Opto Sessions – Invest in the Next Big Idea

AI in Space: How Planet Labs’ Satellites Are Helping Life on Earth

• Haydn Brain & Ed Gotham

Share episode

In this week’s OPTO Sessions, Ashley Johnson, President and CFO of Planet Labs, explains how the company is using satellites and AI to deliver real-time insights about life on Earth. She breaks down Planet’s evolving business model, the move into satellite services, and how its data is helping governments, agriculture, and disaster response teams make faster, smarter decisions. 

If you enjoyed this interview, consider subscribing to our Substack channel for more in-depth insights designed to help you invest smarter: https://optoforesight.substack.com/

Create your Own Stock Index & Invest Smarter with OPTO Folios: https://optothemes.onelink.me/BZDG/ti2lb2fd

X: https://twitter.com/OptoThemes
Instagram: https://www.instagram.com/opto.themes?igsh=MXhwenU4dTk4aDBqMw%3D%3D&utm_source=qr
LinkedIn: https://www.linkedin.com/in/opto-invest-in-innovation-308416193/
Facebook: https://www.facebook.com/OptoThemes

-----

The content in this podcast is for informational purposes only. Opto Markets LLC does not recommend any specific securities or investment strategies. Investing involves risk & investments may lose value, including the loss of principal. Past performance does not guarantee future results. Investors should consider their investment objectives and risks carefully before investing. The information provided is not an endorsement of this product and is for information and/or educational purposes only.

