Viasat Inc (VSAT) 2018 Q1 法說會逐字稿

Welcome to ViaSat Fiscal Year 2018 First Quarter Earnings Conference Call. Your host for today's call is Mark Dankberg, Chairman and CEO. You may proceed, Mr. Dankberg.

Okay. Thanks, and good afternoon, everybody. Welcome to our earnings call for our first fiscal quarter of 2018. I'm Mark Dankberg. I'm Chairman and CEO. And I've got with me Rick Baldridge, our President and Chief Operating Officer; Shawn Duffy, our Chief Financial Officer; Robert Blair, our General Counsel; Bruce Dirks, our Treasurer; and Paul Froelich, our VP of Corporate Development.

Before we start, Robert will provide our safe harbor disclosure.

Thanks, Mark. As you know, this discussion will contain forward-looking statements. This is simply a reminder that 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 said, I'll turn it back over to Mark.

Okay. So we'll be referring to slides that are available over the Web. And I'll start with some highlights and the top-level business overview. After that, Shawn will discuss consolidated and segment-level financial results in some detail. And then I'll go into more depth on each of our segments. Then we will review our outlook and take questions.

So there's a lot to discuss this quarter. But certainly, the highlight of our first quarter was the successful launch and continuing orbital progress for ViaSat-2. That satellite's a valuable source of growth for us across of all our businesses. It's been frustrating to have it sit on the sidelines due to a series of unfortunate launch-related events, but now that's behind us. There will be about a 6-year gap from the in-service dates that -- for ViaSat-1 and ViaSat-2, and that creates some near-term challenges to manage in this quarter. But we, our customers, our distribution partners are all highly energized by the launch and the in-orbit progress to date. So we'll go into more depth on the short-term challenges and our approach to them later in the call.

The second important theme for this quarter and our year as a whole is growth. Government Systems had an exceptional quarter, outperforming an already very strong 5-year track record of double-digit revenue and earnings growth. We'll go into more detail later, but highlights there include 26% growth in revenue, over 88% year-over-year growth in new orders, over 50% growth in EBITDA, and record backlog. Most of that is being driven by what are called Non-Developmental Items, or NDIs, that we invested in to serve targeted high-growth markets with unique technology and value propositions. Those investments are really paying off, with the biggest opportunities still in front of us and gaining momentum.

In-flight connectivity is sustaining strong growth momentum, too. And we're anticipating a surge there as we begin installing our orders for 840 additional aircraft. We continue to see high passenger engagement as a powerful and unique value proposition.

We're already connecting close to 3 million devices a month on our existing fleet of 568 aircraft, which we believe leads the industry. Consumer broadband is challenged by the ViaSat-2 launch delay, but average revenue per user continues to grow to record levels, up 11% year-over-year, driven almost completely by bandwidth services. That did balance subscriber attrition this quarter, this first quarter. We'll go into quite a bit more depth on the factors involved in our growth plans for consumer and broadband for this fiscal year as a whole later in the call.

Overall performance from core business was very strong in the quarter, with a record $154 million in operating cash flow. Strong orders company-wide are up 31% year-over-year and led to a $1.1 billion backlog, lending confidence to our growth plans. The other significant theme this quarter is our investment in our high-growth opportunities, each with quite unique value propositions, yet sharing much technology over several market verticals.

As we've described in prior calls, our total R&D investments are peaking this fiscal year, driven primarily by ViaSat-3 payload pre-flight development and testing as well as commercial in-flight connectivity, STCs and line activity. We're achieving the results we want from those investments and expect them to wind down to R&D spend levels that are more consistent with pre ViaSat-3 proportions at the end of this fiscal year.

We'll give more insight into the objectives and accomplishments later in the call and show how we expect to capitalize on them in our target markets. For now, Shawn will give some more detail on the financial results and then I'll come back and give more color on each business area.

Thanks, Mark. Although Mark already hit on some of the Government segment highlights, I thought it was worth noting right up front the outstanding momentum we've seen across many of our government business units this quarter.

We generated 26% Government segment revenue growth and EBITDA gains of over 50%. This, coupled with Q1 government awards that were 36% higher than segment revenues, created a new record high segment level backlog of $696 million and total company backlog of nearly $1.1 billion. So we're pretty excited about our government business and the enduring growth prospects we see, which Mark is going to touch in on a bit later.

So now I'll hit on some of the other first quarter high points, starting with our Q1 revenues, which continue to reflect year-over-year growth of 5% from last year to $380 million.

In Satellite Services, our revenues were flat year-over-year. But what you don't see is that the top lines grew strongly in our emerging commercial air business, and we continue to see modest revenue growth in our consumer broadband business. However, these increases were offset by a $6.6 million year-over-year revenue decrease associated with the completion of the Loral settlement payments in Q4 of 2017.

In earnings, we also saw the year-over-year settlement impact, plus costs associated with the ramping ViaSat-2 service launch activities and preparations for the large-scale in-flight Wi-Fi ramp we expect later this year, altogether, driving lower year-over-year quarterly segment operating profit and adjusted EBITDA.

In a few slides, I'll spend a bit more time walking through the Satellite Services segment trends as well as highlight some trends we should expect to see in the upcoming months.

In Commercial Networks, quarterly revenues were down about $20 million as a result of reduced fixed terminal bills supporting Australia's NBN initial service launch, which occurred last fiscal year. As I already noted, we're getting pretty excited about the upcoming American Airlines service launch activities and significant ramp-up in Gen-2 installs later this fiscal year.

