Invest Like the Best with Patrick O'Shaughnessy - Will Marshall - Indexing the Earth - [Invest Like the Best, EP. 251]

Episode Date: November 16, 2021

My guest today is Will Marshall, the co-founder and CEO of Planet. Will founded Planet in 2010 with a small team of NASA scientists to build a constellation of satellites that would image the entire Earth every day. Since then, Planet has successfully built and deployed 450 satellites into space, which the company is using to create a time series of images for every place on Earth.   Our conversation covers the untold space story. How space is going through an internet moment where cost reductions and performance enhancements have led to a seismic shift in what’s possible above our atmosphere, and how that can drastically improve life on Earth through unique datasets like the one Planet is piecing together.   Once you listen to Will speak about Planet’s progress and mission, it’s hard to think of a more underappreciated company in business today.   Please enjoy this great conversation with Will Marshall.   For the full show notes, transcript, and links to the best content to learn more, check out the episode page here.   ------   Invest Like the Best is a property of Colossus, LLC. For more episodes of Invest Like the Best, visit joincolossus.com/episodes.    Past guests include Tobi Lutke, Kevin Systrom, Mike Krieger, John Collison, Kat Cole, Marc Andreessen, Matthew Ball, Bill Gurley, Anu Hariharan, Ben Thompson, and many more.   Stay up to date on all our podcasts by signing up to Colossus Weekly, our quick dive every Sunday highlighting the top business and investing concepts from our podcasts and the best of what we read that week. Sign up here.   Follow us on Twitter: @patrick_oshag | @JoinColossus   Show Notes [00:02:57] - [First question] - His thoughts on the renaissance of the space industry [00:05:09] - The earliest days of Planet and why he started the business [00:09:22] - Unique data units captured by their satellites [00:13:35] - The real estate of space and interesting angles to consider  [00:15:59] - How customers interface with Planet and their early use cases [00:20:57] - Thoughts on the sovereignty of space and the laws that exist currently   [00:23:43] - Figuring out the dynamics and pricing of Planet’s business model [00:27:34] - Examples of stress and tensions when working in space [00:29:08] - The future of privacy and concerns we should have there collectively [00:30:29] - Five different types of satellites and their functions [00:31:39] - The most sci-fi potential futures that Planet may unlock someday [00:32:54] - Indexing the Earth and using data to train machine learning algorithms [00:34:02] - What he’s learned about Earth that is most surprising [00:37:12] - Contributing factors to a 70% decline in life on the planet in 40 years [00:38:35] - Ways that going public might impact Planet’s long term goals [00:40:23] - The hardware story of building various prototypes of satellites [00:42:18] - How much is built in house versus outsourced to fabricate their satellites [00:43:48] - Complimentary space trends that are compounding beyond imagery [00:45:32] - Whether or not they plan on making their data open-source [00:47:15] - Democratizing their data and allowing other companies to build on top of it [00:48:30] - The kindest thing anyone has ever done for him

This episode of Invest Like the Best is sponsored by Canalyst. Canalyst is the leading destination for public company data and analysis. Founded by a former byside analyst who encountered friction sourcing, building, and updating models, canalyst is now used by over 400 institutions, including the largest money managers globally and by a number of guests on the show. With detailed company-specific models and data on virtually every public company, panelists clients are able to ramp up faster, update models instantly, and incorporate the highest quality fundamental data into any workflow. If you're a professional equity investor and haven't talked

to Canalyst recently, you should give them a shout. Learn more and try Canalyst for yourself at canalist.com slash Patrick. That's C-A-N-A-L-Y-S-T dot com slash Patrick. This episode of Invest like the Best is brought to you by Watchbox. Whether you're looking for a special gift or something for yourself, at Watchbox, the world's finest watches are available at your fingertips. The growing selection at Watchbox features all the most renowned brands, plus the industry's most exciting independent watch companies, all certified authentic and collector quality. Watchbox's global team of expert client advisors can help you find the watch you've always wanted. Step into the collector's circle at the watchbox.com slash Patrick. Hello and welcome everyone. I'm Patrick O'Shaughnessy and this

is Invest Like the Best. This show is an open-ended exploration of markets, ideas, stories, and strategies that will help you better invest both your time and your money. Invest like the best is part of the Colossus family of podcasts, and you can access all our podcasts, including edited transcripts, show notes, and other resources to keep learning at join colossus.com. Patrick O'Shaughnessy is the CEO of O'Shaughnessy Asset Management. All opinions expressed by Patrick and podcast guests are solely their own opinions and do not reflect the opinion of O'Shauncee asset management.

This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. Clients of Oshonosi asset management may maintain positions and the securities discussed in this podcast. My guest today is Will Marshall, the co-founder and CEO of Planet. Will founded Planet in 2010 with a small team of NASA scientists to build a constellation of satellites that would image the entire Earth every day. Since then, Planet has successfully built and deployed 450 satellites into space, which the company is using to create a time series of images for every place on Earth. Our conference, conversation covers the untold space story, how space is going through an internet moment where

cost reductions and performance enhancements have led to a seismic ship in what's possible above our atmosphere and how that can drastically improve life on Earth through unique data sets like the one planet's piecing together. Once you listen to Will speak about planets' progress and mission, it's hard to think of a more underappreciated company in business today. Please enjoy this great conversation with Will Marshall. So Will, I think the place to begin, since this is such a cool, exciting area of space, the space renaissance that we're going to be talking about today and your company Planet Labs, maybe you could just give us your impression of what's happening in space. And I think it's going to

