How Zipline Scaled to 2 Million Drone Deliveries | Keller Cliffton

26:11Uh, it was like very powerful for me. It kind of inspired me. It was like, you know what? We don't have to be perfect. Like this, like the mission was so important that we were willing to launch things that were not yet perfect, that were actually in some cases sort of bad. And it was still good enough to save lives. And we were able to rapidly iterate from there to something that was extremely reliable and dependable. I imagine over the course of 10 years, you get incredibly good at when shit goes sideways. how do you actually like resolve the issue? What's your type of process when something goes horribly wrong?

26:50What is the checklist that you go through, the steps that you go through to fix it? Yeah, I mean, actually for a hardware company that's operating at Zipline scale today, I mean, Zipline now has 135 million commercial autonomous miles and zero safety incidents. And so we take safety incredibly seriously. That's why we have these large test sites. We're at our first test site. I and and yet like somewhere in the world with a fleet this large doing that many commercial autonomous miles like there is always something going really wrong especially at the test sites and I think you kind of realize that every morning you're going to wake up and whatever is like the most screwed up thing like that's what's going to be at the top of your inbox and so yeah the fire And so I think you have to, you definitely, I think at some point, um, are realizing it's important not to have like an emotional reaction.

27:46Cause I think it's like completely exhausting, right? Cause it feels so bad. You need to like untie yourself. It feels so bad. And just look at the facts. And especially for a company that's trying to go really fast and do something very difficult, like nine out of 10 things you try on any given day are going to fail. Like Zipline is very used to failure at this point. Like we are constantly trying new things, failing in a wide variety of ways, having to rapidly iterate from there. And so I think you definitely learn to instead of like craving, if what you are like sustained by and what motivates you is success, success is actually like few and far between.

28:22Like the vast majority of the experience is failure. and so I think that you have to be very zen about actually being in love with the process of like being in a painful uncomfortable situation and having to rapidly iterate your way out of it and learning by doing and the first 10 times you try you are going to fail and so I think you have to be like motivated by failure and motivated by that like learning and growth rather than learning by like having some big success or getting a bunch of praise or having, you know, those things don't come that often, especially for the first like five or 10 years of a hardware company.

28:59Yeah. Um, we ended up, you know, this wound up becoming, there were times in Zipline's history where, uh, the company's survival 100 % hung in the balance. And every day we needed to be like attacking core engineering challenges or core operational challenges in a very relentless, disciplined, and focused way. Something we actually learned from one of our biggest investors, Valor. Valor was the largest investor, and I think one of the largest investors in Tesla and SpaceX for several decades. They were like on the ground. I think when they were first spinning up the Roadster production, Antonio and one other guy, I forgot his name, Tim, were just flying all over the damn place, basically figuring out all the different suppliers.

29:44And I think it was just a total shit show. It's an incredible story. And someone has to write the book on Valor sometime, but you know, Valor is an amazing investor in that like, you know, they are, there aren't that many, I mean, Zipline has a lot of big, amazing investors. Um, but I would say Valor is our only investor who actually like hands on knows how to build hardware and scale manufacturing and operations in the real world. And so, you know, they were kind enough. They actually like deployed special forces teams into Zipline for a year. Um, this was only a year ago where it felt like, you know we were launching our next generation product which you can now see flying very reliably behind us but it was not flying reliably at the time in fact like i don't know just a year and a half ago i was like here at midnight on a saturday night belaying one of these aircraft because it was so unreliable like it would just turn itself off like every fifth flight and if we didn't want to crash we had to have the the aircraft on belay you know so we had like a high lift and a rope running from me using a climbing harness up to the top of this high lift and then down to the aircraft I mean, you know, this stuff was, it was desperate.

30:44It did not feel like the company was going to survive. And Valor helped us kind of like build this process that we've now turned into our 8.30 a.m. meeting, you know, which is like zip lines, engineering leaders, and some of our ops leaders meet every morning at 8.30 a.m., including Saturdays and Sundays to look at like all of the highest priority issues that we observed the day previously in the 24-7 period. Because these test sites operate 24-7, by the way. So, you know, even if you come in at 7 a.m., you're like catching up on, you know, 12 hours maybe that you missed out on of data and things going wrong.

