Episode #51 -This episode is all about why it is so difficult to build a really big geospatial business and what it would take to disrupt ESRI. My guest on the show is Ariel Seidman CEO and Co-Founder of Hivemapper
This episode is sponsored by GraphHopper
GraphHopper provides web services to add for route planning, navigation and route optimization to your application in the food delivery industry or for traveling salesmen.
We made some socks for you!
You are more than welcome to reach out to me on social media, I would love to hear from you!
In Conversation
From Yahoo Maps to Hivemapper
Daniel: Welcome back to the podcast, Ariel. I’m really pleased you could take the time to do this with me. I say welcome back because you have been here before — in fact you’re the very first person I interviewed on the show. You are the founder of a company called Hivemapper, and Hivemapper has also been a long time sponsor of the show, for which I’m truly grateful. But that’s not the reason you’re here today. We’re going to be talking about the business of geospatial, different models in the geospatial industry, and looking at some of the big players — breaking it down and walking people through what’s working, what’s not working, and perhaps where the industry is heading in terms of business models. Before we dive into that, perhaps you could give a brief introduction of who you are.
Ariel: I started off in the mapping industry probably more than 10 or 12 years ago. I worked on Yahoo Maps and Yahoo Search, when 30% of our queries were all location and mapping based. That was really what triggered my interest in maps and how incredibly challenging and exciting they are to build. I then started a company called Gigwalk, which was effectively mobile data collection at scale — two to three million users who were all collecting local information through their phone for large mapping companies, businesses, the World Health Organization, these incredible organizations.
Ariel: And then I started Hivemapper a little over four years ago now, really trying to build a new, smarter mapping product leveraging video — video that is coming from airplanes, that is coming from drones, that is coming from dash cams — and turning that video, that raw ingredient which is everywhere in the world, into intelligent, smart three-dimensional maps. Really think about it as Hivemapper is doing with video what Waze did with GPS. We’ve been on that mission for over four years, and last year was just this incredible transformational year.
Why the First Three Years Were Brutal
Daniel: So you’ve been doing Hivemapper for four years, and you said last year was a great year. The first three years — they were difficult. Can you tell us what it was about those three years that was so tough?
Ariel: Building maps is hard. I mean technically, just some very deep technical challenges. What we’re trying to do is not just build 3D models, but really we think about it as building this decentralized mapping network. What that means is that a user today who is flying a drone, and another person who’s flying an airplane in the same area, are all adding their map into the same underlying global map. So you can have many, many participants in the map, all feeding a single global map. Without that key piece of technology, what you’re really just building are these little isolated surveys, and they’re all disconnected from one another. There’s no ability to really understand what was there even a week ago if you’re not organizing everything and stitching everything together into a single global unified map.
Daniel: Clearly there are a whole bunch of technical challenges there. One of the challenges whenever you bring something new to the industry is the story around how you sell it in. Has that been a challenge for you? Could people immediately see the value of what you wanted to build? I’m assuming at the start you don’t have a finished product, you have a minimum viable product, something that solves part of the challenge. What did that look like for you — you went to someone and said, hey, I’ve got this thing, we’ve solved part of the challenge, are you in? Was that difficult, or did people jump on it straight away?
Ariel: There were definitely some very early customers that, without them, we would not be here today. They took a bet on us very, very early. The piece that they loved about Hivemapper was two things. One is this decentralized mapping network. These were customers that had all of these aircraft and cars collecting video through dash cams, drones even in many cases, and so they wanted to leverage all of these videos that they were generating on a consistent basis and do something more productive with it — rather than just having these video streams that nobody ever looked at again. That piece of the problem was incredibly valuable to a couple of very early customers, and so they gave us the traction and the revenue in the early days to continue on that mission.
Ariel: But the next set of customers were like, hey, this is awesome, I love that you can build maps cost effectively, I love that it’s decentralized, I love that there’s a single global map — but what I really need on top of that are really powerful tools, so that the end users themselves can get a lot of value out of that data. So that meant we not only had to build the core technology that enabled this decentralized mapping network, where anybody can feed video in and it all combines into the same global map, but then we had to build a set of tools that normal people could use.
Tools Normal People Can Use
Ariel: So it’s: hey, I need a timeline so I can scroll back and forth in time. I need change detection so I can analyze what’s changed here. I need object AI so I can filter out the objects that I do care about versus the ones I don’t. If I’m doing an AGL calculation because I’m landing a helicopter or flying really low, I need an above ground level tool so I can very quickly assess what are the tallest objects in this specific area. All of those tools are technically interesting and challenging problems, but the user experience also really matters.
