 如此多的命运。而在这个季度之后,这已经不再是几年之后的风险了。它正在发生。Google 的 AI 回答已经足够好,以至于很多人根本不需要点击进入底层的 Reddit 帖子。

I'm also not crazy about Google controlling so much of Reddit's fate. And after this quarter, that's not a risk that's a few years out anymore. It's happening right now. Google's AI answers are good enough that a lot of people just never need to click into the underlying Reddit thread.

我想补充的一点是规模。正如我们讨论过的,Reddit 的库每天都在增长,而 Reddit 不用为此支付任何费用。每一段新对话都让档案对用户、广告主和 AI 公司都更有价值。而新来者需要数年时间和一支志愿者大军,才能开始拥有与 Reddit 相同的经济效应。

The one thing I would add is scale. Reddit's library grows every single day without Reddit paying for any of it as we've talked about. And every new conversation makes the archive more valuable to users, advertisers, and AI companies alike.

And a newcomer would need years and an army of volunteers just to get started with the same economics that Reddit has.

所以我会说,在用户方面这是强大的护城河。在广告主方面也相当不错,然后在漏斗顶部确实存在一些脆弱性,这就是为什么最近围绕这只股票有这么多恐慌。但我不认为这对相信 Reddit 平台盈利能力的长期股东来说是生存威胁。

And so I I'd call it a strong moat on the user side. A pretty decent one with advertisers and then there's some vulnerability for sure at the top of the funnel, which is why there's been so much panic around the stock recently.

But I don't think that's an existential threat to long-term shareholders who believe in the earnings power of the Reddit platform.

我相信我们会继续看到双方之间的拉锯动态。但 Reddit 的回应非常坚决,我挺喜欢这一点。是的,我们会看看这是否会搅动 Google 的 AI 概览。

And I'm sure we'll continue to see a push and pull dynamic between both sides. But Reddit is responding very firmly, and I kind of like that. And And yeah, we'll see if that shakes up things in Google's AI overviews.

但如果用定量方法来看 Reddit 的数据,很明显它确实有一定的护城河。它的数据简直惊人。我知道你是研究 ROIC 的,Kyle,专门看资本回报率。所以我不想抢你的风头。这部分就让你来详细拆解。不过我先补充一点背景。

But if you take a quantitative approach to Reddit's numbers, it's clear they have something of a moat. I mean, their numbers are just phenomenal. And I know you're the ROIC guy, Kyle, all about the returns on capital.

So I don't want to steal your thunder. I'll let you do the honors of breaking that down more. But I'll share a bit more on the background here.

稍微想一下,Reddit 并不拥有数据中心,对吧?它租用算力。它不为内容付费,因为用户免费为它创造内容。它也不必花钱雇人管理内容,因为有志愿版主。

And just to think about it for a second, Reddit doesn't own data centers, right? It rents the compute power. Doesn't pay for content because its users create it for them for free.

And it doesn't have to pay people to manage the content because they have volunteer moderators.

所以这几乎是比 Netflix、YouTube 以及许多其他科技公司加起来还要好的商业模式。结果就是,毛利率大约在 91%,Reddit 每位员工带来的收入超过一百万美元。所以它实际上是一家非常精简的运营公司。

So that's almost a better business model than Netflix and YouTube and a whole lot of other tech companies combined. And as a result, you get gross margins that are around 91% and Reddit brings in over a million dollars of revenue per employee. So it's actually a very lean operation.

是的,Sean,那里的运营极其精简。我很乐意在这里谈谈投入资本回报率。所以我一般寻找的企业,最差情况下投入资本回报率大概有 15%,并且有望在近期维持、最好还能扩大这个数字。

Yes, it's an incredibly lean operation there, Sean. And I'm happy to do the honors here regarding returns on invested capital. So I generally look for businesses that worst case have maybe a 15% return on invested capital number and can hopefully maintain or preferably expand that number into the near future.

问题,或者也许我该说 Reddit 的美妙之处在于,它几乎没有投入资本。所以上季度的资本支出只有 120 万美元,这比我所在城市一些人买两居室公寓还便宜。

The problem, or maybe I should say the beauty of Reddit is that it barely has any invested capital. So capital expenditure last quarter was just 1.2 million dollars, which is cheaper than what some people in my city pay for a two-bedroom condo.

但没错,如果你拿 Reddit 资产负债表上的权益,剔除大约 30 亿美元现金,那么剩下的投入资本只有约 5 亿美元,而且其中大部分只是广告商实际欠它的钱。

But yeah, if you take the equity on Reddit's balance sheet and strip out roughly 3 billion of cash, well, you're left with only about 500 million dollars of invested capital, and most of that is just money that advertisers actually owe them.

而这 5 亿美元去年产生了远超 8 亿美元的税后营业利润。所以这家企业每年赚回的利润超过其全部投入资本。但有一个前提。Reddit 目前几乎不缴税,因为多年的亏损留下了相当大的一笔抵扣额,可以抵减今天的利润。

所以是的,这个税盾最终会用完。所以可以说,今天的利润肯定被略微美化了。

And that 500 million dollar produced well over 800 million of after-tax operating profit last year. So this business earns back more of its entire invested capital every single year.

But there's a caveat. Reddit barely pays taxes right now because years of losses left a pretty vast sum of deductions against today's profits. So yes, that tax shield will eventually run out.

So today's profits are definitely a little bit flattered, you might say.

但即使按相当正常的税负计算,Reddit 的资本回报率仍会远高于 100%。资本回报率极高通常带来的问题是,多余现金无处可投。Reddit 坐拥大约 30 亿美元,这是个好问题。

但这也意味着这家企业根本无法再投资。否则它们早就投了,对吧?如果它们能以超过 100% 的资本回报率再投资所有资金,那它们绝对会这么做,一开始也就不会有这么多现金。

But even with a pretty normal tax bill, Reddit's return on capital would still be well over 100%. The catch with the sky-high returns on capital is usually that there's nowhere to put the extra cash.

Reddit is sitting on roughly 3 billion dollars, and that's a good problem to have. But it also means the business can't reinvest at all. Otherwise they would have, right? If they could reinvest everything at over 100% return on capital, then they absolutely would do that and would not have so much cash in the first place.

到目前为止,Reddit 做了一件相当明智的事,那就是拿着多余现金,把它重新分配给股东。2026 年 2 月,他们宣布了有史以来首次回购,规模最高达 10 亿美元,目前已用掉大约三分之一,所以正在推进。上季度还有另外 1 亿美元用于回购。

And so far Reddit has done a pretty sensible thing, and that is sit there with its excess cash and redistribute it to shareholders. In February of 2026, they announced their first ever buyback for up to a billion dollars, and they've used about a third of that up so far, and so they're making progress.

