Planet Labs PBC (PL) Earnings Call Transcript & Summary

[Audio Gap] to be here. Looking forward to our conversation.

Absolutely. Robert, typically, we've just given companies a chance at the start to kick it off with a bit of an overview of trends that they're seeing, things that they're excited about. I'll give you the floor.

Sure. No, once again, happy to be here. Thanks for having us. I'm looking forward to the chat. Planet overview, pushing almost 15 years now as a company, the last 5, obviously, as a public company. I actually joined the company after spending almost 4 decades inside the U.S. government, working inside the Intelligence Committee focused generally in this area of geospatial information and remote sensing. Joined the company in 2021 in the position -- and by the way, to be really specific, what I chair is the Planet Federal Board, which is a subsidiary under Planet PBC. So think of it that way. But I do work with the C-suite on broad strategy, which I'm happy to talk about here. So 5 years as a public company, and I just coincidentally, as a calendar just was with the Planet team over in Munich for the security conference last weekend. And it's a good way to kind of reflect where we are in the movie. So we've -- Planet's core value and differentiated value is our daily scan. We scan the whole earth land mass once per day, and now I'm doing that for about 8 years. So think of thousands of images that are now stacked up literally over every part of the world. And what's most exciting when you ask about trends is that the compute and the computation that we're all living through and experiencing with artificial intelligence is now unlocking that latent value that, frankly, has been difficult to tease out over the years. And we're able now to obviously offer what we think is compelling imagery data sets and what I'd call the raw materials on the left side of your process to going to the other end and offering subscription services and analytics services and alerting services. And we can talk more about the countries that we're talking to and that we've made good progress in. But as somebody who, again, spent 4 decades inside the U.S. government, as awesome as we are in the U.S. government, 4 decades is a lot. And so there's a lot of baggage in there. And when you come to the U.S. government, and by the way, they're a great customer, too, great partner, it sometimes it's difficult to bring a new value proposition to such a legacy organization. But when you go to a country that's just starting, they have a cleaner sheet of paper. Oh, my goodness, that's a sweet spot for us because you can really kind of build a baseline of commercial first and then add government capabilities on top. U.S. has been reversed, right? Government first, commercial on top. I would say it's somewhere in this movie of transition. I wish it would go faster, both in my former government role, my current industry role, but it's still in transition. So very excited about those analytic services. We might want to talk to about the Constellation deals that we've been doing in the past 15 months. That's selling satellites as a service, which has been, again, a very exciting new area for us.

Yes. There's a few things there I want to follow up on. The first thing is the uniqueness of the data, right? Can you talk a bit more about that? You've been -- you've got almost a decade, right, of data, doing it daily. It's very powerful. It's very rich. As you mentioned, there are technologies that are emerging that are helping you mine that data and provide a real kind of value and insights. Maybe you can just talk about how special that is.

