BlackSky Technology Inc. (BKSY) Earnings Call Transcript & Summary

Greetings, everyone. Welcome to the Deutsche Bank Global Space Summit. My name is Edison Yu and I lead the space research here at the bank. Over the next 2 days, our aim is to inform [Audio Gap] hopefully, not confuse all of you tuning in from across the world. We'll be joined by over 30 companies spending 5 time zones and cover a wide range of topics, including earth observation, rocket launch, in-orbit transportation and space habitats. Additionally, with all the conflict and uncertainty we've seen across the industry over the past year, it's become a personal goal of mine to bring increasing more space companies and entities together to generate fruitful collaboration and open discussion for the future. Thank you, everybody, for joining us today. Taking things off, we'll start with our first panel called [indiscernible], featuring 3 companies: BlackSky, Planet Labs and Satellogic. To begin, we'll have all 3 panelists give a brief introduction of the company themselves and then proceed with a more in-depth discussion. Brian, would you please lead us off?

Great. Thanks, Edison. And nice to see everybody. I appreciate the opportunity and have been very much looking forward to this discussion today, a really exciting time for the industry. So thank you for putting this together. So just to kick it off, I'm Brian O'Toole, I'm the CEO of BlackSky. BlackSky is a leading provider of real-time geospatial intelligence. And we do that by providing on-demand access to our high-frequency small-satellite imaging constellation, which enables customers to a software experience to task satellites where and when they want and get information and analytics delivered in their environment that they work on demand, fully automated within 30 to 90 minutes. And this is really a game-changing experience for end users, have the ability to directly access and get information delivered where and when they want it. Our company has been particularly focused on the government sector as we've gone to market. We are now one of the few trusted mission partners of U.S. government and our allies and several ministries of defense around the world for our capabilities. We're integrated into their operations, and they rely upon us to support delivery of mission-critical intelligence on a day-to-day basis. Our strategy and approach was just really heavily validated by the U.S. government in our recent award of a $1 billion 10-year contract with the National Reconnaissance Office. And we also just announced a $10 million 1-year subscription as a follow-on to existing Asian Ministry of Defense customer. So BlackSky is -- we're in -- we've got our baseline constellation up and running, and we're entering into the high-growth phase of our business. So nice to be here today. Looking forward to the discussion.

Thanks. Mike?

Thanks, Edison. I'm Mike Safyan, and I'm the VP of Launch at Planet, and I'm also a part of the founding team, one of the original employees who came to work for the 3 founders when they left NASA to start the company in 2010. So Planet is a [ space ] and data analytics company. We are headquartered in San Francisco, but we have presence all around the world. And we have the world's largest fleet of remote sensing satellites that's comprised currently of 2 satellite fleets. We have our high-res sky sats, which are imaging at 50-centimeter per pixel and then our global daily scan of our Dove series or our flock of satellites, and that's around 180 actively imaging satellites scanning the entire earth every day at medium resolution. And so that combination of data has never existed before in human history to be able to get that level of information and detail and coverage at revisit rate that we're able to provide is totally unique. And we serve customers from a variety of different verticals and end markets around the world, defense and intelligence but also civil government and commercial nonprofit academia. It's a very diverse customer base, and we're making a lot of really exciting progress not only in generating a unique data set by pushing forward all of the analytics on top of that so that we can extract insights from pixels and serve more customers, growing the pie of who can benefit from remote sensing data, not just serving better data to users who are used to working with that information already, which we are doing as well.

Thanks. Hey Matt.

Thanks. It's good to be here with Brian and Mike, and thanks for the participation of Setallogic here. So my name is Matt Tirman. I'm the Chief Commercial Officer here at Satellogic. I've been with the company for just about 2 years now. I joined the [ staff ] of Satellogic North America, which is our wholly owned subsidiary in the U.S. pursuit in work with the U.S. government. A little about Setallogic. We've been around for 10 years. Founded by 2 Argentine entrepreneurs, who had the vision of being able to really map the world with high-resolution satellites at a fraction of the cost of where the industry was at. So again, been around for 10 years, spent around the first 8.5 years really perfecting the technology, developing out the capabilities. So we're about in, I'd say, 1 month 18 of our go-to-market. We went public this past January on the NASDAQ. And really, what differentiates Satellogic from the rest of the earth observation or earth observation peers is proving things. So we have the world's largest high-resolution electro-optical constellation. We only have 26 satellites in our fleet capturing centimeters. That's going to go to 30 by the end of the year. And our goal is to push and remap the world on a daily basis by 2025. And we get there through our superior unit economics being vertically integrated and offering the lowest and most cost-effective price in our observation market. So we're uniquely focused as everyone is right now on the -- where the addressable market is, with the U.S. government and U.S. allies around the world. But we also see massive opportunities, as Mike mentioned, in across other sectors from agriculture, energy, mining. And again, we feel that with our high-cadence constellation at high resolution at the right price point, that's where the market goes from a less interesting $5 billion TAM to a far more interesting scalable market for all. So really excited to be here, and thank you from everyone.