31:21And so 830 a.m. we would meet and it was always like executives plus like the individual contributors on the team who are actually doing the work. And so it's a really good chance to like there is no hierarchy or no like middle management in that meeting. It is like the person doing the work able to give a direct update on what we're going to do to fix it and ask for any resource that they need inside the company and have access to me and the executive team. That became a big part of our culture. And I think it's anyway, it's kind of like how we responded to this feeling of like desperately trying to solve problems day in and day out that are arising, you know, in eight different countries, many different metros, different test sites.

32:04like there's always something horrible going on. Yeah, I remember Sam Altman talked about, you know, the first time that you have this experience, this like company killing moment, it's like the end of the world and you're like cold spots at night, that sort of situation. By the seventh time, you know, you survived the first six. And so you're like, oh, well, you know, if we survived the first six, we'll probably survive this one too. Even you said as of last year, like you had another one of those, what have been the like key like moments where you were kind of waking up in the middle of the night, thinking to yourself, like, are we going to survive?

32:33And how did you manage through those? I mean, yeah, we jokingly called them WIFIO moments at Zipline, which have you heard of that? WIFIO? WFIO, we're fucked, it's over. You have an anger to fuck? Okay, here's another WIFIO moment that we're encountering. You know, I mean, I can certainly remember one in the early days, like when we were actually launching in Rwanda, the president was coming to the distribution center to inaugurate the distribution center and for us to do our first commercial delivery. And we had been kind of like soft launching, trying to do our first commercial deliveries for a couple weeks before he showed up.

33:11And it was a disaster. I mean, the system was not operating successfully. we had used this contract come contractor company to design the launcher for us it was terribly designed the launcher is what would basically catapult the aircraft into the air um and so it would break every like third launch and we're trying to fix it trying to fix it it was like 7 p.m the the night before the president was coming to this huge we had like you know the bbc and the new york times all these big media outlets were showing up for this huge kind of production they built this big tent for the president to give a big speech and he was going to launch the first aircraft and it was 7 p.m and the launcher broke and we were then trying to fix it trying to fix it's like 10 p.m it's not fixed 1 a.m it's not fixed uh and so abdul and i were out like on our backs in the gravel like holding flashlights in our mouths trying to like take this thing apart again put it back together i remember we were skyping back to the team because it was skype at the time we were skyping back to the team in the u.s and it was like you know it was basically like Apollo 11, right?

34:14Sorry, Apollo 13, right? Where it's like, Houston, we have a problem. And they're like, well, what do you have? We're running out of oxygen in one hour. And they're like, what do you have to fix it? We have duct tape and styrofoam. And so they kind of walked us through taking it apart, putting it back together. As we were doing this, by the way, the president has this security team that goes 12 hours in advance to any location that he shows up at. So there are like five Navy SEALs looking guys holding submachine guns just watching us work at like 1 or 2 a.m. You're under the gun. Yeah. And you can tell they're looking at us like, you guys are totally fucked.

34:50You know, like, what are you doing? Like the president is showing up in seven hours and you're taking this launcher apart. And so we ended up putting it back together, testing it once, and then thinking like, you know, what more can we do? And so, you know, all the media showed up, all of the ministers showed up, the president showed up in a helicopter and he gave this big speech and then came to the distribution center and we like showed him to the launcher and he launched the aircraft and it worked perfectly. you know we're all sitting there we were like as we were as amazed we were more amazed than everybody in the audience and um you know this zip flew off made a delivery came back landed and we acted like that was totally normal and expected and then the president actually ended up spending like two hours with us um kind of talking about the future and then he left and then the system like broke again for the next nine months straight you know and so i think it was um again some luck is involved but uh that's a good example and yeah and then just i mean i think hardware companies especially the scope and scale of what you have to do with each new generation platform is so much higher than the last that every single cycle nearly bankrupts the company like i think that you can look at you know um i think the model 3 is a really good example i mean the model s was a profitable product but like the model 3 nearly bankrupted Even 2018, like I think 16 years after Tesla's founded, like that - Nearly bankrupt.