Remember, the sales of those airborne terminals are reported in our Commercial segment. So we should see some strong quarter-to-quarter revenue ramp as we progress through fiscal 2018. Our segment activities will also continue to be heavily focused on the company's ViaSat-3 project, alongside investments and expanded commercial airline platforms.

On ViaSat-3, development will continue for a few more quarters, and we're making some really good progress. Certain modules are moving to the construction and test phases. So we are seeing efforts in this area peak, then start to decline in Q4. In commercial air, opportunities continue to present themselves. We evaluate the long-term potential of each one and expect to continue making STC investments this year, which are both success-based and/or opportunistic in driving long-term growth such as expenses related to line sets. In this quarter, those activities grew our segment R&D by 67% from Q1 last year, which was the primary driver of the lower year-over-year segment earnings performance.

To provide the specifics around Government Systems, Q1 revenues were $183 million and adjusted EBITDA was $49 million, a new record high. This growth was spread across our diverse product and service offerings. Products included JTRS and weapon data links, global mobile broadband antennas and UAB products, as well as government satcom systems such as Blue Force Tracking equipment.

On the service side, we saw increases as well in global mobile broadband services and services associated with Blue Force Tracking. Together, reflecting our widespread growth drivers and the success we've had in taking commercially developed solutions and adapting them to meet the very unique needs of our government customers. Company-wide, we closed the quarter with $61 million in adjusted EBITDA, including the $20 million year-over-year R&D uptick and the $6.6 million reduction due to the completed SS/L settlement. Thus, adjusted EBITDA would have been up about 9% for the quarter without these impacts, again, outpacing revenue growth.

As we turn to Slide 5, we see what I just said in graphical form. Simply, you can see the year-over-year R&D and SS/L settlement impacts I just described. In fact, in order to see how the base business is truly performing, we also layered in impacts of ramping up for the ViaSat-2 service launch and the accelerating commercial air activities later this year, which totaled about $10 million combined. So when you exclude all of these negative impacts, you can see the strong growth in the base business of about $18 million in adjusted EBITDA in Q1.

These are the key drivers we expected and discussed last quarter, and they have aligned to what we expect for the remainder of fiscal 2018: the strong performance of our government and commercial air business; alongside the resiliency in our consumer broadband as we position for the upcoming ViaSat-2 service launch.

Looking at Slide 6, our income from operations reflected the lower EBITDA we just discussed as well as higher depreciation and other noncash expenses. Our GAAP and non-GAAP net income also follow the same relationship, but were partially mitigated by the tax benefit associated with the operating loss this quarter.

In cash flows, we see strong underlying cash flow generation in the business continue. Fiscal '17 was a fantastic year at over $400 million. And this quarter follows suit with a new quarterly record high of $154 million from operations. This figure was helped by a $84 million prepayment by Xplornet for the ViaSat-2 Satellite Services in the Canadian region. However, even without this benefit, Q1 operating cash flow was still very strong and 13% higher than the same quarter last year, thus bringing our trailing 12-month cash flow from operations to $503 million, which again is nearly 50% higher from the same period last year and a significant funding source for our ViaSat-2 satellite and our ViaSat-3 multiset constellation.

In CapEx, Q1 activities were $128 million, down $14 million compared to last year. ViaSat-2 satellite spend decreased as we entered the ground networking phase of that project, plus expenditures on the ViaSat-3 program were relatively flat. Also note that with the bulk of the ViaSat-2 project behind us, we expect total cost of the satellite launch, insurance and the initial ground network to come in around $580 million, which is about $70 million or over 10% lower than we originally estimated, which we're pretty pleased about.

We ended the quarter with no outstanding borrowings on our revolver and $275 million drawn on the Ex-Im loan commitment. Note that in Q2 fiscal 2018, we expect to make the final draws on our Ex-Im facility, including amount of proposed completion fees of approximately $30 million. Semi-annual principal repayments will then begin in Q1 of fiscal 2019 for approximately $19 million plus interest.

Our net leverage is still very low at 2.2x trailing 12 months adjusted EBITDA, reflecting sequentially flat leverage with cash balances growing to over $160 million from Q4 of fiscal 2017. So our liquidity remains very strong at over $980 million at Q1 end.

On this chart, I wanted to spend some time specifically on the Satellite Service segment, as we have a lot of activities underway there. It is important to note from the start that the launch of ViaSat-2 was delayed by nearly 3 quarters since the originally scheduled launch date, with the most recent delay resulting from civil unrest in French Guiana, the home of Arianespace's launch facility. So we find ourselves in a very bandwidth-constrained environment while preparing for our service launch on ViaSat-2.

For several quarters now, we have been talking about how we are driving increases in ARPU through improved plan mix. This is a result of the higher proportion of subscribers on newer, higher-value plans and a decrease in proportion of lower-value and wholesale plans. This trend has grown ARPU to a new record high of $66.61, which you can see on the top chart. And it's important to note that the increase in revenues from higher ARPU more than offset the year-over-year impact from the net subscribers decline, in line to what we indicated on the last call.

In the middle chart, we see how these strategies have driven ARPU growth over the last 2 years, where almost 70% of the gain in ARPU is purely due to an improved broadband plan mix, 20% due to retail wholesale mix with the remainder due to value-added services such as voice, EasyCare and Wi-Fi. So just to be sure, we do think that well-crafted and market-priced value-added services can be an integral part of our consumer services offering. We expect to see ARPU continue to increase in the future as well, but likely at a slower pace over the next few quarters.

And finally, on the bottom, we've provided a walk forward chart for segment revenues to help clarify the core year-over-year performance, which was revenue neutral even with the challenging competitive environment due to the launch delay of ViaSat-2.

