CEO & Co-Founder, Reddit
IST Startup Week 2016 - Steve Huffman - Co-Founder, Reddit and Hipmunk
Steve Huffman is co-founder of reddit.com, a social news site and one of the largest communities online. Steve created Reddit with his college roommate, Alexis Ohanian. Reddit was initially funded by Y-Combinator in 2005 and has grown to over 5 billion page-views and 100 million visitors per month. In October of 2006, Reddit was acquired by Conde Nast Publications.
In 2010, Huffman co-founded Hipmunk, a travel search site that aims to take the agony out of finding plane tickets and hotels online. Hipmunk has been named to Time magazine's Top 50 Websites and Top 10 Smart-Phone Apps. Steve was...
About Steve Huffman
During Reddit’s Q1 2025 earnings call, Huffman described cost discipline as the company’s “single most important management lever” and noted that Reddit’s authentic human content is a key value proposition that the company aims to preserve. He acknowledged that AI and smarter agents represent a new frontier in preventing abuse and manipulation on the platform, and suggested that Reddit’s historically permissive account creation will need to evolve. On the Q1 2026 earnings call, Huffman reported a seventh consecutive quarter of revenue growth over 60%, with an adjusted EBITDA margin of 40% and record cash flow over $300 million, and stated a goal of reaching 100 million daily US users. In a Fast Company interview, Huffman described Reddit’s content as “like oil for the modern internet,” serving as a foundational resource for training large language models. He said Reddit’s policy is that commercial use of its data requires commercial terms, while non-commercial use will be treated flexibly. Huffman also discussed Reddit Answers, an LLM-powered search feature that provides verbatim quotes from users for questions with multiple perspectives. In a separate interview, Huffman expressed views on economic policy, stating that the U.S. treats supply problems as pricing problems, citing rent control as an example that leads to fewer homes and higher rents. He also advocated for unconditional cash transfers, saying “just give people money and let them do what they will,” and commented that “we’re supposed to die” when asked about life extension.
Transcript (78 segments)
✨ AI-enhanced transcript with speaker attribution
S
Steve Huffman0:00
There were probably about a dozen companies doing effectively what we wanted to do, which was order food from your cell phone. The thing was, in 2005, iPhones didn't exist, apps didn't exist, restaurants weren't really on the internet. So the ecosystem didn't really exist yet for what we were trying to do, which was allow people to order food on their cell phones so they didn't have to wait in line at restaurants — basically to make carryout faster.
So we went to Harvard over our spring break to see Paul Graham give a talk called 'How to Start a Startup.' We saw that talk, afterwards we met Paul, we told him our idea. This was actually Alexis's doing — I was basically the engineer and Alexis was the pretty face, the charming guy. Except he didn't have a pretty face back then, he had this like ratty, scraggly beard. But now he's a pretty guy, he's dating Serena Williams, he's doing okay.
Anyway, we met Paul, we told him our idea. Y Combinator didn't exist yet either. And in fact, that weekend for Paul was the impetus for Y Combinator, because he met a bunch of people, a bunch of kids like me basically, who wanted to start companies, wanted him to invest. And he told them all no, and then he felt guilty about it. So they announced YC. Alexis encouraged me — this is lesson number one, this is a basic one but here we go — he encouraged me after we met Paul, after we had gone back home to Virginia, to write an email to Paul thanking him for his time, basically. Which seems like a no-brainer, but I probably wasn't going to do it. I was really shy at the time, I just didn't want to bother him.
And look was like, just email him and say thanks. Paul responded, he's like, 'Oh, are you those guys from Virginia? I just announced this thing called Y Combinator, you two should apply.' And so we did. And we actually thought we had really good odds of getting in, because we had already told Paul our idea, we talked to him for an hour about it. So we headed back to Boston for YC interviews and were ultimately rejected, which really pissed us off, because I just said we talked to him for an hour already, so he could have saved us a trip.
On our way back to Virginia on the train, Paul called us and said, 'If you want to work on something else, come back. We like you guys, we just don't like that idea. We'll figure out what it is and you can work on that.' And so we did. We had a brainstorming session, and that's ultimately what led to Reddit. I still have my notebook, it's in my desk at work, with the notes I took during that brainstorming session. And in there I wrote 'front page of the internet.' And so that's the idea we had going in.
We were basically going to build a combination of Slashdot, which was my favorite website at the time — it was slashdot, 'news for nerds, stuff that matters.' It was one of the preeminent communities online. I don't know if any of you even remember Slashdot, it's still around. It sold for a billion dollars at one point. It was the best, it had the best users, the best content. We were basically going to try to recreate that community with more modern mechanics.
