Good day, ladies and gentlemen, and welcome to the ViaSat Fiscal Year 2013 Second Quarter Earnings Conference Call. [Operator Instructions] As a reminder, this call may be recorded. I would now like to introduce your host for today's conference, Mr. Mark Dankberg, Chairman and CEO. You may begin.

Okay, thanks. Good afternoon, everybody, and welcome to ViaSat's earnings call for our second quarter fiscal year 2013. So I'm Mark Dankberg, Chairman and CEO and I've got with me Rick Baldridge, our President and Chief Operating Officer; Shawn Duffy, our Interim Vice President and Interim CFO and Vice President, Chief Accounting Officer; and Keven Lippert, our General Counsel. So before we start, Keven will provide our safe harbor disclosure.

Thanks, Mark. As you know, this discussion will contain forward-looking statements. This is a reminder that certain factors could cause actual results to differ materially. Additional information concerning these factors is contained in our SEC filings, including our most recent reports on Form 10-K and Form 10-Q. Copies are available from the SEC or from our website. With that, back to you, Mark.

Okay. Thanks, Keven. So we'll be referring to slides as usual and they're available over the web. I'll start with some highlights and a top-level business overview, and then after that, Shawn will discuss our financial results and then I'll give more depth on each of our business segments and then we'll take some questions. So second quarter was pretty exciting across-the-board, and I'll touch on some of the highlights here. We did really well on new orders, over $540 million this quarter on top of a strong first quarter too. So year-to-date, we're at about $880 million in new orders and we've built a backlog [indiscernible]. The single largest order in the second quarter was for the Australian [indiscernible] across-the-board. Those strong new orders on top of pretty solid orders last year is fueling a growth spurt for us. Revenues are up over 25% year-over-year for both the quarter and year-to-date. As we've been expecting, the growth in revenue, combined with steady gains in Exede broadband subscribers, is driving steady growth in our adjusted EBITDA too. Sequential adjusted EBITDA increased $15 million compared to the first quarter of this year, and adjusted EBITDA is up year-over-year as earnings growth starts overcoming the fixed cost that we incurred for the ViaSat-1 Network expansion. We had about 150,000 subscribers on ViaSat-1 at the end of the second quarter, which is just about 6 months after we had the whole footprint lit up on the satellite, plus we continue to make good progress on all the adjacent markets we had envisioned for the satellite too. So we'll take a quick look at the highlights in each of our business segments. We're making steady progress in building up the ViaSat-1 Exede subscriber base. Those adds and net adds both gained sequentially compared to…

Thanks, Mark. While Q2 is marked by good progress financially for us, as Mark mentioned, we achieved records in the quarter for both orders and revenue. This resulted in 27% growth in revenue year-over-year. As we discussed last quarter, Q1 was a low point for us on the adjusted EBITDA front as we launched our Exede satellite service. Overall, the weight of the fixed startup costs and new subscriber acquisition costs outweigh the incremental revenues from our initial launch subscribers. As we have continued to grow our Exede subscriber base and increased revenues each quarter, we have generated some upward momentum in Q2, bringing adjusted EBITDA to $44.6 million. So overall, we're pretty pleased with where our second quarter landed, and some of our key performance indicators are shaping up the year nicely. And looking at our results year-to-date, we have also been very pleased with our order flow for the year. The $880 million in new orders for the year has helped drive our revenues up 25% over last year and generated a record backlog of $966 million. Overall, we've had some very significant wins with our Q1 KACST contract in Saudi Arabia, followed by the Yanko [ph] contract in Q2. Those wins, coupled with strong flows into our situational awareness and government mobile broadband programs in this uncertain DoD spending environment, has started to solidify our near-term outlook. Year-to-date adjusted EBITDA, which carries the weight of the first quarter Exede service launch effect, is almost about even with prior-year results, coming in at $74 million versus $75 million last year. So as we move to the P&L as a whole, our Q2 revenue growth to $283 million was the driver in helping us close the gap on the operating margin impacts of our Exede satellite service launch activities.…

