The Engadget Interview: ViaSat CEO Mark Dankberg on Exede and the future of in-flight WiFi


When you think about satellite internet, dismal speeds, ridiculous latency and astronomical fees are likely first to come to mind. But the only technology capable of blanketing our entire planet in connectivity is now able to provide a fast, consistent and affordable experience, thanks to one market-leading company, and the vision and leadership of Mark Dankberg. Launched from Dankberg's home in Southern California, ViaSat lit up on the map in 1986, and spent the first two decades focused on government and corporate contracts. Last year's launch of ViaSat-1, however, beamed the company straight toward the top of the North American satellite broadband market, bringing high-speed internet to the rural masses. Pair the company's Exede household product with airline contracts for in-flight WiFi, and you have one ever-growing infrastructure giant. We sat down with Dankberg -- ViaSat's CEO since inception -- at the company's Carlsbad, California headquarters, learning about the Exede service launch, a portable newsgathering rig and what the future of in-flight WiFi may look like on airlines like JetBlue and United. Join us past the break for the interview in full, and an inside look at how ViaSat plans to transform the entire satellite internet experience.

Was the Exede launch successful? How many subscribers have joined on at this point?

We feel like it's been successful, and the initial reaction has been very good, in two ways. We wanted to show that satellite could be a better choice than slower terrestrial alternatives. We wanted to see what would happen if we gave people high speeds and had enough bandwidth to stand behind it, and so far the initial reaction has been really, really good. People have been really pleasantly surprised about how fast it is. Not only did we do an introduction at CES, we've also gotten really positive feedback from a lot of our initial subscribers.

There's one school of thought that says satellite is only for people who have no choice and it doesn't matter how good it is. It's possible that that will be the case, but what we'd like to do is show that it could be a better choice for people. If you're not an Xbox Live gaming aficionado, if that's not really important to you, we think that this ought to be a really good service and I think that covers a lot of people. In terms of actual subscriber acquisition rates, we haven't released a lot of data yet. We did say in mid-February, before we had lit up the whole coverage area, that we were averaging a little over 800 new orders a day -- and you'd expect that to have increased since then. But we're still in a start-up mode. Mostly, you have to normalize things to: how many regions we're covering, how many dealers are available to do installs and how many installs they've done so it's tricky to interpret the numbers but we're really happy with them.

Does Exede represent a change in direction for ViaSat?

Yes. Four years ago, we started the ViaSat-1 satellite project, and up until that time our business was selling technology and products. What that really marked was a switch to selling services, and we felt that selling services was going to be more important. People generally buy the service, they don't buy just the technology, and so we never really felt like we could frame it the way we wanted or deliver it in the form we wanted. Now we had the ability to do that. It was really capital-intensive -- it was a half-billion dollar project, so it meant a big investment for the company -- but on the other hand, we felt really strongly about it, we didn't see other people doing it, we felt there was a void in the market, a vacuum in the market, and we wanted to fill it. So, yeah, it's been a big change. Internally, there's a mind shift that goes with doing that transition from technology to services and I think we're making that.

JetBlue and United have both signed on to offer in-flight WiFi. Will other airlines be joining as well? How many aircraft do you plan to have online in 2012 and in 2013?

I think that the first airplanes will go in service in the fall, so by the end of this year I'm going to guess there will be about ten or so. I think that by the end of next year there will be a lot more -- I think it's around 18 months to two years to deploy -- about 300 aircraft. I believe in terms of other aircraft signing on and how it gets used, the market will go through phases. Going back to when we were working on broadband back in 2000, the first question was "Does anyone want broadband at all?" With cable modems coming out, who was going to buy cable modems? Why spend $50 a month for something when you could spend 20 on dial-up? Today, believe it or not, a lot of the airlines are still in the stage where they're not totally sure that people really want in-flight WiFi.

I think the thing that made air-to-ground WiFi service successful in capturing airlines was that it was relatively inexpensive to install on an airplane. What it did was establish that there's demand. Now I believe the next phase is that airlines are starting to realize, well, not only do you want WiFi, but it's got to be good. Then the question is, what does good mean? Believe it or not, a lot of people don't really get that good means a lot of bandwidth for a lot of users. So that's not really the criteria that a lot of the airlines are using when they choose a WiFi technology. I wouldn't be surprised if it takes two, three, four, five years for all that to get sorted out. When it does, I think that people will like the satellite connections because it's a very cost effective way to deliver broadband to aircraft.

What's your business model and how is it different from Gogo? Do you cover the installation cost or is that shared with the airline?