But our motivation at the time was basically: I don't want to look stupid in front of Paul Graham. It wasn't to actually build the front page of the internet, it wasn't to build a big company, it wasn't to change the world, it wasn't even really to make money. It was, well, here we are, we lucked into the situation, let's make the most of it — or at least let's not look stupid. And that's lesson number two. I'm going to stop counting them once I at least run out of the numbers. But lesson number two is basically: think bigger. That alone — like not having a plan, not having a vision — that should have killed us. But we're very fortunate that it did not. And this pattern will come up a lot over the years for us, that there's lots of things we did wrong that should have killed us, but we're very fortunate that it did not.
So we're working on the site with no real plan, no real aspiration, no real design, no product plan — just kind of going with the flow. You know, if you were building the front page of the internet, what would it do? That was kind of our mentality. And so we had the shell of the site. I remember a very clear moment — so we started actually working around June 15th or so, about two weeks in, Paul sent me this email. Now we were Paul's kind of pet project, because of the way we got into YC, because he basically brought us back and helped us come up with this idea.
He sent me this email that said, 'Why haven't you guys launched yet?' We'd been working on it for two weeks. He's like, 'Why haven't you launched yet? It's either because you're waiting until it's perfect, or you're incapable of doing it. And I don't know which is worse.' And you know, I was a 21-year-old kid, this was my idol in many respects, my feelings were hurt to say the least. Remember, my only motivation was don't look stupid in front of Paul Graham.
So I responded to him in this actually kind of weak manner. I was like, 'Well, here's all the things I'm working on.' I was treating him like my boss. Actually, there's another lesson in here, which is: have enough self-respect that Paul's not our boss. I don't owe him anything. He's an investor. Investors aren't your friends, investors aren't your boss. And I was behaving like, 'Here's what I'm doing.' No, I'm actually competent, you know, trust me. I was actually incompetent, probably. But nevertheless, I rushed and rushed and rushed, and we put something online on June 22nd. In fact, we didn't actually put it online — I emailed Paul and said, 'Here it is, it's working, see?' And it was just a link to this website that was running off a machine in our apartment, I believe.
And then Paul wrote an essay — his essays were fairly popular at the time. He wrote an essay about startups or something and he linked to Reddit. And so that's how we launched. I was sitting in our apartment, and I had my screen for my computer where I did my development, and I had a second screen in my field of vision that basically had the logs, the error logs, and it just started scrolling really fast. I was like, 'What the f*** is going on?' And I'm looking at all these errors. And then I dig in, dig in, and find out we're getting a ton of traffic. I'm just like, 'Where? How? What happened?' And I later learned it was because Paul just launched for us.
The website was like super shitty. I mean, if you think it's bad now, imagine how it was with three weeks of work with people who had no idea what they were doing. But that's how we launched. And from that moment on, we continued to just try to make it look like it was working. So Alexis and I had like a few hundred accounts between the two of us that we would submit content as on the site to make it look like it was working. Because we didn't want the front page to look blank, and we didn't want it to look like all 'spez' — which is my username — and nothing, which was his. So we had all these fake usernames that we used, and we had this whole streamlined process for scraping other websites and submitting the content as fake usernames to make that process more smooth for us.
And so we did that for a few months, and occasionally you'd see real users come in. I'd recognize my mother's username, I'd recognize my friends' usernames. But occasionally you'd see a user and be like, 'Oh, who's that?' And that was actually kind of a funny experience for us, like, 'Oh, is that a real person? I guess, neat.'
I remember taking the day off — this is one of the turning points in the company, a really important moment for us. So in August, we'd been online about two months. In August, I took a day off. I don't remember why, I was just tired, burned out, I didn't feel like working. I whatever, I just didn't work. And we had this problem on Reddit where if you didn't submit content, the front page would dry up, it would turn blank, because we only allowed links less than 24 hours old on our front page. And so if we didn't have 25 submissions that day, the page would actually just slowly become blank as things expired.
And so anytime I took a few hours off, I would see this happen. And so after taking an entire day off and not thinking about Reddit, I came home that evening and I was like, 'F***, what am I doing? Reddit is probably blank right now. This is a huge problem.' And so I frantically opened my laptop, checked it out, and it was full of content — front page was full. And that was the moment when we realized that we had a real thing. These were real users.
And so that's when our motivation changed from 'let's not look dumb' — the fear of failure motivator — to, 'Okay, now we owe it to these people, however few there are, really probably a few hundred, maybe thousands, not a ton, to keep this thing online.' Because somebody's getting value out of this. And that was a really, really special moment for us. And that was the kind of modus operandi for the next few years, really — just preserve it, just let it grow. You know, we added features, we changed the design here and there, but it's mostly let's just keep this thing working.