Okay, thanks, Shawn. So I'll go on into a little bit more depth on what's going on in each of the business segments, and we'll start with the Satellite Services. The theme there is steady progress. We're making good progress in refining our approach based on what we learned. In a minute, I'll show you some really interesting data on how we're doing in the broadband market as a whole, which lends some context to the subscriber growth data. We're still making sequential improvements in gross and net adds. Unit economics continue to be in line with our plans. As expected, last quarter was the bottom for us in terms of earnings in this segment and this quarter, we turned about $8 million of revenue growth into about $5 million of adjusted EBITDA gain. The chart of the -- the chart at the bottom of this page shows how our subscriber base and adjusted EBITDA have trended over the last 6 quarters, and it can give you sense of the rate of growth that's associated with current trends and subscriber additions. Adjusted EBITDA margin will fluctuate from period-to-period primarily as a result of marketing initiatives, but generally, we believe will trend in line with our models. ARPU continues to trend up, driven by the stronger market position of the service and a higher retail mix. We're steadily overcoming the fixed cost increase of the ViaSat-1 network expansion based on the unit subscriber economics. Data continues to substantiate that we are growing from the underserved as well as the unserved broadband customer base, which suggests that we have a market opportunity that is far larger than the capacity of ViaSat-1 and the Echo satellite. Our subscriber metrics with the Exede service shows some steady growth relative to our first fiscal quarter.…

[Operator Instructions] Our first question is from Mike Crawford from B. Riley & Co.

On the Exede adds, what percent of the subscribers you signed up came from underserved or DSL markets?

Listen, again, it's kind of in the mid-30s percent, about where we've been before.

And when do you expect that the DIRECTV agency deal to kick in the stride?

That will be this month. Actually, that's when it'll start. It hasn't started at all yet.

Okay. And then regarding -- back to DSL, is there any risk to your model you see with vectoring technology that can potentially, dramatically improve DSL speed at maybe a low economic cost to carriers?

Okay, yes. Vectoring's a technology that's gotten a little bit of play or something. Actually, I don't think it's relevant to our market at all. What vectoring pretty much does is it reduces crosstalk for DSL lines that are in close proximity when that turns out to be the limiting factor. For the markets that we're going after, generally, the limiting factor for DSL is loop length. So it's not that the loops are close together, which is what the vectoring does, it's that the lines are really long and vectoring doesn't really do anything for that. So we still don't see anything really on the horizon that changes the overall outlook for DSL. And I think if you look at how the telephone companies are behaving, it doesn't look like they do either.

And then just last question, switching gears to the mobile broadband. Is that -- it's a little confusing whether you're still branding that service as Yonder's. Is that going to be branded Exede? And I think at one point, that was your fastest-growing business. I don't know if that's still the case and if maybe you can comment on what approximate level of revenue you're getting from mobile broadband today?

So for us, mobile broadband comes in multiple flavors. The 2 -- the main ones that we have going now are at Ku-band, and they used our spread spectrum technology that we call ArcLight, and that's basically what we've called our Yonder services, ArcLight Technology Solutions over at Ku-band satellites. The 2 -- the main flavors of that are the Government business, that's the global mobile broadband business. And the largest part of the services piece of our Government segment revenue is really due to that service, and that's the single largest piece of that. The other pieces that we have for Yonder are business jets and a revenue share that we have in the maritime market that we partner with KVH. Those 2 are probably maybe half the size of the government business together. Then the part that's -- one of the parts that also looks promising, and this is the part that we will be branding Exede, by the way, is the Ka-band part that will be on commercial aircraft. And that we'll be doing with JetBlue and United, pretty much starting in calendar 2013. Our plan is to use the Exede brand name when we can use high-capacity Ka-band satellites and deliver an experience that's like the Exede home experience. So we won't be branding the Ku-bands as Exede because we can't really provide the same economics or performance. So we'll preserve the Yonder name. But overall, what we'll be aiming to do, and this will take years, is to eventually be able to use all Ka-band on those markets on a global basis.

Our next question is from Rich Valera of Needham & Company.

I just wanted to get a sense of how you're thinking about the gross of adds per quarter. You talked before about target, sort of a medium-term target of 90,000 to 100,000. I think you're around 71,000 this quarter. And given your commentary on the last call where you'd seen a pretty good pickup in the month of June. I would have probably expected a bigger gross install uptick than the 5,000 we saw. So just trying to figure out what's going on there from the gross installs, and if the 90,000 to 100,000 target is still the active target over the next quarter or so?