So, from a business model perspective, it's a bit funny, because we're not totally sure what the business model is yet. Our model to start with is that airlines will probably buy the equipment to put on the aircraft and then they'll buy satellite airtime. JetBlue has in-flight DirectTV by LiveTV and the part of United that we'll be on used to be the Continental Airlines airplanes that also had LiveTV. The LiveTV guys would talk about a catch phrase "Coke and a bag of peanuts" -- meaning if they had a service that was inexpensive, they could consider giving it away to passengers. So the idea would be if the cost of broadband, at least for some applications, was sort of bounded, then maybe they could give it away for free, but they don't know for sure yet.

I think that the business model is still to be determined and will probably be shaped by each individual airline, and what we'll do is we'll be sort of behind the scenes fulfilling the bandwidth part of it and then we're going to think about other ways we might package our bandwidth. A lot's gonna depend on what the airlines perception of what the passengers think, because they don't know yet. Basically I think that the airlines will end up gravitating towards things that influence passengers' selections of flights -- that's really what's going to have the impact. What we suspect is that connectivity is actually a pretty big feature and that people will be able to tell the difference between pretty good and really good connectivity. So that's what we're aiming for and basically we feel like things that we do to improve the connectivity value proposition are very similar to what we do for home broadband, which is drive down the cost of bandwidth, yet deliver higher speeds.

Does Gogo have a competing product in the works? Do you see a place for them in this space as well?

Currently Aircell has the old Verizon AirFone frequencies. It's about 4MHz of spectrum, and I think they built out several dozen towers in the US that use that spectrum and point upwards, basically. In round numbers I think they've spent hundreds of millions of dollars on the infrastructure, and they have gigabits of bandwidth. We spent hundreds of millions of dollars and we have over a hundred gigabits of bandwidth. So I think that when you get to an environment where what matters is the cost of bandwidth, I think satellite should compete really, really well. I think that the big attraction of Aircell at the beginning was that because it was an air-to-ground system, you needed a relatively simple antenna, the installation was relatively inexpensive, and so you could offer WiFi with a smaller investment. So now I think the challenge for them will be to drive their bandwidth costs down substantially and it's not so clear how you can do that with ground-to-air spectrum, whether or not there's enough spectrum available to do that.

Are there other segments of the in-flight entertainment space that ViaSat hopes to one day conquer?

I think that entertainment is a subset of connectivity, I don't think that connectivity is a subset of entertainment. So, what I mean by that is that if we give somebody really good connectivity, they can use whatever rights and subscriptions they have to entertain themselves, so they could watch a baseball game, a TV show or movie. If all you give people are pre-programmed choices, that's your only choice at that point, so I think that there's a long-term opportunity to provide entertainment, but to do it through connectivity. That's basically what our value proposition would be. I don't really look at us as good programmers, you know, picking things for people to watch. We don't really have any advantage in getting rights to movies or things like that. But I think we could be really good at providing low-cost connectivity and then let people use the rights they already have. I think people would find that pretty "entertaining," so to speak.

So ViaSat-1 is the highest capacity satellite in the world, but eventually it won't be able to support additional customers. Are there plans for a ViaSat-2?

Yes there are and what we're trying to do is come up with something that's got even better economics. At the stage we're at we have a new design, which has new system capabilities. That's kind of what we did with ViaSat-1, we came up with a satellite design approach that was really unique. We designed the way the satellite worked, and how it interacted with the ground. And that's what we're doing now, again, for ViaSat-2, in a different way, to get an even better result. We're working with some solid manufacturers and we have to wait until we get back a definitive firm proposal we can buy, and we're getting close to that.

Will the Exede infrastructure that customers are investing in now be compatible with a future product? Will customers be able to see a benefit once ViaSat-2 launches?

A lot of the ways that we're going to get improvements in the economics is through new technology insertion, so it's not necessarily true that a terminal you buy today will work on the new satellite. Well, it may work on the new satellite, but it may or may not deliver all of the benefits of the new satellite. For instance, we can use our old WildBlue network equipment on the new satellite if we wanted to, but it just doesn't deliver as much of the economic benefits. So it will be some combination of technology and business that enables the service to improve.

As an example of one of the things we can do now, we have existing WildBlue subscribers on the old satellites and we can make the old satellites better in two ways. We can use the new networking equipment on the old satellites and that'll confer some benefit. That's what we call the Exede 5 service, but you need a new terminal for that. We also realize that the old bandwidth that we have has a lower replacement cost, we can give more of it out when we advance the technology and acquire new satellites.

One of the things that we'll be doing, even for legacy WildBlue customers that don't get new equipment, is over time be able to allocate more bandwidth. As other users migrate from the old satellites to the new ones, there'll be more bandwidth available on the old satellites, and that'll allow us, over time, but not instantly, to improve the service quality and the speed offered at the same price, or the bandwidth offered at the same price. Those are the ways that people perceive better service with this new stuff through the old equipment.

Do you plan to introduce Exede and in-flight internet outside of North America at some point?

