In this episode of On Point with Korem, I’ll be speaking with Matt O’Connell, he’s the former CEO of GEOEYE and an operating partner at Data Collective Venture Capital. Matt, long ago, saw the value in satellite imagery as a play in the data market and is now looking at the value of geospatial technology and imagery data in both the transportation and logistics industry as a form of information arbitrage. It’s a great discussion with someone who is looking at the geospatial industry from a financial perspective.
Joe Francica: Well Matt, thanks for taking the time to do this with me. I want to take you back to some of the old days, where you were a CEO of GOI, before the merger with Digital Globe. And one of my observations, because I came out of that business in satellite remote sensing, and it seemed back then that there were two types of clients; you either had government clients and maybe everybody else. But the business model of the satellite companies back then seemed to be differentiated into very two, and there wasn’t much growth beyond that is that a fair assessment of where the business was maybe five or ten years ago?
Matt O’Connell: Absolutely Joe! I mean, not surprisingly you’re right on the money. You know I remember getting into the business in say 2001-2002 and I wanted to go into the commercial industry but there really was no commercial industry. We didn’t get Google Earth until 2005. I had been in the cable industry and in the cable industry we say “before HBO and after HBO” because HBO literally changed the landscape. I can get movies on TV. Same with Google Earth. Everybody in the geospatial industry should still salute Google Earth, because it was the biggest promotional mechanism any of us had, post 2005. People in the agriculture industry could look at Google Earth and say “well I’d like to see what I have?”… the real estate industry, the transportation industry. So you didn’t get much commercial until after 2005 and even when we sold GOI in 2013, I would say of our billion dollar valuation, two-thirds was government related. And then within government, I would say three quarters was defense and intel, so most of it was defense and intel and then some of it was government civilian, and then you had some commercial. Google was a terrific client. We took Google away from DigitalGlobe early on and I had a great relationship with those guys, but we realized at the time there was a sort of a land grab, they were all trying to map the world and I said “folks, they’re going to taper off. So, let’s find some replacement stuff” which is why I went into products and services.
JF: So, let me pick up on that a little bit because in the intervening 10 years, or so, since that time certainly Google has floated the boat of a lot of the geospatial industry. What’s changed since then in terms of the uptake in commercial? Is there…and the way I want to ask that is, is there really an understanding of the value of imagery as a product? As a data product? As something where you derive information from it, versus the pretty picture that you know we see on Google Earth?
MO: Great question Joe. I think that the industry specialists, like the guy who runs the contract between the US Government and the geospatial industry, I know him, I never mentioned people by name in the intel community, but we were standing at a reception in Monterey and he said “You know, I’ve told the guys at MAXAR and Planet Labs”… which I should say as a DCVC company (where I work), he said a couple years ago “I don’t want imagery anymore. I want products. I want trends. I want analysis.” So he gets it! Most people don’t get it. And we are still educating people as to the value of imagery and what it can tell you. That’s where companies like yours, Korem, come in. Because, you’ve got to help most people realize the value of context. It’s not just a pretty picture. It’s where am I? Where is my product? Where is my customer? How do I get one thing from one place to another and how are the dynamics of that transportation going to affect people? You look at what’s going on in the supply chain right now Joe. We have a historic number of ships off of the port of LA and Long Beach. The port of LA and Long Beach is the biggest port in the Western hemisphere, and it has never had as many ships sitting at anchor for two, sometimes three weeks. So the lockdown came and people had to order stuff more because they couldn’t go out, and shipping has just…(writ large; trucks, trains, boats, etc.)…having people understand how location information can help them solve those problems, is still something we spent a lot of time on educating. And that’s why I’m glad you do these kinds of podcasts, is to get the word out and say to people we can help you solve your problems, we can help you make more efficient decisions. Then of course, tying those systems together, which is what you guys do for a living. You’ve got to tie those systems together so that ordering then knows how to talk to something else. A friend of mine made a huge hit on food delivery in Saudi Arabia, and one thing he did was tie the order to inventory… “oh you’ve made 150 pizzas, you’re going to run out of tomato paste on Thursday if you don’t order more tomato paste.” And that kind of information and decision making is really valuable!