And there's another 100 million dollars on that front that went toward buybacks last quarter.

真正的问题是,这些回购究竟是真的减少了股份数量,还是只是抵消了发给员工的股权激励,因为我知道这大概是会让你困扰的地方,Kyle。

And really the real question is whether those buybacks actually shrink the share count or just mop up the stock-based comp that's being handed out to employees, because I know that'll be something that probably bothers you, Kyle.

是的,你刚好提醒了我对 Reddit 资本配置故事最大的不满之一。我想我们俩都不太喜欢看到自己持有的公司股份被稀释。但这是你几乎不得不接受的现实,重要的是试着把它理清楚。

Yes, you just reminded me of probably one of my biggest gripes with Reddit's capital allocation story here. So neither of us, I think, really like to see shares in the companies that we own get diluted.

But you know, it's just a reality that you pretty much have to live with, and it's important to try and make sense of it all.

上季度股权激励约占收入的 15%,这绝对不算低,而且同比增长超过 40%。

So stock comp was about 15% of revenue last quarter, which is definitely not a low number, and that grew more than 40% from a year ago.

所以,尽管 Reddit 是一台现金流机器,过去 12 个月产生了 12 亿美元自由现金流,我还是必须在心里减去股权激励,因为我认为用股票而不是美元支付员工薪酬仍然是一项真实成本。

So, even though Reddit is a cash flow machine generating 1.2 billion of free cash flow over the last 12 months, I have to mentally subtract the stock comp because I think paying employees in shares instead of dollars is still a real cost.

但即便做了这项调整之后,我算出来的所有者收益仍然接近约8亿美元。这依然是个非常出色的数字,而且我认为这个数字对给这家公司估值相当重要。

But even after that adjustment, I'm still getting something close to about 800 million of owners earnings. Still an exceptional number and a number that I think is pretty important to value the business on.

但好消息是,现在的回购大致抵消了稀释。而且自回购计划公布以来,完全稀释后的股本数实际上基本持平。但仍然,稀释会一直是投资者心头的一根刺。这有点像浴缸开着水龙头、同时打开排水口。

现在两者是平衡的,所以水位没有上升,但这种情况随时可能改变。

But the good news is that the buybacks now are roughly offsetting the dilution. And since the buyback program was announced, the fully diluted share count has actually been pretty much flat.

But still, you know, the dilution will be a constant thorn in the investor's side. It's kind of like a bathtub with the tap on and the drain open. You know, right now they're balanced, so the water level isn't rising, but, you know, that is subject to change.

这是个相当好的比喻,用来理解我们在市场上很多公司看到的股权激励和回购。这个比喻我可能得借用一下。

That's a pretty good uh that's a pretty good metaphor for how to think about the stock based comp and buybacks we see in a lot of companies in the market. I might have to take that one.

不过说到回购价格,他们上季度每股支付了大约150美元,在最近这轮抛售之后,这看起来挺贵的,因为我们今天录制时股价大约在100美元。确实显得偏贵。

But yeah, on the buyback price, they paid around $150 a share last quarter, which looks expensive after the recent sell off with the stock around $100 at the time we're recording today. Definitely seems pricey.

但如果Reddit的价值接近我们认为它可能达到的水平,那它肯定也不是在烧钱。而且这绝不是你预期会从NVR这类公司身上看到的那种捡便宜式的回购。NVR是一家房屋建筑商,以非常策略性地运用回购作为资本配置的一部分来推动股东回报而闻名。

But if Reddit is worth anywhere close to what we think it can be worth, then it definitely isn't burning money either. And it's just not the bargain bin buyback that you would expect to see from company like NVR, which is a home builder that was famous for using buybacks very strategically as part of its capital allocation to drive shareholder returns.

然后AutoZone是另一家同样以此闻名的公司。所以比起回购,我其实更喜欢研发资金花在了哪里。研发支出增长很快,大约每年40%,但营收增长得更快,这告诉我这些钱是花得值的。

And then AutoZone is another company that is pretty famous as an example of that as well. And so what I actually like more than the buybacks is where the R&D money went. R&D spending has grown fast, something like 40% a year, but revenue has grown even faster, which tells me that money is is earning its keep.

其中很大一部分投向了机器翻译之类的东西,它能自动把Reddit翻译成其他语言。既然国际用户和国际营收同步增长,我认为这是个相当明智的决定,对吧?因为你解锁了以前无法接触到这些内容的全新客户,并让更多人能够访问Reddit上的内容。

And a lot of it has gone into things like machine translation, which automatically translates Reddit into other languages. And since international users have grown alongside international revenue, I think it's a pretty smart smart decision, right?

Because you're unlocking brand new customers who weren't able to reach the content before and and making more of Reddit accessible to anyone.

正是如此。我真的认为机器翻译其实就是一个聪明的作弊码,因为内容本来就已经在那里了,它已经存在。Reddit基本上只需要让它变得可读,而他们在Reddit Answers上也在玩类似的套路。

Exactly. I I really think that the machine translation is really just a clever cheat code because the content that it's already out there, you know, it already exists. Reddit just basically has to make it readable and they're running a similar play with Reddit answers.

这是他们的人工智能搜索工具,让用户可以与讨论内容对话。它的周活跃用户从2025年初的约100万增长到同年年底的约1500万。而且他们甚至还没有对它进行变现,这意味着未来某个时候它可能成为他们的另一个增长引擎。

So this is their AI search tool which allows users to converse with the discussions. It went from about a million weekly users at the start of 2025 to about 15 million by the end of that exact same year.

And they haven't even monetized it yet, meaning it could be another growth engine for them at some point into the future.

所以我会说,Reddit在资本配置和研发投入方面做得非常非常好,但在回购或股权激励方面,得分就没那么高了。所以除了用股权激励支付员工薪酬之外,这家企业其实没有多少大额账单需要担心,主要是因为Reddit实际上没有任何债务。

他们资产负债表上大约有30亿美元现金,我预计他们会继续把这笔钱投入回购,但会用经营产生的现金来补充。

So I'd say that Reddit is doing a really really good job on capital allocation and the R&D outlays, but it just doesn't score quite as high on buybacks or the stock comp though.

And so besides stock-based comp to pay employees, there really aren't many substantial bills that the business has to worry about, namely because Reddit doesn't have any debt effectively.

They have roughly $3 billion in cash on the balance sheet and I expect they will continue to deploy that into buybacks, but they'll replenish it with cash generated from operations.