Essentially, in Q1, we see the Lorel settlement done. Commercial mobility grew strongly and consumer revenues were up just slightly, with strong growth expected later this year.

So before I hand it over to Mark, we wanted to share this chart as a reminder of trends we saw leading up to ViaSat-1's service launch in 2012, as we're likely to see similar trends occur as we transition to ViaSat-2.

The key takeaways are: in 2011, we began improving existing WildBlue service plans in advance of ViaSat-1's service launch. On that network, we accomplished this by not replacing [NASA terms] and giving the excess bandwidth to the remaining customers, which led to subscriber count decreases for a few quarters. Financially, we also began to see fixed costs increase ahead of service launch, And we're seeing that again with ViaSat-2, with Q1 activities of about $3 million. Then, when ViaSat-1 went into service, back spends began growing in proportion to our growth-out activities, with about 2/3 being capitalized and 1/3 expensed related to commissions and advertising.

Finally, as we see in the chart, our revenues grew sharply, overcoming the fixed costs and stats, operating leverage took hold and EBITDA grew robustly to about 3x the peak levels we saw before ViaSat-1's service launch.

So I'm going to take a moment here and quickly remind everyone that the new revenue standards will come into effect for us in April 2018, with 2 primary impacts. One, commissions incurred by new subs after April 1 will begin to be capitalized and amortized over the customer life in future periods. And two, we'll record our recaptured commission asset net of amortization associated with the subscribers on the network at that time, which will also be amortized over the estimated remaining customer lives.

So because commissions will be capitalized, the (inaudible) historically should be less pronounced. However, advertising and marketing efforts will continue to be expensed.

So moving forward to today. We see that we're in that transition phase once again. The impact should be more muted this time as we now have a larger, more diversified revenue base in Satellite Services, including our growing commercial air business. And we are already paving the way for our newest ViaSat-1 plan offer and teeing up what's coming up next on ViaSat-2. So with that, I'll turn it back over to you, Mark.

Okay. Thanks. So we've described the high-level Satellite Services segment summary results, which are flat revenue consisting of a slight increase in consumer broadband revenue based on ARPU growth, offset by subscriber count attrition, plus strong growth in commercial mobility driven by more aircraft in service and greater bandwidth usage, and then both offset by the conclusion of the SS/L settlements payments. EBITDA declined due to the end of the SS/L payments and start-up costs associated with the ViaSat-2 network, and planned rapid growth and in-flight connectivity based on installing the existing backlog.

It's important to decompose all the factors to get more insight into our competitive landscape and growth outlook. We'll start with consumer subscriber attrition. There are multiple factors at work in the first quarter in addition to normal seasonal factors that affected ARPU, gross adds and disconnect rates.

First, DISH switched from a branded retail model to an agency model at the end of our fourth quarter. While we're working on mutually beneficial service plans with DISH, under their new agency model, we don't expect those to start until the second half of this fiscal year. We also saw increased disconnect pressure due to both unlimited mobile LTE plans and the new used Jupiter-2 plans. There are Jupiter-2 promotional pricing essentially tilted remaining DVS TV channels for hard and soft cap plans in their favor, and also affected the disconnects for our lower-end retail and wholesale plans.

But our dealer distribution channels and our higher value plans were much less affected by these competitive pressures. Strong growth in commercial and government mobility also meant we allocated more of our bandwidth to those applications, leaving somewhat less for consumer broadband. Finally, all these factors are manifested differently in different geographic areas, so it requires some detailed analysis to reoptimize our financial performance.

At a top level, ARPU increased significantly and overcame attrition because more of our adds came from high-value plans and more of our losses came from low-value plans. Remember that our bandwidth costs are essentially fixed, but our cash variable costs increase with more subscribers. So our earnings and margins are both higher if we can attract the same revenues with fewer subscribers.

We diverted some bandwidth to serve our mobility customers, but we also increased our bandwidth inventory, which creates opportunities for us to more strategically ignite growth in our consumer service ahead of the actual ViaSat-2 service launch, including beginning to add subscribers. So I'm going to go into more depth on that in these next 2 slides.

In many ways, the playbook for rapid growth in our consumer business is very similar to the same as it was for ViaSat-1, but updated to reflect the current competitive environment. The upshot is we expect to be in a stronger market position with ViaSat-2 than we were when ViaSat-1 launched. There are 2 simple main dimensions of value in the home broadband services market other than just price: faster connection speeds combined with sufficient bandwidth usage. The key to optimizing economic value and addressable market size is to balance those to make our service more desirable, all things considered, than the next best alternative in each geographic area. Speed is what primarily attracts customers. This chart, reproduced from a recent Craig Moffett report, clearly illustrates how customers prefer higher connection speeds and his view about the difficulty in serving those speeds with telco infrastructure.

Our Satellite Service infrastructure is designed to compete in the top half or better of the speed market on introduction. You can see that when we launched our 12-megabit-per-second service on ViaSat-1 in January 2012, we were about at the nationwide median for fixed broadband services. In the intervening years, we've generated substantial earnings. A 12-megabits is now a bottom quartile speed. So last year, we introduced 25-megabit-per-second speeds in some ViaSat-1 markets, which were again above median. But due to a combination of bandwidth and network infrastructure limitations on ViaSat-1, we decided to offer only a limited number of 25-megabit-per-second plans even though those have been in relatively high demand, because up until very recently, the 12-megabit plans were good enough to achieve the financial results we targeted given the long time gap between ViaSat-1 and ViaSat-2. But we can respond to market demands by increasing the proportion of 25-megabit speeds on the ViaSat-1 satellite. We already offer plans as high as 150 gigabytes at both 12 and 25 megabit-per-second speeds.