Well, I don't think it's that over the next quarter. We are -- one of the things we're doing is we're just constantly researching and trying to figure out what's going on in the market. And so we -- when we laid out those targets of 90,000 to 100,000, it was based on what we've historically achieved or what [indiscernible] historically achieved working with both DISH and DIRECTV as distributors. So now we're in the market -- we've been in the market for 9 months, and we're just trying to assess what -- both what we can do and what -- how -- what we are doing compares to the market as a whole. If we were -- once we get sort of more towards the blended disconnect rate or churn rate that we expect, if we're doing 100,000 -- just to put things in perspective, if we were doing 100,000 adds a quarter today, we'd probably account for 25% to 30% of all the net adds in the broadband industry as a whole. So that looks pretty ambitious. I don't -- we don't think we'll get that in the next quarter or 2. So to put things in perspective, that would mean us as a service provider with, let's say, 0.5 million out of 80 million total broadband subs would be getting 30% out of, let's say, 300,000 or 325,000 total net adds in the industry. So what I'll say is it makes that look more formidable. On the other hand, when we look at what's going on in the market, we think there's room for improvement for what we're doing. So that's what we're working on. And I'd say there's still things that could help us get there. But we're just not going to -- we don't think we're going to say we're going to do that next quarter or the quarter after. We'll just going to see what happens in the market and report on it.

Is it fair to say you would expect sequential growth in that gross sub add number?

It looks -- so far, it's hard to tell. I mean, I could say we're off to a pretty good start this quarter. I think we'll just have to see quarter-over-quarter. But I think the things, what we're finding is that the value proposition seems to be there in the underserved market. We think that the reasons that subscribers are buying are the reasons that we intended, which is that we offer really good value and speed per megabits per second per dollar, that the speed level that we're offering is the right range, the people are attracted to that. And I'd say what we're finding now, more than anything, is that our adds are sort of proportional to the effectiveness of our marketing spend. That's kind of what we're finding. And so we're trying to perfect that. One thing we could do is just throw a ton of money at it, but that would sort of screw up our unit economics. So what we're trying to do is do it more thoughtfully and purposely, and we're learning a lot on how to sort of optimize that. We're trying things in different markets. And what we feel like is it's going well sort of a way we'd look at it. So we're not -- I guess what I was trying to say is we're trying to not just focus on the absolute numbers but on the unit economics, the returns we'll generate and the reasons that we're getting the subscribers.

I'll take it. This is Rick, Rich. I'll add to what Mark has said. Last quarter, we reminded you guys that [indiscernible] of the market here. We could see some pressure from them coming to the market for a quarter or 2. Didn't think it would affect us overall, it doesn't change the answer. But we could see some near-term pressure and really don't know yet. I mean, they're just getting started.

Yes, it's early to tell. Plus overall, I mean overall, I would say we're trying to look at sort of the strategic environment. That's more what we're trying to comment on.

Okay, understood. And just trying to understand the sequential increase in service revenue. You just take the addition of net subscribers that you got and kind of take the average subscriber base for the quarter times your ARPU, you get a significantly smaller sequential revenue increase than what you reflected. I'm trying to understand what's the other component there? There's clearly sort of a service, just sort of average subs times your ARPU, but there's got to be some other underlying, yes, I guess, maybe hardware component there that I'm not understanding, just to -- I'm trying to understand how we can model that going forward because that increase, like I said, was bigger than I would have expected.

I think one of the things you have to count for, Rich, it's Rick again, is the subs that came on over the last quarter weren't all reflected in the previous quarters, so this builds on itself. So you've got to go back in the last quarter and reflect those subs plus there is some -- there is an incremental margin for some of the equipment sales for [indiscernible]. These guys buy the equipment, but it's certainly not -- it's still insignificant, yes.

Yes, I mean, what we cited here is blended ARPU, and so the incremental ARPU of the subscriber we're getting is much higher than the blended ARPU. That's one of the factors that's helping us achieve the results that we aim for that our ARPU is kind of new subscribers like 20%, 25% higher than we originally modeled.