36:20Yeah, the Model 3 production ramp. I don't even, I don't know. I think Antonio was actually on the Model 3 production line in Fremont with Elon at that time. Yeah, I think they all were. Like, you know, the company was on the verge of bankruptcy. You know, when you are doing these big hardware launches, it is, you know, getting hardware to be reliable and launching into the real world is incredibly hard, obviously. You have this big, complicated global supply chain, and it only takes one. to screw you. It only takes one unreliable supplier. It's, you know, your performance is the worst of your 10 ,000 suppliers, you know?

36:53And so it's pretty harrowing. And the scale, if the company is growing fast and you're building a big ambitious product, the scale and scope of what you're doing means that like a couple months of not producing is probably like bankruptcy for the company. And so we went through that moment with Platform 2. You know, we knew that Platform Platform 2 was serving a market that was 100 to 1 ,000 times the size of the market that Zipline had originally focused on, which was just healthcare logistics in rural areas. Platform 2, which is operating behind us, is designed to really enable teleportation-like logistics from any business, health system, from a hospital, healthcare facility, warehouse, retail store, or restaurant, directly to homes in a way that's 10 times as fast and dramatically less expensive but um you know platform two was a way more complicated product and the scope and scale of what we're trying to do was uh was more ambitious and we wound up being you know six months late shipping the product uh and it felt like the company's i mean it was certainly the case that i think had we gone another six months i think zipline would not have survived and so it was like we just got into this hardcore mode whereas like there was no op there were no other options like we knew what we had to do.

38:09Burn boats. Yeah, the boats were burned and no one was going to give us more money. And so it was like, we needed to launch this product. And this was only about a year and a half ago, or even less than that, like a year ago. And the team worked through Christmas and everybody canceled their vacations between Christmas and New Year's. And then we launched the product on January 15th. How have you kind of thought about the hardware iteration side? Because I know you want like the most reliable system, but you also have to be improving the technology at the same time. Have you kind of married the two over time to be able to do as many deliveries as possible while also recycling in new builds?

38:51Yeah, I mean, that's like that's the main trick, especially when it comes to this kind of an in this industry, automated logistics, drone delivery. I mean, no one has any idea what this is going to look like at scale. It's 12.01 a.m. We're figuring everything out completely from scratch. There are no precedents. When you think about it, at least Tesla knew they were building a car. They knew it probably had four wheels and a steering wheel. But no one has any idea what this looks like. There really is no precedent for how should charging work? What does the business model look like? How does maintenance work?

39:25What's the right regulatory strategy? How reliable does it need to be? How should you price it? No one has any idea. Yeah. And I think one of Zipline's advantages was, I mean, for sure, we were like the smallest team. I mean, Amazon at the time, right? Jeff Bezos had been on 60 Minutes saying they were going to be doing drone delivery to every home in the U.S. within two years. This was in 2013. And so we always thought, OK, well, we're definitely going to be a fast forward Amazon. They were spending like, who knows, you know, hundreds of millions, a billion dollars a year funding that effort.

39:57and we had a tiny fraction of that and a tiny team. And we always thought we were not the biggest team, we didn't have the most money, but we always knew for sure we were the most practical. That was in our DNA. We were extremely practical. We were living in these construction trailers out here, working nights and weekends. And we thought that our DNA should be like we could move super fast in the real world and get things into the real world quickly. And so the whole trick is that when it comes to hardware systems like this, safety is obviously paramount. I mean, you can look at Uber ATG or you look at Cruise, I mean, you can see the impact of a really bad safety event on the company's trajectory.

40:35And it was especially important because Zipline was delivering a life-saving product to someone whose life was like hanging in the balance. And yet we knew we needed to move really fast. And so really the answer to that is like being extraordinarily good at testing.