The ViaSat-2 network is designed for plans up to 100-megabits-per-second or possibly higher. So we have the opportunity to move further upmarket relative to the landscape by offering cable and fiber-like speeds and value to the pool of millions of subscribers who have no other choices at those data rates.

As an aside, our ViaSat-2 network infrastructure is backwards compatible. So you also have the option of using it over the ViaSat-1 satellite and delivering 100-megabit-per-second services there as well. Of course, we have to deliver enough bandwidth at those speeds to make the service attractive and retain the customers we acquire. While volume's important, it's secondary to speed. For instance, an unlimited 5-megabit-per-second DSL service doesn't look very attractive because it will never perform as well as the 25-megabit service for applications that benefit from that speed. As with speed, expectations for volume also grow over time and they're well-understood mechanisms for predicting those volumes and relating them to addressable markets. Finally, don't forget about our objective of creating differentiated value propositions. We'll come back to that in just a couple of minutes.

So here, we show the ways we design our service plans to balance speed and volume in creating useful services. Most subscribers don't relate to gigabyte volumes, but they're still unhappy if they reach their usage cap. Per capita consumption increases over time, so if those caps remain constant, more subscribers will reach them each year. But if we can improve bandwidth productivity on our satellites and bring them into service at the right tempo, we can earn an attractive return while delivering competitive services to a big market. But when you think about that, average bandwidth consumption is a terrible way to think about usage. This slide shows FCC data on the distribution of bandwidth consumption from a 2009 report on the broadband availability gap. While speeds and volumes at that time are much lower than they are now, the distribution of relative consumption is still highly relevant. The left-hand side labeled A of the FCC figure shows that like many people report, a small fraction of users consumes most of the bandwidth. For instance, 1% of the subscribers use 25% of the total bandwidth, 2% use 40% of it and 7% use 65%. On the right-hand side, with the B on it, translates that distribution into percentiles of busy hour offered load and shows how misleading averages are in determining your addressable market.

At that time, the 50th percentile subscriber generated only about 20% of the peak busy hour traffic compared to the average one. And a broadband service provider could have served 80% of subscribers with only about 20% of the average peak busy hour load.

Then the chart in the upper right from Cisco in June of 2017 shows the current average bandwidth consumption, and they divided it into fiber to the home users compared to high-speed DSL and cable. Fiber homes averaged about 28% higher consumption. We're not attacking fiber, so we use a lower number as a benchmark for us. Cisco projects forward about a 20% compounded annual growth rate for bandwidth consumption for the non-fiber homes for the next 5 years.

Taken together, data on the average consumption and distribution models help us frame our offers on ViaSat-2 and beyond, to optimize our return by addressing the rightsized market for our bandwidth supply. While volume caps are helpful for forecasting and managing networks, they're off-putting to customers who don't want to deal with usage constraints.

Also, from a behavioral economics perspective, increasing a bandwidth cap by, say, 20% may not represent a 20% increase in value to a customer. So we want to compete within a competitive landscape that's familiar to consumers. For instance, mobile carriers have equated unlimited to mean a 22-gigabyte-per-month soft cap. That includes only 10 gigabytes per month of tethered data, and that's the bottleneck for Wi-Fi devices in your home such as an Apple TV or a Roku.

For DSL, they generally equate unlimited with about 150 gigabytes, a level we currently offer in a number of places. Despite short-term competition from unlimited LTE plans, it's not clear those plans will have enduring value for fixed home use. While households can save on home broadband by using their mobile device at home, it's becoming evident they can save even more money by actually upgrading broadband and switching to an over-the-top video instead of broadcasting. So by combining the data from this chart and the previous one, it should be clear that our best service plans from 25-megabit-per-second speeds and or higher and softer volume caps are still quite competitive and are likely to have greater attention, although our 12-megabit-per-second plans with 10 gigabyte caps are more vulnerable and are churning at a higher rate. So we've undertaken near-term actions to both reduce churn for a particular cohorts of subscribers and to increase gross adds for higher value plans that we believe will mitigate net subscriber attrition.

And we can use our bandwidth inventory in our third quarter to introduce ViaSat-2 type plans in specific geographic markets and channels to prime our distribution for ViaSat-2 service launch in the fourth quarter, creating good growth opportunities while still most effectively optimizing economic value. But our biggest focus was on the ViaSat-2 and ViaSat-3 services in addressable markets. So let's take a look at that.

When we first introduced ViaSat-1, no one focused on the capacity of satellites. Since then, it's clear, capacities are a very important measure to determine the capital efficiency, quantity and quality of broadband services per satellites. But there's no standard measures of capacity. To be sure, capacity varies depending on the types of use of terminals, their locations within the beams, the way satellite power and bandwidth is allocated, local weather and other factors. Still, apples-to-apples comparison to network capacities is important to investors, customers, distribution partners and others.

So we take a simple, but very useful proxy for capacity that can be independently measured to the total bandwidth of the gateway terminals for a given satellite, focusing on the forward link for downstream bandwidth, because that's the bottleneck. That's the bandwidth from the gateway to the satellite. All users are connected to the Internet through backhaul at those gateways. And since gateways are licensed, there must be filings that identify the locations and spectrum associated with each one.