这是一家创造现金毫无问题的公司,而且他们还有未使用的信贷额度,他们将其缩减到5亿美元,并把到期日推迟到2030年。这有点像企业版的注销一张你基本从不使用的信用卡,或者给它办个延期。

This is a company that does not have any problem generating cash and they also had unused credit lines which they shrank to $500 million and pushed the maturity date out to 2030.

So that's sort of the corporate equivalent of canceling a credit card you basically never use or getting an extension on it.

在我们这次重新审视这只股票的过程中,我花在读Reddit债务契约上的时间,肯定比Reddit实际使用债务的时间还要多。

And on this revisit of the stock that we've been doing, I definitely have spent more time reading about Reddit's debt covenants than Reddit has ever probably spent actually using debt.

是的,我认为,就像你说的,这里没什么好担心的。没有利息支出,利润丰厚,而且鉴于现金不断涌入这家企业,他们真的没有必要出去融资或发行股票。

Yeah, I think, you know, there's just like you said, not much to worry about here. No interest expense, a ton of profits, and no real need to go out and raise money or issue shares just given how much cash just keeps piling into the business.

所以他们最接近债务的东西,是每年2.2亿美元的云计算承诺,而他们用现金流非常非常轻松就能覆盖。所以是的,我认为你说得很到位。这里没什么好多说的。

So the closest thing that they do have the debt is this $220 million a year in cloud computing commitments which they very very easily cover from their cash flow. So yeah, I think you pretty much nailed it. Not much to go into here.

我可能有一个与债务无关的想法,但大概还是值得讨论一下。我们喜欢思考企业是否对社会以及它们所触及的生态系统是双赢的。有些企业对企业自身是赢,但也许对供应商或客户是输。然后也有像Costco这样的双赢局面。

There's maybe one thought I had that isn't related to debt, but it's probably still worth discussing. And we like to think about whether businesses are a win-win for society and whatever ecosystem that they they touch.

And you get some businesses that are a win for the business, but perhaps a loss for suppliers or customers. And then you get win-win situations like Costco.

而说到Reddit,答案就没那么清晰了。用户免费得到答案。广告主得到正在寻找他们所售产品的人。但版主和发帖者创造了平台上大量的价值,他们却没有得到任何价值回报,尤其是与YouTube这样的平台相比——在YouTube上,创作者能分到他们所产生广告收入的一部分。而他们做这些真的是为了社区,不是为了钱。

And when it comes to Reddit, there's not as clear of an answer. Users get free answers. Advertisers get people who are looking for products they sell. But moderators and posters who create a lot of the value on the platform, they don't capture any value in return, especially compared with a platform like YouTube where creators get paid a split of the ad revenue that they generate.

And they do it really for the community, not the money.

而且,这种做法一直有效,直到它不再有效为止,我想。作为这种做法可能反噬的一个例子,2023年Reddit开始向第三方应用收取数据访问费用。而许多版主依赖这些第三方应用来获取上下文信息。

于是作为抗议,成千上万的子版块集体关闭,而Reddit的回应是告诉版主们必须重新上线,否则就会被替换成愿意上线并保持子版块活跃的版主。

And And that works right up until it doesn't, I suppose. And so as an illustration of how that can backfire, in 2023 Reddit started charging third-party apps for access to its data.

And many moderators relied on these third-party apps for context. And so as a protest, thousands of subreddits went dark, and they responded by telling moderators that they had to come back online or they'd be replaced with moderators who would go online and keep the subreddits active.

而且它奏效了,但它仍然向所有人表明,社区确实拥有很大的力量,也许甚至比Reddit愿意看到的还要大。我记得整个过程都让人非常困惑,同时也挺搞笑的,因为一些最大的子版块会用梗图来抗议,连续好几天只发喜剧演员John Oliver的内容。

这就是非常典型的互联网幽默,也是Reddit抗议的一个绝佳缩影。

And it worked, but it still showed everyone that the community does have a lot of power, maybe even more power than Reddit would like to see. And I remember the whole time it just being very confusing and then also kind of hilarious because some of the biggest subreddits would protest with these memes where they would only post John Oliver, the comedian, for for several days.

And that's just very quintessential internet humor and a great epitome of a Reddit protest.

但在这些玩笑背后,我确实认为有一个真实的提醒:版主们实际上并不拥有这些社区,尽管其中一些人花了多年心血来建设它们。而Reddit可以通过做出对业务更有利、对版主却不那么有利的决定,来挥霍它与这些版主之间积累的善意。他们的利益并不总是完全一致。

But underneath the jokes, I do think there was a real reminder that the moderators don't actually own the communities that some of them have spent years of their life building.

And Reddit can squander away the goodwill that it has with those moderators by making decisions that are better for the business and not so good for the moderators. Their incentives are not always perfectly aligned.

是的,而且我认为这一切其实都回到了之前那个问题:内容到底归谁所有,对吧?Reddit仍然拥有内容,而且它对版主们有影响力,但很明显,如果版主们觉得自己在某种程度上受到了怠慢,他们绝对能制造一些麻烦,至少在一段时间内是这样。

Yeah, and I think this all really goes back to the point earlier about who really owns the content, right? Reddit still does and they have leverage over their mods, but it's clear that the mods can definitely stir up some trouble at least for a time if they feel they've been slighted in any way.

所以,既然Reddit依赖大量免费劳动力,确保其创作者和版主们满意就至关重要,因为如果Reddit不得不开始为内容创作者掏钱,它的商业模式就远没有那么好用了。所以这今天还没有体现在资产负债表上,但这是一个真实的风险,因为免费劳动力是Reddit利润率如此之好的一个重要原因。91%的毛利率,我的意思是,这简直离谱。

So since Reddit relies on a lot of free labor, it's vital for them to ensure that their creators and their mods are happy because Reddit's business model doesn't work nearly as well if they have to start shelling out dollars for content creators.

So it doesn't touch the balance sheet today, but it's a real risk because, you know, free labor is a big part of why Reddit's margins are just so good. You know, 91% gross margins, I mean it's it's insane.

但资产负债表是否干净,其实只取决于决定如何使用它的人。而那个人就是Steve Huffman。所以Sean,我知道你对Huffman的过往经历特别有感情。

But the balance sheet is really only as clean as the person deciding what to do with it. And that man is Steve Huffman. So Sean, I know that you've got a real soft spot for Huffman's backstory.

