Zipline ft Keller Cliffton - Reinventing Delivery with Instant Drone Transport

0:05I think, you know, for a lot of our investors, I think they had largely written the company off and And we're like, yeah, the company's just dead. It's not going to go anywhere. Frankly, that was probably the worst and most difficult time I've ever had to go through, like, in my life. It just felt like chewing glass. It's just so scary. It's incredibly difficult. You're totally emotionally raw. You have no idea if, you know, you're going to make it to the next week, let alone the next month. And I think that, you know, we had to get lucky. And it also helped that we were just totally stubborn and hard-headed.

0:35And we're like, we're doing this. No one's going to get in our way.

0:42Welcome to Crucible Moments, a podcast about the inflection points that defined some of the most consequential companies of our time. I'm your host and the managing partner of Sequoia Capital, Rolof Wotan. In this season, we'll explore the transformation of finance for the digital age, the story of the world's biggest cybersecurity company, and how a radical approach to game development created Europe's first Decacorn. Today's episode is about a startup's audacious vision to take on global delivery and logistics.

1:27In 2014, Keller Clifton presented an idea to his current investors and board. He was going to scrap the direction of his consumer robotics company and repurpose the business entirely towards building an autonomous drone delivery system that would bring life-saving medical supplies to remote populations around the world. The response was less than enthusiastic. The challenge of pulling off hardware and software for this type of endeavor would be steep, if not impossible. The investors asked Keller, what did he know about drones? What about healthcare or logistics for that matter? The answer was not a lot.

2:12Today, Zipline is a global leader in autonomous delivery, credited with delivering critical supplies to over 40 million people. Today, we'll look at the pivotal, crucible moments that shaped Zipline's rise from a bold idea to reimagining global logistics.

2:37I'm Keller Clifton, the CEO and co-founder of Zipline. I didn't even really know that entrepreneurship was a thing you could do graduating from college, but I read Delivering Happiness, which was this book by Tony Hsieh. And I was kind of blown away because Tony and his partner, Alfred Lin, had built these amazing businesses coming out of college. And they'd actually both lived in the same dorm that I lived in in college. And so it was like, wow, you know, me seeing people who had built these really cool things completely from scratch, it really caused me to, you know, to think differently about what I wanted to do with my life.

3:18We were about two years out of college and a few friends and I got really interested and passionate about robotics. Around that time, iPhones had become totally ubiquitous. And while there were a lot of people trying to build these huge apps on top of the iPhone, we were looking inside the iPhone and seeing all of these interesting components that would be really, really key to building powerful robotic systems for the world. Very few people were thinking about robotics in 2011, but it felt to us like there was this obvious opportunity to try to build the Apple of robotics, a big robotics company focused on solving problems in the real world.

3:55So, you know, at that time, the idea of building a robotics company was stupid. Like, we're investing in robotics. Nobody was really that interested. Feeling stuck, Keller thought back to the book that had inspired him in college and decided to reach out to its author, Tony Hsieh, for advice. Hsieh saw potential in Keller and offered him and his team an apartment in Las Vegas to start building their company. They called it Remotive. Their first product was a robotic toy called Romo that users could control with a smartphone app. Shea also introduced Killer to his former colleague at Zappos and Link Exchange, Alfred Lin, who was now a partner at Sequoia.

4:37Tony Shea sent me an email about a Remotive and asked me to come out to Las Vegas. and when Tony tells me that he has an interesting company, I listen. I'm Alfred Lin. I am a partner at Sequoia Capital. So I went out there, met with Keller and his team. Keller was a very special founder. He would light up the room. He's such a good storyteller and his effervescence for technology, for robotics really came through in the first meeting. In October of 2012, Sequoia invested in Remotiv. But the team soon faced a painful reality.

5:28We were so extraordinarily naive, you know, when you're 23 years old and thinking like, oh, it would be so cool to build robots. Like, let's go figure out a way to turn this into a business. The team didn't really have a ton of expertise that we were bringing to bear. He started selling the product and people played with it and they were overjoyed to get the product. But inside of somewhere between 24 hours and two weeks, they put the product back on the shelf and they stopped playing with it. And so it was really hard to see how this was going to be a large business. I think ultimately when we started shipping that first product for Emotive, it started to become pretty clear to us that consumer robotics was not the right place to start.

6:13One day, Keller even told me, I don't think this is what we aspire to be. And I told him, did you think that this is the way to end? And I think that inspired him to rethink what he could do in that moment, which was a tough meeting, a tough conversation. But he left, in one sense, crushed, in another sense, trying to figure out exactly what he still had at Remotive and what he can turn it into. Quite literally, everything was at stake during this transition. My name is Ryan Oxenhorn, and I am a co-founder and responsible for software at Zipline. We fully shut down the toy company. We kind of like, you know, fully closed that door before knowing which door we were going to open next.

