All right, we're gonna go ahead and get started. I'm Doug Anmuth, JP Morgan's internet analyst. We're pleased to have with us today, Reddit co-founder and CEO, Steve Huffman. So Reddit is a global digital city built on connecting people to communities. It has more than 100,000 active communities or subreddits, where users go to share information, learn, and engage around specific topics and interests. Reddit reaches around 83 million users daily, more than 300 million users weekly. We estimate the company will do more than $1.1 billion in revenue this year, almost all of which has been advertising until now, but there's also a rapidly growing data licensing business that we'll talk more about. Steve co-founded Reddit in 2005.
He then went on to co-found travel company Hipmunk, where he was CTO, and he returned to Reddit in 2015 as CEO, and has been leading the company ever since. Welcome, Steve.
Morning, Doug. Hi, folks.
All right, to start off, maybe you can talk about how Reddit fits into social media and the broader internet space.
Sure. So look, if I had 20 minutes, I'd give you a fulsome answer here. So we've been around Reddit since 2005, and so we predate social media, we predate the word influencer, and so we've been doing things, you know, our own way for a long time, right? We didn't wake up and say: Hey, there's a better version of doing this. We've just been doing it this way. But I often frame people. I also, I often frame Reddit in juxtaposition with social media. And so social media is powered by the Fs, as I call them, friends, followers, family, famous people, and Reddit organized around community, and so the dynamics and incentives are really different.
Social media also has algorithms that make things popular, whereas on Reddit, things are made popular through voting and through voting in the context of a community. And every post and comment on Reddit starts out at zero points and has to earn its visibility. And so what this means is things don't become popular on Reddit unless a group of like-minded people make it popular. And so the content, if you kind of play all this through, what happens is on Reddit, is the content's just more authentic, and it'll be more vulnerable as well. We don't use real ID on Reddit.
On Reddit, you'll see people talking about not just like the headlines and what's going on in the world, which they do, and not just the memes, which we pioneered, for better or for worse, or GIFs, which were a big part of that too, but everything they're going through in life, right? The ups and downs, relationships, jobs, all those things. And so Reddit, long story short, we describe as the most human place on the internet.
Okay, great. You talked about how you've been around for 19 years. Maybe you could just help us understand the phases of the company, why it's just this year that Reddit will cross the billion-dollar annual revenue mark?
Sure. So the brief timeline, we started in 2005. We actually sold the company in 2006. I left in 2009. Reddit spun out from our acquirer, Advance, in 2012. They're still our largest shareholder. And I returned to the company in 2015. When I came back to the company in 2015, we did, I think, $12 million in revenue that year, and we were about 12 million DAU, I think. We didn't even measure it that accurately back then. And so the business didn't really get going until the last decade or so, and it wasn't until 2018 that we got our own ad tech going. And so it's been a journey.
We certainly have a lot of perspective on the internet, but the business part of Reddit is just beginning to scale now. And so as Doug said, yeah, call it about a $1 billion run rate, where we are today. So we've grown nicely in this latest phase, but it's been a, it's been a roller coaster-
Okay
To say the least.
Great. All right, so you've seen very strong user growth, four straight quarters of acceleration. Can you walk through the key drivers here and maybe talk about how big of a benefit some of the Google algorithm changes have been?
Sure. So yeah, we've been adding about 1 million users a month, so more than 1 million users a month, 8 of the last 10 months, and so we've had a nice, steady clip. The primary driver... and so that's 37% year-over-year, so, we're happy with that growth. The main driver is product quality, and so, about 18 months ago, I told our team to, I just said: "Quit being so academic about the product." We were A/B testing ourselves to death. I said, "Just focus on quality. Make it faster, make it easier to use." And long story short, our work is working. And so one of the main input metrics that we care about is new user retention. So a user shows up to Reddit, do they find something interesting to them? Do they find their home?