Ariel: Just like Excel in the early days really nailed some of the key user experience elements around spreadsheets, and Salesforce, in terms of managing all your sales and all your opportunities, really nailed some of the user experience aspects of it — you’ve got to focus on the user experience aspects just as much as the technology. Because if you build out all this incredible data but you make it really hard to use, guess what, it’s only going to be used by these GIS experts and not by normal people within these businesses.
Daniel: I could completely see that happening, and I’ve experienced similar things myself. You build something, it works for you — but can you teach someone else how to use it? Is it intuitive for them? Are you building it for yourself or for the public? That makes a massive difference to what you’re building, how you’re building it, and how you sell it in at the end of the day.
How Big Is the Geospatial Industry?
Daniel: One of the questions we’re going to be discussing today is how do we disrupt Esri, how do we build a company that’s bigger than everybody. But just before we get into that, maybe we could take a step back and look at the geospatial industry as a whole. How much is it worth each year? Can you give us an idea of the size of the industry?
Ariel: Well, it depends whose numbers you believe, but generally speaking it’s somewhere in the area of 200 billion to upwards of 430 billion dollars. That includes a lot of different dimensions of the geospatial industry. Obviously it includes software, it also includes the actual sensors that are built for purposes of mapping, and it includes a lot of professional services in that mixture as well. These are people who are taking something like ArcGIS, maybe in combination with other types of tools, and creating a custom solution for a customer — solutions that are incredibly valuable for the customer.
Ariel: It also includes all the data collection services, of which there are a ton. These could be mom and pop shops, all the way to much larger companies that are putting satellites up into space. Fundamentally these people are collecting the raw ingredients that then get fed into various software products, which then turn those raw ingredients — the imagery, the video, the lidar data — into something that is incredibly valuable for an end user and customer.
Ariel: You’ve also got to remember that in the geospatial industry you have Google, Apple, and then you basically have everybody else. What I mean by that is that Google and Apple are investing ridiculous amounts of money in their own collection mechanism. All those cars driving around collecting Google Street View are an example of that — that is a capital intensive program, and they’re doing that on a very consistent basis. So just Google and Apple are spending in the billions on an annual basis building and refreshing their map. And then you have the rest of the industry as well.
Why Collecting Pixels Is a Race to the Bottom
Ariel: If you break it down maybe one level further — the guys who are collecting the raw ingredients, that industry honestly is a race to the bottom. Because hey, I got a good pixel, you got a good pixel, my pixel was from today, my pixel’s from maybe a couple hours later. The number of different dimensions that you can actually compete on is limited. It’s actually a really hard company to build and a really hard problem to solve, but from a purely financial, economic perspective, the moats are decreasing, and the competitive dynamics associated with selling one pixel against the other are not fantastic.
Daniel: So what I got out of that is a huge industry, a lot of money being spent, but also a very diverse industry — a lot of different players at a lot of different levels, and some big ones that perhaps people don’t often think about in terms of geospatial. You mentioned Google and Apple there. Then you talked about the raw ingredients. There seems to be a lot of focus on this at the moment, and mainly when we talk about this kind of collection of pixels we’re talking about space-based platforms, which are hugely expensive to operate, hugely expensive to get up there. Why do you feel like we’re racing in that direction to create these raw ingredients, if they’re very rapidly becoming a commodity and therefore a race to the bottom? What is it about the space industry that’s so attractive right now?
Ariel: I think the cost associated with launching satellites is obviously coming down, SpaceX being a component of decreasing costs. Using commodity sensors and other commodity hardware to actually build the satellites themselves — obviously you can create satellites now that are capturing imagery of the world at a much lower cost than DigitalGlobe was 15 or 20 years ago. Anytime you have a significantly lower cost technology that is enabling much lower cost solutions, it creates a set of excitement.
Ariel: But at the end of the day the product that they’re delivering is very similar, and has certain financial constraints around it that fundamentally haven’t changed from 15 or 20 years ago. You’re selling imagery, you’re selling pixels. Regardless of the fact that there’s all this great new technology, that is the end data product that you’re generating and selling — and that’s a rough business. If you go further down into the stack, even independent data collectors, while an enormously valuable and important role — if you’re flying airplanes up and down various cities collecting lidar data or collecting imagery, you’re running an airline business, fundamentally. You’ve got to think about okay, how much gas am I going to use, how much do I have to pay the pilots, how much is my equipment to put on my airplane. That in many ways looks a lot similar to an airline business, and we know the challenges associated with an airline business.