So we throw around a lot of terms, both observation, geospatial intelligence, remote sensing. Could you maybe help the audience understand just what the semantics of those words entails? And how those words actually mean different things in different contexts? And maybe we can start with Brian. I know you've you probably cover all these areas, but maybe simplify those terms down, so when people are looking them up, we know who we're talking about.

Yes, for sure, Edison. I think, [ I need ] to simplify it. Earth observation is -- it's a wider range of different types of sensors that can observe the earth, right? And there's electrooptical imaging sensors like many of us have, those create the images you see on Google Maps, for example. There's radar sensors that can also look at the earth surface and bring back different spectral information. There's hyperspectral sensors that also allow you to see different aspects of the earth. So that's really the remote sensing earth observation side. Geospatial intelligence is when you convert that raw data and combine it with other information to deliver information and insights that end users need, such as looking at changes in agriculture, field conditions, monitoring border crossings, just look at activity, things of that nature. So there's the raw data, which is the earth observation and then there's the transformation of that into geo-intelligence. And that's the primary difference. And just from a market perspective, the market for the raw data is relatively -- it's a decent size, several billion-dollar market. But the analytics market, which is the geospatial intelligence market, is, in some estimates, 10x more or 100x larger than the raw data market.

Great. And then, Mike, can you maybe dive into the difference between optical SAR and something like AIS? Can you maybe just explain to maybe the benefits or trade-offs about using these various types of types of sensors?

Yes, absolutely. So optical is essentially taking a picture with a camera and a telescope, and that's the type of imagery that's most widely adopted and easy to understand it and work with. And as Brian mentioned, that's primarily in type of sensors that the 3 companies represented here operate. There's also different trends emerging in remote sensing, especially on the satellite side, so SAR or synthetic aperture radar, which allows you to send a pulse of information and then gather the reflection of that. And that's how you're sensing what's happening on the earth rather than just taking pictures. And there are advantages of SAR. You can see through clouds. You can see at nighttime. And there are some really exciting companies that are working on that in new innovative ways that haven't been available before. There are also downsides to SAR. It's a more complicated data set to work with. It's not as widely adopted. And so I think it's at an earlier phase of its evolution. AIS is a different type of data altogether. It's a beacon that different -- the maritime vessels of a certain class in size have to broadcast in terms of their name and their cargo information and things like that. And so it's a really great way to track maritime assets. And you can detect that from land if you're close enough to the shore, but there's a lot of open water that you'd want AIS detected from satellites to be able to understand what's going on with maritime activity. And I think that the challenge with AIS is that when ships want to do something illegal or do something under the cover of darkness, they'll turn off their beacons, and then you're not getting that information that you need. So it's a great example of how you actually really benefit from combining data sets. So AIS plus optical or SARR can then give you a picture of who's out there with their AIS transponders off, that's a really great indication of something illegal potentially happening. And finally, I'd even throw in another category, which is starting to grow in this industry, which is RF, radio frequency, signal detection or signal intelligence. So HawkEye 360 is a company that comes to mind, but there are others as well. It's just broadly scanning radio activity as another way to detect what's happening on the earth surface. And things get really interesting once you start combining all these different data sets and get a richer and clearer picture of what's going on in the area.

In terms of the difference between optical and SAR, I think, if you could remind us, is the main benefit that basically you can see at night and through clouds?