Do they find a community that speaks to them? Therefore, you know, do they increase the likelihood they come back? So we've gotten much, much better at that. So the performance on our website, it's 2-5 times faster than it was a year ago. That makes a huge difference. Native apps, faster, you know, startup times, loading times, all of those things, much, much faster. UI is much easier. Our recommendations, our machine learning recommendations, when you're trying to find communities, much more effective. That really started coming online last spring. Now, if you knew me over the last 3 years, you'd know, two of those 3 years, we had a Google headwind, as this algorithm change kind of, they come and they go.
You never hear me celebrate an algorithm change or complain about it, because they tend to wash out over time. As it happens, we've had a few go our way in the last year or so, but the biggest change is the web performance. The Google crawler, Googlebot, likes speed. If your page loads faster, you'll rank higher, and so the work we did on performance has made a nice improvement there. You will see where the algorithms go. You know, the algorithm changes come roughly twice a year, usually one in the fall, one in the spring, but our growth has been much more steady over the course of the year.
As I mentioned, 8 out of the last 10 months, over 1 million users, and so that speaks really to the performance impact and the quality impact versus these algorithm changes that may come and go.
What's your view on Google's new AI Overviews SERPs for both Reddit and the broader internet? You've got, you know, some unique perspective on the, you know, on the broader space. Just curious how you think they could impact the internet broadly in terms of traffic, and then if there's any impact to Reddit.
Yeah, look, again, with all these, like, algorithm changes or UI changes, we'll have to see when we see. Big picture, Reddit has what people are looking for, and Google helps people find things. Like, that's the math there is pretty simple, and especially in this AI era, where more and more content on the internet is written by AIs, there's an increasing premium on content that comes from real humans, and that's what Reddit is, right? Reddit's conversations, advice, reviews, stories about anything and everything from real humans. And so as long as people want that, and I think people will want that forever, then Reddit's, I think, will be a net beneficiary of this. Now, who's gonna get hurt? The spam websites that use AI to write their content.
I think we want that to happen because we want Google to send people to great content, and Reddit has the best content on the internet.
Okay. Sticking with users, we get a lot of questions about logged-in and logged-out users. Maybe you can just help us understand mix on the platform, the relative value of each, and how you're able to target those two different groups.
Sure. So we're about, call it 50/50 logged-out right now, maybe a little biased towards logged-out. The logged-out traffic is mostly from Google. No surprise there. Somebody asks a question, they end up, Reddit has the best answers on the internet, they end up on Reddit. But both are growing, so logged-in users grew 27% year-over-year. And the logged-in users, particularly logged-in users on the native apps, that's the bedrock of our ads business, and so that's the really the cornerstone of our revenue, and that is very consistent. You know, we see some volatility in the logged-out users, you know, as Google evolves and changes, but logged-in moves basically independently of that, and that's been steady and growing for a very long time.
Now, a super common question we get is: How do we convert the logged-out users into logged-in users? And, look, we've been very aggressive on this at times, but we've found that's not super effective, right? Putting up walls, forcing people to log in to see content, things like that. It doesn't work over the long term. It definitely works in the short term. But what works over the long term is just keeping our users happy. So if you come to us with a question and we've got the answer, we give you the answer, it's okay if you bounce.
But what we really care about is if you come to our front page or you open the app for the first time, that you find something like in those moments, you're coming to find a community, to be entertained, to make a connection, whatever it is. In those moments, that's where we put all of our effort, is focusing on new-user retention, and so that's the product quality, the recommendations, the things I was talking about before. And so we do see that mix change over time. Even, like when I came back to the company in 2015, we were probably 80%-90% logged-out, and so the logged-in trend has been shifting more and more, or the users have been shifting more and more logged-in, and that's really the growth of our native apps.
Just nice, kind of steady migration there.
Okay. Reddit's pretty unique in that the subreddits are controlled by more than 60,000 moderators. Then you obviously provide some broader platform oversight. How would you characterize your relationship with moderators, and what did you learn from the June 2023 protests and blackouts?