Sure. So I became the NGA Director, National Geospatial-Intelligence Agency in 2014 in the fall. I should get the date, but around the same time, Will Marshall, Co-Founder and current CEO of Planet Labs, did a TED Talk. I think it's -- and I didn't know Will at the time. I never heard of Planet Labs at the time. But basically, his TED Talk kind of previewed your question. Why would you want to image the whole world, right? But Will basically went out and said, I'm going to build this company and we're going to figure out how to image the whole world once a day. And look, I get it. We're a rather small community, but my head exploded. I said, you can't do that. I mean, we've never done that. I mean that's just not possible. And -- but I was intrigued by -- and because, again, that was a private company with private capital, taking a risk and innovating to meet this objective. And I think it took Planet 3 or 4 years from that TED Talk to be able -- they called it mission 1, right, to do it. And look, in some ways -- and by the way, even today, I run into people to say, yes, but that resolution, the spatial resolution of that is only 3 or 4 meters, so which is 10 feet to 12 feet. It's pretty gross resolution. So I'm not interested in that. But to me, while I understand that because I used to live in the spatial community and everything was about how exquisite of an image you can take, how detailed of an image you could take, that the differentiated value of the daily scan was the temporal resolution. It's Monday, Tuesday, Wednesday, January, February, March, 2016, 2017, 2018. And therein lies insights that just don't exist even at very high resolution because you see and sense trends over time, patterns of activity that just aren't detectable when you're staring in a very high resolution mode at a target. So one way to think about it is, as humans, we have 2 kinds of vision, right? We have our staring vision. You and I are communicating to the other. So you're giving me body language and signals. I'm waving my hands at you to try to convince you to believe me. But at the same time, we have peripheral vision. So that if somebody decided to throw a bottle of water, hey, I don't like what Cardillo is saying, I'm all detected at some point to do this, right? I'll turn my head, I'll be alerted and I'll put my arm up to defend myself. So think of the daily scan kind of like that. While you're doing your daily life, think of us as your peripheral vision. And not more than just general peripheral vision, you set the dial to what you're interested in, maybe it's ship traffic, maybe it's rail traffic, maybe it's military activity. But you could set that dial. And the only thing that will -- I mean, obviously, you're the customer, you can decide what you want, but you could only bother me when one of my dials gets a hit. If nothing gets hit, I don't want to hear from you because then I know nothing is -- you know what I mean, it's the absence of information, which is -- or activity, which is necessary. So I know that was a long answer to your short question, but to me, that really is the difference. And again, to reiterate, no one else is doing this, which is, again, something that's very differentiated.

Yes, in and of itself. And I think ultimately, what's special is that you've wrapped a very scalable business model on top of this. And a separate question is debating the TAM, which we can get to, but talk about how easy it is to scale now that you have all this data.

And again, redirect me if I'm missing the point of your question. But when I think about where we're going vis-a-vis scale is it's less about the data itself and more now about the applications.

Mining it, yes.

That's right. And so Will Marshall has been public with our partnership with Anthropic and experimentation that we're doing with large language models, which, as we all know, are large multimodal models now. Most of us has taken the picture of our fridge contents and ask for help getting, well, you got to be able to interpret pixels to give you answers to that. Now satellite imagery isn't exactly the same, but the fundamentals are the same. And so I mentioned earlier unlocking the value, but really what it's doing is it's kind of elevating, and to your question, it's scaling potential users that we just couldn't before. Again, you're looking at somebody, yes, I'm in my fourth decade of this. So I've been around for a long time, but it was an exquisite profession when I joined it, right? I went to school for 9 months, 9 months, now I went to an Air Force base to do it, right? And it was all classified. But they would not let me do anything for 9 months until I got -- and then when I got trained, I was given the very basic stuff to do because you don't want to make a mistake at that. So think of a very tailored bespoke profession. Now fast forward 4 decades, I won't say it's quite as easy as a ChatGPT query, but we're moving in...

It's getting there.

Yes, it's getting there. And so to me, the -- what's exciting are the scale applications that just literally weren't possible even a year ago, are now possible today.

Yes. And maybe you can elaborate on some of those services that are possible to apply to this data set?