draw a nice contrast to the very obvious rocket focus, billionaire focus thing that ever, going to Mars, going to the moon. I think you're seeing a very different side of space and are a part of building that new industry and that new world. Just give us your overview, how you interpret what's happening in this space renaissance. It certainly is a space renaissance out of the gate. So there's no question, I agree with that statement. We are seeing massive shifts. Rockets have decreased cost about 4x. That's a big deal, mainly through reusability of, especially because of SpaceX. And then, and a proliferation of nanor rockets as well, which is kind of cool. But I think the other side of it is what's happening to satellites, which is we've seen about a thousand X cost

performance increase in satellites over the same period of the last five, 10 years. What the main upshot of both of those trends compounding, and they do compound, is the, that we are seeing vastly new data sets emerging. We are generating and transporting vastly more data, like 10 times more than we were before or more about the Earth and transmitting that around the Earth. And what's that relevant for? It's relevant for everything here on the Earth.

It's about data about the Earth that helps us to transition to a sustainable economy. Everything the countries are trying to do in missions, every company trying to measure ESG targets is relevant for that. And it's relevant for digital transformation, the general trends of AI and big data helping all these industries become more efficient, like agriculture, transportation, government, and so on. And a lot of that data for those digital transformations comes from space.

So space isn't this esoteric, distant thing with rockets, billionaires only, although that is happening. I think the far more exciting story and less sexy perhaps, because it doesn't involve the rockets or the billionaires, it involves. massive new data that is helping us here on the Earth. It's helping the Earth economy, it's helping us transition to a sustainable planet. And both of those things are super important. I think that's the untold story. I think one of the coolest trends in technology generally is making certain things legible to software through data sets, through image sets in your case,

et cetera, and the wild things that can then be built on or learned on top of those emergent data sets. And so I'd love to tell that whole story through the planet, lens in as much detail as we can, maybe going all the way back to the beginning, you have the benefit of literally starting, I think, the first of satellite was built in a garage. So the proverbial garage was a real one in your case. Take us back to the earliest days of Planet, its origin story, and why you started the business. Yeah, I'm happy to do that. And by the way, the investors of DMY, the group that we're merging with as part of going public, his name's Nicola. And he says, all the best software companies build their own hardware. Apple, you

can think of like that. Tesla, he considers a software company, but it's building his own hardware. There's a way of viewing that. I think certainly Apple's, I think, a really fantastic example. We feel that the same. We're not selling the satellite hardware. We're selling data that is produced from that. But by building our own satellite hardware, we great much efficiencies in that. Back to the garage days, myself and a small team, about seven of us left NASA to start Planet. and we literally started building the satellites in our garage. What we were thinking about at that time was, how do we as space geeks help all these challenges the world?

From poverty to climate change to everything that was going on. And we felt that there's a huge business opportunity as well in one and the same mission, which was to image the whole Earth every day. Because we thought that data would spur economic development and it would help with these real huge problems. And no one had ever done that because it required at least about 100 satellites into orbit. Well, no one had ever launched 100 satellites.

We'd have to do things very differently. We had been pioneering at NASA this thing called the small spacecraft office, which was low-cost planetary missions and other things. And we were trying to take that a couple of steps further and say, okay, can we even leverage consumer electronics, like the kind of electronics that's in your laptop or in your phone in space, and thereby, instead of spending billions of dollars per satellite, spend a lot less, like a thousand times less than that or 10,000 times less than that, like a million or $100,000, and do something like imaged a whole planet every day. We figured that we could, and so we left NASA to do that.

And six and a bit years later, we achieved that mission of having launched the largest constellation of satellites in history. We had just over 100 satellites in orbit. SpaceX, by the way, is just overtaken us recently. The bastards. Not that I'm too upset about that. By the way, if you count the numbers of satellites in orbit, they doubled roughly in the last five years, and half of the satellites are now of two companies.

That's planet and SpaceX. So although there's a proliferation, it's really quite concentrated at the present time. But anyway, yeah, so we started building us satellites in the garage, and six or seven years later, we got to our daily cadence, having launched all these satellites. It was much harder than I thought.

I thought we could do it in three years and it took six. But we achieved something that was really cool and we have this brand new data set of an image of the entire Earth every day. So it's a bit like when you go onto Google and you see the satellite layer, that image is maybe three or five years old. We're doing that every day for the whole planet. And we're keeping all of those images. So we have about 1,700 images for every point on the landmass of the Earth.

So a deep stack of images. it's like Google, but with a time access. So it's a recent data and a tall stack of information. And that's how we figure out what's changing on the planet, where all the resources are being moved, all the vehicles, agriculture, shipping. We figure all the changes on the panel over time, and we can train all on machine learning on top of that stack of data.

You ask any machine learning expert, like, what's the most exciting thing? It's training data. Well, we have gobs of it. We have 7,800 images, as I said, very proposed. on the Earth's landmass on average to train all your algorithms about what's going on. Sounds like the coolest thing ever if you were a data scientist getting your hands on this dataset. Tell me a little bit about the unit of data. So if I take one of those images, what is being captured by you or the satellites and what isn't being captured? And maybe

that's something that may change in the future. But give us a deep dive into the actual unit of data itself. Each pixel from our scan is about three by three meters. We collect them in 47 megapixel images. So each satellite takes an image, 47 megapixel. So it's around 20 or 25 kilometers across and 20 kilometers high. Frame image, a full frame image. We take about 3 million of those images per day to cover the entire landmass. In fact, we cover about just over 300 million square kilometers per day.