So we sold Reddit to Condé Nast in 2006, after about 18 months. We sold at the time because we were really dysfunctional. There was four of us by this point — there was Alexis and I, there was Aaron Swartz, whom you may have heard of, he was in Y Combinator that first summer as well, he ended up joining us, we merged companies. And Chris Slowe, who was our first employee, who was also in YC that summer. So really three of those first eight YC companies turned into Reddit in some fashion or another.
The problem we had was Alexis — his mother was ill, his mother had cancer. He found out that first summer and she has since passed away, but that summer and the rest of that year, he was spending a lot of time with her and dealing with that. And so he was kind of gone — if not physically, definitely mentally, a lot. Aaron, him and I had a fantastically productive four months over the winter of that first year, but then he lost interest. We had kind of different visions — more to the point, now that I reflect on it, he had a vision and I didn't really. But his vision was different from anything I could come up with. And so I wouldn't say we had a falling out, but we just kind of stopped working together.
And then Chris Slowe, who was incredibly productive, was working on his PhD at the time. So he would spend all day working, and then he would go home at night — or he'd come home at night and work on Reddit basically while I slept. So he's kind of a superhero. He still works at Reddit, him and I have been working together for over 10 years now. But he would get up before I got up, go work on his PhD, come home, we'd have dinner together, I'd quit working, go to bed, and he'd be still up. So I don't know if he ever actually slept. Now he's got two babies, so he definitely doesn't ever actually sleep.
So that left me at the time, during that first year, as the only engineer working. And now we had to start to confront issues like spam and scaling, and just lots of — every day it was like an emergency, what do we need to take care of today just to keep the site online. And so I was burning out, I was just running out of energy. So when Condé was interested, we were interested. And we actually felt like — our mentality during that negotiation was that we were getting away with something. Like, somebody's gonna pay us for this piece of s***. We're completely dysfunctional, we're pretending, like we put on a happy face when we talked to them, but we didn't know what we were doing, we didn't know why we were doing it. I was completely burned out, and they were going to pay us. So we're like, yeah, that's a good deal, we'll take it.
And we felt like we were giving them a bad deal. We may find out that was a pretty good deal for them. We'll see. So we ended up selling to Condé, we moved to San Francisco. We worked in the Wired office — Wired.com was based in San Francisco. Condé is a big publisher, they have Vanity Fair, Condé Nast Traveler, Vogue — like magazines. And they own some cable companies, they own a few TV stations. But they're basically a big media company, almost entirely in New York City, except for Reddit and Wired, which they kept in San Francisco.
So I moved to San Francisco at the same time as a bunch of these other dudes that you're going to hear talk. Very fortunately, we all kind of lived in the same space. And at the time it was funny, because everybody was in the early days of their company, and I was the only one who had sold a company. So for a time there, I was like this little s*** who had some money, didn't work too hard, just kind of hung around while everybody else was like busting their ass on their startup. So those were good days, those were fun days.
But I worked for Condé on Reddit for the next three years. And you know, Reddit continued to grow. We developed all sorts of bad habits. Here comes another lesson: Reddit just grew every month. It just kind of grew, whether it was fast or slow, or downtime, or adding new features or not adding features, it just kind of grew. And so we had these inside jokes — you know, we would be down for a few hours and we'd turn the site back on and traffic would jump by 10%. And so we'd just be like, 'Man, maybe we should have downtime more often, like maybe that's why we're growing.' Or, 'Maybe if we don't add features, maybe that's why we're growing.'
It was kind of like we knew we didn't know what we were doing, but then we also felt like maybe we were doing something right. And I've seen a lot of companies go through this — they're growing, many of them sometimes quite quickly, and they think they're hot s***. Founders think they're hot s***, the whole team thinks they know what they're doing, like they're king of the world, everybody else is dumb, but we've got it figured out. And that may last even a couple of years. And if those companies don't realize that they don't know what they're doing, they enjoy that nice ride and then they crater. This happens all the time.
MySpace is probably the best example. I don't know if any of you even remember MySpace, but there was a time when MySpace was king of the world — they were the social network, they were the company everybody was chasing. They sold for like $500 million, which everybody thought was a lot of money at the time. I mean, it's a lot of money, but Facebook's worth 200 and some billion dollars, so that's the discrepancy. But then they didn't know what they were doing and they just cratered.