What Makes Esri So Successful
Ariel: Then you go one layer further and you have all the different tools and software packages that are taking all of that data and trying to do something meaningful with it, trying to create data products that end consumers and business users could actually use. The big boy in the room, so to speak, in that category is obviously Esri. Successful company, founded by Jack and his wife 25 or 30 years ago. Amazing story — they never took venture capital money and just grew the business organically to what it is today, which is probably doing close to two billion dollars a year in revenue.
Daniel: So we back up just a little bit here. We don’t want to be in the business of anything that’s a commodity, so we’ve identified that collection of pixels, collection of data — that’s not for us. We look over the fence, we see Esri, they’re doing a great job, they’re an incredibly successful privately owned business. Let’s start with what makes them so successful, and how would we disrupt them?
Ariel: It’s a great question. There are a couple of different elements. Let’s first define Esri — a software-based product, sold at, I think the minimum price for that is roughly six or seven thousand dollars a year for their base product, maybe a little bit less. Usually if you’re an enterprise user you’re paying quite a bit more, because there are all these add-ons that you have to get, so you’re definitely pushing the price up quite a bit once you get into the enterprise zone on a per-user licence basis.
Ariel: A lot of their business, from what you can tell — and there’s not a lot of public information on it — is that they have I think 400,000 businesses and organizations that use or pay for their product. Obviously a lot of those are cities. The city of Pasadena is a great example. I think Pasadena has something like 100,000 residents, and they pay Esri on an annual basis for a licence of 150,000 dollars. So one way to think about their business is it’s a dollar fifty per resident of the city of Pasadena for all their GIS.
Ariel: But there are a lot of layers. There are other businesses in and around Southern California that also use Esri, there’s a county down in LA that probably uses it, I know for sure that various offices of the state of California use it — and they all get to share data amongst themselves. That creates this tremendous lock-in effect that Esri has with some of these organizations, because they’re all using the same data. It would be like everybody sharing and managing and collaborating off of GitHub or Salesforce, all seeing and sharing that same data. That becomes really powerful as a lock-in effect for those businesses.
Daniel: I heard you allude to the network effect and lock-in. These are things that we see in Apple, for example — once you have an Apple product it’s better for you if other people have Apple products as well, it’s better for you if you stay within the environment, things become easier. And I see this is something that Esri has done really well. To be clear here, we’re not bagging on Esri, we’re not giving Esri a hard time — we’re dissecting the business model, looking at what they’ve done really well and why they’ve been so successful. Firstly I think they’ve picked a vertical which was more likely to succeed — it’s not commoditized in the same way that data collection is becoming. And they’ve built the network effect and lock-in into their product. Can you talk a little bit more about the value chain of something like Esri?
Ariel: A couple of thoughts on this. The Esri value chain is not just the software tools that they provide — those are obviously really key — but in addition to that is all the data that they stream in from third-party data providers. Some of it is free, some of it is just publicly organized datasets that they’ve either curated or managed or pulled in and extracted from a lot of different sources. Kudos to them for doing all that hard work in terms of data curation. When you open up Esri, for the most part it’s not an empty shell, it comes with all of this great data. And I do know that they have deals with third-party data providers where if you want you can very quickly purchase additional data, so they make that part of the experience quite easy to use.
Ariel: The other part is that they have a tremendous number of system integrators who can take ArcGIS, the core product, and then customize it in a thousand and one different ways. This is not so different from most enterprise software companies — Salesforce being a good example of this, or Microsoft Office. You have all these value-added resellers who can resell that product, customize it, turn it into their own. There are people who take ArcGIS and say, I’ve created an ArcGIS solution just around ports. So if you’re Long Beach port and you need a GIS solution to manage all your different — the Port of Los Angeles or the Port of Long Beach is massive, so you definitely have a GIS problem and GIS data — you’re using this custom solution that’s been created by a third party around that.
Ariel: But they’ve also done a really nice job in terms of curating and managing this GIS community over 25 or 30 years now, where people bleed Esri in a lot of ways. There’s a tight-knit community within that world, and kudos to them for doing that. That takes a long, long time to do, but they’ve done it, and it’s been a key part of their overall value chain.
Daniel: I completely agree. What you’re talking about there is something that often gets referred to as super fans, or a community. They’ve built a community that really believes in what they’re doing, loves the product, and they’ve been using it for so many years now that they just can’t imagine doing anything else.