Okay, so some context here. When we say the word moderator, these are users. So when other platforms use the word moderator, they're often talking about paid employees or contractors. When we say the word moderator, these are users who create communities on Reddit. So you create a community, a subreddit, I use those words interchangeably, ta-da, you're a moderator. Moderators have complete control over their communities. So they set the, they write the rules, they set the tone, they can ban content, they can ban users, and so they're really the shepherds of community, and they're a really important part of Reddit, right? Without, there is no Reddit outside of communities, and there are no communities without moderators.
In fact, if a community doesn't have moderators, like if a moderator, if the last moderator leaves, we instantly ban that community, 'cause we have a rule on Reddit, no unmoderated spaces. Okay, so what Doug's referring to, last summer, we changed our policy. We had a. Well, we still have a free API. We decided to charge the heaviest users, about 6 users of the API, who are extremely heavy users. They're building Reddit apps. We said, "Hey, you gotta pay for the API usage." Turned into a big fight with our user base. I liken it to like an old married couple. The fight started about something day-to-day about the dishes, and then 10 years of history came up, and so it was my...
Not my favorite two weeks in the company history, but we did get through it. Now, Reddit is unique in that our users and moderators have a deep sense of ownership over Reddit. It's really, really important. One of the analogies I use for Reddit often is that Reddit is a city, and I could take this analogy really, really deep, but think of subreddits as neighborhoods. The same way that the citizens of a city self-organize into a neighborhood that have distinct cultures and vibes, Reddit, the users self-organize into subreddits. Well, in this moment, we had a protest in our city, and I think protests, just, like in the real world, are an important form of expression in a democratic society. We had a protest on our platform.
Lasted about, I don't know, a week, two weeks. There were some important points they raised outside of the API pricing thing-
Mm-hmm.
Which we didn't change, by the way. And so, you know, they made some important points. The moderators wanted better tooling, which we've delivered on. They wanted better accessibility support, in our apps for people who are, like, visually impaired, for example, which we've really closed the gap on. I think we'll be as good as anybody by the end of this year. So I think those are important dynamics. I can't say we'll never see a moment like that again, but I'd say I think we've dramatically de-risked something of of that scale.
Okay, great. Maybe you can talk a little bit about how you think about the opportunity for search on the platform. You're doing more than 1 billion queries per month. What's the potential to make search a key part of the Reddit experience?
Yeah, so, you know, generally, or I'd say historically, we've looked at Reddit as a, like, a real-time platform, right? You're having conversations about what's going on in the world or about what's going on in your life. Now, we've been doing this for almost two decades, so now we have two decades of conversation, of answers, advice, reviews about everything, and so it's one of the most valuable corpuses of content on the internet, and so search becomes really important. It's both important, I think, to unlock this latent value that's in our corpus, but it's also just important for the product itself. A ton of new users, about half, more or less, of new users in their first session on Reddit, run a search.
So if one of our biggest product challenges is taking a new user and helping them find their community on Reddit, half of them are typing into a box exactly what they're interested in. So big opportunity, I think, both to unlock value and to deliver on a great user experience. So, I'd say we're in the like, we have search working on Reddit, right? We're doing over 1 billion queries a month, so it's not nothing, but I think from a product point of view, we've got a long way to go. So I'd say we're still early there on delivering both really quality search results, the back end, and a really quality presentation of those results, the front end. So that's a pretty big investment of ours over the next couple of years to really make that world-class.
Okay. Let's shift to advertising. Your ad revenue grew almost 40% in 1Q. Can you just help us understand some of the biggest drivers of the ad upside in the quarter and, maybe how we should think about impact of broader online ad market improvement relative to some of the Reddit-specific initiatives?
Sure, so I think you can see both those things. But if you look at our peers in the advertising space, you can see they also had good quarters. So, like, the overall climate's getting better. For us, more specifically, what we saw over the last couple of years, our biggest category has been tech, so the big tech companies as advertising customers. That really flattened off over the last couple of years, just due to macro conditions, this or that. We've seen that come back, which I think just speaks to the overall improvement in health in the market. Now, the good news for us is, while tech was flat, we grew substantially in every other category.