So I'll start with 2 big broad areas, and we can unpack those, if you'd like. Planet provides a maritime domain awareness service today, which is exactly what it sounds like. So for example, kind of our largest contract in this area is with the U.S. Navy, who you might imagine is a pretty demanding maritime customer. And the region they've asked us to monitor for them is the South China Sea. You might imagine that, that's a pretty challenging area to monitor. And so we were very proud when we won their contract, their competitively bid contract about 3 years ago, Actually, Maxar was the incumbent at the time. The reason I think we became compelling and ultimately displaced them was, frankly, one, the daily scan that I've been talking about was over the land. We weren't monitoring open ocean because, quite frankly, there wasn't much business in open ocean. And by the way, we're still not monitoring all the oceans. But South China Sea, Mediterranean, Black Sea, Baltic, Gulf of Mexico/America, I'm not sure what we're calling it today, are very interesting areas. So we were adding that capacity, right, through our scanning service. And in this case, we found the right partner, a company called SynMax, a privately held company that specializes in detecting ships from our data sets. So they brought the algorithm, right? We brought the data set, and then we took that to the government and they said, "Oh, that's the solution I want." So that's one bucket of solutions. And the other one is what we call it global monitoring service. So think of what I just described at sea on land. And the one that we've been quite public about and proud about is a pilot that we're doing for the Indo-Pacific Command headquartered in Hawaii. That is the U.S. geographic command responsible for China and North Korea and lots of other trouble spots, but China mainly. So we've been now since last summer, so we're about 9 months into a pilot we've been doing for the command. Think of it as a China monitoring service. Now no, we're not monitoring every site in China every day. We're monitoring the sites that INDOPACOM has told us we want on the dial. So let's face it. One of the major issues vis-a-vis a potential Chinese threat is to Taiwan. Obviously, it's in our interest to keep the Commander of INDOPACOM as alert as possible to any change in Chinese disposition. And Admiral Paparo is the current Commander, and he's been public about saying that he has lost what he calls traditional indication and warning, which means there are signals that military commanders understand about other militaries. And you can imagine kind of the classic ones, multiple aircraft on runways and activity in ports that would transport marines and supplies. Well, the way China is leveraging its ability to mask those, it's done it by kind of elevating the noise level. They're flying a lot. There's a lot of maritime traffic. And so they're trying to kind of keep the noise level up so that should they make a decision to invade, it would be very hard to detect. So what we're trying to do for the Admiral is move what we call left on the schedule. I get it. We're probably not going to get the key indicator at the port or the airfield, but maybe at the place that supplies the port and the airfield, we'll see increased petroleum production or even medical preparations or logistics. You can imagine...

Unusual activity.

Right. And so that's what we're piloting now for the Admiral. We're quite proud of where it's going. Obviously, a pilot is a pilot, and we've got to prove it, and we're obviously learning a lot, too. So those are the 2 big use cases. But let me just give you one more because it's a little different. We have a contract with the country of Brazil. They have an issue with deforestation and not just deforestation, but illegal activity in the forested regions of Brazil, which are quite large and quite remote. We do a daily scan. We provide the results of the daily scan. And one of the things -- obviously, they're looking for tree clearing or cutting, that would be an obvious potential misuse of the land. And obviously, Brazil knows what is supposed to happen with their forest. But the other one is just airstrips that get developed quite quickly, usually associated with counter narcotic or weapons trafficking or human trafficking, et cetera. It's something that obviously, somebody is trying to hide. So when that gets indicated, right, and when they get that alert, they send the right people and maybe they send a drone ahead of time. You know what I mean, maybe they go to -- or they deploy a petrol unit to go look. And I don't have the stats in front of me, but we've done events with our Brazilian customer where they've talked about the reduction in illegal drug traffic and the reduction in deforestation that they've been able to do from that kind of scanning service.

Yes. And as you demonstrate more and more use cases, it does feel like we're hitting a bit of like an S curve, an inflection here. Award activity has been higher, right? Bookings are improving, et cetera. Talk to us about how important it is to show multiple different services, multiple different angles and how all that's combining into a broader trajectory.