And you look at a company like Facebook that also went through that growth, and right place, right time, made better decisions than MySpace, but still didn't really know what they were doing. They hit a few plateaus over the years, but they figured their s*** out. They got their s*** together, they became good managers, they hired the right people, they actually had a plan, and they've powered through it. And so I'm always — whenever I'm talking to young people starting their companies, especially if those companies are going well, I try to remind them, ask them: are you going to be MySpace or are you going to be Facebook? And it takes some amount of humility to realize that you don't know what you're doing.
And at Reddit, we did not know what we were doing. And so I have been very fortunate, Reddit has been very fortunate, in that we still have time. Reddit still grows, and we've got time to clean things up. But Reddit is a 10-year-old company, which is very rare for a 10-year-old company to still be in this position. But that's something I'm very, very thankful for. And if you find yourself in that position on your first company and everything's going right, it's not you — trust me, it's not you. And even if it is, remind yourself: it's not me. It's, congratulations, you started an internet company and the internet is growing. You start a company in a growth market, it's going to grow, that's nice. But it's not going to grow forever. So I'm going to pause there and see, do we have any questions right now at this point?
A
Audience Member18:22
Yeah. So I have two questions. One is, like you said that your first company wasn't the right time to really explore that, and I was wondering what would have — I mean, hindsight is 20/20 now — but what would have been key indicators that it wasn't the right time?
S
Steve Huffman18:40
Well, I think, I mean, I kind of explained it — we were swimming upstream. There's all these problems we would have to solve. So the idea, the very specific idea, is ordering food from your cell phone. Well, we would have to explain to restaurants like what an app was, but apps didn't exist. So we'd have to explain — so you have to go even further — we'd have to explain to consumers that yes, you can use your cell phone for things other than making calls and texting people, which are basically the only two use cases a phone had at the time. You had these kind of crappy mobile websites, but that was it. So that's one kind of element of the current going against us. We'd have to get restaurants online and explain that people can put in orders through things other than fax machines. So that's another one. Nothing was playing in our favor.
And so you want to have as little friction as possible, right? Like especially as first-time entrepreneurs. In hindsight, we didn't know — we were doomed to failure, and I'm glad Paul and YC recognized that. But we would have had to be doing too many things all at once. So, you know, just give yourself a chance to succeed. Don't start in a market that doesn't exist where you have to create both sides of the equation. None of that stuff existed yet. Now, you know, everybody knows what apps are, there's distribution mechanisms, everybody has phones in their pocket, everybody's online, restaurants and stores are very tuned to different channels of customers. People are thinking in a different way. So it's way easier now. And that's why those companies now — Eat24, Order Ahead, OpenTable, all that genre of company — they're doing pretty well. But it wouldn't have worked 10 years ago.
A
Audience Member20:37
Okay, thank you. And my second question is: how do you know your idea is bad? Like, I think Paul told you your idea was bad, but then if he didn't, how would you know it's bad?
S
Steve Huffman20:50
Well, here's the thing: all ideas are bad. Just start with that premise — all ideas are bad. Because if it's such a good idea, somebody else would have done it and succeeded by now, right? Now, it's not that easy. The thing is, good founders can take a bad idea really far and make something of it, and then all of a sudden it doesn't seem so bad anymore. And they can take a company that shouldn't exist just through like raw horsepower — good technology, good marketing, good processes internally. And bad founders can take the best idea — MySpace — and ruin it. I shouldn't call them bad founders because they sold their company and left and, you know, Facebook killed it, whatever. But there are many examples of companies that fail, two companies start at the same time and one fails and one succeeds. The difference is the founders.
So I wouldn't overthink is this a good idea, is this a bad idea. The way I try to frame it as an entrepreneur is: is this an idea that I can run with? Do I have an intuition about the market? Do I understand the customers? Do I have the ability to build it myself? Try to stack up as many unfair advantages as possible, because that's what life is all about, right? At least that's what success in business is all about — accumulating advantages and using them. So who do you know, what insight do you have, what are you good at? Play to your strengths, don't force the issue.
A
Audience Member22:20
Thank you. So maybe in this vein, some people say it's all in the idea, others say it's all in the execution. Where do you stand there? And are the same people good at ideas also good at executing?
S
Steve Huffman22:39
I think you'll find that people who are good at — both of them are learned skills. You can learn to be good at executing. That's why I'd say a great first move, if you're not going to start your own company, is to join another company that's doing well, or that might do well, or people you like, or people you admire, or whatever. It could be a shitty company — either way, you're going to learn a lot of lessons, and that's how you kind of learn the execution. It's just by being around it, trying and failing, trying and failing, you'll learn all of those lessons.