7:04And at least for me, I had a pretty simple criteria that I was weighing when considering different opportunities, which was ultimately which of these problems that if left unsolved was a really bad thing for humanity. We explored a bunch of things together, the three of us as to what that higher impact could look like, one of which turned into Zipline. I'm Kenan Wairobek, and I'm the co-founder and CTO at Zipline. What became Zipline was inspired by a few things. My wife, as an epidemiologist, and kept telling me about these projects in public health, they get stuck on logistics, things like vaccine campaigns, where they have the vaccines, they have the doctors, and they just, the logistics of actually getting the vaccines where they need to be when they're needed just kills the whole project.

7:51And that's one of the things that inspired us to go dig into that and really see what that was. The more we learned about global logistics, the more we understood that, you know, this is a market that really only serves the golden billion humans on Earth well. If you're the 7 billion people who aren't in the golden billion, your access either sucks or is non-existent. As a result of that, millions of people, mainly children, die every year due to lack of access to basic medical products. and we've been making excuses for a hundred years why it's so hard to get things to where they need to go. Robots had already at that time scaled pretty dramatically in warehouses, in some hospitals.

8:31And I remember going and seeing this company, Kiva Robotics, and I got to go to one of the Kiva warehouses and see these orange robots running around inside this warehouse, moving things back and forth. They would go and pick up a shelf and bring it to a human packer. And I remember seeing that thinking, man, somebody's going to do that for outside the warehouse. And that's going to be one of the biggest companies in the world. We thought, man, somebody's going to build an automated logistics system for Earth. And that would be something we would be very proud to spend a decade of our lives working on.

9:04The team's interest in healthcare logistics coincided with the rise of new technology. Amazon had just announced that drone delivery is coming, you know, any day now. The summer of 2013, I remember you couldn't go to a park in San Francisco without seeing someone flying a DJI copter. That was pretty cool. That was an example of like, wow, technology that was previously just a TED Talk and a college lab demo was now suddenly out in the world. The founders began to consider whether drone technology could be applied to the problem of healthcare logistics. As a very huge tech skeptic, I was like, all right, we're going to go spend time in the field and understand a million reasons why we could never make a dent in this problem.

9:50And the deeper we got, the more convinced we were that we actually could solve this problem and that we should go for it, solving medical supply logistics with drones. I think it's worth pointing out that all of our investors thought focusing on logistics was an extremely stupid thing to do. And we were talking about, oh, maybe we'll go to Africa. We'll focus on global health care. People were like, what are you talking about? I remember some of the conversations with Alfred were rather difficult. The idea sounded kooky at the time, which was we're going to figure out drone delivery. We also challenged them that delivery at scale, you can charge a small dollar amount.

10:31But at the beginning, when you're subscale, delivery costs will be very high. And the only kind of applications that would work in that situation are high-value applications. He took that into account and then came up with an even kookier idea, which is to do healthcare logistics. You don't know anything about drones. You don't know anything about healthcare logistics. And then on top of that, we're going to go find a country in Africa where road infrastructure is hard and complicated. And we're going to fly there because they're going to want our application, which is healthcare deliveries. And we can do it instantaneously versus waiting three weeks for delivery.

11:16It's common for startups to pivot, but the ambition and magnitude of this pivot is almost unheard of. Companies make little P pivots all the time. These capital P pivots probably happen less often. But if you're going to be a successful company, you probably will have these refounding moments because inevitably your first product, your first idea is not the product that gets you to the largest scale ever. I think, you know, for a lot of our investors, I think they had largely written the company off. And we're like, yeah, the company's just dead. It's not going to go anywhere. I remember, you know, trying to fund the company during that period was extraordinarily difficult.

11:59There were just a few investors who were willing to believe in us with like this totally crazy vision of building automated logistics. I mean, frankly, that was probably the worst and most difficult time I've ever had to go through like in my life. I mean, it's just so scary. It just felt like chewing glass. It's incredibly difficult. You're totally emotionally raw. You have no idea if, you know, you're going to make it to the next week, let alone the next month. And I think that we had to get lucky. And it also helped that we were just totally stubborn and hard-headed and were like, we're doing this.

12:32No one's going to get in our way.

12:37The team charged ahead with its vision for what would become Zipline. Their next step was selecting a first market to build and test a healthcare logistics system. We knew we needed a public healthcare system. And we also thought we probably needed a smaller country that would be more impedance matching with Zipline in terms of like the velocity we wanted to operate at, the way we wanted to innovate. Hard for a really small company to operate with a really giant, ossified, bureaucratic government. And so we were looking in Central America, we were looking in Africa, starting to talk to governments.