And the ones we're losing in terms of disconnects are more off the low end, lower end.

Right, low end, yes.

Yes, that's it. I would have to take a look at that. And then going back into last quarter, you had talked about replacing, what you call the Sales Only call center, and that it caused some disruption. Just wanted -- if you can provide an update on that, how's that new one working out? Is it working as expected?

Yes. Basically, there were a couple of objectives that we had. One was we thought that we could improve the economics. We were paying double, twice, for some of the marketing spend that we're getting in the way that we went through, particularly Sales Only dealer. The other factor was that we felt we were not getting access to all of the data that we could. What we found is that both those things worked out well. The economics have improved, so that helps us, and we're getting access to more data, which was helping us be effective. It actually worked out well. We did that early in the September quarter. We had some glitches for several weeks. Glitches meaning like backward -- took step backward for the week. But then we've recovered what we had set out to do. And that kind of thing, those kinds of steps that we're taking away from the way we use to do business, I think, are things that we're going to try to do continually over the next several quarters.

Got you. And just one final one if I could. Obviously, a great quarter in the government business, and you kind of acknowledged in your prepared remarks that you had some -- a lot of things go your way. Can you give us any sense of how you're thinking about the step business for this year? I mean, it's obviously a little tough to model when you get that level of sort of volatility, but any way you could kind of help us think about that business on a full year basis or a sequential basis would be quite helpful.

I think, Rich, there was definitely a catch-up in the quarter as Mark mentioned earlier. Shawn and Keven can chime in here too if you have some additional comments. But so there were some stuff that had slid from previous quarters into the quarter which caused a blip. I would go back to previous quarters, and given our growth that we projected, it draws a smoother line.

Yes, I think when you look on a full year-over-year basis, in the past, we've been talking about a 10-ish percent growth. We didn't quite get that last year, but some of that really was sort of deferred to this year. I think this is a catch-up year for us. That kind of growth on a sustained basis would be good.

Our next question is from Yair Reiner of Oppenheimer.

It looks like the rate at which you're adding subscribers is maybe, all things considered, a little bit slower than you'd originally forecast. Given that, would it make sense maybe see how sub trends go for a couple of quarters before signing a contract for ViaSat-2? It looks like the current rate is going to take you quite a few years to fill up the current capacity.

All right. So, okay, one, is good point, Rob. It's we're being thoughtful about what we're doing on ViaSat-2. If you look at the rate we've been in business on ViaSat-1 in a full footprint for about 6 months. We're about 150,000 subs on ViaSat-1. So I think we talked about filling up the satellite to 1 million subscribers in the 3- to 4-year range. That's very consistent with that sort of more on the faster end of that. The other factor is that the satellite doesn't fill up uniformly. Some markets fill up faster than others, you want to reinforce those. So those things, sorry to say that it's time, certainly time to bring ViaSat-2 contract into place. The other thing is I think people should look at the revenue contribution, which is really the product of the ARPU and the ramp rate. And when you factor in the substantially higher ARPUs that we're getting, that basically, even without the adjacent markets, which are shaping up well for us, the returns look good and so that we're encouraged. I think that we look at the following satellite as a way to extend what we're doing.

[indiscernible] that cover all of our fixed costs and become cash flow generating within a year after the launch is very successful.

And if you could talk about, right now, your costs, it looks like quarter-on-quarter costs came down quite a bit in terms of your period cost. Can you talk about the factors there? It seems like you're understanding some of the period costs a little bit better.

I'd say,-- are you referring to the satellite service in particular?

Yes, yes.

Yes, so what I would say is we're going to see some fluctuations in period-to-period. One of the things that we've been looking to do on the unit economics is to be able to turn revenue growth into ARPU. We wanted to -- one of things we wanted is to make sure that people understood that. And that's what I'm saying for this quarter, for this September quarter, on a sequential basis, we turned about $8 million of revenue growth into about $5 million of incremental ARPU, a little over 60%. And that's pretty well in line with what we're aiming. I think we'll do better as we scale when we get more economies of scale. That's pretty good and there's really not anything grossly unusual about the quarter in terms of our spend.