JF: So in the satellite business and this is where I think the rubber kind of meets the road in particular with today’s cloud native processing machine learning, which is another interesting buzz word we used to call it “image classification” when I started, now they come up with the term “machine learning” which I think is always interesting, right! We gotta change the nomenclature just to think that we’re in the know. Sorry, that’s an editorial response.
MO: You’re the consultants! You know, you do this for the rest of us.
JF: But anyway, it was always hard back then. You know, you do image processing you don’t want, as you just said, you don’t want the pretty picture, you don’t want the raw data, you want the answer right! I got to have the answer, you want to know how am I going to get that tomato paste over to my pizza parlor and there’s a disconnect between the data, the raw data, and the answer and I think that that’s still a main problem for our industry and at least educating the commercial side of the business. The government guys get it but the commercial, maybe not.
MO: I think that the smarter operators do. If you read the quotes about different people going public… I’m limited in what I can say about Planet Labs because they’re a company of ours and they’re going through the SPAC [Special Purpose Acquisition Corporation] back right now… and one thing they said is “well we don’t think we’re going to spend a lot more money on satellites. We’ve got a good constellation up there, we’re generating a lot of valuable data, we’re going to spend more on products and analytics”. And when I first got recruited from Seraphim to DCVC, just about a year ago, literally right as COVID started, odd timing! But I had gone to see Planet in February before that, February of last year, and all I wanted to talk about with Robbie Shingler, one of the two founders, was products and analytics. You got a lot of beautiful data, how do you make it more usable? When you think about it, what Steve Jobs did was he didn’t actually develop stuff that wasn’t around. I saw a small music device like the iPhone, before Apple came out with it. Jobs made it easy! He made technology easy. If you can make technology easy, that’s where you make a ton of money. And I think we’re still working on that in this industry. The industry was founded by engineers, for engineers, and one the reason why GOII grew so fast is, we tried to simplify it. We also tried to figure out what the client wanted and then work with them on that, and that’s not easy. So we’ve gone from the industrial age in this industry where we were bending metal. Now we’re in the information age where we’re bending pixels, and it’s not how can you bend the tin to get it to do what I want, can you do an orbit at this level? Can you have this kind of camera? Now, it’s okay I’ve got these pixels. How do I tie them together or how do you tie together different sources of information? I hired a guy to be my CTO, Brian O’Toole, Brian is now running Black Sky and he’s doing something very clever, he takes feeds from the social networks and other people; like Hawkeye 360 (but those are not my companies by the way but I like them both), and so as they’re going over the world, their satellites say “Hey, the Twitter feeds are way up over Yemen, let’s use that to tip and cue the satellite”, this points the satellite at Yemen, and maybe we’ll derive some information. And then automatically, this is something I created, and I said to Brian, “I want to look at social networks and I want it to be automatic.” So automatically an email goes from Black’s guy to NGA; “Hey, Twitter feeds up, shout out in Yemen, you want it? And one machine talks to another, they go through different parameters and they say “ah that’s interesting.”
We have connected Black Sky and Hawkeye, both of them with Capella, which is one of our companies. They have small satellites for SAR radar, and they can see at night in storms. We’ve also connected them with Hawkeye 360, Inspired, I mean we’re talking about trying to get them to talk to everybody. It’s when you combine different sources of information, that’s when you get some really interesting products. That’s also when you get complexity, which is great for a company like yours, Korem, because putting all these different products together ain’t easy. SAR imagery is different from EO, is different from drones, but you put them together and that’s when you get some real razzle dazzle.