Let me answer your question this way. So I'm going to generalize to make my point, and if you want to dive into any of these, we can. But if you think about the early days of Planet and its offerings, predominantly it was an imagery company, had imagery as a commodity. You are a buyer of imagery. Let's see if we can meet somewhere on this cost chart where you will pay this much for that imagery. We shake hands. I task the satellite, I download the imagery and I send it to you, right? Now by the way, we love those customers. We loved them then, we love them today. The NRO, the National Reconnaissance Office is a very large customer of that kind of model, right, that kind of service. Then I think the second set, and this is a newer set is the people that say, "Yes, look, I don't have an imagery factory to use your inputs. I just want the answers, right, or the outputs, right." So those are those dials that I set. And so we're seeing more and more people come to us and say, I appreciate the imagery, but just send me the analytics. So that could be a subscription, right? Or it could be the maritime service or the GMS service I described before. That's clearly growing. I talked about kind of the weight of history in the U.S. model. When we go and talk to companies at the Munich Security Conference that don't have all that, they tend to go here, or they go to the third category that I'll talk about, which is what we broadly call constellations as a service. So Planet -- by the way, that global scan is accomplished through satellites that we call Doves and SuperDoves. There's about 180 of those in orbit at any one time to do the -- to get the whole world. We have a smaller number of SkySats, which are a higher resolution, I think, 40, 50-centimeter resolution that can be pointed and tasked at a specific area. Those are being replaced with Pelicans. We like bird names at Planet. And Pelicans fly lower in real life. So we're going to fly lower to get better resolution. So we're going to get 30-centimeter resolution from our Pelican fleet, which has 6 up now, but will grow to satisfy the market. And by the way, we've also launched a hyperspectral sensor, which is more a bespoke capability to really go after the methane detection and kind of climate mitigation issues around that. And we have announced that we're going to take the Doves and build an Owl constellation, again, back to the birds. And -- but the big difference there will be 3 to 4 meters to 1 meter. So much higher resolution, but still the daily scan. But back to my point about satellites as a service. Up until 14 months ago, the only way you would have gotten access is either buy our imagery upfront or buy our analytic services here. Now we offer this satellite as a service. And we've announced 3 of those deals. Japan was first 14 months ago, Germany 6 or so months ago and Sweden just 1 month ago. They're all very different. Well, they're all bespoke to what the country. But basically, what you should understand is what we've -- the need that we've gone after is those countries interest in having assured collection, assured capacity, guaranteed. So essential ownership. Now again, each model is a little different. I can tell you a little bit more about the Japanese model. We've talked more about that. We haven't talked as much about the last 2, the details. But in the Japanese model, it's going to be 10 Pelicans that they will own when they're over their ground stations. So think of it as fractional ownership, timeshare, if you will, of the whole orbit, but they get the part that sits over Japan, which is a large part of the world, but when it flies out of their area, it comes back to Planet to commercialize and monetize. So it's a win-win. They get their dedicated capacity where and when they want it. We obviously get revenue and returns on that. And then we also get to monetize the rest of the orbit. So it's a growing model. When we had our Investor Day, Analyst Day in October, we talked about 20-ish, I think we actually used the number 20, so let's just say 20 that were kind of in the broad pipeline to head to a potential satellites as a service. That included Sweden. Of course, we didn't announce that because it wasn't done. So think of 20 minus 1 now out there that we're pursuing in that third category.

And the pipeline is still robust? Or is it growing? Or how would you characterize it at a high level?

If you haven't noticed, the world's a little crazier than it was. It always seems to be, but it's especially crazy this year. Again, I was in Munich. Europe, predominantly because of the Munich conference is quite nervous right now. They had made a long-term bet on cheap U.S. security. It turns out that bet is not paying off right now. As a matter of fact, it's reversing. The U.S. administration has told Europe time to pay up. And again, I'm not -- we're not going away, but we're looking for you to step up here. Planet comes in and says, I understand your anxiety. I understand that you've got issues with your national budgets and parliamentary votes and difficult questions about how you're going to raise your investment here. As you think about owning more of your own security, you're going to think about long-term production and manufacturing and talent and development, all these things that you would need to create a defense ecosystem, which probably will include eventually tanks and planes and submarines and ships. But everything I just mentioned is a multiyear, minimum of 5, but more likely 10-year proposition. You're just not going to build a fighter jet on your own. However, if you want to increase your local awareness and your understanding of the threat and your regional activity, have I got a solution for you? And I'll just use the German example, the time delta between when they signed -- we signed that contract and they had access to one of our Pelicans, dedicated access was 2 months. So we can move that quickly. And the reason we can is because of the proud history of Planet, which is agile, innovative, economically efficient development of satellites, and we're proud of being able to do that. And we have an active factory in San Francisco. By the way, we love showing it off if you're ever -- and by the way, I mean downtown San Francisco, Harrison Street, so really San Francisco. But we're also proud we just announced the intention to build a satellite factory in Germany. We'll build Pelicans there. Again, those details are being working out, but lots of good reasons to do that. We need more capacity. We need to spread our manufacturing risk, and let's face it, it helps us, quite frankly, in Europe. I mean we had and have a large team in Berlin, about 300 Planet employees, but this obviously helps our local investment and local recognition, if you will.