So that's the unit measure of sort of the area coverage. And the Earth's landmass is about 150 million square kilometers. So we cover all the lowest landmass and some ocean terrier. like Mediterranean, Caribbean and the South China Sea and a few other hotspots and coastal areas around all the landmass. And in some areas of the land we do multiple times, you have a fleet of higher resolution satellites that can have a resolution of about 50 centimetres, that three meters.

They don't scan the whole earth. They're tasked to specific locations. So if you say, where do you say you were in Connecticut? Connecticut. If you want to take a picture of downtown there, we can do that up to 10 or 12 times per day. but you have to tell me I want that picture. And actually these systems work together. So we have the scan that finds changes around the planet.

Oh, this ice boat move, these ships move, this whatever, this field's got tilled. Then you can use the high-res system to point at those changes. We can say, oh, there's a big new building construction here. Let's have a look. There's a lot more ships in this port. Why is that? Let's take a deep, to look, what sort of ships are they? So the system's working complement.

Really, I think, of three legs, the stool of planet, which is the sky. scan, the zoom in, and the analytics, the machine learning that enables us to understand it all. Is there a future where you start to measure more than like a visual image? So you start to measure infrared or I don't know what the other spectrums would be or the other information. But if so, what does that roadmap look like? In 15 years, what might we know daily about the Earth that we don't today? Generally, the answer to your question is the data is going to get better and better resolution.

That's the size of those pixels. Temporal resolution, that means more frequently. and better spectral resolution, which is more spectral bands, which is what you're getting at. Right now, our satellite fleet, the main one that does the scan of the whole Earth, which is 180 satellites, does that in eight spectral bands. So three that are familiar to the eye, the red, green, and blue.

But we also have five other spectral bands. So there's a near-infrared band, a red edge band, and they're for various different things. One helps us to tell biomass. So basically we can tell crop type and yield for agriculture, Another one that enables us to tell the top of the atmosphere haze level, which helps us to calibate all the other data and so on. Yeah, with time, we will increase that as well. So it's not just pretty pictures, actually.

You can't even see these things. We do to do false color images of them. But for example, that near-and-red band, as I said, we can tell it's wheat and it's doing this well. And that is how it gets into this agricultural application, which is actually our biggest market. We can tell crop-time yield in every farm's field, in every bit of the farmer field, that three-by-three-meter box. we can say how well is the crop doing? It doesn't need fertilizer. Does it need harvesting?

When does it need water? And so actually, yes, we're already getting other information. And then that's before we build up analytics on top of that, that allow other things that automatically detect certain things. Like we automatically, with machine learning, pull out roads, buildings, ships, planes. So you can now, if you're interested in shipping, you could circle the top 10 ports of the world. Just tell me how many ships there are in these ports over time.

time. That's all I want. I don't want to look at the pictures at all. You can now do that on our platform. Now, you can't do everything. You might ask some other questions we don't yet answer. We're trying to get better and better so that one can actually not have to have a PhD in geospatial science to understand this imagery, but rather actually everyone could get information and value out of the imagery every day. What is happening in space literally? Where are these things? Are they in low Earth orbit? Is it getting crowded up there? Give us a little bit of education on the literal real estate of space and the angles there that are interesting. If this is just going to keep compounding and it's going to get more and more higher resolution

and more satellites, what implications does that have? What's the story up there? These satellites are in low Earth orbit, about 400 kilometers up. And that's considered quite low. Most of the satellites are in 800 to 1,200 kilometers orbit altitude in low Earth orbit, and then there's the higher ones in geostation orbit, to mainly during communications. And a few others like Mio, which is sort of in the middle of like GPS. that you use every day for your navigation.

Our ones, we care a lot about the space debris problem that you're referring to. There is a challenge of space debris. It's mainly not the satellites. There's about 3,000 satellites in orbit today and operating. There's about 100 million pieces of debris. So the vast majority of it is bits of old satellites. We're not talking about asteroids. We're talking about man-made stuff circling the Earth,

but it's all the fragments of previous launches. We didn't know about space debris early on in the space industry. So the Russians and the Americans in particular did a lot of stuff that created a lot of junk in space, like exploding bolts on rockets or they leave fuel in the satellite. Some spark happens after the satellite's dead, long years afterwards, and it blows up the satellite accidentally. Instead of being one bit, there's now thousands and thousands and thousands of bit. And then occasionally, countries have blown up satellites deliberately.

recently the US did, China did, India did. And that's mainly to say to the other countries, hey, we can knock out your eyes and ears if we want to, really frustratingly from our perspective because it's like a bloody great mess. There are challenges, mainly not the satellites. It's the debris. So what do we do about the debris? We can maneuver the satellites out the way of each other.

No problem. What do we do about the debris, which don't have propulsion systems on that? We actually came up with a scheme when I was at NASA to deal with that using a ground-based laser. It's called a system called Lightforce. Anyway, it's a longer story, but we keep our satellites way down low to make sure that out of the way of debris, not creating any more debris. They come down naturally at the end of life after about three years and burn up in the atmosphere and get out of the way. Let's talk now about how customers interface with you. So I love the story of starting in the garage, building a very low-cost satellite,

the dove satellites, getting the constellation up there, mission accomplished. It's a going to keep compounding and getting higher and higher resolution. What was the first commercial intersection that you have? Did you plan to cater to the agriculture industry to start, or did that emerge as sort of like an unpredictable or unpredictable use case? We had thought about that use case. It was a very early one. One way to think about it is look at the areas, the earth and how are they used? About a quarter of the landmass of the earth, 25 percent, is agricultural land. So we went into that. A quarter of the landmass of the earth is forestry. So we went into that.