40:52ZipLine invested a lot, not just in this large team that was doing software in the loop testing, but we also built these really powerful HIDL systems. HIDL stands for hardware in the loop testing. We learned a lot from SpaceX on that front. SpaceX may be the best company in the world at HIDL. And we learned a lot from them and built these complicated HIDL systems that meant that we could deploy new software onto the actual avionics of the aircraft, which is then connected to a simulator. So the plane thinks it's flying, even though it's basically sitting on a desk, is how to think about it. And it's a really powerful thing because it means that 95 % of software bugs we can catch without sending something to a test site and accidentally crashing an aircraft due to some stupid software bug.

41:36I'd imagine it's way more expensive and way more time-intensive if you have the latter. Yes. But even so, 5 % of the problems you can only catch at the test site, which is why Zipline has invested a lot of money and time and resources into building these large test sites that operate 24-7 in the US. We have three major test sites. One just opened last week. These test sites enable us to get access to a ton of different real-world conditions that would be impossible to replicate in simulation. And some stuff you are only going to see it break in the real world. You're looking for the really weird edge cases of like some bird nesting in some part of the hardware or, you know, um, like solar weather, for example, you know, solar weather wound up being a pretty big consideration for us in terms of getting the GPS reliability that we needed, the number of nines we needed.

42:21Solar weather winds up being like a really major consideration, something that most normal people don't even know that much about, but like is really important for these kinds of systems. Was the fact that the like weather in Rwanda is so bad and there's like lightning storms and all this sort of thing, just like the perfect environment for the initial tech. Well, of course, we were complete idiots and we had no idea. You didn't know at the time. No, we had no idea. Like we launched in Rwanda with zero context on like what we were getting ourselves into. Had we had any, you know, ounce of common sense, we would have like looked at the weather forecast before we signed that contract.

42:51I don't think we did. Also, Rwanda is actually a very high elevation country, which means it's like, they're just - The mountains are - It has some of the most volatile weather on earth of any country. It has challenging terrain. I mean, there were a lot of things that were really challenging about Rwanda, but it was hypergravity training. Like we had to build the system in a way that 10 million plus people could rely on with their lives and the lives of their kids. And obviously failure was not an option. And so we just had to make the system, we had to harden the system to these extreme weather events.

43:19And that work has obviously become immensely valuable to Zipline over time because Zipline is now the only autonomy company that operates at this scale in all these gnarly weather conditions day in and day out, 24 seven. but we were sort of like forced into that by the nature of the original use case that we chose. I think Antonio was interviewing you at some point and he said that you guys are basically the only company that was able to just collect all the data through just doing all these tests. I think it was like 40 million autonomous miles flown before you even started doing commercial operations in the US at the time.

43:55Yeah, that sounds right. Yeah, 40 million, 50 million. What was it like when the FAA came over there and basically went and talked with the authority in Rwanda and basically talk to them about potentially, you know, working with you guys on the U.S. operation. What was that like? Well, we always, I mean, you know, the big vision here was to build an automated logistics system for Earth. We think that in the same way that the Internet can move things back, move information back and forth very efficiently, no matter where you are in the world, we thought, man, logistics, it's like all built on top of humans and gas combustion vehicle and unions.

44:28And it's just this like very complicated, antiquated system in certain ways. And we kind of felt like ultimately this is all just going to be automated. It's going to be dramatically less expensive, dramatically faster. And of course, the advantage of that is that logistics today only does a good job of serving like maybe the golden billion people on Earth. But if you're not in the golden billion, there are seven billion humans who are not in the golden billion. And those seven billion people, their access to logistics either sucks or is non-existent. And so we felt like by far the most exciting thing here would be to build a logistics system that could serve all people equally.

45:05That was the promise of automation and robotics being applied to this problem. And so in 2019, when the U.S., I mean, it wasn't the FAA. I mean, it was really a lot of these big U.S. companies. A lot of big U.S. health systems were seeing what we were doing, saying like, hey, we have all these similar problems. We're trying to advance telepresence. and we need ways of delivering directly to customer homes faster or we need ways of meeting like home health care nurses where they are at and then we were also hearing from big companies like Walmart which is you know it was one of our main partners in the U.S.

