Scaling Product | Fireside with Joe Gebbia and Reid Hoffman

It is my uh privilege and honor to be on stage with Joe, who um actually in fact um I have learned a bunch of different interesting uh product and design things from. Among other things, I haven't done this yet—Is your furniture stuff out yet or no? Next month? Next month? Okay, there’s some furniture stuff that's coming out that I recommend to you, and whatever degree we can talk about that, we'll get to.

Part of the reason I wanted to do this section with Joe was that uh kind of a key theory around scaling is actually, in fact, first doing things that don't scale across a set of different key uh places in the business and then working out to uh how to get that part to scale. Airbnb is like, you know, the canonical example of being super design intentional on this. And so let's start a little bit with kind of like how you guys got to thinking about, you know, going door to door in New York—what that experience was like, why you did it. Clearly not scalable having the founders go do door to door in New York or anywhere.

And then what's the design process that you run for beginning to think about everything from that initially super handcrafted into a kind of more scalable process?

Yeah, cool. Well, first of all, thanks for having me. It's great to share these stories um from like the early days of Airbnb, when it was like truly like, you know, you're at war trying to get product-market fit, trying to get people to adopt this crazy weird service that no one would invest in. Um, I think, you know, reflecting back on that time of like right before we tipped into scale, um, it's like how do you push so hard to get that, I guess, inflection point right?

Um, for us up until that point, I feel like we had subscribed to this um, this myth of Silicon Valley, which is that you have to solve problems scalably, right? Because that's like, that's the beauty of the internet; that's the beauty of code—one line of code can, you know, touch an infinite amount of people. And for many, for far too long, we tried to code our way through these problems that we were facing um, and we were getting nowhere. It was the quintessential Paul Graham mantra—like totally flat. I call it the Midwest of analytics—like there was just zero growth, um, and it was really depressing.

Um, it wasn't until we had this um, this session with Paul Graham where was the first time somebody said to us like, "It's okay to do things that don't scale. It's okay to step outside that myth and to try any and everything." And the minute that that happened, it was like the creativity just started to flow and we're like, “Okay, well what could we do?”

And it was looking at this pattern on our search results that hosts in New York, of which we had about 30 at the time, had really bad photos. Like they just didn't know how to take a good photo—there were pictures at night, pictures with dishes in the sink, pictures of toilet seats up. I mean you're looking at it, you're like, “I wouldn't want to stay there.” So no wonder we're not growing, no one's—why would somebody pay for this kind of accommodation based on these images?

And that's where, like we're in the session with Paul Graham; it's Nate, Brian, myself, and we're all just like huddled around. We're like, “Well, if we can do things that don't scale, why don't we just go take better photos?” I've done photography my whole life, took some classes at art school—like let's just go to New York for one weekend. And we have no idea what's going to happen. We'll email our community in advance and we'll say, “Hey, you know, we want to come meet you. We want to send a professional photographer to come shoot your place.”

Um, so we’d fly out; we rented this wide-angle lens camera from a place in Union Square; and you guys were the professional photographer. That was the aha moment for a lot of hosts—um, you show up at the door like, “Hey, I'm Joe, it's nice to meet you,” and they're looking over your shoulder like, “Where's the photographer?” And you go, “Oh, uh, that's awesome. Me too.” And then they put this look of disbelief on their face.

Um, but then something really interesting happened. We go in, we take photos, and we’d show them on the back of the camera and say, “Hey, what do you think?” And they were blown away. They're like, “Oh my God, my apartment looks so good! Do you want to stay for some tea or coffee?” And so there I'd sit down on the couch and our earliest customers, much like yours, were our zealots. These are the people who were willing to take a risk. Try a new weird crazy product, to be kind of outcast amongst their friends, um, which meant that they had a lot of knowledge, they had a lot of insights into this kind of activity because they were already doing it on other websites. They were on Craigslist, they were on sublet.com, they were on HomeAway, blah blah blah.

And we're sitting down with them; they pull out their laptop, and they start showing us and they start asking questions like, “Show me how you log into your account. Show me how you evaluate a guest if you feel like you can trust them enough to let them into your home. How do you update your calendar?” And all of a sudden, right in front of my very own eyes, I saw UX—user experience—train wrecks of people trying to navigate our beautifully designed product but not being able to do nearly what we intended them for to do.