And then you'll become more savvy about ideas. A good idea is really just the process of: am I in a good market? Is it a big space? Can this market support a large company? I think that's the mistake a lot of first-time entrepreneurs make. If I were to ask — you all being college kids — what company do you want to start, I bet a third of this room wants to make a company that makes it easy to exchange used textbooks, and probably wants to start a company that makes it easy to find what's going on in town tonight, and probably wants to make a company that helps me discover new music. Because those are the problems you have as a college kid. Probably some dating too — how do I find girls, right? Which is similar to the 'what's going on tonight' problem. Like those are your ideas. I'm certain of it, because I've talked to thousands of college kids and they're always pitching me those ideas.
And so it does require a little bit more experience and savvy to think like, okay, what is an actually good idea? What has a big market? What do I have a chance at succeeding at? And what do people want? But I tell you, those ideas all suck. You know why they suck? Because every single person has those ideas and fails at them. Anyway, not every single person, but the odds are not in your favor. So experience and time, I think, is more important. But if you have an intuition, if you are the customer, it's a lot easier. That's where I would start with if you're a first-time entrepreneur — try to find a situation where you are the customer. Just don't do one of those things.
Okay, so I'm going to continue on with my story for another minute or two. 2009 was the end of my contract at Condé Nast. I left because I wanted to do something new. I'd been working on Reddit at that point for about five years. There were only six of us. The site at this point probably had around, call it, 30 million monthly users. So it was good, it was nice, we're growing, we're always growing. This is kind of one of the bad habits — I took that for granted. I mentioned that before, I didn't even realize it. It just didn't feel like a success to me because we had always — that's all we ever knew was growth. It was never fast growth, it was just kind of slow and steady growth.
Meanwhile, companies like YouTube, which were started after Reddit and sold before we sold — in my mind, that's what success was. So Reddit didn't feel like success, it just felt like a nice thing. So I left in 2009 because I wanted to learn. I felt like I was too young to be the best engineer at this company, which is true. I wanted somebody to look up to. I wanted to do a business that was involved in the exchange of money, because on Reddit I felt like we were always trying to contrive business models and it just felt like we were swimming upstream. So I left.
I started another company called Hipmunk, with this fellow named Adam Goldstein. It's in the travel space, so we help people find flights and hotels. Hipmunk was one of the best experiences I've ever had. Travel is a shitty space. It is tough. It is very, very competitive. It is very cutthroat. There's basically two players in travel — Priceline and Expedia. Every other travel company I've heard of in the United States is owned by one of those two companies, except for Hipmunk. And so you're in this situation where you're just kind of stuck in a duopoly, and those companies are very, very wealthy companies.
So I had the experience — and I would suggest if you can put yourself in this position — we didn't get it at Reddit — of working at a company that every day has to consider its own mortality. I learned a lot of really good lessons there. Like I said before, Reddit always grew. Hipmunk, its natural state was death. Hipmunk would prefer to have zero users, and if you turn your back on it for one...
Hipmunk taught us really important lessons. We had to get savvy about who our customers were, how to reach them, what experience they'd have, and how to optimize every revenue channel. Hipmunk was known for having a really nice product—very pretty, easy to use—which contrasts nicely with Reddit, which is perhaps neither of those things. Same people, but we were forced to actually care about those things in a way we didn't before.
Over five years at Hipmunk, we raised about $50 million, hired about 80 people, sold a lot of plane tickets and hotels. I was the CTO—that's where I finally learned how to actually manage people. Reddit was only six people when I left. I got to hire a lot of people, unfortunately had to fire people, build an executive team, learn how to operate a company with discipline.
After about four years, I started to get that itch again. We had hired really well to the point where I didn't have much of a job left. Our VP of engineering and VP of product had basically taken my job from me, in a good way. I'd come to work, mentor younger engineers, do code reviews—but I wasn't learning anymore. Meanwhile, Reddit starts to go through turmoil. The CEO, Ellen, left. Reddit CEO is a tough gig. She'd been there a couple of years, joined when Reddit spun out of Condé Nast around 2012. Reddit became an independent company again, with Condé Nast being the initial investor. Condé wisely realized Reddit was suffocating inside it.
I'm seeing Reddit there, independent, doing okay, but I still got Reddit on the brain. Coming back to Reddit was something of a dream job for me. I'd do events like this and talk about the Reddit story, but I wasn't actually at Reddit anymore. I always felt a little silly answering for a company I didn't have any control over anymore. Reddit would occasionally check in: 'Hey Steve, do you want to come back?' I'm like, nope—well yes, but I can't. I'm at Hipmunk. I made promises to these people. I hired these people. We said we're in it together, we're going to take on the travel industry, we're going to succeed come hell or high water. I loved those people. Hipmunk was very cultish too. The thought of leaving was so painful. I would only consider it for about five seconds at a time. No, not leaving. That goes on for almost a year.