13:15No idea really what we're doing or like how to sign an infrastructure contract with the government. But it ended up being Rwanda, this first country that really took a bet on us saying, yeah, you should build that here and we would use it and we would pay for it. One of the innate things about Rwanda that made it a perfect first country for us is that, first of all, they were entrepreneurial like us. And they really wanted to become a leader at the intersection of health care and technology, which is very much our goal as well. My name is Maggie Jim. I'm the chief of staff to our executive office at Zipline.

13:53The nature of Rwanda's topography is that it's super hilly, even though, you know, if you like look in a sky map, it's like 40 miles away. It could be anywhere from three to six hour drive, depending on the weather condition. And so the conventional logistics system is insufficient in solving the problem. And the challenge that we could help with in Rwanda would make the impact undeniable. And so all of those things made it the best place we could have started with. I remember this very specific meeting with the Minister of Health where I was saying, hey, we'll use these autonomous aircraft. It'll deliver all different medical products to every hospital and health facility in the country.

14:37And she looked at me and said, Keller, shut up, just do blood. And I was kind of blown away by that, but she had a lot of conviction. She was like, look, 50 % of transfusions are going toward moms with postpartum hemorrhaging. 30 % are going toward kids under the age of five. This is a really hard product to get to where it needs to go when it needs to be there. You have all these different components, packed red blood cells, platelets, plasma, cryoprecipitate, different storage requirements and shelf lives. You have different types of the blood, A, B, A, B, and O, positive and negative RH factor.

15:08It's a really hard thing to get the logistic system right for. and I actually only discovered like seven years later, seven years after that conversation, I was meeting with this minister and she told me, oh yeah, Keller, you never knew, but the reason we said yes was that just that morning, this email had gone around the Ministry of Health about this mom who had lost her life because it took us like nine hours to get a blood transfusion to her and she had passed away by the time we got it to her. And I remember thinking like, this is so screwed up, there has to be a better way. And that afternoon you happened to walk in.

15:43But, you know, she really believed in us. The president of Rwanda believed in us. They took this big bet on us and asked us to start delivering blood to 21 different hospitals. And we were like off into the races.

15:56With Rwanda's blessing and a goal of delivering blood via drone to 21 hospitals across the country, the team got to work on bringing this technology to life. I was sure when we started Zipline, we could just buy a drone. Just like, why wouldn't we buy a drone? Even if it wasn't the best thing to scale on, it would get us out to serve customers sooner rather than later. And so let's find a small aircraft style drone and modify it for delivery and we'll be off to the races. And when you're making things that fly, at the time and still today, the rage is vertical takeoff and landing. But if you want to go really, really far at the lowest price point, you don't do that with a vertical takeoff and landing vehicle.

16:34You do it with an airplane. We had that insight, and we were all aligned to go do it, despite everybody around us, including some investors, being like, what are you doing? Don't do an airplane. You've got to do VTOL, vertical takeoff and landing. The moment the vision for Zipline really clicked for me was when we got past the idea of a quadcopter and really started to think deeply about an airplane. At first, it seemed totally preposterous. Like, wait, how are you going to make an airplane work for this? and I think it was only a few days of thought experiment and whiteboarding out that we started to realize, oh, hey, the number one thing we were hearing from our customers is each additional kilometer that you can get outside of a region's biggest city is that much more value.

17:19We started talking to some experts and realized a pretty small little airplane could fly pretty fast, extremely efficiently, And range estimates for that, you know, 10x, 20x, 30x, what a quadcopter could do, was pretty mind-blowing and pretty obvious in hindsight. Zipline built what it called Platform One, a compact, unmanned aircraft designed to cover long distances and deliver small packages by paper parachute. While the company had done preliminary testing in California, it soon found that operating a drone in a testing area is one thing. Flying it out in the real world is another. We didn't know this until we launched, but Rwanda has some of the most volatile weather of any country on Earth.

18:06And so we were operating in lightning storms, rainstorms, insane dust storms. We'd be flying very low over mountains. And so before we went to Rwanda, we engaged with our friends at NASA, the world experts in this low altitude, and they helped us figure all this out. How are we going to simulate it? How are we going to tune this? And we go to Rwanda and we see that the air over these mountains in thunderstorms is way worse for the aircraft, way more turbulence than we were expecting. Not like 10 % more, like 10X more. And so we go back to them. We're like, hey, what do we do wrong here? We're in a hurry.