Welcome to the show, Ashley. So how are things? Well, they're great. It's obviously interesting times where I kept very busy following the news, but generally speaking, everything's good. Thanks for having me on the show. times. Yeah, no worries at all. Glad to have you. So I usually start with any interview where we're interviewing the kind of leader at a public company just for them to give us an overview of what the business does. So for anyone unfamiliar with Planet, what do you do? Sure. So Planet is a company that was founded by two founders out of NASA. We're headquartered in San Francisco, but we're a global company that has approximately 200 satellites orbiting the Earth, taking images of the Earth every day. So we have a fleet of about 150 to 180 satellites that are in a north-south polar orbit and effectively line scanning the Earth every day, creating the world's most robust data set of the change that's happening on the Planet. And this data set has approximately at this point, 3,000 images of any place on earth so that you can look back and see how things have changed over time, even in those areas you didn't know to look for. And then we complement that fleet with two other satellite fleets, one of which does high resolution imaging. So it allows you to really zoom in. and understand that change at a more granular level, so approximately 50 centimeters per pixel. And then we have a new fleet of satellites that images with a hyperspectral sensor. Effectively, this allows you to see things that you can't see with the naked eye. For example, one of our partners is using this data set to identify methane emissions around the globe. With the multimodal fleet of satellites, we effectively can capture all kinds of information about the change that's happening on Earth. And our business model is we sell this data to companies and to governments globally. Yeah, fantastic. Really fascinating overview. I want to dig into a few of those key points. But before we do, I start again by asking anyone that we speak to just for a bit of information about their background. I read that you've worked at companies like Wealthfront and Service Source as well. What lessons from those experiences have you been able to apply at Planet today? Sure. So I am a rather unusual background. I'm both the president at Planet and I'm the CFO at Planet. I started my career in private equity, which meant that I was investing in technology companies with all kinds of business models. And that was the thing that always attracted me to a company was having a really unique and interesting business model. So Service Source... is a company that really focused on recurring revenue streams for companies and actually outsourced the management of those recurring revenue streams. And Wealthfront is a fintech company. So it helps people that are relatively early in their careers understand how to prepare financially for different goals that they have in the future. So much more B2C company versus service sources, very much a B2B company. But I'd say in both cases, the experience taught me that you really need to have the customer at the center of how you think about your business and whether or not you're driving the right value to that customer where they are. So service source is all about looking at recurring revenue streams and the information that you get from that data, like net dollar retention rate, on time renewal rates, tells you whether you're meeting the needs of that customer or whether there are changes that you need to make to the product, to the way you're delivering the service or understanding the changes. and the competitive environment. And all of that is relevant to us because our business is recurring revenue as well. It's selling data streams to our end customers. And as I said, with Wealthfront, it was about knowing the customer and being a trusted brand, especially when you talk about managing people's long-term savings, that trust is absolutely critical and understanding where the customer is on their financial investment journey. so that you can meet them where they are and really help them in the way that they need to be helped. If they're an expert, they don't need the basics. But if they are nervous, then you need to really be there and help them understand at each stage what might be going on, especially in moments of volatility. And since we were purely a digital company, it was understanding how to be that through a digital interface. And so all of these things are concepts that I bring back to. to plan it, which is we work with different types of customers, government customers that are very experienced in using satellite data to solve problems and to be there for their constituents or to think about intelligence applications. And then we have customers that haven't had the budgets for GIS and satellite data. And so we need to bring those incremental capabilities to them and help them understand how we can solve important mission critical problems for them, all the while knowing that we have to be a trusted brand. If governments cannot rely on the data that we're providing and the advice that we're giving to them, especially when you're talking about intelligent services, then they won't work with Planet. And so all of these things are incredibly relevant to us and quite frankly to really any company that's solving important customer challenges. Yeah, absolutely. Really useful context. I think now we've kind of introduced Planet and yourself as well to the listeners. Let's move on to understanding the industry in which Planet operates within. The Earth observation industry is rapidly evolving. I think that's fair to say. So where does Planet sit itself within that evolving landscape and what are your key differentiators? Yeah, so I think Planet has really been a change agent in this industry. Historically, Earth observation satellites were giant satellites. Even in the commercial side, they were very, very large satellites that cost hundreds of millions, if not billions of dollars to design, build, launch, and operate. Planet came into that market and said the net result of having such an expensive satellite is that Very few companies can provide for it or can afford it, but also the technology ends up being relatively old when you think at the pace at which technology is changing. So if you're building a satellite that's going to cost a billion dollars to design, build, and operate, you need to make sure it's absolutely fail safe. And so that means you're using technologies that have already been proving you're not using the bleeding edge. It means you're designing it to be in space for 10 years so that you can amortize that cost over the longest possible period. If you can think back to the phone that you had 10, 15 years ago, would you want to use that phone, much less take pictures with it? And yet these are the types of delays in technology advancements that Earth observation satellites were suffering. And so Planet came in and said, if you want to be able to strap Mars a lot of