JF: Yeah you said a couple interesting things there, I want to pick up on. One of which is the source analytics side of it which is putting together other location-based data with satellite imagery. I’d never thought of that before but that’s an interesting application where you know you want to derive certain information from certain geographies because other things are happening. And I can certainly see that even in the commercial space with mobile trace data, where all of a sudden you’ve got this collection of people at downtown Starbucks, “What the heck is going on there? Oh well, you know Pumpkin Spice latte just came out!” I mean it’s a weird analogy to what the intel space is doing but it’s similar in respect.
MO: Yeah I mean you look at what did Waze teach us? If you want to map roads and if you want to know where traffic is, satellites are great, but the cell phones better! It’s tracking everybody! You know, I don’t want to get political but a friend of mine said he wasn’t going to get a shot because he thought people could be tracked if you had a shot and I said “you got a credit card and a cell phone, trust me they know exactly what you are, and they know how many lattes you’re buying.” I told the guys at Spire “you track ships”. I told the guys that at Black Sky “one of your products is pictures of ports, i said let’s get together”. Then we tell the ship operators “hey, throttle back when you get to the port of Los Angeles, you’re going to have to sit for two weeks. Why burn up the oil? Sit down. So there are a lot of interesting developments. One of the things that is very gratifying is, so launch costs have come way down, SpaceX was a great help. Blue is helping (a company we’ve invested in). Rocket Labs is helping. If one of the biggest costs comes down, launch, then you can put more up there, then you can try this new stuff, and then you can say “hey this works. This doesn’t work, and you come up with new products. So that’s very exciting for people at the same time. You know Covid kind of unsettled the capital markets but there is a whole lot of money going into space right now. It’s a great time to be an investor, if you have stuff! So we’re glad we got into Rocket Labs early. It’s nice to be able to share their upside in the SPAC, same with Planet. One of the things that we are constantly challenged by now is everybody comes along and they may not have even gone to space and they say well we’re worth a billion dollars.
JF: Yeah so one of the things, and I know you can’t say too much about the companies you’re invested in, but when I heard Will Marshall come out and say “hey, we’re a data company“, that was a real change in focus of the satellite image companies and to me it was a repositioning of; we’re not just gathering pixels anymore, we’re really gathering information. And frankly, I think the whole geospatial industry has transitioned from the software phase and now data is more important. So, I was happy to hear Marshall say that and that’s where the money is, right! It’s in that data. It’s in the information.
MO: I totally agree but then I said it in 2002. I got into GOI, I took the rockets models and the satellite models out of the lobby and I put them in the satellite operating center, and there was this riot, literally! The employee said, “what are you doing” and i said, “we’re not a satellite company” and they said, “what”! I said, “we’re a data information company” and it was a big challenge. Now I will say I was gratified to hear Will say we’re an information company, I was like yeah I said that what 20 years ago, yeah 20 years ago. Hearing him say we’re going to invest in products and analytics, that really appeals to me because people have been talking about being information companies, you know; Digital Globe, Max R, but they’ve never really committed the funds to do that and Walter Scott of Digital Global, very smart guy but he’s a rocket and satellite guy by training, not a software guy. I hired Brian O’Toole, who’s now running Black Sky because Brian was a software guy. You have to have the software products and then you have to have people who will stitch it all together. Because all the different systems have to work with each other and then they have to work with the client’s information. I’m looking a lot at logistics right now because the supply chain is so messed up and frankly, it’s a much bigger industry. You know mapping is nice but it’s what, a 10 billion maybe 20 billion a year. Logistics is somewhere between 150 billion and a trillion, depending on how much the infrastructure you can. So I heard something interesting, somebody has technology that will make trucking yards more efficient but one of the operators said, “yeah but I’m not replacing my yard management system”. So this little company came up with a smart way to do it on a SAS basis and said you don’t have to replace your system, just drag our tools down to your legacy system (even if your legacy system is 15 years older than a piece of junk), but I think that you have to invest in the data and analytics, and then you have to figure out in your go to market plan, how to make it easy for the customers. I’m sure that’s a lot of what you guys are doing at Korem because it is not easy to put all these different image resources together, all the information sources, and then have the different systems talk to each other.