That's how I frame the benefit. Yes, I think that, again, the challenge is that it's a much more complicated data set to work with. And so it requires a whole different type of like processing pipeline. And so I think there's a reason when you open up Google Maps, you're seeing optical imagery. You're not seeing [ simple ] imagery. But I think that the industry is evolving, and it's a really powerful and unique data set. It requires a very different type of satellite design. And so that still is also growing and maturing. And so I think that as things go on, we'll see more SAR integrated and used, but where things stand and kind of the momentum of the industry is still heavily focused on optical.

Great. And Matt, you have several satellites up with hyperspectral capabilities. Can you tell us what this means exactly? What's the difference between multi versus hyperspectral?

Yes. So I think unique to Satellogic, we're aside from a handful of Chinese providers. We're one of the few commercial satellite earth observation companies that actually [ supply ] hyperspectral sensors right now. So our satellites are a bit different. They are about 45 kilos, about the size of the small refrigerator. And on each satellite, you have a multispectral sensor, which is the primary sensor. And that's at 70 centimeters across world, and it's red, green, blue. And then this is multispectral imagery, this is great for detecting changes to infrastructure, the natural environment at sub-meter resolution. We also have a full-motion video capability as well on the satellite, which kind of adds a new dimension to kind of the geospatial intelligence collection. But we also have hyperspectral. So hyperspectral for us, we have 4 dedicated satellites using across 29 different bands across the spectrum at about 20 meters [ GSV ]. So with this capability and wide swath it has, it means that we can remap the world on essentially a biweekly basis with our hyperspectral capability. And the use cases of this are really interesting, especially at a global scale product and can be pretty profound for users across the defense intelligence world, academia, as well as commercial applications. So we've been working with a U.S. government lab customer on doing everything from nighttime analysis for population growth and density-adjusted urban development indices, looking at global change detection. So being able to tip in queue to higher-res assets like our multispectral or other higher higher-res constellations out there is a really unique application or even different modalities like synthetic aperture radar. And then the creation of a global image base map for land use and land cover. So HSI just provides just a new range of opportunity kind of outside that visible spectrum available for a wide range of customers.

Another difference about the constellations is sort of the purpose. And Brian, I was wondering if you could dive into this a bit more. You have these constellations out there that are scanning and then some of them that are passing or the majority of which are cascade. Can you go over the differences there? And what would be the benefits and trade-offs of each?

Yes. I think Edison as you point out, there's a couple of different strategies that have been taken in the industry. One is to have hundreds of satellites and they're in a scanning mode where they're -- as they orbit, they take large swaths, collect large swaths of imagery with the purpose of collecting the earth every day and every part of the U.S. -- of the world's land mass every day. So scanning sensors are really good at doing that because they can collect large amounts of data with satellites, but it requires hundreds of satellites to do that. The other approach is a tasking-based approach. That's the approach we've taken. And what that means is we use software to only collect what people are interested in. So in our case, we've deployed these satellites in what's called the mid-inclination orbits. So there are plus or minus 50 degrees around the equator. That's where 90% of the global GDP takes place. And we use software and AI to task only the most important strategic assets and economic and strategic activities in that area and do that at very high frequencies because we're in that type of unique orbit, we can see things reliably every hour. So from 7 in the morning, 6 in the morning to 5, 6 at night, we're going over the same location, at least every hour with our satellites, and that's where we differentiate in the market, which is this high-frequency dynamic monitoring. Most scanning systems that are painting the whole earth are typically only -- they're in an orbit where they typically only take pictures at 10:30 in the morning or 01:30 in the afternoon. So if things are happening at 9 in the morning or 4:00, those satellites don't see that activity. But it's really set up for two different applications. The global mapping is kind of a static mapping application, tasking is around dynamic monitoring. And so that's what you see is different in the market. And in our case, we put the power of that tasking in the hands of the users, so they can see where they want, when they want. And we can do that very efficiently with a rightsized constellation that we can monetize each of those collections many times.

Great. And just a follow-up. Can you explain the types of orbits? I think you mentioned that you're mid-inclination. What are the other types that people would use?