Until the summer of 2015—about eight months ago—Reddit, for lack of a better description, was actively committing suicide. The community was in open revolt, really mad at the company. The moderators were mad at us, the community was mad at the moderators, they both hated the company. It boiled over to the point where the moderators basically shut down the site—the largest communities were so angry they shut the site down for about three days. When the site went down on July 2nd—called the blackout—every single person I knew texted me and said, 'You should go back to Reddit.' And I was like, 'Who the fuck wants to do that?' But I was in really close contact with Alexis, my co-founder, who had returned to Reddit, and with a lot of the Reddit investors, trying to give advice on what to do.
I took a trip to Sonoma, an hour from San Francisco—wine country, nice bougie area. Spending time with friends at a birthday party, and this is weighing very heavily on me. Half of them are like, 'You should do this,' and the other half are like, 'Why would you run into a burning building?' I remember this moment very clearly. It was Saturday morning, July 4th. I was laying in bed thinking about Reddit's options, playing through all the different scenarios of things they could do, changes they could make. And I had this realization: they can't do it. There is no path out. I'd seen this before when our biggest competitor, Digg, committed suicide around 2010. Digg was always bigger than us in the social news space. They fucked up—they did this redesign, pissed off their users, and the users all came to Reddit. Literally in one day, Reddit's traffic doubled. I hated Digg so bad. That company had no soul. So I was really pleased at their demise—it was nothing we did right, they just fucked up.
That morning I was like, 'Reddit can't do it.' So I started trying to get a hold of Sam Altman. He was the president of YC. He had led Reddit's raise of a $50 million round in 2014 personally. I was trying to tell him, 'Yes, I'll do this,' but I couldn't because I was in Sonoma and I'm a T-Mobile user and my phone didn't work. There's two places on the planet where T-Mobile doesn't work: Sonoma and State College, Pennsylvania.
We're at a winery getting a tasting for my buddy's birthday. My phone rings and it's Sam. I go outside to answer. Sam is the most annoying person to negotiate against—he's just relentless, very persuasive, he will use every angle of attack. I answer the phone, I could literally hear him inhale, and he's ready to make his last plea: 'Will you please do this?' And before he could say anything else, I'm like, 'I'll do it.' And he was just like, 'Oh, thank God.' That's when we started game-planning. Between that phone call, I thought I might go back as early as Monday. I hadn't told Hipmunk yet. Reddit didn't know. The only people who knew at this point were me and Sam.
So I ended up not going back until Friday. Telling Hipmunk that I was leaving was easily the most difficult thing I've ever done. Not a dry eye in the room. Hipmunk was very much a cult that I had built, and I just felt like I was letting them down. But they were incredibly supportive. It was simultaneously one of the hardest moments of my life, but that affection was very powerful.
I go to Reddit Friday afternoon. It takes a while for them to negotiate things with Ellen. Ellen and I have always gotten along—it was an awkward situation, but there's no animosity between us. We talked to the Press together, which was really tough—I don't know how she did that. Then we addressed the company together. She read a statement saying she was leaving, and then she left the room. So I'm sitting in front of the company—it's about 75, 80 people—staring at me. It was almost the same size as Hipmunk, the same size room almost. But instead of a wave of affection, it was a wave of anger. These people had been through so much. I was the third CEO in a year. The company was getting ridiculed in the press, was potentially actually dying. These people were just tired and angry and confused. That was tough. I usually feed off that energy of people looking at me, and it just wasn't there. So I'm like, 'Hey, I'm here. We're going to fix this thing. I'm going to make the world proud of Reddit. I'm going to make you guys proud to be at Reddit. I'm in it. We're going to do this together.' And it went okay.
Everybody goes back to their desks. I'm on my laptop doing an AMA with the community, wheeling around the office introducing myself to people because I didn't know any of them—all the old-time Reddit folks were at Hipmunk. The first guy I meet, I'm like, 'Hey, I'm Steve.' And he looks at me and he's like—goes back to his computer. I was like, 'Oh fuck. I should have just fired him on the spot.' I fired him two weeks later.