About 10% is urban development, a suburban or urban. Then there's called large areas of marine. Anyway, but agriculture was an obvious one just from a sheer area. We knew that we could do something useful in agriculture with a near-infrairate band. It's a known thing from Lansat, which has been doing this for 40 years, except it's just doing it in a slower cadence, which makes it less useful growing the crops and helping the crop development. And it's more like for the annual survey of how do we do, which is useful at the end of the year. But it's not helpful for the grow in the crop grain season.

Because what the farmer wants is intelligence about their field a few times a week during the crop growing season. And so by having daily information at that three by three meter area of every field of that farmer, every one of their fields, we can help them do what's called precision agriculture. And that can improve their crop yield by 20 to 40%. And at the same time, decrease their use of things like fertilizer and other resources by similar amounts. That's a big deal. If you think of that, we can do that across a trillion dollar industry. This is, as I said, the general notion of big data and AI enabling the digital transformation of industries. Ag is a canonical example.

We do that, for example, with Corteva, they use this to image about a million farmers' fields every day. This is not a minor operation. and a lot of those kind of companies that can use our data to help a precision act. The other areas, we help civil governments, normally things like disaster response. It's also science. So we work with NASA on a lot of science, like climate science. But we also work with state and local governments and federal governments on things like floods and fires. So we've been helping recently with the floods in Germany, where they had a lot of big floods,

the biggest floods there for 60 years, with the fires here in California. helping the wildfires. We can both detect where the fires are to help the firefighters, like where's the wind blowing, where's the edge of the fire, where is it compared with the whatever hill. And so we can help the real time and we can help with preventative work. So we can actually, and we have, map every tree in California and where's the fodder for growth for future megafires? And that can help them to then make clearing or do a fire lane in the future to stop future megafires. So it's preventative work. Same with flasers. Like we can see the flood the day before, the day after the flood,

see apples-a-wows, which bridges down to help the relief effort. But we can also then model where's the floodplain going to be? Which buildings are in the wrong place? Where should they build and not build? Especially in developing countries. They often build in the wrong place. So we have a big project with Google where we're doing this in Bangladesh and India. It's because the flood planes change.

They can then advise governments, don't build here, do build here. That's important for getting assets. out of the way of the flood. And it's also important for finance, by the way, because they also want to understand which assets are at risk from these various disasters or potential disasters.

So civil government is another area. We also work with mapping. That's a fascinating use case. And this is where it gets into what consumers might actually see, because we're not selling the imagery to consumers. We're selling to big businesses. But it does affect day-to-day people

because the maps that you have online kept up to date with our data, the news you see about world events. We're sort of shedding light about everything going on on the planet. Pulitzer Prize this year was won by some journalists who used our data to discover about 200 potential Uyghur, that's Muslim detention camps in Western China that had previously not been known. And that sort of thing is really important journalism that happens using our data to uncover

what's going on around the world. And also, if you've got up in a disaster, our data might be helping the disaster responders to help you. So it actually affects people day-to-day. But in mapping, that one case, we work with Google, for example, on updating the maps that you see online all the time, whenever they find any indication that map is going out of date, they automatically task on our stuff like,

it takes a picture of that location, extracts out that new road, that new train station, that new building, whatever it is, and then it updates the map you see online so that your directions and everything stay up to date. So there's a lot of different applications. I'm just giving you a sample. The story that's emerging is like this incredible thing where so much the world is dictated by legibility around data that we've surfaced. And this is like what's more important

than the place where we live. And it makes me start to wonder about big questions like defense or national sovereignty, the famous spy planes from the Cold War or something. Like, this is that every day of the entire earth. So how do you think about certain nations not wanting you to see certain things? What is sovereignty in space? How will the rule of law and nations interact in this new frontier? Good question. Firstly, it's interesting. The U.S. and Russia established early on in the space age

that they would not be able to fly planes over each other's territory. Famously, the Russians shot down that U-2, spy plane, Gary Powers, and Kelton captive. But they would allow satellites to fly over each other's territory. And partially, there's because satellites can't just maneuver around countries. If you're coming up to Russia and turn left, you're in a little. You're going to go over Russia, right? unlike the plane, you can turn around. So you're going to go over Russia.

And then the only question is, are you going to allow them to take a picture or not? Well, how are you going to stop them taking a picture? So then they were like, shit, this doesn't work. But we're going to allow satellites to fly over each other's territory without permission. In fact, they felt that this was a good thing, and they cemented it. And it became international law that up to 100 kilometers altitude is your territory, and beyond that is space. And once you're in space, you are not in anyone's sovereign territory.

and you're allowed to take pictures from that of any territory without their permission. So we can take pictures inside North Korea, and the North Korean government may not like that, but they have no choice. But vice versa, in principle, the North Koreans can put up satellites and take pictures of America, and they, in principle, can do that. They don't have that, but, like, they could. It's a bit of a thing that they decided early on in the space age was going to be something that everyone could do, and that transparency was better.