我觉得他是个很有意思的人物。是的,Steve Huffman,Reddit的创始人兼CEO,确实有着相当有趣的背景。基本上,2005年,Steve和他的大学室友Alexis Ohanian向Y Combinator推介了一款通过短信订餐的应用。

而Y Combinator,这个著名的孵化年轻初创公司的机构,最终拒绝了这个想法,但真的很喜欢Steve和Alexis,让他们去尝试做点别的,然后再回来。

I think he's an interesting character. Yeah, Steve Huffman, Reddit's founder and CEO, does have a pretty interesting background. I mean basically in 2005, Steve and his college roommate Alexis Ohanian pitched Y Combinator on an app by ordering food by text message.

And so Y Combinator, the famous group for developing young startups, ended up rejecting the idea but really liked Steve and Alexis and told them to go try building something else and and come back.

于是他们把这个建议记在心里,带着他们所称的“互联网首页”回来了。但大约一年后,他们把在Reddit的股份卖给了Condé Nast,价格大概在1000万到2000万美元之间。

And so they took that advice to heart and came back with what they referred to as the front page of the internet. And about a year later though, they sold their stake in Reddit to Condé Nast for a ballpark price of maybe 10 or 20 million dollars.

于是Huffman后来实际上离开了,联合创办了一家旅游网站,然后他远远地看着Reddit成长到超出他预期的程度。

And so Huffman actually ended up then leaving and co-founded a travel site and then he watched from afar as Reddit grew beyond what he ever expected.

然后到2015年,公司确实陷入了一团乱麻。它遇到了真正的麻烦。于是Huffman被请回来担任CEO,从那以后他就一直在这个位置上。

And then by 2015, the company was really in something of a mess. It was having some real trouble. And so Huffman was brought back in as as CEO and that's where he's remained ever since.

天哪,我简直无法想象。我在IPO后股价上涨50%时卖出股票,然后看着它从那里又涨了大约六倍,就已经够难受的了。而Huffman,正如我所说,以大概1000万到2000万美元卖掉了整个公司,然后看着它变得价值数百亿美元。

所以区别在于,我猜他得到了回来修复它的机会。说实话,从某种意义上说,我也得到了,只是规模小得多。

And gosh, I just I can't imagine. I mean I felt bad enough selling my IPO shares after a 50% pop and then watching the stock go up something like sixfold from there. Huffman sold the whole company for maybe 10 or 20 million dollars, as I said, and watched it become worth tens of billions of dollars.

And so, the difference is I guess he got the chance to come back and fix it. And honestly, in a way, so did I, but on a much smaller scale.

嗯,Sean,我觉得也许你和Huffman该成立一个Reddit早期卖家互助小组。我相信某个地方肯定存在这样一个子版块。不过,他担任CEO的第二个任期相当令人印象深刻。所以,自2020年以来,Huffman把Reddit的收入从大约2.3亿美元提升到超过30亿美元,营业利润率从深度负值一路提升到接近30%。

Well, Sean, I think maybe you and Huffman need to start a Reddit early seller support group. I'm sure there's a subreddit that exists out there somewhere. But, you know, his second term as CEO has been quite impressive.

So, since 2020, Huffman has taken Reddit from about 230 million in revenue to more than 3 billion and operating margins from deeply, deeply negative all the way up to nearly 30%.

是的,收入增长超过12倍,利润率从非常负到接近30%。而且在我看来,他已经挣到了他的报酬。去年他的薪酬略高于300万美元,看起来相当 modest,但这也有些误导性,因为他在2023年确实获得了一笔巨额股权授予,当时价值1.92亿美元。

Yeah, revenue up more than 12-fold and margins going from very negative to nearly 30%. And I mean, that is as far as I'm concerned, he's earned his pay. And at a little over 3 million dollars last year, his pay looks pretty modest, but also it's a little misleading because he did have this giant equity grant that he received in 2023 that was valued at 192 million dollars at the time.

所以,还记得我们说过股权激励在Reddit是个真正的问题吗?所以,尽管他的年度薪酬看起来微不足道,但薪酬本身一直非常不均衡。

So, remember when we said that stock-based comp is a real problem at Reddit? And so, even though his annual comp looks puny, the compensation itself has been very lumpy.

是的,所以他在2023年确实大赚了一笔。他最初的薪酬方案在股价达到约90美元的目标时兑现,但Reddit在IPO后最终远远超过了这些目标。于是董事会最终制定了一项新计划。

Yeah, so he made quite the windfall back in 2023. His original package paid out at stock price targets up to about 90 dollars, but Reddit ended up blowing through those after the IPO.

So, the board ended up putting a new plan into place.

现在,我认为新计划还可以。它与三件事挂钩:日活用户、收入和调整后EBITDA,后者基本上是一个现金利润指标。而日活用户是下个季度可能萎缩的那个数字。所以,Huffman有充分理由让它重新增长。

Now, I think the new plan is okay. It's tied to three things, daily users, revenue, and adjusted EBITDA, which is basically a cash profit measure. And daily users is the one number that might shrink next quarter.

So, Huffman has every reason to get it moving again.

所以,在他远超最初的基准之后,董事会基本上重置了记分牌。我觉得这我可以接受,因为业务确实表现超出了预期。但我觉得更难接受的是治理结构。Advance,也就是Condé Nast的母公司,持有约62%的投票权,然后他们有一项投票协议,Huffman的阵营控制着大约四分之三的全部投票权。

所以,像我们这样的公众股东对公司发生的事情真的没有发言权。

So, the board basically reset the scoreboard after he blew past the initial benchmarks. And I think I can live with that because the business genuinely outperformed. And what I find harder to accept though was the governance structure.

Advance, which is Condé Nast's parent company, holds about 62% of the voting power, and then they have a voting agreement where Huffman's block controls roughly three-quarters of all the votes.

So, public shareholders like us really have no say in what happens in the company.

关于Huffman的一个引人注目的细节,也许能窥见他的心理,就是他去年在个人安保上花了110万美元。这大约是他报告年度薪酬的三分之一。信不信由你,我不认为这完全是为了他自己的自尊。

我的意思是,当你面对一大群Reddit用户时,事情可能会变得有点棘手。在某种程度上,他对很多人来说是一个既被爱又被恨的人物。

And one fascinating detail about Hoffman and maybe an insight into his psyche is that he's spent 1.1 million dollars on personal security last year. So that's about a third of his reported annual compensation.

And believe it or not, I don't think it's entirely for his own ego. I mean, when you have a large group of Redditors out there, things can probably get a little dicey. He is a loved and hated figure in some ways to a lot of those people.