So we've been growing nicely the last couple of years, even in choppy conditions. The conditions have gotten better. Now, that's. Let's, let's call that, like, half the explanation. The other half is our ads product is getting better, and so we've got a near infinite roadmap of work that we know will work. It's pretty straightforward. There are a number of other ad platforms on the internet, big and small, and we know exactly what works because we've seen them do it, and our customers tell us exactly what they want that other platforms have that we need. So our roadmap literally writes itself, and we grind our way through it, and just about everything we do works because they're known winners. Now, there's another dynamic that I think is really powerful, which is the uniqueness of Reddit itself, right?
Reddit, every hobby and interest is on Reddit. Everything you're thinking about doing or buying is probably on Reddit. So what should I watch? What should I wear? Where should I go? What car should I buy? How do I buy a house? What video game should I play? It's all there. The interesting thing about hobbies, and even I've just come to appreciate this in the last couple of years, is how latently or sometimes obviously commercial discussions about hobbies are. If you take up golf, you're gonna spend a lot of time, it turns out, thinking about what clubs to buy, what clothes should I buy, all this stuff, right? Every hobby is like that. And so more or less, every company's customers are on Reddit somewhere, and there's a high likelihood those customers aren't on other platforms.
You take the biggest platforms in the world, like Facebook and Instagram, 30% of their users, or 30% of our users aren't on those platforms at all. And the numbers just get bigger as you go to the other platforms. So every company's customers are on Reddit, and in many cases, they are uniquely and distinctly on Reddit, and they're literally talking about, "What should I buy?" And so it's a really natural commercial environment, which is funny for me to say, because 20 years ago, when we started Reddit, I didn't like ads at all. I've, I've come around.
A little bit. All right, you're coming off the deal with OpenAI, announced last week. Maybe you can talk about the data licensing opportunity on the platform more broadly, and what makes your content valuable to large language models?
Sure. So, you know, as I've mentioned, Reddit has got nearly two decades of conversation about anything and everything, and so it's a really valuable source of, first, just information, and then two, data and content for training large language models. And so, at the beginning of this year, we announced a deal with Google to license our data for training and search. And, just last week, we announced a similar deal with OpenAI, licensing our data for training. And, this is a new business for us. We really just started closing deals this year. We're talking to lots of folks in the space, big and small.
But again, this is, it's another opportunity for us, in addition to the core platform we have with the ads business, unlocking the value of our corpus. The way we look at the landscape right now is if you are building a search engine or something that's like a search engine, and so I would put ChatGPT into that category, right? You've got a text box, you type in what you're interested in, and you get answers. How can you be the best or be competitive at that without the largest source of human-generated content there is, or that's accessible, rather, at least? One unique thing about Reddit is that we're open, and we're open for business. The other platforms with our scale of content are all closed.
The reason that we want to be open and we're in this business at all, as opposed to hoarding it for ourselves, is because we're believers in the open internet, and we've been beneficiaries for a long time of the open internet. You know, just looking at, for example, the traffic that we get from Google Search. And so I think it's both important for consumers, for our data to be out there and be accessible. I think it's good for the internet, and I think it's good for Reddit because we bring in a lot of people to Reddit. So we're very considered in these deals about what our data is used for, how it's displayed, but our default position is we'd like it to be out there.
Now, the products in this space are evolving rapidly, and so we'll see where this goes. That's why we're being very considered about who we work with and about who will be respectful of our users and their privacy. Also, a little over a week ago, we announced our Public Content Policy , which is kind of a new concept in the internet. Everybody has a private privacy policy, which governs what people do, what platforms do with their, the private information they have. Ours is very short. It doesn't leave Reddit. We don't sell it, we don't license it, it doesn't go anywhere.