Yes. Changing gears a little bit, but building on the same themes. Can we talk about how AI is being leveraged or could be leveraged in the business more and more? And what services or opportunities that might unlock?

It's terribly exciting. We've talked publicly about our partnership with Anthoropic and the Claude tool. We've got research teams dedicated to answering your question what can happen. Will Marshall has been public about how impressed he has been. And again, Will is an astrophysicist, has very advanced understanding, obviously, of satellite technology, but also AI, about how good Claude was, is out of the box. So no fine-tuning, right? Just take the core Claude AI tool, load up some Dove imagery and say, tell me what you're seeing that's interesting. I mean, you could be more specific, interesting around Chinese military facilities in January. Claude will come back with an interesting answer, but just like your experience kind of get some things right, kind of get some things wrong. You go back, you just said, "Hey, I like the first part of your answer. Not interested in the second part, please go back and rerun it." Now we're not ready today to put that out, download the app, Claude, Planet app kind of thing. However, to me, it's clear that the development of those models, which obviously are multimodal now, is going to accelerate the unlocking of what the value that has been latent, hard to tease out of our archive. So I'm very optimistic about this. Please don't ask me how we'll market it, how we'll price it. I have no idea. They don't tend to ask me those questions. But what excites me, though, is, again, just as a personal user that I've -- when I've interacted with kind of the prototypes, it just -- it has that same kind of like, oh, my goodness, and look, I'm a lifelong professional here. I've been very pleasantly surprised with it just -- it lowers many barriers to admission here. And so it feels like to me like we could broaden use cases and users, I think, in a very exciting way.

At a high level, though, it seems fair that a year from now or make up a date in the reasonable future, we're looking back and saying like, all right, Planet Labs really unlocked some interesting services, interesting technologies that were AI driven. Is that fair?

I think so. Yes.

And marketed it to customers.

Right. And again, I just don't know how to do that. Look, I mean, again, I'm not Will, but I've been around him and Robbie and another co-founder, they formed the company to be that, right? The original -- I think the original tagline for the company was use space to help life on earth, right? Just make life better by using space, right? And make access to space cheaper and more -- and make the world more transparent and all these things. And by the way, that is still the DNA of the company. Have we shifted or added focus to defense and intel over the past few years? Of course, we have. And you've seen those results as we report out quarterly quarter. Let's face it, the budgets are there, right? And we're obviously competing for those budgets. But yes, I do think that there will be a broader democratization, which again, goes to that core value proposition. But again, I'm glad we have smarter people in the company that can figure out how do you do that while still protecting your differentiation and making sure that there is a monetizationable, I make up words now, process.

Yes, sure, how to commercialize it. My point is just that there is urgency to deploy AI and create real products around it.

There is definitely urgency to do so. Yes.

Got it. Okay. Can we just check through a few of the strategic highlights from the last quarter. And I was just hoping you could elaborate? So what I'm talking about are things like the Owl constellation and just double clicking on that. Just update us on the progress there and what that unlocks?