How You’d Actually Disrupt Esri
Daniel: Okay, so me and you, we’re going to build something that’s going to disrupt Esri. We need to build something that has lock-in in some way, we need to build a community, we need something that has an add-in or plug-in environment, and we need a network effect — and it needs to do the job amazingly well. This sounds like they’re going to be difficult to disrupt. Is there anyone out there who’s on the verge of doing this? What about open source, for example — are they going to disrupt Esri?
Ariel: I don’t think open source will. I think it requires a business to be able to manage it. There are — QGIS obviously is open source — but I think we’ve all tried to use that, and I personally find it incredibly frustrating just because of the user experience. It has a lot of capabilities, but those capabilities are not packaged in a way that makes that tool incredibly easy and accessible for normal people to use.
Ariel: So I would argue that one of the ways to disrupt Esri is to actually go beyond the GIS user. The GIS professional is incredibly important, but a lot of the workflows today in some organizations all have to go through a GIS person. So how do we rethink this, where the GIS user can empower a hundred people, 500 people in the organization to effectively be self-serve? If you can create a tool that’s easy to use for normal people who maybe are not dealing with GIS data on a daily basis, then I think you have a really good shot. And what that means is the business model has to be different, right, because you can’t charge 5,000 bucks per user per head if you’re trying to get your product in front of 500 people. So you have to rethink the business model, and you also have to rethink the delivery mechanism. It can’t be a Windows desktop only product, it has to go way beyond that. Web browsers are everywhere, they’re incredibly powerful — there are Photoshop and Illustrator level tools now that are in the browser, so we can definitely build the power of a mapping GIS product in the browser now. I don’t think that was the case maybe six or seven years ago, but that’s definitely true today.
Daniel: I completely agree with a lot of what you said there. Web mapping is that distribution mechanism you’re talking about — not only can I send it out to lots of people in a format they’re comfortable with, in a browser, but I can also customize it. You don’t need to see everything; I can limit the functionality so you can just solve the problem you’re interested in solving. So I guess this begs the question: is that what open source is missing if it’s going to be a competitor? For me it has a whole bunch of functionality. Perhaps it’s not as shiny as Esri products, they’re very polished. It has a ton of functionality, it has super fans, people that are dedicated to open source, there’s more and more data available and it’s becoming easier to source and drag and drop into these products. But it’s not all bound together in the same way that Esri is — that’s what I see as being the biggest hurdle here.
Ariel: I think also businesses want something they can rely upon. GitHub is obviously built on an open source project, but even developers who are incredibly technically sophisticated — that market of users demanded and needed something like GitHub to manage many key parts of their workflows, in terms of managing the repositories, managing the check-in and checkout of code. All of that stuff needed to be managed in an easy to use, reliable, consistent way. Businesses need that and require that. If that’s such a core part of your workflow, you don’t want to be relying on the whims of an open source project.
Daniel: What I hear you saying there is: we’re a business, we want stability, we want to know that you’re going to be here tomorrow and the day after that, and continue to support these lines of products. Personally my feelings on open source are that it’s getting better and better, and I can see it slowly creeping into the industry, into larger organizations here in Denmark. But I think for a long time the story that Esri could tell was: we are here, we’ve always been here, we’re going to continue to be here, we are stable, we are reliable. And that is a huge plus for a business.
Ariel: Absolutely. I think that’s one of the reasons why I don’t think the disruptor to this will be open source. The other thing with Esri though is that Esri is clearly a very successful company — like we talked about, probably approaching two billion dollars a year in revenue. But if you look at the overall macro geospatial market, it’s still a relatively small component of that. If you contrast that with other big enterprise software companies like Salesforce or Microsoft Office or even Adobe — Salesforce, don’t quote me on this, but I think is doing like 10 or 15 billion dollars a year in revenue, if not more. Why don’t you have a company in the GIS mapping space, a software-based business, that is doing 10 billion or 15 billion dollars a year? To me that is the big opportunity that exists, and that is the thing that we should all be asking ourselves — why doesn’t that exist? That’s definitely something we’re very focused on here at Hivemapper: what will that product and business look like that can hit an even bigger market than what Esri showed us is already capable of doing.