But the public content on Reddit, wanted to be clear that we're open to licensing it, open to even to giving it away for free to researchers, but only if you do so on our terms, which include provisions around not reverse engineering the identity of our users, for example, you know, not reselling it or rehosting it, things like that. And so, we drew the lines, or kind of put out the guardrails, and then companies that are willing to abide by them, we're open to doing business with. So, so far, Google and OpenAI are the big ones.
And on OpenAI, you mentioned it seems fairly similar to the Google partnership. Can you walk us through the different components? Just how should we think about relative financial impact?
Look, it'll start. This is the lamest answer I think I get to give now, is you'll start to see the numbers next quarter. Like I said, it's very similar to the Google deal. We try to keep things simple. You know, so we get money, they get a license of our data, subject to certain constraints, for training and other stuff. We'll also, there's some other smaller things in there. We'll be an API customer of OpenAI. They'll be an advertiser with us as well. So far, the deals that we've done and the deals that we're working on, they're all kind of bespoke. They almost feel like M&A deals in their complexity, which is fine, but it does mean they take, t hey do take some time.
Okay. If there's a mic, we're gonna open up to some questions. Let me ask you, I guess just to on continuing data licensing, those are two big deals. Is it fair to think you're in conversations with other large language models here, and there's other opportunity beyond this?
Yes. So we're talking to lots of folks in the space. It is a new market, so we're kind of developing it as we go. As I mentioned, most folks who are holders of large content sources like Reddit are closed. We're default open, or we would like to be. We'd like to have relationships with as many people as possible. So we're in the market doing that now. Now, we also have another class of customers that would be smaller companies for non-AI training purposes. So think like financial institutions, investors, social listening services. There's a lot more customers in that space. Basically, anybody who wants a real-time feed of what's going on in the world, what people are talking about, what products are interesting, what brands are interesting, things like that.
And so we're also doing deals on kind of the other end of the scale spectrum. Little smaller deals, but many more customers.
Questions in the audience? There's one in the back, way back near the door.
Steve, question for you about... Is there any traffic envisioned in this part of the deal with OpenAI, the way you do get with Google? Will there be any reference, credit, traffic that comes back to Reddit? And, how do you protect your source data from others who are not paying for it today?
Sure. So good morning, good to see you. The expectation would be yes. Now, they're at the kind of beginning of their journey on their own product. But one of the things we look for in these deals is, you know, you have to link to Reddit, basically. Now, we're confident, we're confident that the machine here should work well just because we've seen it for so long with Google. So the argument that one would make is, hey, if your content's out there being used for training or this or that, does that mean you're actually gonna see less traffic? Is somebody gonna just kind of intercept you?
I think that is certainly a dynamic for some, but just look at the last year, our traffic's growing faster than it has in years, even, or in particularly with logged-out traffic from platforms like these. And so I think, you know, trying to find the right balance of using this as effectively a marketing channel can be really powerful for us, and help consumers generally on the internet find what they're looking for, and train them, as we've seen, knowing that whatever your question is, Reddit probably has the answer. Now, we run the company day-to-day on DAUs, so it's 82 million in Q1. Our monthly is over 500 million, and so our top of funnels is absolutely massive.
And so I think the idea that Reddit is just out there, pervasive on the internet, on people's mind, is really powerful. You know, we don't do consumer marketing. That is how we grow Reddit and market Reddit.
Another question in the audience?
Yeah, thanks. Hey, Steve. Could you talk a little bit about the opportunity overseas? I saw that you're using ML to translate some of the Reddits and, you know, is that a big opportunity in the next few years for you? Thank you.
Thanks. Yes, great question. So today, we're about 50/50 U.S. versus non-U.S. If you look at our, just kind of peers in this space, we would expect that, 50% international number to actually be 80%-95%. I'm confident we'll get there even just naturally, but we want to get there faster. One of the things we're doing that I'm very excited about is using large language models to translate our entire corpus. And so a year ago, if you were to do this, it would be okay, but with large language models, you can do translation at human quality, and so you can pick up on, on humor and nuance and all these things. And so, we're in the middle of testing this right now. So French is the first language we're doing.