And in fact, it reduced tensions between the Cold War. We're actually doing a whole bunch of work with government's intelligence agencies that do use our data and you think, oh, well, they already have satellites. Well, they do, yes, but they don't have nearly as many as we do. We have the most by far. And so we see a lot of stuff that they don't yet. We do the scan and they only task at locations they already know to look, which are very interesting to them.

We find new things that they didn't know to look for, and you missar base in Eastern Iran, we just recently found, for example, and that's kind of interesting to a lot of people. I think this is generally a good thing. We're not, by the way, ever exclusively giving that to one government or another, and the more that we give that to lots of governments, the better it is for peace and security, just like it was during the Cold War, the Russians and the Americans,

too much of the United States. But it's also a significant business opportunity for us because that data is relevant to a lot of countries, and they all want to know. Copy through a little bit the way you figured your way through a business model, because most businesses you can think about like cost-based pricing or value-based pricing, just to really oversimplify it.

Using the crop example, you're making 20-something percent improvements in a trillion-dollar industry. Well, that sounds like a lot of pricing power, right? How do you think about pricing your various products and services? It seems like a real challenge that there's enormous investment to get this ongoing thing into space and happening, which took six years, you said. Now you've got to figure out how to charge for these things. It's not like someone can spin up a competitor easily. So how do you think through those dynamics of the business model itself and things like pricing? Yes, so we basically sell imagery on a per unit area basis,

kilometers, hectares, mainly just area. And the more area you get, there is a volumetric discount, but obviously you pay more. So we do have a lot of pricing power. I agree. Right now there's no one else that does this. And to your point, it's very hard to get this data set. You have to erect a huge satellite fleet and ground stations,

mission control systems, data processing and all the rest. So it's certainly not for the faint of heart. And of course, so it take many years for somebody to build such a system. And of course, we're not going to sit on our hands. We're going to improve it. And that data archive, it's impossible to go back and get. So it's actually there's some really important notes around what we've built. But yes, we do have a fair bit of pricing power.

Some level it's value-based pricing. But I do think that what's most important for us is where we're going next is not just the imagery and selling the imagery, but selling information products derived from that imagery, what we call going up the stack. The main thing we're doing today and the reason we're going public, one of the myths, of course, we have capital to then deploy, and the main areas we're deploying it,

one is just sales and marketing to go after the vertical market, we already know work like agriculture, as civil government once I was describing. The second thing is that there's loads of potential to other markets, but they need more than images. Hedge funds could get huge value out of our data. I mean, we know how well crops are doing before anyone else. The whole world's soy year.

We know the output from all the world's copper mines before anyone else. So presumably those things are really valuable to those commodity prices. However, they do not want those pictures. They want a time series calibrated data. We have the data to underpin that, but we haven't built the analytics that enable us that yet. So the other side of what we're doing is investing in what we call going up the stack. And then we'll be charging more and more for just information, products derived from the data.

Anyway, we've had a long-winded answer to your question, but so mainly it's volumetric based on the amount of data, but also as we go up, the value stack will be charging more for those, if you like, smaller bits of data that we can derive from the imagery. How much of that is push versus pull in terms of deciding where to go with that movement up the stack? Are you basically letting people tell you what they want, or are you anticipating what they might want and then trying to sell it to them or some combination of the two? I think it's a healthy combination of the two. We pride ourselves on thinking more than just what the customers immediately are asking for, because often the biggest opportunities from business perspective are doing things that people don't know they want yet. No one asked for an iPad or an iPhone. When they got it, they were pretty excited. So they were thinking ahead. No one asked for a daily image of the whole earth, thinking ahead. The same with analytics, but we also listen to our customers. So, for example, a lot of our ag customers are saying, can you just fuse this data with this data? And then it would be more of

Well, we're doing that ourselves right now, but we don't really want to. We want to focus on the farming. Great. Sure. We'll do that. And we'll follow their lead and add that stuff so that it takes that off their plate and adds value to us. And it helps the customer eases their use and helps them to focus where they want to focus. What does tension feel like in this business? When you're stressed out, what are typically the reasons? Like it seems like the problems that you encounter are just of a completely different type than a typical business would, given the unique nature. of what you've built. What is crisis, stress, tension? What does that feel like for planet? I would say we're a company like any other. There's growing pains where you have to shift different stages of company where not everyone knows each other anymore and then this happens or you need this sort of leadership and not that kind of leadership and this. There's so many growing pains as we go. I think we've done fairly well, but we're not immune to those things just like any other companies.

I would say it's the nuts and bolts. There are some things that come up when a certain actors ask us for data and we're like, should we give this to them? And so we have an ethics committee that we firstly review against the embargoed list that we can't sell to North Korea, we can't sell to the Taliban or something. But also we have an ethics committee that even after that goes, okay, we're allowed to sell to these guys, but should we? Our data is not generally relevant to militaries targeting or anything like personal privacy because we can't identify a person. But there are some cases where we worry and avoid those use cases. In general, our data really helps this wide-scale transparency and helping companies and

countries understand resource movements that's really very positive. And we wouldn't be doing this. We didn't think the net was massively in the direction direction. But we are trying to occasionally, a stressful moment is when someone asks for it that we were like, I'm not sure we should give it to these people. And we have often refused in those cases. Data privacy has become a really interesting big issue in the world that's proliferated in data and so many companies that know so much about us. What do you think will happen in privacy over the next couple of decades when it comes to your business or just the idea that assuming all these resolution curves keep going on the same direction? Presumably, you're going to be able to see like into my house on a second by

second basis potentially. How do you think about the future of privacy and concerns that we should have there collectively? Firstly, right now really can't see who I identify a person from space. And I don't think anyone can, even with a huge satellites. And the reason is that you're just so far away. You're 400 kilometers away. So yes, we can see your house. But firstly, it's impossible with current technology to see through it. And I don't know how we would do that. And secondly, the pixel size is really dark.