Total gateway bandwidth, which is the number of gateways times the bandwidth used by each, is a good proxy for total capacity as a result of Shannon's Theorem, which says capacity is directly proportional to bandwidth but only increases with the base 2 log of signal-to-noise ratio. Modern satellite networks get very close to theoretical capacity for signals and noise ratios in typical ranges. A satellite system with only 1/3 the spectrum of the competing [one metal] would need 8x the power to have equivalent capacity, other factors being equal. And ViaSat-2 and ViaSat-3 are among the most powerful commercial satellites ever built in terms of total bus power and user-linked effective isotropic radiated power, or EIRP. The number of gateways is proportional to the total of bandwidth, which is why we place such a high emphasis on R&D that allows us to drive down gateway size, footprint and cost, and allow compatibility with 5G applications of millimeter wave frequencies.

So the left-hand chart shows total forward link gigahertz for ViaSat-2 compared to ViaSat-1 and other spot beam satellites. We've also cited the FCC filings used to calculate the data. The Ku- and Ka- values that are in the left-hand column are based on websites or published data. Then the right-hand chart shows forward link bandwidth for the first ViaSat-3 satellites in the context of the current state-of-the-art.

We also show the total forward link bandwidth for a new used piling at 95 West, that might or might not represent their next-generation satellite. Different satellite architectures may also apply aggregate gateway bandwidth in different ways that affect user capacity. But we still believe this is a simple, powerful and consistent way to assess capacity.

ViaSat-2 clearly has a very big advantage and it gets even better with ViaSat-3. Just to be clear, we're not saying satellite broadband is a 0 sum game and that we'll be the only winner. In fact, the total market for satellite broadband in the U.S. grew substantially beginning in 2012 with the launch of ViaSat-1 and Jupiter-1. But we do think there's a substantial competitive advantage from satellite networks that are much more capital efficient as well as more flexible in matching supply and demand than other alternatives. We currently believe the best way to leverage that advantage is to position our service at are higher into the broadband market than others with higher than median speeds and plans that are reduced or reduce [gear] caps. That plans that are positioned to reduce [fare] caps, sorry. As opposed to offering the same plans to more subscribers or the same plans at lower prices, we believe we can ultimately earn a better return on the investments in the satellites, being more resilient to terrestrial competition and address a larger market. And now we're just a few months away from bringing this to market.

Invariably, we've spoken about how our in-flight Wi-Fi service revenue is growing fast. And a recent reason is gate-to-gate service, allowing passengers to connect while on the ground. Gate-to-gate can be a good way for airlines to differentiate, especially on shorter flights among busy airports, such as New York, Boston or Washington for a long time spent on the ground. Our gate-to-gate performance is only one of the factors that makes our service the benchmark for speed and value. We've got 840 additional aircraft under contract and our installation rate is planned to accelerate beginning later this quarter. There's rapid industry-wide growth in Wi-Fi plans, but we think there isn't enough satellite capacity to support other services as the number of planes grows. Pretty much any service works with only a few planes. But we expect that in the next year or so, it will become clear there's significant network congestion issues at the busiest airports served by major carriers.

While there's a lot of focus on cost per bit of bandwidth, the failure mode that passenger's see is that there isn't enough bandwidth near busy airports and Wi-Fi becomes slow. That's the same failure mode that occurred with air-to-ground. The remedy there was to raise prices, which reduced demand in its attempt to improve performance. While that might be good for the service provider, it's not for the passengers and that has not been good for the airlines. So to help illustrate the problem, we analyzed scheduled commercial flights for the major airlines in the U.S. and Europe to show how much bandwidth would be needed over some of the busiest airports. The number of simultaneous planes to be served near an airport depends on the service radius for your transmission system. Think of that as the size of a spot beams. Larger beams have to support more planes. And with larger beams, nearby busy airports compound the problem, such as in the U.S. Northeast. The upshot is that the bandwidth needed to serve major airlines at busy airports will be in the gigabit per second ranges. And that problem is even harder for Ku- and global Ka-bands spot beam satellites because their beams are so large. You can see how big those spot beams are up by looking at the beam maps on their websites. While different providers may have only a portion of the airlines at each airport, the bandwidth needed is so high that even a single major carrier can't be served if the passenger take rate gets much above single-digit percentages.

So the way to look for this effect is to test in-flight Wi-Fi service at or near busy U.S. airports such as Atlanta or the New York-Boston-Washington corridor, or in Europe at Frankfurt or London. We expect the effect to be increasingly evident as the number of planes in service increases. Higher prices or other barriers to contain usage would likely come up.

So our strategy is pretty straightforward. We believe, in-flight Wi-Fi is one of the most cost-effective ways to improve passenger satisfaction, but that only works for passengers that actually use it. So we aim to help our airline partners deliver greater customer satisfaction at lower cost, not by just -- not by merely having a good Wi-Fi system, but by taking advantage of the power of the Internet to engage more passengers. That means appealing to different passengers in different ways, including business use, social media, stored or live entertainment, or en route travel services. To the extent that passenger engagement can be monetized by the airlines, we believe we can achieve substantial competitive advantage. Our airline customers so far have been the ones most interested in increasing passenger engagement. We're doing our best to make them successful in their markets and we're also investing now in [line fit thin] and STCs to get on the most valuable planes into scale to support substantial growth.

So our Government segment had a stellar quarter on every front: revenue, new orders, backlog and profits. As I mentioned before, it extends and may help improve a 5-year growth trend that has seen revenue grow at 11% a year and earnings at 17%. Our fastest-growing areas have been in mobile broadband and tactical data links. Both of those areas benefit substantially from R&D initiatives we've been investing in for several years to create Non-Developmental Items with unique proprietary products that have delivered capabilities outside those identified by the program of record based acquisition system. Our current rapid growth is due to distinct trends. New products and services are being adopted by different, but functionally similar government organizations and different platforms or use cases within those organizations.