是的,显然如此。所以就控制权而言,我认为,当你觉得自己可以信任开车的人时,仅仅当个乘客是可以的。所以在Huffman的情况下,我在这里不太担心。他似乎了解这项业务,并且把股东的利益放在心上。

Yes, apparently. So in terms of control, I think, you know, just being a passenger is okay when you feel that you can trust the person who's driving the car. So in Hoffman's case, I'm not too worried about it here.

He seems to understand the business and has shareholders' interests at heart.

所以他还持有约500万股股票和期权,即使经历了我们这次的大幅下跌,这些也价值数亿美元。所以他显然是有切身利益的,我认为这与股东形成了相当稳固的利益一致性。

So he also owns about 5 million shares and options, which are worth several hundred million dollars even after this massive drop that we've had. So he definitely has some skin in the game, which I think creates some solid alignment with shareholders.

但我要说,有一点确实让我有点不舒服,就是锁定期结束后出现了大量内部人减持。而股价表现相当不错。所以一方面,我确实理解,

But I will say that one thing that does bug me a bit is a large amount of insider selling since the lockup ended. So the share price has done quite well. So on the one hand, I do understand that, you know,

而这还是在公司回购自家股票的情况下发生的。你会想,如果股价便宜到公司都愿意买,那么一些内部人士或许也会跟着买。所以我会关注这次下跌之后有没有人出手增持。

And that's despite the company buying back its own stock. You'd think that if the stock is cheap enough for the company to buy, well, then some insiders might also join in. So I'll be telling whether anyone steps up after this drop.

所以我对管理层的总体看法是:作为运营者,Hoffman非常出色。作为资本配置者,他一直表现稳健。这里没有搞帝国扩张,看起来资本也没有被浪费在愚蠢的收购之类的事情上。

So my high-level view on management is that as an operator, Hoffman is excellent. As a capital allocator, he's been solid. There's no empire building going on here and it doesn't really look like capital's being wasted on silly acquisitions or anything like that.

在信息披露方面,我会给他们一个低于平均的评分,因为他们计划不再披露登录用户拆分数据,而我认为这个指标才真正真正体现了这项业务的走向。而这恰恰是Reddit转向仅登录用户模式后我们最需要的数字,这就引出了风险。

Now on the disclosure side, I'd give them a below average grade for planning to drop the logged in split, the one metric that I think really really shows where this business is heading.

And that's exactly the number we'll need most as Reddit goes logged in only, which brings us to risks.

还有一个我们没提到的风险是监管。我要说,对某些其他社交媒体公司来说,监管是更大的问题。Reddit确实运营着互联网上最大的公共论坛之一,随之而来的是日益加重的监管负担。

And one risk we haven't touched is regulation. And I would say regulation is a bigger issue for some other social media companies. You know, Reddit does run one of the largest public forums on the internet and there is a growing regulatory burden that corresponds to that.

在美国,人们多年来一直争论的第230条款在很大程度上保护平台免于承担用户发帖相关的责任。所以如果有人在你的平台上发布仇恨言论,你未必需要承担法律责任。欧洲执行《数字服务法》,英国有《在线安全法》。

澳大利亚现在禁止16岁以下儿童使用社交媒体。这些规则每一项都给一个真正建立在数十万个虚拟社区匿名发帖基础上的平台增加了成本和复杂性。

In the US, section 230 though, which people have been fighting over for years, largely shields platforms from liabilities associated with what users post. So somebody posts hate speech on your platform, you're not necessarily legally liable.

Europe enforces the Digital Services Act and the UK has the Online Safety Act. Australia now bans kids under 16 from social media. Each of those rules add costs and complexity for a platform that's really built around anonymous posting across hundreds of thousands of virtual communities.

而这实际上正是Hoffman给出的转向仅登录模式的理由之一:要求人们想访问Reddit就必须创建账户。当你知道网站上是谁的时候,跟上年龄限制之类的新规则和监管就容易得多。

And that's actually one of the reasons Hoffman gave for going logged in only, requiring people to create an account if they want to access Reddit. It's just a lot easier to keep up with the new rules and regulations like age limits when you know who's on the site.

然后还有一点风险,就是这只股票仍然对新闻高度敏感。7月份我们又一次看到,《华尔街日报》报道Reddit在续约谈判中考虑限制Google对其内容的访问。当天股价下跌约9%,而据报道这笔交易价值仅约6000万美元,约占收入的2%。

And then there's some risk just that the stock continues to be highly sensitive to news. And we saw again in July when the Wall Street Journal reported Reddit was considering restricting Google's access to its content during renewal talks.

Stock fell about 9% that day over a deal that was reportedly only worth about $60 million or 2% of revenue.

所以这是一只波动剧烈的股票,这既创造了机会,也可能让一些人在应对时感到情绪上的挑战。而且我认为投资者关注的并不是那6000万美元。他们真正看的是与Google更广泛的关系,他们更担心的是与Google的争斗会如何影响流量。

So it's a jumpy jumpy stock which creates opportunities but can also be emotionally challenging for some people to navigate. And I don't think investors were focused on the $60 million.

They're really looking at their relationship with Google more broadly and they're much more worried about how a fight with Google would affect traffic.

而对我来说,最让我担心的风险仍然是Google。监管是一种缓慢而稳定的成本。但Google可以像我们见过的那样,一夜之间改变Reddit的流量。

And for me, the risk that worries me most is still Google. Regulation is a slow and steady cost. But you know, Google can change Reddit's traffic overnight as we've seen.

而我反复回到的问题是:在整个流量漏斗可能消失之前,Reddit能否把足够多的Google访客转化为真正在Reddit上创建账户的人。

And the question I keep coming back to is whether Reddit can turn enough of those Google visitors into people who actually create accounts on Reddit before that entire traffic funnel potentially goes away.

是的,我觉得监管这一点让我想到拥有一个巨大的公共公园。大多数日子里,家庭会来野餐、遛狗、在公园里玩耍。但当有人往凉亭上喷漆或踩踏花草时,你也要负责。而且每年城市都得增加几条你必须监控的规则。

这实际上就成了这项业务的一种永久性税负。而且我看不出短期内它会变得更便宜。

Yeah, I think the regulation point kind of makes me think of owning, you know, a giant public park. So, you know, on most days families are going to picnic, they're going to walk their dogs, they're going to play in the park.

But you're also responsible when someone spray paints the gazebo or tramples the flowers. And every year, the city has to add a few more rules about what you have to monitor. And that really becomes a permanent tax on the business.

And I don't really see it getting any cheaper anytime soon.