Yes. So the primary type of orbit that's been used in the industry for decades is what's called a sun-synchronous polar orbit. In simpler terms, that means satellites are orbiting around the poles, they're going north to south, ascending and the earth is turning underneath them. It creates this unique and reliable ability to map the earth. But when you take a picture, it's always at the same time. So -- and over time, in the beginning, that was important as you were using this data to build maps and you wanted the visual elements to be very reliable. We're in a mid-inclination orbit, which is counter to the earth's [ location ] around the equator. And that's how we get very, very high frequencies because we're traveling over these locations many times a day with a fleet of -- our current baseline is 14 satellites, we will grow that to 30. But that gives us 1 hour revisit over any location within that band. And that will improve every 30 minutes as we build out our next-generation constellation. So two different -- very different types of orbits. But frankly, they're set up and they're used in a way to support the different business models for the different companies. And so -- and it's a big market. So there's a very big market for global mapping in a very large market for what I'll call high-frequency monitoring and geospatial intelligence.

Great. I think we have a much better understanding now of the lingo of EO. Could we maybe dive into some practical use cases, some commercial use cases? Mike, I know Planet has a lot of commercial customers. Could you go over a very big use case and probably one that you can call the customer?

Yes, of course. And just to add on really quickly to Brian around orbits, I think you can -- there are different advantages to those different orbits, but you don't necessarily have to pick one or the other. So for example, Planet's constellation, we have both sun-synchronous satellites in the high res and also medium res and also high-res satellites in those mid-inclination as well. And you can start to get the advantages of all those orbits if you deploy your fleet in that way, which isn't unique to Planet, but is one way where we're able to kind of expand the number of users that we supply. And so Edison, you're right, defensive intelligence is a big customer base and also civil government. But in the commercial space, we've also had a lot of really exciting traction. I mean agriculture is probably the biggest one. And a lot of the big ag players around the world are used to working with remote sensing data, and they have the processes set up, and they're able to ingest it. And they just appreciate a better data set or a lower-cost data set and more timely data set. So Bayer comes to mind as one of our customers. But another customer example -- another customer announcement we made recently that I find particularly exciting is a collaboration with SynMax, and they're monitoring energy infrastructure so that you can give intelligence reports around oil drilling or other types of energy manufacturing, which doesn't just have to be in the U.S., could be all around the world. And there's places where reporting around energy output is inconsistent or unreliable. And to be able to have an objective data set that you can take a look at and make intelligent decisions on, and so this is kind of a theme I think we'll hear throughout this webinars, is that we're moving up the stack of not just providing pixels but providing intelligence or actionable insights. And so I think our partnership with SynMax to be able to more smartly, in a more timely way to monitor energy infrastructure, that's a really big deal and one of the bigger growing markets that we're seeing.

Matt, is there a use case, commercial use case that you can highlight from Setallogic's perspective?

Yes. Look, I think the -- one of the exciting value propositions that we can bring to the table is again around our unit economics. And when you have a fleet of our size at 70-centimeter resolution at our price point, we're able to open up markets for countries and companies that previously they didn't have access to Earth observation or other geospatial data. So I think one that comes to mind is we recently announced that the government of Albania a multiyear deal supporting their National Civil Protection Agency. And this is really exciting, too, because I think as Brian rightly pointed out, right? I mean there's no shortage of defense and intelligence users and use cases here. But it's really exciting when you're able to work with the sovereign government to kind of expand their use in their application of our observation data. So it's a really cool opportunity we have with them. So we're helping them with a couple of primary use cases. The first is looking at cadaster data developing out, land use land cover. They're very keen on understanding legal building around the country, so they can go and properly account for tax record. So this is an opportunity that did very well, turn into, I would say, a revenue windfall for the government of Albania in terms paying for the service itself. And then you have more traditional use cases for the government. They'll be able to use the data for border security and disaster planning and emergency management, given the fact that they do sit on fault lines and experience forest -- an increase in forest fires over the past few years. So if you think about a country of Albania's size and multiply that out across all the different friendly foreign allies around the world, again, it's within that sovereign government addressable market, but these are countries, again, that have never had that access to the space data before. It's incredibly exciting and [indiscernible].

So I've heard the words, analytics, intelligence, software used several times so far. It almost sounds as if you guys are more software companies than satellite companies. So I want to dive a little bit more into that. Mike, Planet has a really great reverse pyramid chart that you use often. Can you maybe describe or go over the layers of EO that you touch or that you hope to touch in the future?