For instance, we've got strong relationships with early responders to troubled situations worldwide. Over the last few years, our mobile broadband and tactical data link products have helped multiple different types of these organizations for different missions in different locations.

But there are some big opportunities for us to grow very significantly if or when some of these products and services transfer to more mainstream organizations. We're starting to see some of these opportunities now. If we're successful, we believe our growth will accelerate meaningfully beyond what we've experienced so far. The opportunities are really big. They're also very lumpy. Our target markets are also attracting large competitors, though that has always been the case.

While we primarily sell network transmission products and/or services, our success has mostly been driven by helping our customers apply the capabilities of those products and services to improve specific measures of mission effectiveness by more effectively sharing information under difficult conditions, being able to deliver benefits such as improving response times, improving situational awareness, increasing the accuracy, increasing resilience of threat environments, reducing mission costs, reducing collateral damage and saving lives.

Okay. So we'll switch our outlook in summary, and I'll cover some of the major factors for the balance of fiscal '18 for each of our segments, and we'll start with Satellite Services.

So again, we've already described the major factors affecting EBITDA for the satellite services segment in comparison to last year. The SS/L international property settlement payments completed in our fiscal '17 which accounts for about $26 million revenue and EBITDA reduction for this fiscal year compared to last fiscal year.

Shawn showed that we incurred a total of $10 million in ViaSat-2 services and network start-up costs in the first quarter between consumer and mobility services in anticipation of rapid growth later this fiscal year. Those costs will continue and some portion will actually increase during the fiscal year. We expect sustained and accelerating mobile services growth as we install more aircraft and see additional bandwidth usage.

We do expect that net consumer subscribers will likely decline in the second quarter, though to a lesser extent, based on retention and distribution initiatives. We believe that subscriber growth can resume meaningfully in the third quarter and we deploy technology and bandwidth with ViaSat-2-like plans in select markets and prime our distribution channels.

Tax expenses will weigh on EBITDA initially as gross adds and subscribers increase. Also, some bandwidth and associated revenue earnings will migrate from our Satellite Services segment to Government Systems as our government mobility services scale.

Overall, we think the medium- to long-term outlook for investment satellite services growth is very positive, given the services plans we can offer with ViaSat-2 compared to the current competitive landscape. The history for ViaSat 1 services that Shawn reviewed is a good way to think about trajectories for revenue, EBITDA and subscriber growth as we transition to ViaSat-2.

For our Commercial Networks segment, we expect meaningful revenue growth from sales of airborne terminals for our expanding in-flight connectivity network. There is significant backlog of aircraft, stretching over about 3 years with a lot of opportunities to grow that backlog, some of which are at a very mature stage. The second half of the year, we should see reductions in R&D investments on ViaSat-3 payloads as we transition to flight hardware production. We anticipate that R&D in supportive commercial aircraft STCs and our [line set] will continue based on the opportunities we see in the market.

We're especially seeing more opportunities for long-haul aircraft given the launch of ViaSat-2 and as ViaSat-3 gets closer. Then as we discussed earlier, Government Systems has excellent opportunities for continued strong growth and momentum. We've got a record backlog now in the early fiscal year '18 and good opportunities in mobile broadband, tactical data links and satellite system applications and in cyber security. So with that, we can open it up for questions.

(Operator Instructions) And our first question will come from the line of Mike Crawford with B. Riley & Co.

Could you provide an update on the status of the 2 JVs you're creating with Eutelsat?

Sure. The first JV is the one that uses their existing -- 2 joint ventures: 1 for infrastructure, 1 for retail, those are the 2 JVs. And the first application of those joint ventures is using the existing Eutelsat case at satellite. So the first things we've been doing there basically, working with them to analyze what their current performance is and to make changes to the distribution system, the service plans to get financial results that are more like what we've gotten from ViaSat-1. And I'd say that is going well. We've got basically, top level agreements with Eutelsat to make changes in both the plans and the way we distribute them, and to introduce some of our newer services like in-flight connectivity and government applications to KA-SAT that will also help with the financial results. It's going to take a little while and right now, the changes there aren't really material to our results. For the next stage of the joint venture is to incorporate the European ViaSat-3 bandwidth into the 2 entities, and we're still working with Eutelsat on that. I think we both talked about the end of this calendar year is kind of a good target. We're trying to reach an agreement on that part of the joint venture.

Okay. And then just to get back into this satellite capacity explanation, where you helped provide the Shannon's Theorem. How else is that capacity shaped by some of the other technologies you bring to market, like some of the beam forming techniques you have in ViaSat-2 and 3 that aren't -- that weren't in ViaSat-1?

Okay. Yes, so there's really 2 parts. We think of it in 2 parts. Part #1 is, how did you get a lot of capacity? And the whole point of that discussion is the way you get capacity is through bandwidth reuse and it's exactly the same in terrestrial mobile. If you want to increase capacity of a terrestrial network, you either get more towers, you get more bandwidth per tower or both. So that's the first part, but the second part is really being able to apply that bandwidth in the places where the demand is greatest and where you can get the greatest return for it. So think about that as moving the bandwidth around. And in some cases, what you'd rather have is actually less than the peak amount of bandwidth if the peak bandwidth is wasted in places where there's not demand. So that's what our technologies around bandwidth flexibility or bandwidth portability are about. We didn't really even go into that part because the first part was such a huge difference. Well, we're pretty, pretty confident in is that there aren't any other satellites that have any of the bandwidth flexibility that ViaSat-2 and 3 have at the scale of bandwidth that they can deliver. We think that flexibility is just another kicker on the value of the satellites, but the point of what we're trying to go to is just really to compare the total amount of capacity.