Even as I said, the biggest spies, adenact to the biggest countries cannot identify a person from space. Their resolution is just not sufficient. And so it's really a matter of distance. So if you want to get in personal privacy, really drones might get into that. But satellites, it's a long way off. I can't even imagine that. It's just because you can't fly. recently in a satellite lower than about 300 kilometers altitude. Otherwise, you're just starting

to burn up in the atmosphere. To go below that, you really need a plane or a drone. And so I think space for a long period isn't going to get really into the thick of privacy. Maybe you're going to do it all, but I'm curious just to talk then a little bit more about the technology of satellites themselves. If you had to group satellites in like some sort of taxonomy into different categories, how would you do that today? So there's imaging satellites that, mentioned communication satellites. There's roughly five times today. There's imaging satellites. There's communication satellites. Those are the main ones where there's a commercial sector, although both of them have military versions of those things. The other three are navigation,

which is pretty much only governments at the present time, Galileo and the EU GPS here. Early warning, which is countries scan for anyone's missile launches. And then there's what they call signals intelligence, which is listening into people's communications, to listen to your cell phone or listening to the, and again, this is sort of CIA stuff. So those are the main five types of satellites that exist today. But the commercial ones are really just communications, and now SpaceX with the Starlink system is putting up a big communication fleet and Earth observation, which is planet has put up a big fleet up. If you stick in the Earth observation Planet category, what do you think are the most far-fetched sci-fi potential futures for what the

technology might enable, say that, you know, the rest of your career or something like that on some timeline that's long but reasonable. But I do think that it's possible in the long term that we would more or less be able to go have a live image of the earth. I think that that's possible. It's not where we are now. It won't be for a long time, but it's possible. And then the second thing is, I imagine that you should be able to just query that. You should just be able to write instead of just imagine like a search query box on it, and you can just say, hey, how many houses are there in Pakistan? Give me a plot of that versus time.

Tell me where the trees were deforested, the latitude and longitudes of the trees that are deforested in the Amazon in the last three weeks. It should be able to just tell you the answer to those questions without ever you looking at the images or it may highlight that in the background or something, but you should be able to get answers just like you can from Google. So I think a lot of what planets doing in the long arm is a little bit similar to Google.

Google figured out how to search the internet, to index the internet, sorry, and make it searchable. And we're figuring out how to index the Earth and making it searchable with the combination of the data and machine learning that sits on top. Can you say a little bit more about that machine learning and that actual exercise? That just seems like a Herculean task to go index the Earth. Pretty cool idea, but sounds pretty hard as well. Well, there was a hard task to index the internet. But then it's to do it. Well, just imagine a picture of the earth and machine learning,

especially computer vision that has been developed, especially by companies like Google and academia and others, has been particularly strong in computer vision, where it does things like extract out cats and dogs from pictures, you know, that you use online. The same technology could be used in our image to extract out a road or a building or a ship or a plane or a train or a tree. So we just have to do that for every image that comes down.

I said we have about 3 million images every day. So we have to do that for every image and identify all the different objects. And then you have basically a database of every object on the planet over time. And then you should be able to exactly do that. Query on top of it. I actually think it's intellectually relatively straightforward.

Of course, there's a huge amount of work behind that. With this unique data set, moving up the stack that you've done, what knowledge about the Earth has most surprised you that's resulted from this data set so far? Well, the degree of calamity of the destruction of ecosystems is just staggering. So we are wiping out forests, mainly to put cows on them so we can eat beef burgers. We are destroying the fisheries without ocean trawling. we are seeing huge transitions and mainly because it's not actually climate change, it's mainly these things like deforestation and illegal fishing.

So it's been driven by humans deciding to change the use of those territories. And that is just really obvious and sad, but hopefully our data can help companies and countries and individuals better manage the planet. We've lost 70% of life on the planet in the last 40 years. It's gobsmacking to think about. We are just whittling it away still. Our raison d'etre in many ways is to help stop that, is to help the governments to see the deforestation.

So we have a project to map all, which we do, all the deforestation in the 64 tropical countries with government of Norway paying for the data, helping that data be available to those forestry ministries to stop deforestation. Huge project. We have another one to map all the world's quarries, and we just released that a few weeks ago,

which first map of all the world's corals classification of different types and showing early signs of bleaching or if there's any illegal fishing going on, we can alert governments in fact that already six countries have established marine protected areas around coral reefs that we mapped for them. So those are the kind of things that can really help us protect and stop the ecocide that's happening on the planet. And so that is both a thing that's super important for the planet and it's a massive business opportunity because what is really happening is that global economy is going, ah, we cannot any longer presume that natural capital is free. The trees you cut down and it's free for the landowner. They can just cut down that tree and it's cost nothing.

Or you can just put gases into the atmosphere. It doesn't matter. Or pollution is into the river isn't matter. There's an externality of that cost. We've got to integrate that externality into our economic system. And that is the country is saying we're going to measure these emission targets and set these limits and its companies doing their ESG targets and saying, hey, the environment piece I'm going to measure and make sure that my resources came from a sustainable source or make sure I don't build those assets in a flood risk zone or these sort of things. And what does that mean? All of that means those countries and the companies have to measure all that stuff.