Yes, absolutely. I think you're right in that we are operating more and more -- at least from my perspective, everyone on this panel is operating more and more a software companies. It's a very different business model to build satellites or design missions for others to operate. We plan that -- we feel that the satellites are part of our core capability, and that's what enables us to do what we want to. But ultimately, it's about the software, it's about the data. And so when we think about this pyramid or data stack at the bottom, you have the basic infrastructure that you need. So the satellites, the ground stations, all the regulatory approvals, mission operations, things like that. And then you start collecting data, and raw pixels are the basic form of that. You can start cleaning up that data and do interesting things to make it easier to analyze. So creating cloud-free base maps or stitching together data or getting daily alerts of when new data is coming in. But then you start moving up the stack even further, and you can start extracting basic information like soil temperature or a detection of a road or a detection of a tree, and then ultimately, what you really want is insights. And so after all of that kind of groundwork and preprocessing, maybe you just want to forecast of what corn output is going to be like in this region in the world versus that one and be able to compare that to historical output. That's based on third-party data sets or satellite data as well. And so more and more, we believe that it's going to be machines that are analyzing this data rather than human eyeballs. I mean we are generating terabytes and terabytes of data a day. There's no one government organization or team of remote testing scientists that can really understand everything that's happening with that. We are [ across ] the threshold of what humans can really do to fully extract the value of that data, which is why we're leaning so heavily into all of the analytics. And I think that's why having a richer data set, so in our medium resolution on our flock constellation, we've increased that data set, our PlanetScope data set to 8 spectral bands. We're also working with NASA JPL on a mission under the Carbon Mapper program, which is a hyperspectral satellite, specifically tuned to monitoring methane emissions, but there's 400 spectral bands that will be monitoring. And so providing a richer data set to these automated algorithms, I think, will be a huge benefit, so that we can really extract information that we wouldn't be able to see with eyeballs and then ultimately produce that into insights that nonexperts can use to make informed decisions about energy, agriculture, disaster response, all kinds of things.

In that framework, Brian, I know Spectra AI is very near and dear to your heart. It's what the company has worked on. How does that play into this value stack?

Yes. I think this goes back, Edison, to your -- to the prior topic on commercial use cases. I think what is important for everybody to understand is that up until recently, this type of data that all of our companies produce was really never accessible to a large market before. And we sell all -- there's been 3 barriers. It's been the economics of the data, the accessibility and then the frequency of getting the information where and when you want it. And we've always felt from the beginning that software is going to enable -- is really going to be the platform that enables massive adoption of real-time geospatial intelligence to markets that never were able to capitalize on this before. And we started investing in our software platform 8 years ago, in fact 5 years before we had satellites. And we're focused on that machine-to-machine and end user experience of actually delivering actionable intelligence. And because of the nature of what we've done in space and because we have a real-time platform, we can integrate our software with third parties, which is enabling access of this information to a whole new user community. So for example, our software was quickly integrated with Palantir's platform. And that enables any Palantir user, whether they're in -- using it for financial analysis or for security analysis or things like that to task our satellites and get analytics delivered directly in the platform that they're working in every day. We just announced the same type of relationship with ESRI, they're the largest geospatial software company in the world to get 300,000 organizations, 4 million users. And now we've unlocked on-demand access, just base data and analytics, through our software platform. So this is changing the user experience, software will be the enablement. And as Mike outlined, it will be the machines that are going to be analyzing this information and delivering that actionable intelligence to a very broad market. So I think -- the -- even when you think about -- Edison, back to the commercial use cases, just the ability of anyone in any of those platforms, whether it's utility, someone managing utilities, somebody managing a construction project, someone doing agriculture analysis, we don't need to go to all those markets. We just need to go through those platforms is where those people already work. And that's really where we see the massive opportunity in the market. And I'll finish by saying the platform we built is agnostic to sensors and AI algorithms. We're able to plug and play other electrooptical satellite constellations, synthetic aperture radar, RF, all kinds of data sources and plug in other analytics capabilities from our partners. So we see the platform as a massive enablement for accelerating the products and services that we're bringing to market through a large-scale machine-to-machine adoption.

And then Matt, how do you see these various layers evolving? It sounds as if we have a raw input sensor layer, you have refined -- more refined data than you have software analytics. What's your view on that?