It would take them 12 steps to do something we thought was only going to take two, for example. Um, people didn't know how to update the calendar; they were like, “We reviewed the wrong way.” All of these train wrecks that we just had no idea were even a problem, right? Because at this point, we're just in San Francisco trying to code our way through problems. We're shipping code around the world; we never took the time to go out into the world and actually meet the people using our service.

Do you think it's worth talking at all about the kind of experience—not just being the website but the entire experience around it—as part of that?

Sure. I mean, 'cause this is one of the things I learned from you guys. I think even going back to the very origins of the kind of that core weekend with the three guests in our living room. Um, I won't go through that story other than to say from the very beginning we thought about this more than just a place to put your head at night, a place to lay down. Even from that very one weekend, we're like, “Well, what if we picked them up from the airport? What if we provided breakfast in the morning? What if we gave them, you know, a Bart pass to get around the city with a map to San Francisco?”

And even from the beginnings of this weird weekend project, was this idea of like how do you think about everybody's experience from end to end? And that’s, I mean, often like a designer's job. Like this water bottle is a great example of that. A lot of people might look at this like, “Oh cool, design of the water bottle.” A designer would look at this and say like, “Actually, this is just one instance of a much bigger design.” The design of this started with thinking about the, you know, the shelf experience inside the store or maybe even further than that—the first time you heard about the Fiji brand.

And you think about each sequential step that somebody goes to, and this just happens to be an instance on a much larger holistic kind of arc of how people experience a product. So it's considering all those things—what you call it—design thinking. That's all that is; it's just a consideration for each of the touch points that leads up to the moment when someone consumes your product, including what happens to this afterwards, right? Like a good designer would consider, well, the materials in this and the labeling to show what you can do with it, which bin it goes in, so that it has like a life cycle beyond its consumption—would be like how we think about design at the company.

And then how do you move from those initial steps? Because obviously, the reason why Silicon Valley has this kind of classic only do things at scale is because you're worried about getting trapped in like, “Oh, the founders have to go take all the pictures” or “We have to hire all the photographers on to do this,” and you get worried about, “Well that'll never scale and so that'll never work,” and so you shouldn't even start it. And they get the kind of learnings about really talking to your customers, really getting a sense of it, helping build that initial kind of like just like love experience—yes, but that kind of thing.

And so how do you go through the process where you then, because you have another design experience, you say, “Okay, we figured that out; now we're designing the scale”?

Yeah, well, you know, there's something that magical that happened during that first weekend. When we're taking these photos, we're seeing our customers use our product in their natural habitat, right? We're in their living rooms. They call this ethnographic research, right? You go out into the world and you study something in its natural environment. And, uh, it was through those interactions that we saw the distance—the gap—between our product and our market.

It was this invisible gap that we couldn't see, and by going and talking to our earliest, most passionate customers, they helped shed a light on the distance between those two gears of a product and market coming together. And so, a couple of things happened. We take the pictures, we’d come home and upload them. We’d also take their feedback and actually implement changes very, very quickly, right? So Nate would, thank God, he's a genius—he would take these ideas that would come back from guests and hosts and start implementing them like the same night.

So imagine you're like this customer of this website and you get this email that says, “Hey, it was great to meet you. Um, here's the photos of your place— we uploaded them for you. And by the way, the idea you have for the calendar—we've implemented it! Tell us what you think.” And these 30 hosts that we met over the course of a couple weekends doing this started to use the word Airbnb and love in the same sentence.

And something magical started to happen. A couple things. The first is that revenue doubled—from all of $200 a week to $400, which was huge. Um, but we started to see more listings popping up in New York. And this is at a time we had no press, we had no online marketing; there was no possible way that somebody heard about it from us. This was like the most pure word of mouth we probably have ever had.

Um, because I feel like when people fall in love with something, you tend to tell others in your life about it—your neighbors, your colleagues, your co-workers, your family, your friends. And so we started to see that 30 listings become 42, become 57, become 65. And what was really cool is that these new hosts coming on board would see the quality of listings and be like, “Wow, I guess I need really good photos.” So they would upload really good photos.