So again, Owl is the follow-on to the Dove, SuperDove constellation. It will be the new daily scan, so 3.5 meters down to a meter. We intend to have a tech demo end of the year, early next year, so around a year from now. We haven't announced when we think the full constellation will be up. But if you look at our past, tech demos tend to -- when they work, right, tend to -- I mean, that's our bread and butter, right? Once we figure out the technology, we can ramp up production. So I'm optimistic that we will have an Owl constellation in a handful or less of years for sure. And look, as a professional, I'm very excited about what we're going to be able to sense and detect at a meter versus 3.5. We've gotten great reception from government customers about because they, too, understand what will be detectable. And look, we're also experimenting with onboard compute, right? We've announced the edge compute that we're doing with NVIDIA. So think of the algorithms that we're running on our desktop today running in space. And why would you want to do that? Because you could get to the answer sooner, right? Get the signal, run the algorithm, just send down the answer. And then the other thing we're working on is optically connecting our satellites to also move just data faster. So in a world in which what have you done for me in the last 3 minutes, it's better to be able to send you the answer in 2 minutes. And so we're setting up a series of links so that -- whereas today our data is down linked to a ground station. And by the way, we have a few dozen of those around the world. So there's plenty of them, but it still takes time to go from Norway to San Francisco and back to the customer. So with the mesh network, we could compress those time lines. That, too, will be part of the demonstration that we'll look with Owl so that Owl can be that mesh network and very timely responses to analytic detections.

Yes. And then the manufacturing facility in Germany, what opportunities does that unlock?

It's Pelicans to start. That's what we've announced. It's -- I mean, if you come to San Francisco, and I've already invited the audience, they're invited a second time, you'll see SuperDove production, you'll see Pelican production, and you might even see Tanager production. Have we talked about Tanager yet? That's our hyperspectral sensor, yes. So all done the same shop floor, right? There's just different corners of it or in different cubicles, et cetera. Berlin will be a Pelican factory in the beginning, but I think the market will decide whether or not it becomes more than that. What it -- let's face it, because of the geopolitics I mentioned earlier, as a U.S. company, we come with baggage because we're a proud U.S. company. We're a proud server, a partner of the U.S. government. But let's face it, when non-U.S. countries are thinking about how am I going to grow my own capability, they don't always think about, well, I'm going to go buy that from a U.S. company first, right? They would -- might want to buy locally or develop locally. So -- but we didn't wander into Germany yesterday, right? We've had almost 10 years of presence there. Now that was with a satellite command and control station and our 300 employees. So we're German already from that sense or European already. I think this is also going to help because we'll hire German engineers and metalworkers, et cetera, to populate the factory there in Berlin. And obviously, we want to be competitive going forward and we think this is a way to increase our competitiveness there.

Recognition that there's a lot of demand for services in Europe broadly.

That's correct.

Yes. And then Project Suncatcher, that was another thing that came up recently. Would you mind just elaborating on that?

So Suncatcher is our research and development agreement with Google, which, by the way, is a long-term partner and investor in Planet Labs. Actually, the SkySat, I haven't talked much about those. Those are our legacy -- they're still flying, but those are our legacy high-resolution satellites that Pelican is replacing. Those we bought from Google a few years back. They were called Skyboxes for them. So we've had a pretty tight relationship with Google. Suncatcher is very early days exploration of a potential compute and space proposition. And look, I'm the wrong guy to have the debate about how real this is and if it's real, how close it is. I appreciate there's vastly disparate views about compute and space. So you should think about Suncatcher this way, about us doing very early exploration of the core R&D that would be required at a very small scale. And I think a lot of the debate is about not the basic R&D. I think people agree about the engineering science. I think where people disagree is how can you scale that. So I guess our message is while we're excited about it -- well, we're obviously excited about our partnership with Google in general. While we're very interested in it, we'd recommend caution here, early days. Sundar, by the way, talked about a 10-year horizon kind of is when he thinks about it. I'm not going to question that. I'm just going to throw it out that there's somebody that's pretty thoughtful about delivering services like compute and space. But please don't think about it as a quarter-to-quarter kind of effort.