No, look, I think -- I mean, Mike and Brian hit the nail on the head in kind of describing the kind of the picture as it is today and where it's going, right? Even today, it's less about delivering an orthorectified beautiful 70-centimeter image to a customer, and it's more about delivering an image chip or the metadata to ensure the customer gets what they need because they are looking, as Mike said, for actionable intelligence and actionable insights. So we've worked a lot with partners and some of our government clients around pushing images directly from our satellites right down to the end user. So processing on the edge on the satellite. That's where this is going. I think it's a little bit different on the Satellogic side compared to our colleagues in the industry is we're taking very much a data-first approach, okay? Our hypothesis and the reality that we see now is that there's a lot of value-added analytics players out there. There's a lot of platforms out there that are ingesting multiple sensor modalities and ours -- Planet's and BlackSky's among them. We feel that's only going to grow and that's going to grow as more satellites come online, more data comes up online and that cost goes down. It's going to create [ many ] industries within the insurance vertical and shipping and industrials to where we feel we'll be in an excellent position to fuel those industries with our data along with other different sensor attached as well, too. So we're very much focused on making the access to our data as seamless as possible through flexible APIs, the ability to task our constellation and gain the data in the way you shape and manner that the customer needs it.

Great. Let's switch from software to talk a little bit about the hardware since that's what usually people see in the sky, so you know watch the sky. I know [indiscernible] is not in a standout to kind of defend itself, but could we go over historically the business model or the go-to-market for a traditional satellite and how that has sort of evolved now? And maybe we can start with Mike, what are the big differences between the way things were done and how they're going to get to we're getting done right now?

Yes, absolutely. So the traditional approach to building a remote sensing satellite is typically a very large satellite. So you can imagine, sometimes up to the size of a school bus because the focus was really on high res. There are kind of 2 ends of the spectrum. You had really high-resolution satellites, very large, which also end up being quite expensive because once you start building a very large and expensive satellite, you get this feedback, we will -- if it's so large and it's so expensive, you really want to make sure that it works because you only have one chance to launch it. And so you end up with a lot of redundancy, a lot of really high-grade, very costly components and a lot of extensive testing on the ground, which again -- which ultimately doesn't necessarily 100% predict success. You've also seen on-orbit failures of fairly expensive assets from government agencies who are very advanced. The space is hard. And so you can get very high-resolution data or what we've also seen is government assets like Landsat and Sentinel, also very large, very expensive that's more about scanning. But at a [ course ] of resolution, but they focus more on spectral bands and kind of scientific quality of that data. I think that what the companies on this panel have done differently is go much smaller by order of magnitude and also take advantage of commercial off-the-shelf components and be able to rapidly iterate and innovate on the satellite platforms, so that they're constantly improving and not necessarily relying on any one single satellite that's kind of holding the bulk of their business. It's more about a distributed constellation, which gives you better coverage, better frequency of revisit. And so -- and also, I think, again, from all 3 of the companies, we're seeing this very rapid iterations of different generations of the satellites that are coming out every handful of years, so that you're improving on the capability and performance of the satellite, responding to market demand. I think it's very different than if you're trying to build a satellite that's supposed to last 15 years in orbit, which means that it's already flying old technology because you're typically relying on stuff of space heritage. And it's a pretty big paradigm shift when it comes to the kind of new space commercial approach to smaller satellites, rapid innovation and better and better capability as time goes on at lower cost. Yes.

Matt, you have some very good slides about the cost -- on the cost side. Can you put some numbers, some real numbers? I think the one slide that got to me was the world view for cost, I want to say, $800 million or something like that. So could you maybe put some numbers or discuss the numbers so people have a sense of just how big a difference is?

Yes, for sure. And Mike did a really nice job of kind of painting the state of the small sat industry compared to legacy players with Maxar and Airbus. And he's absolutely right. It's going vertically integrated. It's developing unique partnerships with existing bus developers, it's cost or specified [ paths ] for your manufacturing, your supply chain. I mean those are the absolute keys. And again, I'll stress the other point that you mentioned, too, which is the ability to rapidly iterate. We're starting to put up our RMR [ clouds ] next year. And then right after that will be our next gen. I know our colleagues in the industry are doing the same with their future iterations. So I think that's the exciting piece, right, better resolution, better compute coming out on a very frequent basis. But when you look at the Satellogic, our bill materials included in that launch and everything, put all together, you're looking at around $1 million a [ station ]. You then compare that to some of the legacy space providers, and that's in an order of magnitude more expensive. So again, it's the ability to be far more flexible on the software side, upgrades to the sensor, upgrades in the computer, just something that legacy space can't compete with. And it's again -- look, it's very complementary in different use cases, but it's a -- I think the companies on this virtual panel are certainly the ones that are kind of paving the way for the future here.