45:40that they wanted to be able to enable teleportation from their stores directly to homes and we were hearing from a lot of these amazing restaurants that we work with like Chipotle or Sweetgreen or Mendocino Farms or Wendy's. And they also wanted to be able to do teleportation directly to customer homes. And so, and again, and some of the biggest health systems like Cleveland Clinic in the US and Michigan Medicine and Memorial Hermann, they were all like really, really excited to launch. And so it kind of became this issue of like, well, we got to get regulatory permission. And the FAA didn't have a precedent for how to do this.

46:14I think it's just a really challenging chicken and the egg problem because they, you know, drone delivery companies had been asking the FAA to do this. But the FAA was like, cool, prove that it's safe. And these companies were like, well, we can prove that it's safe once you let us operate. It's the catch-22. Yeah, it's a catch-22. And so Zipline had this big advantage, which is we had 40 or 50 million commercial autonomous miles outside the US with zero safety incidents that we were able to provide to the FAA and say, here's how you know it's safe. And that wound up being really, really valuable.

46:40And so the FAA was able to move pretty quickly. We've always kind of had this strategy of just being an amazing partner to them. Like the FAA wants to innovate, especially all the people, I would say, doing the actual like individual contributor work, like they want to win. They're often constrained by like crazy legal bureaucracy and just these like crazy processes and rules. And, you know, don't get me started on NEPA and all these things that actually prevent federal agencies from doing what they want to do. But we ended up creating this really productive partnership with the FAA over five years.

47:19And Zipline has now been able to basically like year over year over year, kind of like push the envelope, expand what's possible. And the FAA has been our partner every step of the way in making that happen. So I think that, you know, maybe there was this narrative that like, oh, the FAA is going to be impossible, or it's like too bureaucratic, it's not going to get the job done. I mean, it's like, it's been the exact opposite over the last three to four years. Like there is no regulatory constraints on Zipline's business right now. Because once we were able to show that the thing was safe, the FAA, you know, was a partner in accelerating this.

47:49I was talking with Denton beforehand, and he mentioned basically flying to Rwanda recently, and he was like having trouble with the border patrol or something. And then the guy heard that he worked a zipline, and it was just like this smile that came across his face. And he's like, awesome, you know, go right through, no problem. If you deliver a million bags of blood or vaccines or so on. Maybe you're only touching a million people's lives, but then you have this like ripple effect where there's not only, you know, it's a million people with families, right? And so you create a bunch of like customer love and goodwill with, with an entire like nation, honestly, um, by, by creating something that's super valuable for a single person.

48:24How do you think about just like kind of creating the right customer brand and, and like love for drones in general, uh, in your mind? I mean, yeah. And by the way, you know, we've now expanded from rwanda to ghana nigeria cote d 'ivoire kenya um and and then obviously the us and and um we'll be announcing new countries next year too but like i think that in all of these countries i mean yeah you're definitely right if you if you spend time in these countries like you will for sure meet people who are alive because of zipline or they have loved ones who are alive because of zipline and like you hear stories all the time and so like you know people at zipline they'll be in like cabs and they'll be talking to someone they're like oh yeah my sister my mom my son my grandpa whatever like there are a lot of stories i mean zipline now um you know these global public health institutions estimate that zipline i mean university of pennsylvania did this big study showing that we've been able to reduce the maternal mortality rate in rwanda by 51 percent In fact, the Minister of Health of Rwanda was just named to the Times 100 list this year for improvements to maternal mortality the country achieves.

49:37It's incredible progress. Do you know how many lives that's saved? Many thousands, yeah. I mean, these other big studies estimated that Zipline is saving about 17 ,000 lives a year. Now we're about to triple or quadruple the scope and scale of our Platform 1 infrastructure in the world as part of this big contract with the U.S. State Department. and so that'll go from 17 ,000 lives a year to 60 ,000 lives a year. That's definitely a huge part of why we do what we do and it's why like logistics, I mean people often think of logistics as like a luxury for the rich and the reality is it is much more than that.