space, then you've got to bring the cost down. And so the satellites that we use for Earth observations, the ones that are taking the daily scan of the Earth, cost roughly $300,000 a piece. So if you think about that tectonic shift in the size, the shape, the cost of those satellites, it also means that we are constantly iterating. Every time we send new satellites up, we're sending satellites that have new capabilities embedded in them. And in fact, one of the most recent satellites that we launched are Pelican satellite, which does high resolution imaging. included in Nvidia chip so that we can start understanding the ability to do AI at the edge of space. And so that's one of the core innovations that Planet brought into the market. And in doing this, in moving to a model where we're selling data that is usable by anybody and we're scanning anywhere on earth means that we don't have to just sell to the highest bidder. We're actually selling, we're creating a data set that's a one-to-many data model. or business model, which means that we can offer much more affordable data. We can experiment in industries that aren't sure yet how this type of data is going to help them solve problems because we can price it at a level where they can engage early, experiment, and then as they find that value, really scale their engagement with us, which is very different when you have a finite data set. You have to point and click in those areas of highest value because you've got this giant satellite. cost of which you have to amortize over a very large revenue stream. So these are some of the core changes that we've made. And you see those changes impacting not just Earth observation, but other industries as well. Comms has moved away from the idea of a single communication satellite to a distributed satellite fleet for many reasons, but also because this brings down the cost of each single satellite and it allows you to be more innovative with the satellites that you're building and launching. So you've referenced your customer base a couple of times. You serve both commercial enterprises and government agencies. How do you navigate the difference between those two segments in terms of sales cycles, product needs, revenue predictability, et cetera? Yeah, so again, coming back to the importance of having the customer at the center of everything that you do. And this is actually a change that we made in our own operating model last year. And we talked about this with our investors. A government sales cycle is going to look very, very different than a commercial sales cycle. And even within government, have sophisticated in the sense of experience with GIS data, customers on the intelligence side who have spent years, decades working with this type of satellite imagery. And then you have parts of government that this is going to be relatively new for them in applying GIS data to solving challenges like planning and building departments or implementing oversight for new types of programs that they want to offer to incentivize the right type of farming practices for sustainability and other purposes. So we've actually realigned our sales team behind three different vertical segments. So one is the defense and intelligence sector, one is civil government, and then the final one is commercial and understanding within commercial various segments as well. Agriculture is going to be much more experienced in working with this type of satellite data, but some new markets that we're really excited about. like insurance and financial services, this is a relatively new investment for them. And so we need the commercial teams that are engaging with these end markets to understand where the companies, where those customers are in their experience working with this type of data set and their ability to handle this type of data set. And then making sure that we also understand the sales cycles. So as you said, government's going to be very different. from a commercial procurement process. Commercial procurement processes generally rely on very clear understanding of ROI and how this gets incorporated into a budgeting process. What spend you might be replacing that was existing there before. Whereas on the government side, it's really understanding the procurement processes, what type of sourcing practices they need to adhere to per government regulations. where the government is in their budgeting cycles and how far in advance you need to lay the groundwork for a procurement around this kind of data set. And then the services that you need to bring to bear for each one of those end markets are going to look different again. And so this is why aligning our company around this operating model where you have a very deep understanding of that end customer market and those customer needs becomes really important to how. how we operate and see success. Yeah, so you've given us some examples of industries that you currently work with and some of the applications of your technology as well. But are there any untapped markets, areas of big opportunity that Planet are looking to explore moving forward? Absolutely is the short answer to that. So some of the areas that we, some of the new solutions that we've been building with partners using AI are opening up new markets to us. So one that we're particularly excited about is being able to be there both for disaster response. You may have seen some of the work that we've done with our partners at Microsoft and their AI for Good team in providing building damage assessments immediately following. a crisis situation. But some of the ways that we're working with insurance providers and with civil governments is looking to what happened before those disasters occurred so that you can understand were there early warning indicators that would have told you that a wildfire is about to break out and that it could be a disaster level, not your typical annual wildfire, but actually something that can be as devastating as what you saw in Lahaina or in Los Angeles. and what were the factors that caused that to be so significantly different this year versus in prior years. That information is obviously very interesting to insurance providers who are trying to recalibrate their risk models. Also very interesting, obviously, to governments everywhere that know that wildfires are part of the things that they are thinking about and planning for, but the magnitude of change that's happening means that they need to understand this at a very different level than they have. previously. That's also obviously interesting to parties that have been involved with these types of incidents. So PG &E is a really important customer to us that's using our data for their risk models to understand where they need to prioritize the bearing of power lines so that they don't run into the types of situations that have caused them to go through bankruptcy in prior years. And when you think about these kinds of early warning indicators or post situational impact analysis. That's really interesting to the financial services