Yes. And I just -- I wanted to jump in for a second. And I think Maxar and Airbus do deserve some credit in that, that they are kind of moving in that direction where you see the difference between [indiscernible] constellation that Airbus is developing. But there's like another big step until that you get, once you get to the new space players that are on this panel in terms of cost, capability, different type of characteristics of the constellation that the bigger satellites can excel in one area and then the smaller, more agile satellites excel in others, and there's enough business to go around. And it really truly is a growing part.

Yes, just, Edison, if I could just jump in. I think there's another factor here that's, I think, changed the industry in the last 5 years, and that's the cost of access to space. So 10 years ago, 15 years ago, if you're going to launch a satellite, you needed to procure a $90 million rocket, the types of launch, access to launch capability that's available to the industry now really is just coming into its own right now. So the reduction in the cost of access to space is a huge enabler for the small-satellite constellations. And if you -- when you think about that from a big-picture perspective, we are getting very closer, we can start applying Moore's Law out to space. And because of the rapid -- most companies have built vertically integrated capabilities like we have, where we control manufacturing through launch operations and final product delivery. We think of our satellites much like iPhones. Every 18 months is going to be a better camera, there'll be better processing, there's going to be more apps, and that's enabled by access to space that wasn't there before. So we're just in the early days of what I think is a super exciting time.

Before I get -- wind down, I just want to note to the audience out there, you do have a slot on the browser to ask any questions. So if anybody wants to take some in the next 5 minutes, by all means, do so. We highly encourage this to be a dynamic discussion. The last thing for me is we've talked a lot about the industry, the details, the software and now the hardware. How does this translate into a market opportunity? I think some of the third parties out there are saying $6 billion, I've seen estimates from your companies that are much higher than that. what's a proper or sensible TAM that you think your companies will go after? Maybe start with Matt.

Yes. Look, we've been very upfront with the fact that we feel that Satellogic is in an amazing position to help really expand the TAM for beyond the $5 billion current investment market that we existed now to one that is well north of $100 billion in the coming years. And that's purely because Brian mentioned it before, right? The limitations have been around access, have been around costs. And those are -- the cost of data is coming down. There's more capability. There's more there's more data available. And our contention is that this is -- I think one of the things that our industry has done, has not done very well, is be able to translate what we do really well and the data that's thrown off of our satellites and making that part a valuable part of the enterprise. And I think now is the time to where this has to move out of the domain of the geospatial intelligence experts have been defense and intelligence and move much more towards data that comprises broader sets in the Global 1000 and Fortune 500 that actually helps move -- that helps make the decisions for some of the most valuable and impactful companies in the world. And I think we're just scratching the surface on this now, and that's why we're so optimistic as to be the growth of the TAM.

Sure. Brian, any thoughts on that?

Yes, I agree. I think there's lots of studies, the size of the data market or the data market for Imagery, several billion dollars. The geospatial analytics market is anywhere from 5 to 10x that. But I think to Matt's point, those studies are based on an old model and just as an example, look what happened when geospatial intelligence was enabled for developers through Google Maps. And there's 5 million plus platforms that run on that now and applications were developed out of that, that no one thought of. And so this industry is at that point where the -- I'll dare say is that Internet moment, where a lot of investment in launch, space infrastructure, satellites, we're just getting started on the software enablement and easy access and affordable access to data that no one's had before and access to thousands of sensors that are going to change the way we see and understand our planet. So the applications that are going to come out of all that, I don't think anybody here can dare to imagine how big that can be. But we're just getting started. And I feel all of us combined are pushing this forward very fast. And it's really going to be about software enablement, and that's going to grow the TAM tremendously.

Fantastic. Unfortunately, we're out of time. The good thing is we will be welcoming [ after-review ] back later in the conference for some deeper dives on the companies. And so thank you all for joining. Also a reminder to everyone listening, we did just launch coverage on Earth Observation earlier this week. If you're interested in getting the report and the details about all this, please reach out to you to me or your representative. And we will now take a short 10-minute break. And next panel will be the space. So thank you, everyone, and look forward to speaking again very soon.

All right. Thanks, Edison.

Thanks, Edison.

Thank you.