It was tough. I wish I had kept a journal for those first few months because it's such a blur now. Total chaos—external chaos, internal chaos. I did keep notes though. I had one-on-ones with every person in the company over the next couple weeks. I'd ask two questions: first, 'Who's good here? Who should I be talking to? Who can I rely on? Who are the best people?' And then, 'Who's not so good here?' I'd write those down. I thought I'd build up this list of bad people and fire them. But the problem was the lists overlapped. Different people would tell me different things, and it was really confusing. In hindsight, now that I know everybody and I've been there a while, it all makes sense—what everybody's agendas were. But this was the first time I'd been in a company where people had agendas. It was very surreal. I wear my emotions on my sleeve, pretty straight shooter, I'm not a great liar.
But we made it through. We've done a lot of hiring. The culture at the company is so much better than it was before. The morale is much, much higher. It couldn't not be higher because the morale at Reddit when I arrived could not get as low as it was at most companies—it's not possible because people would just quit. But people didn't quit Reddit because they loved Reddit, the idea, the thing. They put up with all this internally just to be a part of it. But they were really, really unhappy. Now we're out of that period. It's been a slog, but we've made a lot of changes. Now we're getting back to a state where I think we're actually getting productive and we can do new things. We know our place in the world. The mistake I made in the early days of not knowing our mission—that's not a problem for me anymore. It's not a problem for us anymore.
Reddit fulfills this very basic human need of connection, of allowing people to share themselves—to be their real selves. Not their real name human carcass on a stage self, but my interests in video games, my desire for relationship advice, my interest in sports—all these other facets of my personality. You can do that on Reddit.
We've got 240 million users right now, and we're chasing a billion. I can see a very clear path to get there because we haven't internationalized—we're almost entirely English-speaking, so we have this massive opportunity. I feel so very fortunate to have this chance a second time. Not many people get one chance, and I've had two—to be part of something that can actually change the world, that can actually make a difference. We save lives, we provide entertainment, we provide people a place where they can be their true selves. People come out on Reddit, get relationship advice, participate in hobbies, connect with other people. They exchange kidneys. There are all sorts of wonderful things that happen on Reddit, big and small. To do that on an even larger scale would be really incredible, and that's what we're chasing.
M
Moderator44:11
No questions? Yeah, I thought so.
A
Audience Member44:16
Hi, sorry. What did you do to fix the culture at Reddit specifically?
S
Steve Huffman44:24
Okay, so culture—a funny thing. We started by defining it. What are our values? I used to think missions and values were kind of these bullshit MBA business things. For context, I've been an engineer my entire career until this CEO job. But those things aren't bullshit—they're tools. They're tools for making decisions, whether it's a product decision like what should we be building, or tools for making hiring decisions: does this person uphold our values? Are they going to make decisions the way we want them to?
First we started defining the values. Let's see if I can remember them. The first was Evolve—Reddit's got a long history, a lot of baggage, good and bad. A lot of product decisions 10 years old, you can't not have baggage. But we got to look towards the future. The second is Remember the Human—I've never liked that wording, but it's really important. There are people on the other side of the computers. It's our duty to make sure those people are treated like people, are not harmed on Reddit, they feel safe, they don't fear for their personal well-being, they have a good experience. The third is Listen, Debate, Ship. At Reddit it's really easy to get in these conversations about what the community does, maybe we should do this, well this is how we've always done things, we might make the community upset. You can have those debates, but when it's over, when we leave the room, get to work and ship it, even if you disagree.
Finally: What Would Snoo Do? That's my favorite value. Snoo is our mascot—it's kind of a grab-bag value, but it's actually worked out really well because everybody who works at Reddit has an intuitive sense of what our mascot is. Snoo is an optimist. I was interviewing a guy, and part of my job as CEO is to repeat these things over and over to get people to buy in. We were talking about values and he asked what's important to you. I said, 'What would Snoo do?' And he was like, 'God, you're the third person to say that to me today.' And I was like, 'Fuck yeah, it's working.'
So you create those values and hold people to them, and you hire people who believe in those things. That's how you change culture. You reward the behaviors that are good and you don't accept the behaviors that are bad. It takes time. There were definitely times a couple months in where I'm just like, 'I miss Hipmunk. I miss that culture, I miss those people, I miss that feeling at work, and I don't know if I'm going to have that again.' But now, eight months in, I'm really proud to say that yeah, I have that again and it's really great.
I remember some of those moments. The first time we had a party—there was no drinking at Reddit when I arrived. It was a dry company, had incidents in the past. I was like, 'I can't work at a company where I can't party with these people.' When I proposed a party, people were really nervous: 'You can't do that, bad things will happen.' I'm like, 'Look, we're adults. We have to be able to treat each other like adults.' We had that first party and it was fine. People came out, had a good time, went home. I remember being really happy. People came up like, 'I didn't even know Reddit was going to be able to do that.' It seems like such a little thing, but little moments like that where people realize, 'Oh, we can do this together. We can change the culture.'