market. If you want to understand how an incident is going to disrupt supply chains, it's going to disrupt financial markets. Being able to apply these kinds of analytic models on historical data that tells you a lot more information about the situation on the ground, that will help you get that incremental edge that that allows you to make much smarter investment decisions. we're excited about all of those markets, which are relatively nascent for us right now, but offer incredible potential, especially as we move away from just selling raw GIS pixels and moving towards working with partners and building our own capabilities on top of the data to make the data more useful and immediately relevant to these end customers without them having to make huge investments in building GIS teams and procuring GIS systems. Yeah, fantastic. The financial services application sound particularly interesting. So interested to see what Planet do there moving forward. I want to just dig into your technology and AI integration, given that's something you've referenced a couple of times. And then we'll circle back and cover financial growth and expectations there. How does your partnership with Anthropic change the way Planet processes and delivers geospatial intelligence? I'd say what's really exciting about the partnership with Anthropic is we've already seen that these large language models can help an end user get value from, from GIS data. And that's what these models being literally chained on cats and dogs, right? They're, they're trained on images that are on the internet. They're not trained on this very deep stack of data that we have, which the deep stack is both the historical archive. which nobody else has of this daily scan of the earth. It's the multiple data layers that are embedded in the spectral bands that we deliver. So a different spectral bands means we're capturing things that aren't something that you look at, but it's rather something that you analyze the underlying data from. As these models get trained on this type of data specifically, the question is, what capabilities could that open up? Because today, what large language models really do is it accelerates that time to value for the end customer. It can help them identify something has changed. We can see from looking at the data stack that something appears to have changed here. And because we're a large language model and we're built on all the data that exists on the internet, we've correlated that with something that has happened that has made it into the news. So we see this change. We've correlated it with this information and that exists out there. And we know now to tell you that we should flag this as something that an analyst should dive deeper into. That acceleration of value, which again used to take months, years to build these types of ML capabilities is now happening in virtually seconds. mean, they're able to do this corollary capability so incredibly quickly. And the question is, if these models get trained on Planet's really robust data set, which is built for machine learning, I mean, we designed the data sets that we were bringing to market from the start to be about machine learning, not about an analyst looking at a picture, because the volume of data that we're providing just makes that impossible to even contemplate. As these Language models become much more sophisticated on our data set specifically. What could that open up in terms of capabilities that could then drive solutions answering really robust problems that companies and governments are trying to tackle today? Yeah, fascinating. You're obviously pairing that with advanced satellite technology as well. You're constantly evolving that side of things as well. You mentioned the Pelican 2 satellites earlier. How did that compare or how do they compare to the existing constellation? Yeah, I mean, in every possible way, this next generation satellite is improving the capabilities for our high resolution fleet. So we purchased our current high resolution fleet from Google. Actually, they had acquired a company called Skybox or Terrabella. And they then sold it to us when they realized Google can do a lot of things, but space is really hard. And that wasn't a capability that they had in house or were going to continue to focus on. And so when they sold this business to us, these satellites to us, we put a data deal in place so that we could get finished the launching of these satellites and really make them produce a data set that was helpful to them, but also for a broader market opportunity. And this was back in 2017. eight years ago, you can imagine these satellites had been designed and built in the years prior to that. So we... basically took every component of the satellite and applied latest capabilities to it to make the reaction times faster, the imaging capabilities much stronger. So ultimately we want to get these satellites from imaging in the 50 centimeters per pixel, which is where our SkySat fleet is and get it to be able to image at 30 centimeters per pixel. So very high resolution capabilities, faster reaction times with the various components that we've designed into the next satellites. I've already talked about the ability to do AI at the edge. What this enables is understanding in the moment, did you catch the image that you wanted to be able to capture or was it obstructed by a cloud, for example? And so we want to pass that information onto the next satellite to make sure that we capture that image. Or did we capture that image and we've got everything that we need and we can send the data back down to Earth? even before we've processed the image. if what you wanted to know from that image is how many planes were on that airstrip, we can send that information down ahead of processing the image and getting that image presented to an analyst. Other capabilities is around revisit rate. So ultimately our goal with this fleet is to have approximately 30 pelicans that can image at 30 centimeters per pixel. that can have a 30 minute latency. So speed at which we get this data back down to earth and 30 times a day revisit. So if there's change that's happening in real time and you want to understand the pace at which that change is happening, having revisit on that area of interest becomes really important. And so all of these have been designed into Pelicans capabilities. And again, putting my CFO hat on now, the other thing we did was roughly half the cost per satellite. So whereas a SkySat was roughly $10 to $12 million per satellite, all in with design, build, and launch, the Pelicans are designed to be roughly $4 to $6 million per satellite. Well, well, while your CFO hat is still on, let's talk about financial growth and business expansion. you reported earnings results last month. I read that you achieved adjusted EBITDA profitability, which is fantastic. So what are the primary financial levers that will drive sustained profitability and cash, free cash flow generation in coming years? Yeah, so one of the key differentiators for