50:16It is a necessity for the poor and instant delivery particularly is like a necessity. like the less time you have in your life to you know drive yourself to somewhere or that you have access to like get to that good hospital or get to that place or you know show up for that appointment you know being able to meet people where they are is incredibly important from an from an economic perspective and from a health care perspective so anyway as we've expanded to all these countries, like, I think that one observation is that obviously that's really, really kind of critical for the brand of the company and for like community acceptance where we operate.

50:58If you ask people, a lot of times they'll say like, oh yeah, like our drones save lives. Our drones deliver these blood transfusions. So I think there's like an extremely strong sense of local ownership. That's because Zipline hired entirely local teams. We created all these really high paying jobs of like, you know, flight operators and fulfillment operators who ran the distribution centers. And then we promoted those people into positions of global leadership and global executive leadership over time as they were like, you know, prove themselves. And, and so I think there's this very strong sense that this is not like an American company or, or a African company.

51:35It's, it's a global company. Yeah. And anybody in the company can rise super fast if they are extraordinarily talented. And if you're in San Francisco, I think the perhaps sad reality is that the places that have like the most access to technology are the places that appreciate it the least. And in fact, are often the places that have these really strong like NIMBY or kind of like anti-tech, anti-progress strains and ideologies. you can see like the coning of autonomous cars in some of these cities and like you know there's this i mean waymo just hit a cat right they're like ban waymo now yeah there's this like big new york times article about how like they should ban waymo and obviously only at the very end of the article do they mention that waymo is not is is 30 times less likely to hit a cat than a human driver you know so it's like do we care about the cats guys or is something else going on here I think the reality is that like some of these cities have had have had it so good and get so much tax revenue from technology companies and from this kind of like technology boom that they actually have become pretty negative on it.

52:48And maybe just think that like no matter how hard they try to strangle the golden goose, like it'll always be there. But like the reality is 99 percent of the world isn't that they're dying for innovation, dying for a technology. startup to show up and help them solve problems in a really innovative way. They are dying for entrepreneurship and technology and high paying jobs. And there are entrepreneurs in all of these countries and in all these other parts of the US that are dying for an opportunity to work on something really cool. You have only to open the door a single inch and these kinds of people will do the rest.

53:27And I think that that's kind of been the story of Zipline. It's like we went to parts of the world that people, you know, ignore. And we did that outside the U.S. And then we even did that inside the U.S. You know, when we launched in the U.S., we launched in Bentonville, Arkansas, and then we launched in Dallas. And we're going to announce a lot of other metros that people don't think of as like, you know, top technology metros. You would think that like San Francisco is the top place, but then, you know. San Francisco, New York, Boston, you know. So anyway, I think that is the cool thing about, I think a lot of this next generation AI and robotics technology, it's like a lot of this stuff requires operating in the real world.

54:05And when you have to operate in the real world, you actually have to do it in partnership with government. You have to do it in partnership with regulators. You have to do it in partnership with cities. And that is going to lead, I think, more companies to do what Zipline did and go and find government partners that are as aggressive and startup oriented as they are and really want to create change and move super fast. They exist. In fact, they might be the majority. It's just that they're not the metros that most people think about when they think about advanced technology. Yeah. If you had to go back and you were given the same set of cards, but you have the knowledge that you have today on how you built a business and survived, what would you have done differently?

54:44It's hard to know the counterfactual. I think ZipLine got incredibly lucky at a lot of key moments. I you know always like the biggest mistake you make is is around hiring like I think probably everybody has to learn these hard lessons for themselves like you can hear it but you have to eventually go make the mistake but I think that the biggest mistake that I made on hiring again and again was that I would fall in love with specific experience over innate characteristics and whenever we hired for specific experience over innate characteristics it was a huge mistake Like the reality is like a startup that's moving super fast and rapidly iterating.