So this is a massive business opportunity because we have the data set. that's pretty foundational to the measurement of all that natural capital. It's measuring all those things. It's not just that we can see every tree in stuff deforestation. We can count the carbon stock in all those trees. Our data is foundational to underpinning the transition to a sustainable economy, which is a multi-trillion dollar transition as well. The 70% number that you quoted over the last 40 years was just kind of a staggering,

horribly depressing number. What's the attribution of that 70%? what specifically has been wiped out that makes out most of that decline? Well, there's pretty much everything. 82% of wild mammals have gone, 75% of insects, 70% of the fish in fresh water, rivers and lakes, over half the coral systems. These are all the things that have gone already.

We've been wiped up more than half of the forests. There was roughly double as much forested land of the earth 40 years ago. It's just staggering. So we're basically like an invasive species. is one way to think about it. Yeah. This data set is step one in any change, which is awareness. Exactly.

More than just awareness. I feel like satellites have already been bringing awareness about this problem for about 50 years. That's why the climate scientists have been yelling from the top of the mountain tops as best they can and telling us, this is what's going to happen. And no one's been listening. This is the data set that enables us to take action. You see, the difference between what had been happening before with satellites was every few months or every year we would take a measurement and go, this is what's going on, guys.

now it's like there is deforestation happening there, stop it. There is legal pushing there. You know, it's the real time action and the real time measurement and the real time enforcement that is actually going to enable us to, I hope, turn the page on tackling this massive challenge. One interesting thing, just where the rubber meets the road, you already mentioned the company's going public. How do you think about that transition as it relates to this big mission? I mean, if everything we've talked about holds true, this sounds like it could be both in business terms and in terms of impact terms, like one of the most important companies on the planet, or off the planet, I guess. How do you think about the relationship that you build with the business and investing community to make sure that you

maximize the odds that that's true? Well, look, I'm really excited about going in public. I think it's the right step for planet. Planet is ready to be a public company. You know, we've launched and operate this satellite fleet. It's ready. That technology to risk is retired. We've got a mature business. with over 100 million in revenue last year, we're ready. And what we're feeling is the pull. We're feeling everyone needing our data now. Well, then we need to scale up to do that, to address all of that market opportunity.

And I think from what I've seen, investors really like it for a number of reasons. One is that our data is completely unique. No one else can get it. It powers on a lot of vertical markets. It's not just relevant to agriculture. It's relevant for energy insurance, finance. They love the fact that it's a data business. We've got this hard moat to crack in terms of the hardware that enables this.

But the data business, we're selling data. And we can sell each image multiple times. So the profit margins are really large. The incremental cost of us selling an image to a second customer is very, very low. So the direct margins are just huge. And then everyone recognizes how important the transition to a sustainable economy is. And so it's space, it's sustainability. is a data company.

It's exactly what the market wants at this time. I'm confident this is the right move for planning this time. I'd love to learn a little bit more about the hardware. You said that great quote at the beginning that the best software companies also build hardware because it enables a very specific one-to-one relationship between the two. Tell me what's evolved since that first satellite you built by hand in the garage. What were the components like then?

And what has happened? Give me a sense of what it costs to make one of these things. What's going into it? What is the hardware story here that's, most interesting. Yeah, since that time, we've had 18 design-build iterations of our satellites. So just like the iPhone, iPhone, iPhone, one, two, three, four, five. And those satellites have increased in capability per satellite per day in terms of area coverage of imagery by 10,000 X over that period. I am not kidding you, that is what's happened from the first

satellite to the last. Just in the last year, we double the number of spectral bands, exactly to this point of best software companies built their own hardware. Our customer said, we needed these spectral bands. So the next generation of satellites, we added four more spectral bands that addressed various customer needs. So two years ago at our user conference, I announced the Superdove, which is the next generation of Dove. This one, I announced the fact that we'd already completed daily imaging with that Superduv fleet, we'd already had seven launches, about 100 of these satellites, and they're now operating and getting eight spectral bands that help these extra markets. Those satellites produce five times more than the previous

generation of satellites per satellite per day for almost exactly the same cost. Great. That's the kind of trajectory we're on and we're going to continue on. We call it strapping space to Moore's Law. Every time there's a better sensor that comes out, every time there's a better processor that comes up, there's a better hard drive that comes out. We step it into the satellite and put it up. That way, just like you don't want a three-year-old phone in your pocket, you don't want a three-year-a satellite in space. We iterate them really fast. We treat them like a server. We one that goes out of date, we replenish it. That's why we are on a very rapid trajectory on the satellite side. How much do you control versus outsource the component section of a given satellite?

In one of the satellites, how much is something that you're taking the best, whatever sensor from some other company and incorporating it? And you're kind of just the chassis that pieces the best in class together versus the Apple M1 ship or something. I'm trying to think of the right analog where your end-to-end controlling the component itself. I think it is a bit like Apple. I think that's a good way of thing about it. It's all our own bespoke boards. It's all our own bespoke designs of optics, of camera systems, of radio systems. We do leverage chips. At the chip level, we're leveraging the latest quad-core computer. We're leveraging the latest Y-Max Wi-Fi components and so on. So we leverage at the chip level, but we've got our own designs at the subsystem level. You're leveraging those chips,

but which are totally our own radio design, and they're vastly more capable than anything you can buy online. We can get 1.6 gigabits a second over 1,000 kilometer range from our little satellites to these little dishes that we've built all around the planet,

48 ground station. I can't believe we can get 1.6 gigabits a second at 1,000 kilometer range with these radios. But that's how much effort we put over 10 years now building the better radio systems. I can tell you that about every subsystem, the camera system, the telescope, the power systems, all the makeup.