So we had quantity and quality going up at the same time. And when that happened, people started booking. They're like, “Oh, I see what I'm paying for,” and they have the locations that I'm looking to stay in. This idea of like traveling around the world to take photos of apartments, obviously, is like that can't last forever.

Um, and so it was always our intention to really, uh, understand what that operation was from the inside out so that from any dimension, we would understand the needs of, like the timing—like how long does it take to shoot an apartment? How long does it take to get between two apartments in the city? Uh, what kind of system do you need to upload photos quick? Like all these different things that you have to figure out, we did ourselves.

So eventually, over time, like we threw product and engineering at it, and today we have, God, like tens of thousands of photographers all over the world—contract photographers that shoot in every major city around the world. And it's a free service that any host can click, and a photographer shows up in about two days to take free photos of your place. And that all started with just us like bumping into walls and stumbling through it and trying to figure it out—the professional photographers showing up, “Hey, we're the three co-founders. We're going to talk to you.”

Um, and actually, and that was another lens of the kind of the photography service—was another lens of saying, “Look, yes, we're investing in this heavy cost human touch way of doing it, but it leads to a compounding effect within the marketplace and the network.” And that's another angle of the unscalable things that then later lead to scale as a way of doing it.

On your um TED Talk, you talked about designing for trust, right? And I actually think that's one of the things that I think all the companies is super important for because part of encountering a new company, a new experience, going and staying at someone's home, etc. But all of us have this experience. What are some of the takeaways you think that the, uh, the kind of founders, CEOs, and so forth when they're saying, “Look, I am trying to provide this new experience.” What are some of the principles, either as stories of Airbnb or general principles of designing for trust?

Wow, yeah. I mean, uh, I think people probably know the early stories if we tried to raise money; it didn't work out. Um, and we were just like scratching our heads like, “Why is this not working? Why will people not see what we can see?” Um, that this is like, there's a much more authentic way to travel. And I think what we ran up against—looking back, I know what we ran up against—was the age-old Stranger Danger bias that we've all been taught since we were kids. We just didn't realize it at the time.

And like we had to figure out some way to navigate this bias in the world that—well, of course, like what investor would invest in a company whose business is about having strangers sleep in each other's homes? Like that's crazy, right? Back in the day. And so we really had to get smart and clever and try a ton of different things to figure out how do you create a condition where people feel comfortable saying yes?

Right? Like what information do you need to provide? Um, I mean, we had to create, you could say, Olympic-sized trust between people. And it was uh, for us, you know, I think when you face a challenge or you face a problem, you fall back on what you know. And we've, we fell back on design because that's what we know.

And again, when you think about design, it's about the whole experience—not just the look and feel of something. And so for us, it really begged that question: can you design for trust? Can you actually create a condition where people feel comfortable saying yes to something like this? And we weren't actually sure if you could.

I think through a ton of iteration, there's some classic stories of um, the amount of disclosure that somebody provides when you're uh, exchanging a message with a potential host has—there's a lot of leading indicators to tell how much that host might accept your reservation request into your home. For example, if you send a message that just says “Yo,” the acceptance rates drop. If you had a message on the other side that was way too long—like “I'm having issues with my mother,” and uh, like all these things—like acceptance rates also drop. So there's this sweet spot in the middle that's like, “Hey, I'm coming for vacation with my family. Love the art on your walls,” etc., where there's like the right amount of disclosure so people feel comfortable saying yes.

And something that, like once you understand that, you can look at the data and look at the design, you can actually create, you know, guidelines to help encourage people to enter that kind of middle part of the bell curve. Uh, I think smart companies, and Airbnb is at the very top of this, realize that uh, customer service is not just a, you know, is luck, there's cost and else, but it's your branding moment.

And obviously, we all I'm tracking what's going on with United—I think enough said for what a disaster there. Um, so one of the key things that I think also is an interesting part of scaling that this touches on is how do you, uh, essentially preserve a culture as you're scaling?

And one of the things that Airbnb has done an exceptionally great job at is preserving this kind of design and trust culture. What are some of the things that you're doing as a way to say, as we scale our employee base? Now obviously some of it's how you interview and who you hire and so forth. And interesting if there's more on that, but what are the kind of some of the key kind of ways that you operate to keep that focus on designing experience, making sure that everyone's putting trust at the very top of their thinking?