Yes. No, makes sense. And we have a few minutes left. I wanted to talk about M&A and the broader strategy there. You have had a couple of successes. Bedrock comes up, right? Maybe you could talk about the landscape as you see it and how you think about the M&A philosophy?

I'm just happy I got this far in our conversation without saying, I don't know a ton about that. Guess what? I don't know a ton about that. Now look, I met Bedrock. I'm very impressed with them. I'm glad they're on our team. But it's not -- while I do strategy, I don't do that part of the strategy. So I'm sorry.

Fair enough. Maybe I can open it up to the audience for any questions.

[Technical Difficulty] and sensing within space, the other things going on in the space. Are you going after that?

So non-earth imaging, which, by the way, I just have to say is exciting for me as a civil servant to say in public because for years, I couldn't talk about it, right? It was so sensitive, right? Nobody talked about it. We called it sat-squared in the days, satellite-to-satellite imaging. So this is just fun for me. It's an area that we have discussed various partnerships with companies that are pursuing this as kind of a primary delivery service. But I would describe it as more of R&D, test and evaluation versus something that I would put into those 3 big buckets. So exploring, willing to be surprised perhaps at what it might become, but not a key business area for sure for now. Yes, sir.

Two quick ones. So number one, you've got factories in -- a factory in San Francisco, you're building another one in Germany. I'm sorry, if I missed it. Like what quantum of satellites are we talking about here? How many do we need? Number one.

So since Planet was born, we've launched around 700. So it's a lot. It's the most earth observation satellites anyone's ever launched. Elon is launching a lot more comm satellites, we won't compete with that, but for earth observation. So -- and we launch about quarterly. I mean we launch, we subscribe to a Falcon 9 or a Rocket Lab. And every launch is obviously different, but I think our last launch had a Pelican and 36 SuperDoves. So we launched them in flocks. So that's -- those are the kind of numbers. Yes, I know it's a bird thing. Those are the kind of numbers that -- and again, if you come to Willy Wonka land in San Francisco, you'll see it's a busy place. And we can turn them out quite quickly. I mentioned the responsiveness to the German demand, 2 months. And the way we did it was basically they took a satellite off the line. So think of an automobile line. "Oh, I want that car." Well, it will be done next Thursday. You can have it next Thursday. Now we're not quite that fast, but very proud of that kind of volume. Germany, again, will be focused on Pelican to begin with, and then we'll see from there.

And are they principally Low Earth Orbit? Or...

Yes.

And then finally, it sounds like the level of redundancy around sort of denial of -- in a more -- do you have the redundancy in a sort of denial of...

Yes, space is a competitive environment. Our -- the Planet approach is proliferated architecture, right? Meaning not that we want people to attack our satellites or to jam them or to lase them or whatnot. But if it happens, we don't just have 1 or 2. We have 180. And again, we're not asking anybody to do anything to any of our satellites, please. But we understand that it is a risk. And just -- and I know we're over time, but just one of my early introductions to the Planet was one of the launch failures. This is back in 2014. Remember, I was just learning about them. It was a launch from Virginia, which is an interesting place to launch from anyway. It was an Antares rocket at the time, and it was at night. It was a dramatic explosion, right? Just one of these fire balls that go up. there was like 3 dozen Doves on the thing. Some of them landed on the beach, right, because that's as far as they got. And they started to communicate. They said, "Hey, I'm in space. I'm ready to start taking pictures." And we said, "No, you're not. You're here on the beach." What I loved about it was, and I can't remember the exact number, but 3 months later, I think it was October to December, some of the same Doves that either landed on the beach or maybe they stayed intact, were launched on a SpaceX rocket. Do you know what I mean? So to me, that's resilient, right? And that's 10 years ago. And so we're very proud of our ability to brush off the sand and let's go again.

Listen, that was fantastic. Robert, thank you for joining us. I really appreciate it, and thank you for the audience for being engaged.

Thanks. Appreciate it.