Planet from a business model perspective is the fact that we are a one-to-many data business. So going back to Earth Observation previously, it was about having effectively a very large camera at space that you could point and click on demand from that customer. What we're doing is scanning every day so that you don't have to know that you needed to look there. The data set exists, whether we have one customer for that area of interest or tens of thousands of customers. The other thing that that enables from a financial lens is very high gross margins because the single satellite can generate millions and millions of dollars. And so our target gross margins are 70 to 80%, which is very much in line with, for example, a SaaS company or any type of data as a service business. And we've been making steady progress towards that goal. Last quarter, we were at 64 % gross margins. The other levers that we have are the fact that from a sales and marketing perspective, we have one aligned to those end markets so that we can have teams that understand the government sector and can be very focused on that one-to-one sale with very high dollar potential and customers and contracts. And we're doing the same in the commercial space where we have sales reps that will engage directly with very large potential and existing customers. But for those markets where a customer may just want to experiment with the data or understand how this data might enable their application to work better, or they know, for example, in an education and research institute, they're going to need the data for this period of time. and then they'll come back when they need it again or need a refresh of the data. And we're servicing those customers through a technology platform and building out self-service capabilities so that we can eliminate that high cost of a high touch sale and move to a low touch engagement in those areas of the market where we're either dealing with experts that know exactly what they want and how they want it, or they want to experiment and they just need those APIs and capabilities embedded in the platform. that allow them to do that experimentation in a relatively low cost way for them and low touch way for us. And that enables us to have a very scalable cost of acquisition for the sales and marketing line, which is important in our journey towards sustained, not just EBITDA profitability, but also cashflow profitability. And with the book of business that we've built to date and the large backlog, that we've established, especially last quarter and really growing that backlog significantly. What we were able to share with Wall Street was that not only do we see a path to sustain debita profitability, but actually line of sight to cashflow profitability in the next 24 months. And that's a really important milestone for any company, but especially in a skittish Wall Street market, they want to know that you're not going to be coming to market to dilute them. frankly ever. So that's one of the commitments that we've made since we went public and put a large tranche of cash on our balance sheet and we've adhered to those goals and commitments. Fantastic. OK, exciting times ahead. I want to finish the interview by looking ahead. I've got two questions here, one to do with your technology and then the final question just to get a better understanding of where you see Planet in the next five or 10 years. But let's start with the technology. What's the biggest shift you expect in the Earth observation industry over the next decade or so? I mean, I think AI is really changing this landscape as much as it's changing any, potentially even more. As I said at the onset, planets been a disruptor to the Earth observation industry, moving away from this idea of a single large satellite to a distributed network of satellites, but also moving away from the idea that you only look in the area where a customer knows that they want to look and instead you're scanning the Earth every day. And with AI, this is just going to completely change the way that anybody is using satellite data because you no longer need to have an analyst that's looking at an image and telling you what's embedded in it. But instead, you've got machines that are telling you things that you didn't even know were in that image in the first place. some of the changes that I'm excited about when I think about the potential that we can unlock with AI. is really getting that potential in the deep stack of spectral bands. As I mentioned, our tanager satellite has 400 spectral bands that are today telling us where methane leaks are. But what can that tell us tomorrow in terms of impacts to biodiversity, impacts to water quality, specialized agriculture that we can be doing that is uniquely designed for the ecosystem that you're growing those crops in? There's so much data inside of these spectral bands that I don't think we've come close to unlocking today. And as AI gets trained on the data sets that we have, I think we're going to learn things that tell us how we can exist on this Planet in a much more sustainable way and tell us those places where we are on a trajectory that's going to be harmful ultimately to human life. So Planet's mission from the get-go has been using space to help life on Earth. And I think the data sets that we have combined with AI will really unlock the ability to deliver against that mission and understand things about our Planet that we're just not looking at today because we don't even realize it's there. Fantastic. with that in mind then, and perhaps it is simply to deliver on that mission over the next five or 10 years, but what does success look like for Planet over that time period? I mean, for us, it's about moving from just geospatial data analysts digging in and understanding what's embedded in our data set to any data scientist anywhere using this data set to solve problems that we're not even thinking about today. you know, when you think about the success of a company and using, for example, Jeffrey Moore's analogy of crossing the chasm from those early adopters to the early majority, A big part of that is having a whole solution that enables those more skittish early majority customers to get value quickly from the data. And that's the shift that we've been making. Our early adopter customers are some of our biggest fans and they work with us to make us better in service of this next wave of the market that we're capturing, which is a harder market to capture. but much, much, much more valuable. So I'm excited about the wave of new customers that we're not tapping into today, really getting value from the data. And that's both an excitement around the financial value that we're going to unlock with that, but also the impact that we can have through those next wave of customers. Yeah, fantastic. And I think a perfect insight on which to end the interview. I think that just leaves me to say thank you very much for joining me on the podcast. It's been a real pleasure. It's been a pleasure from my side. Thank you, Hayden.