It's a great question. I mean, I think we do a pretty good job; there are areas where we don't, you know, and um, this is a really tough topic, especially when you're scaling. If you're in hyper growth—if you're blitzscaling, whatever the term you use.

Um, I immediately think of advice that um, Tony Hsieh gave us way back in 2009. He said, “If you care about culture, make sure you get it right from the beginning. Make sure that's something that you invest in upfront.” And he used the metaphor; he said, “It's a lot easier to mold concrete when it's still wet than to chip away at it later when it's dry.” And so the DNA that you set in the early stages of the company can have tremendous impacts on how the culture plays out over time.

So we really took that to heart. I think we were really impressionable around culture at that time; we really wanted to get it right. And um, we just started to put an enormous amount of emphasis into the people that we hired. His other piece of advice is that “culture doesn't make your people; it's the people that make your culture.” So get the people right and the culture will follow—not the other way around.

Yum, and so I mean, Brian and I were involved in every interview for a certain period of time until a couple years in the recruiting team wanted to have our throats. Um, we were slowing everything down, and then we needed to figure out a way to scale that. Well, how do we make sure that our voice and our um, our presence is still involved in the recruiting process and the interview process?

So we created this thing called core values interviewers, um, which has scaled phenomenally well for us, where we train people inside the company how we would interview, um, and they get to effectively sit in interviews on our behalf, um, and really screen candidates to ultimately make sure that at the end of the day, people believe what we believe, which is the most important part, I think, of hiring—that you know, we don’t want people who are interested in the dollar signs around the company or the next big technical challenge for themselves. We want people who are specifically geared towards our specific mission in the world.

Um, and so how do you find people who believe what you believe? Um, which is what Brian and I would do in every interview. And I think by training others to uh, to look for the same signals or ask the same questions, we've been able to scale that. Now we have, I think, hundreds of core value interviewers in all our offices around the world.

Uh, one of the things that I think is a funny little hack but subtly very important that I learned from you guys was making the conference rooms and the meeting spaces into exact, you know, kind of—we pretty close to what do key hosts and experiences look like because the thing that I—I had this kind of just amazing kind of epiphany walking into it the very first time, and you guys hadn't told me about it, hadn't showed up in any of the board meetings.

You know, Strange Love boardroom is not designed that way, which is the one that I normally go into. And I went to a meeting room—it was you or someone else—and I went, “Oh my God, this is awesome!” Because part of what you're doing is you're reminding everybody in the meeting space, “You're in these are the people we work with; these are the things that matter; these are the experiences that matter; and you're in them too, right here as you're doing them.”

Right? And I don't know which, you know, because such an obvious design insight that I've then been reflecting into. Matter of fact, sometimes you have to come by my LinkedIn office; I'm beginning to change my office more in this direction.

And um, part of the thing is like, what other kinds of like design hacks like that? You said this is how it keeps us focused on what's important in terms of what we're doing?

That's a great question. I think, um, for those who don't know, we turn any available space we have into a replica of an exact listing somewhere in the world on our service. Um, part of it was just to differentiate and say, “How do you know when you walk into our office that it's our—like where you are in the world?”

Yeah, um, and we thought, “Well, our listings are pretty unique to us, so maybe we'll use those to um, to inform our spaces.” And it also just reminds us that while we're behind computer screens all day, uh, like probably most of us here, um, that there's a physicality to what we offer that at the end of the day we’re coding and designing to facilitate a connection for people to enjoy really magnificent spaces around the world.

Um, and so that's like the other kind of subtle reminder behind this, um, and I have to give Brian credit because he came up with that idea. Yeah. But it was awesome! Yeah, it was a good one.

Yeah, um, so what else do we do? Um, I think dogfooding is really important, so making sure that we give opportunities for any in the company to continually stay engaged with the service. We highly encourage anyone in the company to be an active host—not everybody can for various reasons.

Um, but we are constantly trying to find ways that you can be in the customer's shoes, um, so that you're connected to the people that you're serving at the end of the day. I think if you're not careful, then there have been times—I have to admit—in our company's history where we have this gap has emerged with our guests and hosts and we've lost touch.