55:23Like you'll hire someone and then they'll have a role and have a job description. But, you know, two months later, it's like, oh, my God, everything's changed. New direction, new strategy. Now we need you to work on this completely different thing. If you hired someone because they were like just super, super deep in this one thing, they're instantly going to be irrelevant as soon as the company is like changing. And the company is changing direction so often. whereas we realized that there were these like four innate characteristics of people who when we hired them they practically couldn't wait it's like we couldn't get it wrong like these people would always end up winning at zipline and these four innate characteristics were practical problem solver fast learner low ego mission driven those were things that we felt like were very hard to teach someone like we didn't really think we could teach it to people but if we could hire people who had those four innate characteristics, you know, they might fail a couple times while they were figuring out how to do the thing, but eventually they would succeed.

56:20And that was, um, it's crazy how many times I had to make that mistake. What were the like most memorable and how did you kind of like shift that over time towards just focusing on these four things? I mean, every single bad hire you make is extremely memorable. Um, firing people is no fun. And also, you know, when you're trying to lead a startup, you're having to constantly ask everybody to believe in you. You're like, okay, everybody, I have this great idea. We're going to hire this person, and they're going to be absolutely amazing for all these reasons. And so they're going to lead this team, and these four people are going to report to them, and it's all your credibility.

56:59And so then three months later or six months later, you're like, all right, guys, bad news. That didn't work out at all. I was a complete idiot, but now I've learned my lesson. Now we're going to do this new thing. I mean, actually leading a startup is kind of shameless. You're You're making all these stupid mistakes and having to like constantly ask the team to, okay, now you should believe in me because now I have the right idea. And now we're going to go do this other thing. And, um, yeah, I think it is, it is kind of shameless in that way. Uh, but I do love this line of, uh, I think I forgot exactly where it's, I don't think it's from startups, but it's like, um, winning is just going from loss to loss or failure to failure without a loss of enthusiasm.

57:38maybe uh yeah i mean there was a lot of losing and we hired a lot of people who didn't work out and i think that you know to me those are all like the if you if i look at the mistakes from like oh man you know the things i could have done differently it's all around like how could we have bet bigger on the people who have wound up you know being at zipline for a decade and have grown by leaps and bounds and become executives in the company how could we of you know you want to hire as many people like that as you possibly can and then um just treating hiring like a sacred duty and like no amount of time spent with a candidate is too much time because the cost of getting the decision wrong is so is so high so i think the leverage of getting it right is so high yeah and and so those are the biggest like when i look back those are always like the biggest things it's like could you move faster to build the team but i don't know maybe everybody has to learn those lessons for themselves.

58:34How are you kind of deciding when the right time to scale up is, whereas, you know, just focusing on the thing that you're currently doing and getting it absolutely right is the best focus? I think that largely we've just like listened to the market, you know, for five or six years, it was kind of clear to us that the U.S. probably wasn't open to us from a regulatory perspective. We just needed to prove what we were doing. We were expanding all of our kind of like healthcare logistics business outside the U.S. And we had our hands full. But in 2020, I think it started to feel like, okay, we could see three or four years into the future, never more than that.

59:19But like three or four years, it felt like automated logistics was going to dramatically accelerate. And we felt like if we could build a product that would really compete against cars in every way, be dramatically better, that we had enough regulatory progress that we could get regulatory permission. This would work in every country, including in the U.S. I think we kind of had a good enough understanding of the problems. We also had these customers telling us what the problems were. We had big partners like Walmart willing to sign big contracts with Zipline to provide this kind of technology.

59:52and so you know at every stage you're kind of like 10xing or 100xing and you're having to build hardware that is dramatically more cost effective dramatically more reliable dramatically higher performance like it's usually every product is much more complicated than the last um you can look at like spacex going from falcon 1 to falcon 9 to starlink and then starship um like we platform two is like 10 times as complicated as the original version platform one of the technology that we built and uh yeah as i mentioned like it almost killed us trying to get that product into the real world but the thing was like we knew that the market on the other side was like incredibly vast it's like one of the biggest markets on earth already instant delivery and we actually think that if you can make it well actually we now have data showing that if you make it 10 times as fast and less expensive and way more reliable and safe for customers to use, they'll consume like 10 times as much of it.