我要补充的另一个风险是算力。Reddit上个季度的计算成本增长了40%以上。这仍然慢于收入增长。但Reddit Answers、机器翻译和AI个性化都会消耗算力,而Reddit是租用其服务的。

So the other risk that I'd add is in compute. So Reddit's computing costs grew more than 40% last quarter. And that's still slower than revenue. But Reddit answers, machine translation and AI personalization all burn compute and Reddit does rent its service.

虽然这对资本效率来说是好事,但事实是仍然有一个房东在为他们定租金。所以尽管毛利率仍然很高,关注它们希望朝哪个方向变化肯定很重要。

Now, while that's fantastic for capital efficiency, the fact is there's still a landlord who still sets the rent for them. So while gross margins are still massive, it's definitely important to pay attention to what direction they're hopefully going to move.

如果非要给风险排序,我大概会把与Google的关系排在第一位,对用户数量的敏感性排第二,监管排第三。算力似乎相当可控。而且考虑到现在有这么多算力正在上线,确实有理由认为长期来看这项成本可能会下降。

If I had to rank the risks, I'd probably put the Google relationship first, sensitivity to user number second, and then regulation third. Compute seems quite manageable. And with how much compute power is now coming online, there's a real case to be made whether this could actually go down over the long term.

但即使把这一切都考虑进去,这家公司指引下季度收入增长约40%,同时又警告用户数量实际上可能萎缩。这说明增长有多少来自更好的变现,而不是更多的人。

But even with all of that taken into account, this is a business guiding for roughly 40% revenue growth next quarter while warning that its user count could actually shrink. Now, that tells you how much of the growth is coming from better monetization, not more people.

Reddit每周用户超过5亿,其中近1.9亿在美国。只有约5000万美国人每天使用。也就是说,大约有1.4亿美国人已经在某种程度上使用Reddit,但并不是每天都回来。

Reddit has over 500 million weekly users and nearly 190 million of them are in the US. Only about 50 million Americans show up on a daily basis. That's roughly 140 million Americans who already use Reddit to some degree, but don't come back every day.

而且我认为Reddit不需要说服数百万人开始使用这个平台来推动增长。它只需要说服那1.4亿人中的一部分更频繁地回来。

And I don't think Reddit needs to convince millions of people to start using the platform to drive growth. It just needs to convince that chunk of 140 million people to start coming back more often.

是的,我甚至不会把自己算作每周都用 Reddit 的用户。如果把通过 Google 找到的 Reddit 帖子也算上,那也许算。但多年来,Reddit 自带的搜索实在太差了,搜索 Reddit 内容的最好办法就是直接用 Google。

这几乎就像你拥有一座图书馆,但它的卡片目录坏得离谱,人们只能跑到街对面的图书馆去查你的书在哪个书架上。

Yeah, I wouldn't really consider myself to be even a weekly user of Reddit. Maybe I am if you count Reddit threads that I find through Google. But you know, for years, Reddit's own search was just so bad that the best way to search for Reddit was to just use Google.

It's almost like owning a library with a card catalog that's just so broken that people go to the library across the street just to find out what shelf your books are on.

幸运的是,他们开始解决这个问题了。到2025年底,每周约有8000万人使用 Reddit 搜索,高于一年前的6000万;而 Reddit answers——有点像内置于 Reddit 的 LLM 或聊天机器人,用来帮你在 Reddit 上查找信息——现在已经内置到应用里了。

They are starting to fix that fortunately. About 80 million people a week used Reddit search by the end of 2025 and that was up from 60 million a year earlier and Reddit answers, which is sort of like LLM or chatbot built into Reddit for helping you find information on Reddit is now built into the app.

所以,每一次从 Reddit 开始的搜索,就少一次依赖 Google 的搜索,这能让用户留在 Reddit 上、留在 Reddit 广告主面前,而不是跑到 Google 的页面上。

So every search that starts on Reddit is one less search that depends on Google and it keeps people on Reddit in front of Reddit's advertisers instead of on Google's pages

说实话,我最初在推介里看到 Reddit answers 时,我觉得自己并不太买账。我的意思是,我当时称它是为一个并不存在的问题提供的解决方案,因为只要在 Google 搜索末尾加上 Reddit 就行了,现在看这显得非常天真;但那时 Google 还乐于把用户引向 Reddit,而现在两家关系已经破裂,Google 的 AI 真的会自己回答问题,Reddit 拥有自己的 AI 来检索自己庞大的对话和内容档案,这看起来比我当初给予管理层的认可要重要得多。

and I'll be honest when I first looked at Reddit answers in my original pitch, I don't think I was very sold. I mean I called it a solution to a problem that didn't exist because it was just as easy to add Reddit to the end of a Google search and boy that seems really naive, but that was when Google was happily sending folks to Reddit and now that we've had this break in relations between the two and Google's AI really answers questions itself, Reddit having its own AI for scraping through its own archive of conversations and content it seems like something that's much more important to have than I gave management credit for.

是的,这完全合理,而且我认为这些搜索流量目前都还没有变现,这大概正是它更有意思的地方。我不会在基准情景里计入哪怕一美元的搜索收入,但我觉得这算是那种我更愿意免费获得、而不是花钱去买的上行空间。

Yeah, that's totally fair and you know I don't think any of that search traffic is monetized yet, which is probably what makes it even more interesting. I wouldn't put a dollar of search revenue in a base case, but I think it's kind of the upside that I'd rather get for free than to actually pay for.

不过,我大概要给自己这份热情泼点冷水,因为改变人们的搜索方式是在互联网上最难做到的事情之一。微软花了数十亿美元想让人用 Bing,而我老实说都想不起上次用它是什么时候了。Google 只是我长期养成的纯粹习惯,我自己就是完美的例子。

我甚至都不知道 Reddit 搜索已经变好了,即便现在知道了,我大概还是会出于习惯继续在 Google 搜索里加上 Reddit。

Still, I think I'll probably push back on my own enthusiasm because changing how people search is one of the hardest things to do on the internet. Microsoft spent billions of dollars trying to get people to use Bing and I honestly can't remember the last time I used it.

Google is just a pure habit that I've created over time and I'm a perfect example. I didn't even know Reddit search had gotten better and even now that I do, I'll probably keep adding Reddit to my Google searches just out of habit.

所以搜索在改善、数据也支持这一点,这很好,但把它转化为真正的收入显然仍在推进之中。作为 Reddit 的股东,我得说 Bing 这个对比有点扎心,但 Reddit 确实拥有 Bing 从未有过的自身优势,对吧?