Daniel: For me that’s still an open question, I have no really good answer to it. One of the reasons I could see is that geospatial, even though it’s baked into so much of what we do — into our phones, into the way we navigate, the way we receive messages, it is all somehow located, in an effort to personalize things and segment audiences location is a big part of that — so it’s all around us, it’s happening in lots of different products that we use today. And yet we still think of it as being a very specialized thing. We still think of products like Esri as not tools for everyone — this is tools for people like this over here, and not the masses. I think there’s going to be a big shift that’s going to have to happen before we see a company like Esri really grow to some of these other examples you’ve given.
Ariel: I think that’s right. The other part that you were alluding to is this idea that, unlike let’s say payroll software, where the number of use cases is somewhat finite, the number of use cases within the mapping GIS world — there are obviously some very dominant ones like navigation, but even navigation, if you break that down, has a lot of different use cases and nuances to it. The macro point here is that the number of use cases in the GIS mapping space is incredibly large. If you’re going to build a product that can get to that size we’re talking about, 10 or 15 billion dollars a year, you’re going to have to have a suite. And that suite is going to have to touch many, many different use cases in very accessible and simple ways. That is not only because maps are so technically hard to build, but that’s also an incredibly hard user experience problem — how do you not lose control of the product, how does it not become this massive mess of a hairball, while still addressing a significant percentage of the big use cases in the industry?
Daniel: When you frame it like that, what I could see happening is a company like Esri powering a whole bunch of these use cases — so not necessarily seeing Esri on the label of whatever it is that you’re using, but that Esri engine is somehow buried in the background providing the analytics, providing the maps. I could see that would be a solution where it could scale, where Esri is in the background doing some of the analytics but a different product is sitting in front of it. Is that what you were alluding to?
Ariel: I don’t think that’s what’s going to happen. You will definitely have pieces of that happen. I think more likely — trying to think of a good analogy here, this is maybe a little bit older — Salesforce disrupted Siebel Systems, which was the early CRM company in this space. Siebel is not really existent now; I’m sure there are companies that still use it, but most people have never heard of it. The reason they got disrupted was two things. One was business model: fundamentally Salesforce as a SaaS-based product was different from a licence-based product. So I think business model is going to be a key disruptor of Esri. Unclear exactly which specific business model is going to be the one that disrupts them, but I undoubtedly believe that business model will have to be a key component.
Ariel: And then the other thing was just the movement from desktop software to web-based software. You didn’t have to install a desktop, you just ran Salesforce in the browser. That was actually the only way to run it — the only way was in the cloud, there was no on-premise at all. So I think those things in combination, with a lot of great marketing and a lot of great salesmanship and a lot of great execution, was the thing that disrupted Siebel. I do think that it won’t look exactly like what we just described, but once that happens I just think Esri doesn’t exist, or ceases to exist at the same scale that they exist today.
Daniel: That’ll be a really interesting time.
Ariel: I think that will take a while, to be clear. But I do think in the next six to eight years you will have a public company that is three to four times the size of Esri.
Daniel: Wow. Okay, let’s catch up again in a couple of years and see how we’re going with that.
Why Geospatial Businesses Are So Hard to Scale
Daniel: Let’s move on now and discuss a bigger issue that we’ve kind of alluded to. Why is it so difficult to build businesses that scale in the geospatial industry?
Ariel: I think we touched on it a little, but it’s worth getting more precise about it. For starters, the underlying technology in the mapping space is really challenging. It’s not like you’re building — I keep picking on payroll software — a SaaS-based product where you have a set of rows and tables and a database, and there you go. For the most part you’re building a product that oftentimes has three dimensionality to it, these incredibly large datasets. From a technical perspective the challenges there are 10x what they are for a normal SaaS application.
Ariel: Then you have a data problem, which is: showing up as a tool without any data is a non-starter. You have to either give customers the tools to build their data, or show up with a lot of meaningful data. So you’re in the situation where you have to amass and aggregate, or build the tools that enable customers to build their own data, and then deal with all the nuances of that data — whether it be really heavy lidar data, or satellite imagery data, or data like we have at Hivemapper, video that transforms into maps. This data issue is incredibly challenging, and it takes time to aggregate it and accrue it and curate it. And then you also have to address the problem of making your tool easy to use, and then you have to finally take it to market, price it, package it, put all the marketing material around it. So the level of effort and time that it takes to get some sort of traction in the market is fundamentally just longer than other types of businesses and other types of products.
Ariel: I think the only one that maybe bears some level of similar complexity is some aspects of payments. I’m definitely not an expert in payments, but from my friends who are, as they describe the plumbing behind the scenes that needs to happen in that industry, it is incredibly challenging. So I think that’s probably the only one, just from a technical perspective. But that one — you’re dealing with a situation where it’s pretty data light, it’s low bandwidth data. The data problems that we have are incredibly high bandwidth, incredibly dense data. So you have that other aspect of it as well.