And then just one more, if I could. You've talked quite a bit in this call about the underserved market. Some of the stats out there suggest that the unserved market is really big. Why haven't you focused more on that given that it seems to be -- it would have seemed to have been a ripe opportunity and perhaps there might be some pent-up demand there, as you mentioned earlier?

Yes. So this is one of the things that we've dealt with or addressed for quite a while. What we think is the unserved market represents to a lot of estimates in the 7 million to 10 million home range. We think that's a good market but the underserved market is probably 3x to 4x bigger than that. And that's a much more attractive market for us. We also believe that if you run a business that's predicated on selling to people who have no choice, that you don't really develop all of the skills and tools you'd want to effectively address that 30 million to 40 million -- 30 million-ish market of underserved people. So we pretty much bit the bullet at the beginning and said, "Hey, we're going to position this thing to address the larger market." And a lot of that was to deal with this question that you asked earlier about ViaSat-2. If we are dealing with the total market of 7 million or 8 million, we're dividing that among 2 service providers, you look at sort of what the penetration rate would be for that market. You could say, "Okay, well, maybe we should get 1 or 2 more satellites." You position those in places that address those unserved markets. But it's not at all obvious that you can just say "Okay, now I'm going to go take that and go after the underserved." We think we're learning a lot of skills both in marketing, customer service, number of areas. And we're also learning a lot about geography and why people choose our service. So that's what we've had our eye on, and that's why we keep talking about sort of where our subscribers are coming from, what it costs to get them, what the value proposition is, our ability to keep them, what customer satisfaction is. And I think that -- and that involves a lot of learning in our services organization, which I think is going well. But we don't really look at it as a, step one, you get the unserved and then you just say, "Okay, now I'm going to morph immediately to the underserved." I think that there's a lot of issues associated with doing that, so we kind of do it holistically to start.

Our next question is from Chris Quilty of Raymond James.

First, a clarification, Mark. Understand branding the JetBlue service with Exede. But from an accounting perspective, do those revenues still fall within -- do they fall in Satellite Services or government networks in terms of your breakdown of product and service revenues?

I can answer that for you, this is Shawn. So the product side of developing the onboard equipment shows up in our commercial side of the business. And then as we launch the service, that will come into our satellite service area.

The recurring airtime revenues.

Yes.

Okay, fair enough. And initial shipments of hardware or should some service actually be turned on in the fourth quarter?

Services won't be turned on in the fourth quarter. What we're doing is we're really working now on the FAA pipe-worthiness certifications. Those are sort of a gating item to revenue and initial aircraft being in service in the first quarter.

Okay. And a follow-up on Rich's question about the average ARPU for the quarter. And I mean, I ran the same calculation as him and came up with $60 million of revenue and you reported $67 million. So is it fair to assume that from the beginning of the quarter to the end of the quarter, your ARPU moved up significantly to account for that incremental revenue?

You can't just take the new subs you had in the current quarter. So you take a previous quarter ending subs and then average the new ones in at the new ARPU level. And then there were some product shipments that -- there were some product sales in the quarter in that business. And that's growing as an absolute value, probably reducing as a percentage, but growing as an absolute value of a couple million dollars also.

Yes. But just to be clear, I mean, what happens is you sort of go back to the prior quarter, figure out what the midpoint is for that quarter. You'd have to add revenue based on that, and then you do the same thing on a cumulative basis. But the other thing is to make sure you sort of model in this effect on a churn basis, remember in the older days, we were largely wholesale. Now we're largely retail. The wholesale ARPUs before were in the low 20s. So now retail ARPUs are above 50, so we're replacing at a pretty high rate lower revenue subscribers of newer revenue subscribers. It's filling up the blended ARPU, but the incremental ARPU is quite a bit higher than that.

Okay. And how about just amongst new Exede customers, are you seeing a lift in the ARPU now that you're enforcing the FAP of people moving up from basic to improved plans?