它拥有自己的内容。Bing 一直想在 Google 自己的游戏里打败 Google,而 Reddit 只是请人们去答案本来就存在的地方搜索。

So it's great that search is improving and the numbers back it up, but turning it into real revenue is still definitely a work in progress. As a Reddit shareholder I have to say that the Bing comparison stings a little bit, but Reddit does have its own advantages that Bing never had, right?

It owns its own content. Bing was always trying to beat Google at Google's own game where Reddit is just asking people to search place where the answers already live.

另一个效果不错的增长杠杆是国际业务,对吧?机器翻译为那些以前无法消费任何内容的人解锁了 Reddit 的内容档案。现在一个说法语的人可以用法语读到关于修理自行车链条的最佳英文帖子,然后用法语回复,而原帖作者会以英文看到它。所以这是一个相当强大的进展。

And the other growth lever that has worked well is in international, right? Machine translation unlocked archive of content on Reddit for people who couldn't consume any content before.

And a French speaker can now read the best English thread on fixing a bike chain in French, and then they can reply in French, and then the original poster will see it in English. So this is a pretty powerful development.

本季度,国际日活用户增长了约25%,国际收入增长了约三分之一,这意味着他们显然在扩大用户数量,同时每位用户带来的收入也在增加。

And in this quarter, international daily users grew about 25% and international revenue grew by about 1/3, which means they're growing the user count obviously, and then they're generating more revenue per user.

这也将是 Reddit 首次测试仅限登录用户体验的地方。所以未来几个月这一点非常值得关注。我非常期待下一份第四季度财报,听听相关情况。

And it's also where Reddit will first test the logged-in only experience. So that is going to be very important to watch over the next few months. And very keen for the next Q4 earnings report to hear about that.

然后是数据授权安排。据报道,Google 和 OpenAI 的交易每年价值约1.3亿美元。如果其他 AI 公司想获得这么多人类对话的访问权,它们可能也得付钱。而这几乎都会是 Reddit 的纯利润。

And then you have the data licensing arrangements. Google and OpenAI deals are reportedly worth about $130 million a year. And if other AI companies want access to that much human conversation, they may have to pay up too.

And that would nearly all be pure profit for Reddit.

而且我想说,登录墙会让 Reddit 难被爬取得多,这肯定应该让它在谈判桌上拥有更多筹码,至少面对这些 AI 公司时是这样。

And I would say a login wall is going to make Reddit much harder to scrape, which definitely should give it more leverage at the negotiating table, at least with these AI companies.

是的,我认为你刚才谈到的国际扩张非常非常好,但我也认为美国与国际之间每用户收入的差距永远不会完全弥合。我们提到过,美国的广告主知道美国客户就是更有价值。但我也不认为它真的需要弥合。

如果国际每用户收入在五年内只是翻一番,同时这些用户继续增长,那也是一笔几乎不需要资本投入的巨额收入。

Yeah, I think while that international expansion you just went over is very, very nice, I also think the gap between the US and international revenue per user will never fully close.

So we mentioned that advertisers in the US know that the US customers are just more valuable. But I also don't think it needs to really close. So if international revenue per user just doubles over five years while those users keep growing, well, then that's an enormous amount of revenue requiring almost no capital.

不过,重要的是要记住,目前主要靠美国广告定价在挑大梁。管理层此前给出的本季度增长指引约为48%,实际交付了53%。所以超预期是真实的,但下一季度的指引更接近40%,而且用户数可能萎缩。

我认为正是这种组合让股价一天之内下跌了约四分之一。

Still, it's important to remember that the US ad pricing is doing most of the heavy lifting right now. Management had guided to about 48% growth this quarter and delivered 53%.

So the beat was real, but next quarter's guide is closer to 40%, and the user count could shrink. And that combination I think is what sent the stock down about a quarter in a day.

不过,一家增长40%以上、毛利率90%以上且没有债务的企业,并不会让我觉得它陷入了很大麻烦。在我最初的推介里,最大的问题是 Reddit 最终会更像 Meta 还是更像 Snapchat,我们今天已经讨论过这一点,我认为答案已经很清楚。

Still, a business growing 40% plus with a 90% plus gross margins and no debt isn't really telling me that it's a business that's in a lot of trouble. Back in my first pitch, the big question was whether Reddit would turn out more like Meta or more like Snapchat, and we've covered that today, and I think it's been clearly answered.

我当时假设到本十年末营业利润率也能攀升到30%左右,而 Reddit 不到两年就做到了。所以他们让我显得非常看空,而每用户收入,尤其是在美国,增长得比我预期快得多、快得多。所以在变现方面,我实在太保守了,保守得离谱。

I assumed operating margins could also climb to around 30% by the end of the decade, and Reddit got there in less than two years. So they made me look very bearish, and revenue per user, especially in the US, has grown much, much faster than I expected.

So on monetization, I was way too conservative, wildly too conservative.

至少目前看来,我过于乐观的地方是用户增长方面。我当初把 Google 当作一个相当可靠的、时有时无的顺风,我甚至当时就说 Reddit 并不完全依赖 Google,就收入而言这仍然成立,但就用户而言显然不是。而这个假设我大概需要下调。

Where I have, at least at the moment, seemed to be too optimistic was on the user growth side. And so I treated Google as a fairly reliable on and off again tailwind, and I even said at the time that Reddit wasn't entirely dependent on Google, and that's still true for revenue, but clearly not for users.

And that's the assumption I would probably need to mark down.

所以我今天不会给出一个精确的数字,因为坦率地说,结果的范围已经宽了很多,这让准确给公司估值难了很多。而且老实说,我仍在消化最近这个季度业绩的很多消息,以及它与 Google 之间的紧张关系。

So I'm not going to throw out a precise number today because, frankly, the range of outcomes has gotten a lot wider, and that makes it really a lot harder to value the company exactly.

And I'm still, honestly, digesting a lot of the news of this most recent quarter's results and the tension with Google.

但我可以告诉你的是,我的想法在某些方面已经改变了。所以我现在假设用户增长会放缓,尤其是未登录流量方面,会更依赖定价和已登录的核心用户。然后我想用更低的估值倍数,来反映这家公司五年后可能以什么水平交易的不确定性上升。

But what I can tell you is my thinking has changed in some ways, right? And so I'm assuming slower user growth, especially from logged out traffic, leaning more on pricing and the logged in core of users.

And then I wanted to use a lower valuation multiple to reflect the increased uncertainty for where the business might trade five years from now.

把这些放在一起看,我对公允价值的判断肯定比一年前更低,但同时也低于每股100美元。就我个人而言,我很乐意买入。

If you put that together, my sense of fair value is definitely lower than it was a year ago, but at the same time below $100 per share. Personally, I'm a happy buyer.