1:00:56So we actually believe that the instant delivery market that you see in the U.S. today, for example, five and a half billion instant deliveries being made every year. It's actually very obvious that there's demand for 50 billion instant deliveries. And there is definitely not enough roads, not enough cars, There's not enough humans who want to do that job in the United States to deliver 50 billion deliveries a year. But that's where the world is going. And so we need to do it using technology that actually takes cars off the road, that gives humans high-paying, next-gen technology jobs rather than low-wage minimum jobs that people basically, as soon as they can get any other job, they abandon it and take it.

1:01:40And we got to do it in a way that's quiet and serene and beautiful for neighborhoods. and like reduces pollution in our neighborhoods. Yeah. And so that's the vision. And I think that it's only really this year that it's become clear. I mean, looking at the way that Zipline has been scaling in the US, it's only now that it's suddenly clear that that's all gonna happen over the next, you know, three to five years. What was kind of the day where you woke up and you were like, this is gonna work? I don't know, last week. Last week. Maybe I haven't had that day yet, honestly. You live in a state of constant paranoia And, you know, I don't think I ever tell myself, like, it's definitely going to work.

1:02:17But I definitely think there have been eyebrow raising moments over the last couple of weeks. You know, we are hitting like new record numbers of deliveries practically every day. And if you just look at like platform two deliveries in the US, since about April, we've been growing about 15 % week over week. And so when you look at that on a graph, I mean, it is like super vertical. Like we're in a completely vertical part of the graph. And it made me, it makes you realize when you're getting into like thousands of deliveries a day, 10 ,000s of deliveries a day, it's like, this is, even though it's still small scale compared to UPS or FedEx, this is something that is going to grow a lot.

1:02:58And it is gonna be completely normal for our neighborhoods. I mean, I went and visited a grandma in Bentonville. she's 82 years old she's ordered from zipline 350 times in the last year uh you know there are the service right now has a net promoter score of 95 which i don't know if i've ever heard of it yeah i mean normal delivery platforms actually have really bad promoter scores of like negative negative 10 negative 20 um yeah a lot of these delivery platforms have scaled really really fast with negative with net detractor customer bases kind of an amazing thing and i think it's honestly a testament to the opportunity and the scale of the market, like how much people want this.

1:03:39It's that they're willing to use a service that's actually super expensive and pretty unreliable and sometimes unsafe. And where the experience of the product is usually bad, where the food is ice cold and gross and doesn't taste nearly as good. But even so, these platforms have scaled to extraordinary scale. I mean, it's definitely a testament to the operational know-how and performance of those delivery platforms. But ultimately, it seems very obvious that these systems need to be automated. We should not be using technology that's 100 years old to solve a problem that a brand new industry that we've only really understood in the last five or 10 years.

1:04:24Basically, to solve these problems of instant and delivery to solve these problems of global logistics, we think we should be using technology, AI and robotics are gonna transform that. And that's the sort of simple idea that Zipline has been pursuing for 12 years. And when we are now seeing customers adopt the service, adopt the, and ordering everything through the Zipline app, it's kind of crazy to see the user behaviors. I mean, people are, you know, people are ordering at their offices. People are ordering at universities. We deliver to apartments. We would deliver to townhomes, deliver to hotels.

1:05:02We've been delivered to public parks. And I think people realizing that you can just pull out your phone and press a couple of delightful experience. Have something delivered quietly and precisely and safely 10 to 15 minutes later, no matter what your GPS coordinates are, um, is a, it definitely changes the way that people live. Like people change their grocery shopping behaviors. And we talked to families who are like, oh yeah, now we grocery shop like once a week and then we do three to four fill-in orders with Zipline a week. And so it enables them to like eat healthier food. You know, they can, yeah, they can eat fresher food.

1:05:37They can, they waste less because they don't have to batch process and like plan ahead. And most importantly, it's like saving these families three to five hours a week that they don't have to be battling in traffic, parking in a parking lot, dragging kids, kicking and screaming into a store to get something. It's like everything is delivered to you. So it's like three to five hours that you have to just enjoy time with your kids or loved ones rather than fight in traffic or in a store.