I'd say that it's a modest effect at this point. And what I would say is, this sounds a little bit silly, but really, the purpose of the fair access policy isn't to drive revenue. I think in order for us to compete well in the unserved -- underserved market, what we really are trying to do is offer a good value. So if we have tricks that drive up revenue, that don't make it a good value, that's not a good outcome for us. So the real purpose of it is to preserve the integrity of the service and try to match people to the right plans. So I don't think we're going to see a really big lift there. But we'll see the effects of it and it sort of helps us achieve network performance that we're trying to achieve, make sure we get the unit economics that we're trying to achieve. And the other thing, for instance, when we introduced in this past quarter our unlimited overnight FAP-free zone. So that gives people an opportunity to consume a bunch more bandwidth off-peak hours without raising their bill. And that people perceive that as a good value as well. So we'll be trying to do things that sort of preserve ARPU in a range that's competitive. And this is one of the things that we'll go into more depth on in the analyst days, how we look at revenue, how we gauge the value of our service. But the idea is not to drive revenue up to really high levels but to remain competitive and preserve network integrity.

Okay. And I may have missed it, did you give the mix of subs coming on ViaSat-1 wholesale versus retail?

Yes, we've been pretty consistently around 2/3 retail, 1/3 wholesale.

Okay. And just a clarification from the language earlier around the rate of net adds and hitting that 90,000 to 100,000. It sounds like we should back off over the next several quarters perhaps with the entry of Hughes back into the market. The rate at which you're picking up, is that a fair assumption and perhaps push out a couple of quarters when we get to this sort of that run rate net adds?

Well, I don't know. What we did is a little over 70,000 installs in this quarter -- in the September quarter. The December quarter, we have multiple effects going on, including DIRECTV. We'll start during the course of this quarter, so that will probably have some impact. DISH Net went live, that had some impact. And we're still providing wholesale subscribers to DISH Net users in the market, so that's providing some diversion as well. I think that, personally, what I think that is to the extent that the USENET services and the DISH net services are the same as ours and reinforce the value proposition of satellite in a sort of a dollars per megabit per second basis. And ultimately, that's not bad for us. But all those things are going to take a little while to play out. So what we find is, I think that our growth in installs will really depend a lot on the way we market, which is not surprising at the end of the day. It's sort of how effective are we at marketing. And so what we're doing is going about that what we believe is strategic and deliberate way that I think actually gives us a good shot at getting to that target. I just don't want people to be sitting around every quarter going, "Did you do 90,000 this quarter?" Because I think what we're really trying to figure out is how do we get there, what is the business look like at that point, and what are the economics when we're at this -- at the point that we're at now, which are actually pretty good.

And speaking of DISH, at what point should we know whether we're going to get a renewal on the contract? And will you come out one way or another when that negotiation and contract is renewed or not?

We haven't made an announcement about that yet. What I would say is I think things will continue with DISH for a while. How about that?

Okay. And final question, the government business. I mean, I think you had some record EBIT margins in the segment. What's a better normalized margin to assume?

In the Government business?

Yes.

It's like I said earlier, Chris, I think I would draw a smoother line between what you had expected for the year from our previous quarter until what we had modeled for the overall year. We had a catch-up in the quarter. And although the margins aren't substantially different, we had good margins in the quarter, but that was really driven by revenue.

Yes, if you look at our historical margins in the government business, I think that those are -- we've been able to, over reasonably long periods of time, sustain pretty consistent performance and we don't see that changing right now.

So sort of low to mid-teens?

Yes.

[Operator Instructions] Our final question is from Tim Quillin of Stephens Inc.

With regards to DISH Net, are there any particular beams that you would service as opposed to Hughes? Or how is that business going to be split up between the 2 underlying providers?

Of that, so we have right now this sort of very funny situation where the Hughes satellite looks pretty much exactly like ours. So they have coverage and capabilities and plans that look just like ours in places that we have ours. I think that, that is a little bit of an artifact of the way we acquired our first satellite. It probably won't occur again. But it probably will be that way for the next couple of years. So that means that DISH will really sort of exercise their judgment and discretion about how they divide them up. And so far, we have a good relationship with DISH. Obviously, there's going to be multiple factors that they'll put in there, but I think we have a good relationship with DISH, and I think they are treating us fairly, pretty fairly. And I'd see that continuing for the foreseeable future.

Would you expect a change in the wholesale retail mix in the current quarter?