Ariel: I was actually doing a study — Waze was starting in 2007 or 2008, and Waze was really the last mapping company that really stressed this idea of collaborative mapping. Yet since 2006, 2007, 2008, if you look over all the other industries that are doing some sort of software productivity tool, you have GitHub in the developer space, you have Figma in the user and design experience space, you have obviously Google Docs in the docs and spreadsheets category — all these companies building these collaborative tools. But in the mapping space, just given how challenging it is, you didn’t have anybody else really take the mantle of what Waze was doing for almost 10 years. That’s a long, long time in the technology industry.
Daniel: And I guess that proves your point a little bit — that it’s hard. If it wasn’t hard someone would have been there, because the opportunity was big. I just want to highlight one thing that I hadn’t thought very much about and thought was a brilliant observation: data. A mapping product without data is a hard sell. We can see other companies opening up their platforms and saying, okay, it’s going to come with some sort of default background maps, or access to background maps — and to supply those or to curate them, that’s really difficult. If I think about another piece of software I could compare to something like Esri, I would choose Photoshop. There are a lot of technical aspects to that and they perform difficult processes, but you as the user show up with your own data. You’re not expecting Photoshop to provide you all the photos that you need. A test image isn’t that difficult to manufacture. So I can definitely see the data side of it — I thought that was a brilliant observation.
Ariel: The data part is, I think, one area for a lot of differentiation and innovation that still needs to happen. What I mean by that is, today what do you do when you want to create data in Esri? There are definitely different apps that they provide to collect data, forms that enable field people to go and collect data when they’re out and about doing their jobs. But imagine a world where you didn’t have people manually typing into a phone or typing into a desktop, and all of that data was just being generated for you automatically. I don’t think anybody’s nailed that. Obviously at Hivemapper that is one of our core aspects of what we’re trying to do — how do you take all these video streams from car dash cams, airplanes and drones, and how are they passively generating data for that customer? So that customer, literally just by driving or just by flying, is generating data that is feeding into their GIS system.
Ariel: That changes the equation. How do I collect data? I just drive. How do I collect data? I just fly. It just becomes part of the things that I already do, that then gets fed into your GIS system. That’s incredibly valuable to a lot of customers, and if you can do that and do it incredibly well, then you really have something.
Daniel: I completely agree. I don’t think we completely escape the problem with data, because that just moves on to a new problem — how do we get the data from there to over here, to our desktop, to wherever we’re going to do things with it? Do we take the code to the data, or do we take the data to the code? I think there are still a lot of really interesting problems to solve there.
Ariel: Until 5G is everywhere and it’s super cost effective, you’re definitely going to be in this issue where how do you transmit the data becomes really tough. So at least from Hivemapper’s perspective it’s going to be a combination of the two. There will be some customers who can bring the data to the software, just because of where they’re physically situated, and there are other customers where that simply is not possible and you have to bring the software to the data.
Daniel: Ariel, I’m conscious of the time. I have a feeling we could talk about these kinds of things forever — at least I could. Before I let you go, could you let the listeners know where they could go to find out more about you and hear about what you’ve got going on over there at Hivemapper?
Ariel: Check out hivemapper.com. We’ve updated it a lot since we last chatted — a ton, actually. Now we have this amazing gallery with all of these great example maps that customers have created. The other thing we’ve done is we opened up our global map. What that means is that when you upload video you can obviously create your own maps, see what you’re building, but that also becomes part of the global map that then everybody else can leverage and hook into. So you could have a situation where a bunch of people in the same city start to grow their map together, and they can see how each of those different maps are coming together into a single global map, which is a lot of fun honestly. That part of our product has been growing like gangbusters over the last three or four months. It’s always exciting to see what people are doing — we have users doing some really cool stuff in India, a whole bunch of other users down in South America, in Chile, doing some really awesome stuff, and then some folks out in Norway doing some stuff as well. It’s fun to see all these different climates, all these different people doing different stuff, but they’re all sitting on that same map.
Daniel: I just want to say congratulations on your success with this. I think it’s really great that you’ve built this company, that you’ve done something different. I appreciate that it can’t have been all smooth sailing, but you’ve done it and it sounds like it’s working out really well for you. Thanks again for coming along and taking the time to walk us through some of your thoughts on the industry. It’s been a fascinating conversation for me.
Ariel: Thank you.