Yes. We work as a technology provider in the US, we have our own satellites and we operate as a service provider. We're also providing global services internationally for business jets and defense users, using other people's satellites. The technical performance of the Exede service is largely an attribute of the networking equipment. That's a lot of what makes it be able to go so fast and be so responsive. But the economics are really driven by the satellite and how much bandwidth the satellite has. So we can recreate a similar user experience on other people's satellites, but we can't recreate the economics if they don't have a satellite that's as cost effective. There's very few satellites around the world that are even close, but there is one, a company called Eutelsat in Europe has a satellite that's 70 gigabits called KA-SAT. It's about half the capacity of ViaSat-1.

One of the things that I'd say is pretty interesting about the Exede service is we took all this bandwidth and packaged it in a way that we think people will find reasonably appealing. People haven't done that in many other places because they're sort of aiming lower. A lot of the times they're aiming for people who just don't have any choice at all and they're somehow trying to make the price point be very low --- because the service isn't appealing compared to alternatives. What we're trying to do is make it be a really good value, at a price point that we're used to, which is not the same as making it really cheap. For instance, provide a good ratio of megabits per second per dollar.

I'd say, the opportunity for us would be to go into other markets and try to use some of our packaging and marketing and retailing expertise to introduce it in some places using other people's satellites, and then if that can be successful we can then follow on with satellites that are like ViaSat-1 and ViaSat-2. I think there will be really good markets outside of the US. In a lot of ways I think that the US is actually one of the most difficult markets, mostly because of the cable network infrastructure that we have. We have very high penetration with cable TV in the US, and the cable infrastructure is good. In a lot of countries, they never really had high cable TV penetration because satellite TV got there first, and a lot of places don't have good telco infrastructures because wireless got there first, and so they just don't have very high landline penetration. So those markets, I think, are really good potential markets for satellite broadband.

What other products and services do you have in the pipeline?

One of the ones we think is pretty exciting is the satellite newsgathering capability we showed at CES. It's really just a way to get high-definition video and in some cases multiple high-definition cameras from remote locations pretty cost effectively. The airtime costs are much, much lower than conventional satellite costs, and the terminals are a lot smaller and much less expensive. You could carry one around in the trunk of your car. For many live events, where there's a lot of interest and you have multiple media people there, it's gonna be really hard to get enough wireless coverage to reliably stream high definition, so I think there's going to be demand for high-capacity satellite. It's a pretty interesting product and there's definitely scale opportunities there. Basically what we're looking to do is deliver bandwidth cost-effectively to places that are hard to serve otherwise, and I think that there's a lot of applications for that.

The other one that I think people will ultimately find very interesting is restoral bandwidth or backup bandwidth for cell towers in the event of emergencies or disasters. One example was after Hurricane Katrina. There was basically no telecom infrastructure available in New Orleans. We ended up providing emergency service just using conventional Ku-Band satellites, older satellites, with much more expensive airtime, and we would basically set up little satellite terminals, attach them to a few important cell towers and we would land cellular traffic here in Carlsbad. We did this with Qualcomm -- they had portable base stations, and so traffic would come in to either our switches here or Qualcomm's in San Diego. But we could only do it on a relatively small scale, in limited areas, and really just for voice and low-speed data. We're talking about only megabits of capacity for a whole city. You can imagine using a ViaSat-1-like satellite where you can provide hundreds of megabits or eventually even gigabits to a city that has a situation like that, and I think that would do a lot for rebuilding and getting things under control more quickly.

That's a relatively interesting application, and there are others for things that are more security oriented. The US spends fortunes on things like border patrol or maritime patrol -- those are both really important missions. Some parts of the world, there's maritime piracy, so if you can do better maritime patrol well, that's very important. Also, more and more things like oil exploration or gas exploration are very data intensive and they're in remote places, so these are things that have typically been good satellite applications. But they're not done well if the bandwidth costs a hundred times what it otherwise would. So that's really what we're trying to do is to bring satellite bandwidth costs more in line with what you might expect from terrestrial services.

What are you using at home to connect to the internet?

I have both our broadband service and a cable modem and I compare them. I have two WiFi zones, and in one part of the house it's closer to the cable modem, and the other part of the house is closer to the satellite one, and so when I'm wandering around with my iPhone or iPad it just hooks up to either one and usually I can't tell which. That's mostly what I'm looking for, that I can't really tell the difference on a portable device. That's mostly the case, not a hundred percent -- sometimes we find things that we need to fix.

One of the things we've got that really accounts for the responsiveness is a web-acceleration technology. We essentially compensate for webpage practices that are more sensitive to latency than they need to be. As things move to HTML5, I think webpages will be less sensitive to latency and congestion. People are becoming more aware of that now because of mobile devices where you get congestion in networks, and congestion is basically just like latency. So anyway, until that's pervasive, we basically test how our acceleration technology is working and we find those sites that don't respond well and we tend to fix them and most of the stuff works really well.