And then just last question would be, you said you're investing in line-fit. So on what platforms are you line-fit now?

Yes, we've announced the line-fit on the 737 max. We have others in process.

And the next question will come from the line of Andrew Spinola with Wells Fargo.

Mark, I was wondering if you could sort of dig a little bit further into the subscriber result in fiscal Q1. I think on the last call, you felt relatively confident that fiscal Q1 would look a lot like fiscal Q4 and obviously, net adds were much worse. What changed? And maybe, I know you gave a couple of reasons, whether it's DISH or sats and LTE unlimited. Can you kind of break down some of these things in terms of what impact they're having? And then from there, take us into the next quarter.

So I think we've always said in the last quarter was that we didn't think that subscribers would continue to decline in the first quarter, but that the growth in ARPU may compensate for that, and that's what happened in this quarter. We put the decrease in subscribers and that was really all the factors that we described. Some of it was seasonal, some portion of it was due to more competition from unlimited mobile wireless, a little bit of it was from competition from the new Jupiter-2 satellite. The Jupiter-2 satellite really only entered service in the April time frame which was this quarter. And then also, talked about some of it being because we migrated bandwidth away from the consumer service into the mobility services, and we basically accumulated some bandwidth inventory.

But over half of it came from wholesale.

Yes, yes.

Basically, (inaudible) because they have no gross adds in (inaudible).

(inaudible) we talked about the DISH thing and then the DISH switching from wholesale to this agency model. And then the other one was in the other TV channels where they're basically just comparing our services to the new Jupiter services, they got a much bigger fraction of those gross adds than we got in the previous quarter, in the fourth quarter. So some of those factors will continue. On the other hand, the things that do continue to work for us are fewer distribution in our higher-end plans which by definition, are the higher-end plans they use more bandwidth. They yield higher ARPU and that was what drove the ARPU growth. So we'll see -- we expect to see kind of the same trends in the second quarter, but we believe that we'll have some initiative that will mitigate churn in this quarter and that will increase gross adds relative to the prior quarter which means about we probably won't have as much attrition but we do expect there to be attrition this order. Third quarter, we think things will be significantly different than they will be in the second quarter.

And maybe can you expand on that? It seems like my assumption is ViaSat-2 is not available until January. So how could things be so different in the third quarter that you could turn this around?

Oh, because like I said, we're getting some bandwidth inventory back and we're going to repackage that inventory into plans that are more ViaSat-2-like and then also take advantage of some of the technical improvements that we're going to do on the ViaSat-2 services which will lead to higher bandwidth yield for that bandwidth in terms of some combination of subscribers and ARPU. But think of it as, what we'd really like to do is scale up our services, basically our distribution channels and our gross adds going into our fourth quarter which is when ViaSat-2 is really available. So to do that, what we want to do is package that bandwidth inventory that we're getting into these ViaSat-2 forms and basically just go out with those services as opposed to dribbling it out beforehand.

We'll also increase advertising in kind of September and October and November.

Yes, leading up to those services, and that was a factor as well. But the way that the amount that we're investing in advertising based on the staging of the new heads that we'd like to get.

Understood. Just one last one for me. I think you made the comment that the R&D associated with the satellites will dissipate. Can you be any more specific into fiscal '19 from fiscal '18, any magnitude of what that step down in R&D will be in an absolute sense?

Sure. I think -- so just to kind of reiterate what we kind of said for you, Andrew. What we said is that there's kind of 2 parts, right? There's the (inaudible) on ViaSat-3 as well as the mobility. On ViaSat-3, I think we'll see those start to at the end of the year, come down a little bit. Hopefully, with some of the good opportunities we have in commercial air, we'll see commercial air step up a little bit. I think if you continue forward looking into '19, you're probably going to see levels that are kind of pre-ViaSat-3 kind of type levels.

I guess, let me just push on that. I mean, pre-ViaSat-3, your R&D was sort of $77 million in fiscal '16 instead of something like $145 million, $150 million I mean -- or $160 million even. Are we -- I mean, are we talking about that type of step down?

I think you have to think about it as a percent of revenue and it steps back down too. And

that was a more normal level as a percent of revenue. So I think that's -- what do you say, Shawn?

Yes, I think that's right.

Like 5% or 6% of revenue?

It won't happen, it won't just snap from March to April, but it should be down considerably by March.

And the next question will come from the line of Rick Prentiss with Raymond James.

A couple questions following on some of the earlier ones. When you think about the STC is obviously investing and important to get those, what percent of the aircraft type do you think you've already got the STCs for? What's your target as far as what you're shooting for over the next 1 or 2 years, kind of piggybacking on that mobility investment you were talking about?

Okay, yes. So the 737 max is going to be the -- probably the single most popular mainline single aircraft going forward. So that's great that we've got that one done first. We're on a bunch of Airbus 320 series which is kind of the Airbus analogy to that. We're not line-fit on that one yet and that clearly would be valuable to the number of our existing airline partners. So that's an example of one that we would target the -- and that those 2, from a line-fit perspective and an STC perspective, cover a lot of the markets that we've been going after which are sort of North American markets, Europe -- intra-European markets and the Australian market, for example. But with ViaSat-2, we're really getting in position for the transatlantic market on the ViaSat-3 for these global on home markets. So that's going to be an increasing focus with the timeframe for getting those STCs and line-fits kind of lining up with the ability of us to serve those airplanes.