Yes, we buy the chips, you like, but the integrated system is really complicated at the subsystem level, and it all has to work together. How should we think about complementary space curves that are compounding in the same way? Maybe rocket launches the obvious one. What are the other, like, driving tech improvements that are riding one of these kind of Moore's Law-like trends

that are interesting and important that will affect the space renaissance beyond imagery? I already mentioned that rockets have come down and costs about 4x. Satellites increase in cost performance about 1,000 X over the last 5 to 10 years. That's already pretty dramatic. Very few industries go about 1,000x change in cost performance over a few years. That's like the mainframe to desktop computer transition in computing.

This is like the internet moment for space. There's a lot of opportunities. But it's really complicated. Like I said, it's not for the faint of heart. I would say that rocket piece is quite distinct from the satellite piece. The satellite piece has been driven by Moore's law and general miniaturization of electronics, whereas the rocket piece was just Elon's will to have a reusable rocket, and that's enabled reusable rockets, and that's kind of amazing, but it's a different trend.

But I would say, again, the thousand X is what's dominating here. They are compounding. So we have a 4,000 X in cost reduction overall, if you like. But the thousand X is what dominates, and that's really about the miniaturization of satellites. that's taking bus-sized satellites and making them the size of a loaf of bread with the same sorts of capabilities. That's staggering. And that's what's really unleashing. As I said, the upshot of what's happening in space is massive new data sets that we collect about the Earth and transporting around the Earth, in the case of Starling and ourselves, respectively. And that's

enabling new and better economic systems, more sustainability. And that's why I'm excited about. As you go up the stack and more and more the business is about pre-processed analytics that become more and more valuable, like getting further away from data and closer to knowledge or information, however you want to describe that change moving up the stack. Will you then think about almost open sourcing some of the data at the bottom of the stack? Like it strikes me that if this was a nonprofit and you just made this all publicly available, stuff would get built on top of this that would be fascinating if there was no friction or cost to doing it. How do you think about that? I don't think that's the right strategy. I think that the data always has value. And by the way, there's another data company around the corner.

It's called Google. But you will notice they have open-sourced a lot of their algorithms, their analytics, TensorFlow or the computer vision modules of this and so on. They haven't outsource open-source any of their data. There's a real reason for that. The data is where the value is. Algorithms alone have zero value. Data alone has a lot of value.

The economists quipped the data is the new. oil. Obviously, there's good things and bad things about that analogy. I don't think that data is dirty like oil and I hope not. And by the way, oil can only be used once, you know, any bit of oil whereas data, you can pass off to lots of users. So there's an inherent better business aspect of data. But in this sense that data powers like oil does, a lot of different sectors, you have to refine it before it's useful to them. And it even could get commoditized, but still has huge value. I think that's where we would end up with data. Those that have the best data are going to drive the world, and algorithms are super important

are going up the value. But without the data, you can't do shit there. And with the data, you can. And so the power is in the data. It's very asymmetric. Do you anticipate that a lot of businesses build on top of you, almost API style like you would build on top of Twilio or Stripe? And is there a good example of that already?

Yeah. The second part I was going to get to, but answer your previous question, because I think that we do want to democratize it so that as many people can use. it and build those ads. But rather than give it the data for free, we'll still charge the data, but we do want to make it easier and easier to do that in small amounts, get going in an easy way. You have a free trial, but we want to enable them with the tools that can get going. You can imagine an SDK and thousands of apps being built on top of this data. We are building that right now, and that's why we're going to public, is to have the capital to do this.

The other reason, by the way, to get to another question you asked earlier, is that a lot of people still don't know about planner. You know, you seem visually just a tiny bit shocked by the story. And a lot of people are when we tell people out there, they're like, holy shit. This is going to be amazing for my industry, whatever industry is hedge funds, insurance, whatever. Most people still haven't heard of it. You know, I'm telling the story the first time. Do a lot of people. Okay, it was the other advantage of going public is we're going a bit more on a bigger stage. It's not the only reason. That was the only reason. I don't think it would be worth going public, but I do think that's a good reason for going public that is a side benefit.

Will, this has been such a fascinating conversation. I wish we had hours more to learn about what you're doing. I ask everyone the same closing question. What is the kindest thing that anyone's ever done for you? I should think about that more, but the first thing that comes to my mind is mentorship. I've said it has some incredible mentors that helped to guide me to where I am today. My PhD supervisor last year won the Nobel Prize in Physics. So you're just amazing. guide. My head boss at NASA, who helped me to understand space systems and lots and give me a lot of freedom. And I think of people that have mentored me. That was a gift. Well, there's been fantastic. Safe travels today. Pleasure to meet you and learn a lot about planet. Thanks. Bye-bye.

If you enjoy this episode, check out join colossus.com. There you'll find every episode of this podcast complete with transcripts, show notes, and resources to keep learning. You can also sign up for our newsletter, Colossus Weekly, where we condense episodes to the big ideas, quotations, and more, as well as share the best content we find on the internet every week.