是的,说得有道理。Reddit 一年产生约十亿美元现金流,所以关于这家公司更像 Meta 还是更像 Snapchat,我认为答案非常非常清楚,正如你刚才已经指出的那样。

Yeah, fair enough. So with Reddit generating about billion dollars a year in cash flow, I think the answer on whether the company is more like Meta or Snapchat is very very clear as you already kind of pointed out there.

但我认为反直觉的一点是,一家企业可以大幅超出你的预期,而它的价值变动仍可能比你想象的小,因为市场已经提前消化了其中很大一部分成功。

But I think the counterintuitive part is that a business can massively beat your expectations and its value can still move less than you'd think because the market was already pricing in a lot of that success.

而这个季度之后,我认为最大的问题不是明年的数字,而是如果 Google 继续把越来越少的人引向 Reddit,十年后的 Reddit 会是什么样子。

And after this quarter, I think the big question isn't next year's number. It's what's Reddit going to look like 10 years from now if Google keeps sending fewer and fewer people its way.

好消息仍然是,在每股100美元左右的价格上,我认为你在任何财务建模中都不需要激进的假设。剔除现金和股权激励费用后,你实际支付的,大约是我所说的所有者收益的低20多倍,而这家公司仍在指引40%的增长,关键在于这个故事本身应该就能说明问题。

The good news again is that around that $100 share price, I don't think you need heroic assumptions in any of your financial modeling. Once you strip out the cash and the stock-based comp, you're really paying something like a low 20s multiple of what I would call owner's earnings for a business that is still guiding to 40% growth with the point being the story should speak for itself really.

20多倍,是我所说的所有者收益,而这家公司仍在指引40%的增长,关键在于这个故事本身应该就能说明问题。我的意思是,如果 Reddit 继续增长其登录用户,并从每条广告中获得更多收入,我认为这就足以推动盈利增长,你不需要押注市场重新爱上这只股票。

20s multiple of what I would call owner's earnings for a business that is still guiding to 40% growth with the point being the story should speak for itself really. I mean, if Reddit keeps growing its logged-in users and keeps getting more out of each ad, I think that will do more than enough to drive earnings growth and you won't need to bet on the market falling back in love with the stock.

我想问题是我们该何去何从。对我来说,在我、你和 Daniel 三个人里,我大概是对 Reddit 最乐观的那个。所以当我看到 Reddit 以17倍营业利润交易时,我更倾向于出手。

但我也认识到,这家公司的增长存在巨大的问号。所以如果你认为想放弃这家公司,或者暂时不给我们在这家公司的持仓加仓,我肯定不会因此反对你。

And I guess the question is where that leaves us. For me, I'm probably the most optimistic of me, you, and and Daniel on Reddit. So I'm I'm a bit more inclined to pull the trigger when I see Reddit trading at 17 times operating profit.

But I also recognize that there are huge question marks around the company's growth. So if you are thinking that you want to pass on the company or pass on adding to our position in the company, for the moment I I certainly don't follow you for that.

是的,谢谢 Sean。我真的很喜欢 Reddit,我们在节目之前也聊了很多这方面。我认为这是一家非常好的企业。Sean 后来更详细地讲了他喜欢什么、认为这家企业能走向哪里。我仍然认为它处于非常非常好的位置。

Yeah, thanks Sean. And I think I really like Reddit and we talked a lot about this kind of before the episode. And I think it's a really good business. Sean ended up going even more into detail about what he liked and where he thought this business could go.

And I still think it's in a really really good position.

不过我仍然认为,即使它有这个非常好的增长数字,对我来说仍然有很多问题。Sean 肯定更了解这家企业。所以我实际上认为,从他的角度、大概也从 intrinsic value 组合的角度来看,暂时不动这个持仓大概就是合理的。

I still do think though even with this really good growth number that it has, there's still just a lot of questions for me. And you know, Sean definitely knows the business more.

So I actually think from his perspective and from probably the intrinsic value portfolio's perspective, it probably just makes sense just to leave the position alone.

也许吧,我认为 Sean 也暗示了,鉴于管理层暗示下个季度的数字可能会更难看,下个季度甚至可能有机会以低得多的价格加更多仓。

Perhaps I think Sean also insinuated that next quarter given the fact that management insinuated that numbers are probably going to be even uglier. There might be even a chance to add even more at a much lower price.

我要说的是,我对这家企业非常非常感兴趣。我认为我大概需要多花一点时间,也许真的登录上去用一用,更好地感受一下,这样我希望能稍微接近,我不认为我真的能接近 Sean 对这家企业的理解,但至少能更接近一些,因为我确实认为这是一个非常非常有吸引力的机会。

I will say that I am very very interested in this business. I think I probably need to spend a little more time, maybe actually just logging on and getting a better feel for it so I can hopefully somewhat approach uh I don't think I'll ever actually approach Sean's understanding of the business, but at least get closer to it because I do think that it's a very very compelling opportunity.

好了,今天就到这里。和往常一样,我想用一句话来结束,这句话来自 Reddit 的创始人兼 CEO Steve Huffman。如果他们拿走我们的内容并在此基础上建立业务,那是个问题。

如果他们建立的业务导致人们更少来 Reddit,那也是个问题。这句话是三年后说的,我认为这句话读起来几乎像是对今天正在发生的事情的预言。Reddit 肯定通过让数据获得报酬解决了第一个问题,但第二个问题,也就是人们更少来 Reddit,恰恰出现在这个季度的数字里,而只允许登录用户访问,就是 Huffman 押注他也能解决这个问题的做法。

Well, that's all we have for you today folks and as usual I want to leave you with a quote and this one is from Reddit's founder and CEO Steve Huffman. If they take our content and build businesses on it, that's an issue.

If they build businesses such that people come to Reddit less, that's an issue. Now this quote came out 3 years later and I think that quote reads almost like a prediction of what's happening today.

Reddit definitely solved the first problem by getting paid for its data, but the second one, people coming to Reddit less is exactly what showed up in this quarter's numbers and going logged in only is Huffman's bet that he can solve that one too.

观察点

仅限登录用户的测试中登录用户与国际用户的增长情况,将在下一份第四季度财报中披露
登录用户恢复增长而非萎缩
下一季度下一份第四季度财报
支持在100美元以下买入的论点

这个频道对 $RDDT 的历次判断

The Intrinsic Value Podcast 在这只股票上共 2 条判断,这里只列相邻的几条。

跨洋听风