And so there's this constant battle to try to like always stay connected. So we give people um, stipend credit to travel every quarter on the site just to make sure people have, you know, no excuse—yes—to be using the product and service. And ultimately, when people come back, everybody has some insight or some bug that they found in the product that is, you know, it's part of like you can say like our intelligence network, right? Like constantly honing and iterating, making the product better by having our own people go out into the world and use it.

So I'm going to ask uh, two more questions and then we're going to open up for questions from uh, the audience. And if you guys don't have any, I only got to three of the 15 questions that I'd written down and I have some for you too!

Oh great, I have to do that too. Um, in fact, do you want to start with one of those?

Before I'm just curious because you've been involved in the company since uh such an early stage. Like I don't—I don't think we had quite hit the hockey stick at that point, though you could sense it—we were about to. What have you seen, from your—because you have such a unique lens into the business—what have you seen that would be relevant for the conversation here, in addition to the things you've said?

So I think that one of the things that, um, well we've gotten a little bit well—a little bit of the question is, is uh you guys collectively, it's part of the design thing—this is the multiple stars thing, which is you go to, if we were to design the perfect experience and everything would be perfect, what would that be? That's clearly not economically deliverable and not scalable, but then start ratcheting back from that.

And actually, in fact, I think that's a good design exercise for everyone to do, right? Because it—it kind of gives you a product and a product-market fit North Star that then you begin to try to answer what is economically feasible, what is scalably feasible—which are not exactly the same things but related—and work back from that, and that also gives you an iterative and dynamic road chart for doing it.

And that was actually one of the things where I had always been more of a kind of build up—like start here and yes, I know where I'm kind of roughly going, but I didn't design the perfect experience; I kind of went, “Yeah, yeah, we're going to need to be there eventually, now what are the things we can do now?” And that kind of question was one of the things.

Um, so then I actually, I'm going to only ask you um, uh, one of my questions before I open it up, uh, and then you can obviously ask as we go.

Um, describe how you do your vacations—like how you do the kind of the walkabout. Remember the thing that we were—we had dinner and you were about to to go and travel? What's that and why do you do that? And how does that, how does that lead to how you think about design?

Interesting. Um, so "re" is referring to um, every December between uh basically between Christmas time and uh New Years, we give the company time off um, it started a couple of years ago and—started in 2009 or 2010, we've done it every year since. It was really the first time where it was like, “Oh my God, you have a week to go travel! Where could you go?”

And um, I started this thing about six years ago where I, on December 26th, I'd pick a destination and on December 27th I'd be on a plane somewhere! And you get to take one bag and you get to, you know, book one Airbnb at least for the first day.

Um, and it's sort of this kind of stepping into the unknown. You kind of—you don't know where you're going to go; you get one bag so you can't really pack that much, you can't overthink it; you just have to just have to go.

Um, so I—it’s now fondly called jetpack. Um, it's a thing that happens every year, um, and I've stayed with amazing hosts—from, uh, in Uruguay to Japan. One time I ended up in a—I ended up staying with a Japanese Buddhist monk in his temple on a, uh, tatami mat for $60 a night. You can rent this guy's tatami mat in his temple in southern Japan, the middle of nowhere, barely a cell phone connection.

Um, and it was one of the most phenomenal experiences I've ever had, and it took like five minutes of planning! Yeah, the two things I think are great is it's a real deep way of walking in your customer shoes, but it also like I actually think I learned was always to ask you afterwards what was the last one you did.

CU, I learned a bunch about the human condition, uh, as a part of it. So now let's open up for questions because we have a few minutes in case there are any questions. There's one here and one here, so great.

Hey, uh, thank you! So earlier, Brett was talking about how he scaled engineering for quip, uh, in particular around engineering managers, tech leads, and sort of how the structure of the team changed over time. I'm curious for Airbnb, how has the product process, uh, changed over time?

Good question. Uh, I think in the early days we were, if I remember back, we were really adverse to hiring too senior too soon. That was something we like—we would often say to each other, "Man, if I could go back, I would have I would have changed that."

Um, I think that bringing in senior leadership, bringing in senior management, um, when you know that you're going to scale—like it's actually, I wish we had gotten ahead of that. Like knowing that we were going to scale, I wish we had just said, “Okay, let's like hire like a R or two. That seems out of reach today, but because of how quickly things are moving in six months, like that's going to be the perfect role for what we need in the organization.” So that would be one piece of advice I give myself if I could go back is like, um, it's okay to hire a senior even in early stages of of the business.