So we've got this online in France, where you can access all of Reddit's history in French, and it's a native quality translation. So we saw an instant improvement in testing with new user retention, and as an added bonus, Google indexes the translated content and serves us back users from French Google as well. So that one, we didn't, we didn't anticipate, so that was a nice bonus. So we're doing the product polishing as we speak. As I said, we're in testing right now. It's going well. The next stop after French is Spanish. There's 1 billion Spanish speakers, and then we'll go from there. But this is a very exciting opportunity. Look, Reddit is interests and passions. Everybody has interests and passions.
Community is universal, and so I think it is inevitable that we grow to that sort of scale outside of the U.S., but our goal is just to do it faster, and so the translation work really helps there. Now, outside of translation, we also do what we call program work, which is focusing on target markets. We have five target markets, making sure that communities that we know will work, local sports teams, big cultural things, exist, are well-moderated, discoverable in our product, good safety support. And so we're seeing nice results there as well. So, both on the translation and on the program work, we're optimistic about the future.
All right, good. An area you're very excited about is the Developer Platform. Maybe you can talk about that, what it is, what kind of compelling use cases could come out of it, and, and how would it be monetized?
Sure. So Developer Platform, this is the project that I've wanted to do on Reddit for a very long time. We've been working on it the last couple of years. It's coming online as we speak. And so a post on Reddit today is, a text, a link, an image, or a video. That's basically it. And so with the Developer Platform, what we're allowing users to do is create custom posts, posts basically that are code, so like a post that is almost like an app. So a post that can be dynamic, that can have memory, that can interact with each other and with users. And so this will allow, I think, users to radically expand what Reddit is used for.
And so communities on Reddit today are already conversations about any topic imaginable, but some of the communities are almost like marketplaces, some are like games, some are for like storytelling. Our users are constantly pushing the boundary of what Reddit can be used for, which is really interesting, 'cause fundamentally, it's, it's like forum software, and so they're constantly trying to stretch Reddit and use it for interesting things. With the Developer Platform , they'll have now the ability to do that in ways that we never imagined. So the stuff that's online today is really simple. It's like, like scoreboards and stock tickers.
But, you know, beyond just a scoreboard, for example, in the Super Bowl, there was a special, like, Taylor Swift scoreboard that tracked not only the score, but also, like, Taylor Swift appearances, and had all these other little Easter egg-y features in there. Today we have a couple hundred developers using it in testing, but over the course of this year, we're gonna go through our wait list, which is thousands of developers. So we'll get that completely open, get this completely out in the wild, and then that starts to unlock not just the kind of the fun things our users will do with it, but then the economics.
And so, think of users being able to pay for these experiences or subscribe to these experiences, or have experiences that are generally involved in the exchange of money, and so that unlocks. It's a big part of what we call the user economy, which is, long story short, users making money from other users on Reddit, with access to Developer Platform apps or posts, but also subscribing to exclusive content or exclusive areas as well. And so there's a big family of work here that we've been working on for a while, and it's starting to come out over the course of this year. So I think a lot of interesting things to come there.
Okay, great. We're gonna wrap up with a quick word association. Rapid fire.
Okay.
Whatever comes to mind.
Okay, let's do it.
Uh, DAUs.
Up and to the right.
Advertising.
Up and to the right.
Google.
Up and to the right.
Hoping there's a different answer here somewhere. Moderators.
I go back to Google. I actually, Google, I'm gonna say innovation. I think we're in what I hope is a new golden era of competition in that part of the internet, which I think will be really interesting for everybody.
Good. Moderators.
Um, scaling.
Gen AI.
Spam.
Search.
Innovation.
OpenAI.
Search.
X.
Who cares?
Developer platform.
Uh, expansion.
And community.
Uh, underappreciated.
All right, cool. We'll leave it there.
Okay, thanks, Doug.
Thank you, Steve.
Thanks, folks.