18:40We got to get this figured out. And they sort of were scratching their heads for a minute. And then like, you know, now that we think about it, We've never actually had the data from flying low over mountains in storms because no one does that. General aviation pilots would never do that because that would be suicidal. And medevac helicopters stopped doing that 50 years ago because way too dangerous. And so that's probably why we didn't accurately model what the wind and air is doing low over the mountains in these stormy conditions. We had to harden the system to over the course of many years of like just learning by doing and learning from the real world and observing what was breaking and then fixing it, making it more reliable.

19:20Perfecting the drone's hardware was demanding, but it was only a small part of the challenge of delivering blood to a remote location. Probably the biggest thing that we realized, maybe the biggest crucible moment for us, was understanding that the vehicle is only 15 % of the complexity of the problem. We had thought, oh, we're going to design this really cool robot. We're going to design this really cool autonomous aircraft. It's going to deliver things autonomously. But it turns out that in a national scale logistics system, there are a lot of different kinds of software and systems that all have to be interacting in concert to actually have an instant delivery service that people can rely on with their lives.

20:00We didn't really know anything before we had a package. So I ended up building the first half of the operation, which is what we call fulfillment operations, which is everything that has to do with how we interact with the hospitals, how they place order, how we actually manage and store the inventory and how we do pickpacking. We thought the hospitals can place order via a nice iPad interface. I visited some of the health facilities. They barely had cell coverage in most parts of the hospital. The other thing was there's no standard operating procedures for how you like pack and fly and receive blood that was just delivered from the sky.

20:37And so we had to do testing with the Ministry of Health as well. You know, drop tests to understand and make sure that the product's still intact. There's no hemolysis. For the first maybe three or four years of building Zipline, I remember every single night going home with the sensation that my brain was melting and that every single day the scope of the problem ahead of us had just grown. So, you know, we worked a long 14, 15, 16 hour day, but the list of to-dos only grew. I remember the first time the Rwandan Civil Aviation Authority came to visit our distribution center because we have to get approval to fly beyond visual line of sight.

21:19Otherwise, how do you deliver to the hospitals? Because we never had regulatory to fly beyond visual line of sight. We've never actually had a zip disappear into the horizon. You know, we had to like walk them through all of our procedures, do all the pre-flight checks and everything. And then the moment the zip launched and then it, you know, 10 seconds later, it disappeared into the horizon. I think the whole company stopped breathing in that moment. And you just never seen it before. It was the very first time we were seeing it. And we just kept looking down and up, down and up. Is it coming back yet?

21:55Is it back yet? And then finally, when it like came back, it felt like the longest, like five minutes of our lives. and to see on the iPad and team members on the other side, just be like, it delivered. We ran back inside the tent, made sure they couldn't see us. And then we started just screaming and jumping up and down because that was the very first time any of us had ever seen an actual beyond visual line of sight flight.

22:23The team's hard work was finally paying off, but Zipline was still struggling to support just one hospital. And the Rwandan government was growing impatient. We had done a soft launch, and so we had done, like, a few deliveries to this first hospital that we were serving. But the president was coming to inaugurate the distribution center. He was bringing all of this, like, international press, because this was going to be the first autonomous logistic system on Earth. It was an event that was being broadcasted to the entire nation. The president was very proud to show that he's worked with a technology company to bring this to Rwanda.

23:03And we were awake at 3 a.m. because the launchers for the system were completely broken. I mean, it was fundamentally unreliable. The rain came in and so it kind of like ripped through the floor. It was like muddy. It was water, puddles everywhere. This event is supposed to start at 8 a.m. Nothing is working. The launcher is a total piece of garbage. And to make matters more interesting, whenever the president goes somewhere, they send this like special forces team of soldiers like to secure the area beforehand. So there are like four Navy SEAL looking guys with like submachine guns sitting there watching us trying to fix this launcher.

23:40And you could just tell they're looking at us like you guys are totally fucked. After a couple hours, we somehow got that launcher like to a state where it could like ostensibly function. The president showed up. System operated absolutely perfectly for the three hours that he was there. It was magical. Like everything played out perfectly. And the package just like magically floated down super gently and landed in the middle of the banner. It was like picture perfect. And I got the opportunity to actually go and pick it up, open it up to show the president like, yes, blood has been delivered.

24:13And it was intact. It was an amazing moment. It was like a total miracle. and then he left, all the press left, and then everything broke again for the next eight months as we had to like fix things and iterate. So yeah, that's really what the early days felt like. We live by the motto at Zipline that we don't do these things because they're hard. We do them because we think they'll be easy. During that period of time, the Minister of Health started sending us photos of mothers who were passing away, who were bleeding out during childbirth because they were trying to provide us our motivation to get this done.