Dylan, that was one of the questions I was going to ask too. Uh, so given, uh, given the right kind of supply, it seems like at least from my perspective Airbnb is a supply-limited marketplace. So curious how you think about the different types of supply—from folks that have, for example, a spare apartment they want to rent out versus people that are running out their own home—kind of how you think about the user experience for the different groups and how you think about scaling each one up?

Yeah, since the beginning, we've been supply constrained—um, because if you think about the demand-supply side, demand side people have been booking travel online since the internet was around, right? Like since Expedia and the ear, it orbits, etc. So that was like a known behavior, but people were not listing their homes on the internet to rent out to people they haven't met before. So that was like a whole can of worms to try to unpack.

Um, so we've always traditionally been supply constrained. Um, and there have been, you know, all sorts of twists and turns over the years of how we think about getting supply—what our best supply is. Um, I have to admit, in the very early days of Airbnb back in, gosh, 2009, um, we thought that these property managers were like the best—the best thing on earth.

Because in one phone call you could have 50 properties all at once, and for any marketplace—like trying to grow—like you need to get that momentum going, right? And so like, um, so we built all these tools, because if you have 50 properties, it's a very different need than if you're renting an extra bedroom around, you know, down the hallway.

Um, and so we had to build all these special tools, and like, um, what we discovered pretty quickly is that customer service started to ring off the charts because these property management companies and big property groups were in it for a very different reason, right? These were businesses. These are people that needed to maximize every penny for their dollar.

Um, they needed to like squeeze out every bit of efficiency of that unit that they rented or purchased, um, and they cared less about the actual experience of the guest. And we heard it loud and clear, um from U. Things like bait and switch and like all kinds of crazy stuff that these guys were doing, and we realized pretty quickly, um, over the course of a couple of months that yeah, it's great for supply, but it's not great for the customer experience.

And if guests are staying with us and they're like having a horrible stay, like that's not good for the long-term R business. So um, we—after a couple of months of experiencing this, we dropped that and we went back and said, “You know, our core is really people's primary homes. Maybe they have a second home up in, you know, Tahoe or some other vacation destination, um, but it's really about the core primary home experience.”

And we'll keep the marketplace open and wide enough such that kind of like eBay—like if anybody lists any kind of thing, like whether that's a treehouse or whatever—they could do that. But I think like really having that experience, knowing what we didn't want, refocused us on what our core offering is. It's like, um, the beauty of the service is you get to experience somebody's home in a local neighborhood, um, versus, you know, uh some sort of more touristy kind of part of town that's more of a generic experience.

So I think, um, from a supply perspective, it's constantly figuring out how do we—how do we find those those primary residences? People who want to, you know, need to make ends meet for various reasons. Um, actually one of our fastest growing segments right now are empty nesters, which kind of makes sense. You know, they've got an extra room, their kids are off; maybe it's a little bit, you know, less lively in the house. Um, maybe they've always dreamed of opening a bed and breakfast one day that they could retire into, and suddenly in five minutes, they can have the bed and breakfast experience in their own home.

Actually just very quickly because it follows on that, and we're almost out of time, how did the sense of the whole magical trips thing come out of that kind of social connect—like how did you—when did you guys really start thinking this is what we have to launch next?

Um, so we knew since 2012 that accommodations is just one of many decisions that we all make when we're putting a trip together, right? You think about where do I want to go; what's the purpose of the trip; where am I going to stay; how am I going to get there; what am I going to do once I'm on the ground and I'm actually in the city that I'm traveling to or the country?

Um, and so we've known for a long time that again, like designing the whole experience—accommodations was just like one instance of a much bigger journey that people go on when you take a trip. And so it was always our ambitions from starting in early 2012 that one day we wanted to design the whole trip. Like what if we could imbue every part of the travel experience with the same values that we've put into the accommodations experience?

Um, and so that's—that's informed what launched in November, which is, uh, experiences. So when you travel to the city, um, you can now within the Airbnb community find really cool local things to do provided by hosts sharing their passions, uh, skills, and interests with people.

Exactly! All right, well, let's thank Joe for coming with us! Cool! Thanks, re. Appreciate it.