I would say DISH, when you bring DISH Net on, DISH Net is a pretty powerful brand. DISH brings on 750,000 gross adds a quarter. So they can definitely move the needle. We've known that. So I think it's just -- we'll just have to see. It depends on how aggressive they remain in selling it and how they divide up the business. But it's possible that, that could skew us in the short term a little more wholesale. And then we'll have to see what happens with DIRECTV. DIRECTV, our agreement is an agency one, and that is retail. And that one hasn't kicked in yet. So I think what we're trying to do is make sure we have a good understanding of long-term competitive dynamics. And it's very hard to say what's going to happen in the quarter when things are changing, moving around internally in the quarter. It'll take us 2 or 3 quarters, I'd say, to sort out all of these effects on an ongoing basis. But overall, I think and this is one of the things that we're trying to focus on is sort of why do people buy the service, what represents a good value, what are the ways in which they do that, and how the marketing campaigns of each channel blend together to form a whole. And it's still pretty early to tell, but what we think we're learning a lot about is sort of what the overall growth rate is in the market, the churn rate within the players in the market. And we think we can sustain or improve the rates at which we're going and the revenues that we're generating doing that.

Right. In terms of the -- you talked about the return on invested capital assumptions, which above cable, I think, and below satellite television. Would that be at the current pace of net adds? Would that ROIC still be attainable?

Yes. So basically we described it, it's sort of in the range of the cable industry and, plus or minus, could be up, and then aiming more for the satellite industry as a whole as we build scale, which is one of the reasons we're aiming for this larger underserved market. And than what we said was when we -- so when we started the business 5 years ago, we had a spreadsheet, no basically a forecast that said, "Hey, here's what we can get in ARPU. Here's the rates to which we can get subscribers. Here's how big the market is," all these things. And when you played with all those things, we got a return on invested capital that looked like that range that we're talking about. Now here we are a lot later and we know what the actual rates are, we know what the actual ARPUs are. We know that we had delays in the satellite construction. So we put all those things in, and then we project out from where we are now. And like I said, if we were to go kind of at the rates that we're at now, blending all these things in, the return on invested capital, all factors considered, taking into account all these markets and factors, would be in the range at what we thought it was at the beginning. So I think that's a good outcome. And that range, that's a good range. I mean, that's the kind of range you'd say "Wow, knowing what I do now, would I go back inside that project?" I would say yes, sure.

Is there a difference between the churn rates on wholesale and retail right now?

Yes, there is. Right now, the churn rate differential is really more about the legacy network than it is about the Exede network. The Exede network -- the new subscriber churn rates are fine. I mean, they're right in line. What I would say is the issue that we've had on churn on the legacy network, a lot of it has to do with the wholesale base because we don't own or touch those subscribers as much. And so that problem will sort of sort itself out over time.

And then lastly, in your press release, you talked about the wholesale agreement to provide coverage in Northern Mexico, presumably with ViaSat-1. But maybe can you talk about that if it's meaningful in terms of the percent of capacity it might consume. And then you alluded to ViaSat-2 potentially targeting new geographies as well. And I'm wondering if you can kind of fill us in on what you're thinking there.

Yes, so talk about, I would say, first on the legacy stuff and the existing satellites into Mexico. We think that there's underlying root competitive advantages for satellite that should play in multiple markets. And we're looking for ways to prove that, just to have improve it. And some of that, we'll do ourselves, some we'll do with partners. This is one that we're working with partners on. We'll see how that goes. I would say in the longer run, what we'd like to do is create additional satellite capacity that gives us options, sort of 3 options to test, either expanding existing products into new geographic markets or trying new products in new geographic markets. And I think we'll do that in a pretty prudent way, but we can't really talk about what those geographies will be and what the services will be yet. But I think we will once we have more insight into -- or we've established the contracts, the satellite procurement contracts, then it will be more appropriate to talk about that.

And I'm not showing any further questions in the queue.

Okay, good. So there's a lot of going on, a lot of discussion, and we really appreciate everybody's time and interest. Once again, I'd really like to encourage people to think about attending our Analyst Day at the end of the month. We will have an opportunity to go into a lot more depth on issues that were touched on in some of these questions. Until then, I look forward to talking to you all next quarter. Thanks again.

Ladies and gentlemen, thank you for participating in today's conference. This concludes today's program. You may all disconnect. Everyone, have a great day.