JF: Well, I think you’ve raised a good point. I mean the transportation logistics industry is, I think, a key industry. Certainly a key industry for us at Korem. But more to the point it gets downstream. It gets the information downstream from the collection device to the actual use and analytics. And you’re right, we so depend from an e-commerce standpoint about ETA, right… I want to know exactly when my stuff is going to be delivered. And here we are, approaching the holiday season where e-commerce is going to explode again, we’re still in the Delta variant situation. And that’s where I see the potential. You obviously see the potential in the transportation field and you’re right. It’s right for the picking for geo.
MO: Yeah, I read the Wall Street journal logistics column every day now and some might say that it’s depressing. The BBC just had an article about how shipping has never had more bottlenecks. Well that’s not depressing. I see opportunity! That’s what entrepreneurs do, is look for pain, how can I solve the pain? You know when that ship got stuck in (the canal), the word finally got out but it shouldn’t be “finally got out” it should be “bam”! Like Waze, “hey, don’t go on the canal. This is gonna take us a while. Even though it seems longer, you’ve got to go down and around Africa.” So solving those problems I think, is going to be a big challenge but it’ll be fun and I think very lucrative when people get it done.
JF: So let me ask you one final question. Certainly, from a from an investment point of view, I mean you just said that the geo industry is a 10-billion-dollar industry. Relatively small in comparison to others. What’s it going to take to get to the next level, do you think? Where do we need to break out? Where do we need to position companies?
MO: I think products and analytics, as I said before. There is a lot of information in the stuff that the geospatial industry gathers but it has to be made user friendly. Even some of the leaders in software are not sufficiently user friendly and they think more from the engineering or technology standpoint, than from the user standpoint. I don’t want to cite somebody who might be a competitor, but IBM had a great ad for years which said, “Stop selling what you have, start selling what they need” and I took that as a mantra at GOI and I said let’s stop selling what we have, let’s go figure out what they need. Larry Page of Google, has a great phrase “figure out something that somebody is going to use every day like a toothbrush, so if they’re going to use it like a toothbrush every day, you’re going to get a lot of uses”. So we have to help people figure out how we can solve those pain points, like in the shipping industry, how can we make it more efficient? Why are there these bottlenecks? By the way, think about example; before ships coming forward along LA and Long Beach and it’s crowded, so throttle back. If they throttle back they’re burning less oil, “oh by the way that helps our emissions picture”. “Oh well, maybe I can get a better ESG credit?” And there’s a whole lot of information going on right now that’s derived from satellites and drones. DCVC just invested in a company called Kairos that flies planes over oil and gas manufacturing facilities, refineries. Every refinery emits some methane, it’s just a fact of life, it just comes out. What these guys do is help the refineries find that and plug it. Now, they would like to plug it anyway because no one likes leaks. With all the focus on ESG it becomes more compelling for them to plug those methane leaks because then they can say, “hey, you know you may feel badly about oil and gas but at least we’re doing the best we can to produce it in a non-polluting way”. Whereas, DCVC prefers stuff that is oriented toward alternative sources of energy or more energy efficiency. We’re happy to help the oil and gas guys become more efficient, if it then reduces the bad aspects of the production facility. That’s a geospatial product that, I don’t know if people would have thought of a while ago and that’s good on all sorts of levels.
JF: Well Matt, thanks! This has been a really fascinating conversation. Your insights are always appreciated and again thanks for taking the time to do this.
MO: Joe, it’s always a pleasure talking to you. It’s fun talking to somebody smart! It’s like playing good tennis, you know what I mean! This is like the U.S. open of geospatial right here Joe!
JF: I love that! When I get my tennis game in shape, I’ll come back, and we’ll play together.
MO: Okay, great. Thanks!
JF: Thanks Matt. Thanks again for joining us on another On Point with Korem and if you like today’s podcast, please leave a comment in the comment box where this podcast is posted, which could be Apple podcast, Google podcast, Spotify or YouTube. I hope you’ll join us next time for another On Point with Korem, where we’ll get on point.