So a very long answer, and it's a tricky thing, but I think it starts with just knowing what your culture is and holding people to it.
A
Audience Member48:55
Thank you so much. Thanks for the question.
M
Moderator49:01
Hey Steve, we've got about four minutes left, so more questions.
A
Audience Member49:06
Quick question. Thinking back to the early days, a lot of places got caught and taken over by trolls, and Reddit doesn't really have that very much.
S
Steve Huffman49:20
Why do you say Reddit doesn't have trolls?
A
Audience Member49:23
You know what, I think they get modded down, and it seems like a really quite friendly community, compared to what it was elsewhere.
S
Steve Huffman49:35
Okay. So I will answer your question honestly, but I'm still amused at the premise—as somebody who has a hundred accounts and fucks with people all the time. But we have done a good job in certain areas. In hindsight, we made a couple of really good decisions. The way our comments work, the sorting algorithm, the voting, and the hiding of low-scoring stuff—all that stuff is public, it's all open source, but there's some subtleties in there that I think have played out really well.
I've always wondered when I look at our comments, which I think are the best on the internet—YouTube comments, or really comments anywhere else online, even Facebook, those guys are really smart—I just wonder: do we have better users or do we have better technology? When we were smaller, I thought we just had better users. We've just got the best users—these people are funny, they're smart, they're with it, and the rest of the internet is dumb. But now we've got 240 million users. There's not that many smart, funny people in the world. So I can't claim it's our users anymore because our users are the same users on the rest of the internet.
So there's, I think, a number of subtle things in the product and on the technology side that work. And then there's also this cultural inertia where people come to expect what's acceptable behavior on Reddit and what's not. Users enforce that, moderators enforce that. When there's a violation—which we see from time to time, where you'll get a big community of people behaving in a way that's incongruent with how people feel Reddit should be—the whole thing erupts. We had some of these big communities last year that were really bad, some of these really vile horrible users, the worst kind of trolls. We didn't have the tools and technology to deal with it, but the whole thing felt wrong, so we stepped in with a very blunt object and banned some of these things. There is a cultural inertia. It's not always accepted, and you can kind of feel when it's not right, and that's when we'll take extra steps. But I do think a part of it is just the simple technology of promoting the good stuff and hiding the bad.
M
Moderator52:21
Okay, do we have time for one more question? Let's do it. One more—make it good. Let's send on a high note.
A
Audience Member52:30
So you mentioned how one of the risks you have is that people could just wholesale go somewhere else. The question is: what do you do to monitor things to make sure you're not pissing people off? And are there things about your technology that make it sticky, so that even if you piss them off a little bit, you have time to find out?
S
Steve Huffman52:43
That's a good question. Why don't people leave, basically? I've got a couple of stories there. When I got back to Reddit, the community was pissed off. We had some of these bad communities—really racist ones—that we were trying to wrestle with. We're wrestling with this free speech idea, and users are just angry about a ton of stuff. I told the company: when users are complaining about the front page, that's how you know we're doing a good job, because users always complain about the front page and there's nothing else left to complain about. And sure enough, after about five months, users were complaining about the front page algorithm because we had made a tweak. In the company I was like, 'See, we're doing our job. The users are mad about the front page.'
To answer your question—I don't really monitor where people are leaving. People are always like, 'Oh, I'm going to go to Voat.' It's like, whatever, okay, do that. I think if you spend too much time thinking about your competitors, you're in trouble. We spend a lot of time just thinking about how do we make Reddit the best. The frustrating thing is we don't always have the time or the resources to actually execute on these ideas. But in our heart of hearts at Reddit, everybody that works at Reddit loves Reddit, loves being on the site, loves using our apps, loves using the product. We try to think: as long as we enjoy using it, we're probably doing something right.
There are ways of measuring sentiment. We do run polls with both our community and the moderators—it is actually something we measure. But it's not something we take action on, like, 'Oh, users are upset today, therefore we have to do something.' It's more just one of those numbers where it's so big and fuzzy that it's hard for any one thing to move that needle on a day-to-day basis. But it is something we're aware of, so we watch it. Then we just go back to work and think, 'Alright, what are the most important things we can do to make our community and site a little bit better today?' You just keep chipping away at it. We have actually seen sentiment rise over the past few months, and that's been nice to see. It's not perfect—I know we've got a long way to go, but...
M
Moderator55:13
Sweet, thank you, Steve. Let's give Steve a round of applause. Thank you.