Okay. And that's kind of why maybe even a multiyear process on the mobility R&D staying up high, at 100% of revenue?

Yes.

Okay. Obviously, you've spent a lot of time and appreciate all the charts in the presentation, talking about the bandwidth and the market share on tac and not fiber homes. But obviously, you're going to be targeting kind of above the medium level, I think you mentioned. How should we think about that as far as your target ARPU range then? Because you would think higher-value plans mean higher ARPU? What ranges should we be thinking about where ARPU could have in this transition pre ViaSat-2 and then as you roll out ViaSat-2?

So pre ViaSat-2, it's a little bit unclear, probably because a big part of the attraction of the Jupiter-2 plans has been the promotional pricing that they've used which makes a lot of sense for them given that they can add a lot of subscribers. For us, we'll have to think of -- we're going to have a response to that but it has to be a little more muted. But commercial pricing is a factor in our ARPU based on how that blends into our existing base. So I'll say, in the next quarter or 2, it's a little bit unpredictable whether or not it will go up much, or conceivably, it could come down -- not sure about that, but it's conceivable. But in the long term, we think ARPU is going to go up meaningfully because these higher speed plans which are really attractive carry much higher retail pricing. So I don't want to put any numbers on it because I think it will evolve over time, but I do think that ARPU will continue to grow. And I would say that rates that we've seen over the last few years, you could grow those forward by the time we get ViaSat-2 in full swing.

Okay. And my final question is cash went up for the quarter. I think Shawn, you mentioned something about a prepayment from a Canadian area. How should we think going forward, the rest of this fiscal year into next fiscal year, kind of a cash burn rate and when we might target free cash flow positive out into the future as you invest for the future right now?

Yes, I probably -- I think you guys -- hopefully, we've given you a lot of color on what our EBITDA trends are and what we see the generation there. So really, the solver there is CapEx and it's probably going to be -- it really tends to bump around based on the milestone payment. They can move $25 million, $50 million one quarter to the next. But our range of around $550 million, $600 million in CapEx total, all more -- a good more than 50%, 60% of that being the satellite. So that might be a good kind of '18 range.

And then longer-term, as far as free cash flow positive targets, any long-term thoughts out there? I know you've got ViaSat-2, you've got this government work that's pretty exceptional, but when should we think about free cash flow positive on the horizon?

Yes. Well, I think honestly, it's really pacing on what our activities are and when we start the APAC satellite. I think if we look outward, it happens pretty quickly. But it can be real lumpy with the investments that we're making in stats and how accelerated we grow.

Rick, on a free cash flow basis, what we've been saying is that'll really occur when ViaSat-3 -- the first of the ViaSat-3s are in service. That's kind of the timeframe.

Yes, there's a lot of comfort around that, yes.

And is that showing like kind of a 2020 time frame then?

Yes, well, a couple of years past that because you have to really -- you're ramping on as you go up there. But as they get into service, yes. That's an important time when operating cash flow should overcome the CapEx.

And then at that point, even with additional satellite builds, free cash flows, it should be very, very strong and stay in the positive.

And the next question will come from the line of Andrew DeGasperi with Macquarie.

So just the first question on the savings you achieved on ViaSat-2. I was just wondering, that 10% number, do you think it's achievable on ViaSat-3? I know it's a different satellite but just wondering if that's -- if there's a possibility there. And then secondly, I was just wondering, for ViaSat-2, I know you mentioned some allocation of the capacity and bandwidth to different verticals. Do you have a rough idea of where you think residential in-flight and government will likely fall?

Okay, I didn't quite understand the first question. Could you repeat that, please?

I believe you mentioned the satellite (inaudible) came in at $580 million versus the initial cost estimate. And I was just wondering if that's potential to achieve that as well on ViaSat-3?

Well, it's a case-by-case basis, so it's way too early to tell what the ViaSat-3 outcome will be. A lot of the ViaSat-2 stuff turned out -- at near the end, we achieved certain program objectives, especially in gateway and infrastructure, and that's where a lot of those cost savings materialized. We're not far enough along on ViaSat-3 to predict that.

We did the same thing on ViaSat 1 but what we've tried to give you guys is numbers that we feel like we could hit. So there's a little bit of conservatism in there.

Got it. And sorry, on the allocation bandwidth, how should we think about that for VS-2?

Well, it's going to depend on how the rate of growth for some of these different services, including government and the commercial in-flight. But when we first did ViaSat-1, we were thinking about, at least at the beginning, about 90% of it would go towards consumer broadband. And I think when we start ViaSat-2, it might look that way but I think it'll evolve to be a greater fraction than that, that will go towards mobility and government applications. And then we're also just starting some enterprise and pay-per-use Wi-Fi services that also could end up being a disproportionate -- it'll contribute more towards margins than they do to bandwidth use.

And last question. As far as Inmarsat and what's going on there in Europe, I'm just wondering if you could add any color as far as your viewpoint there and what do you think timing is as far as what you can do?

Our viewpoint is that we don't think that the S-Band air-to-ground network is consistent with the objectives of the -- and the requirements of that S-Band application. So we've made that case and we're still waiting for that to play out. I think it's a pretty clear case. I don't think that it -- we don't think it had been considered in the way that we've framed and we'll find out.

And the next question will come from Philip Cusick with JPMorgan.

It's actually (inaudible) on for Phil. I just had a quick question. In terms of how we should think about EBITDA trending for the remainder of the year. I think Satellite Services revenue flattish. I think you've communicated that as Loral kind of settlement payments kind of go away. You also have additional start-up costs related to the ViaSat-2 broadband offerings, investments in the fiber and ground networking. So I just had a question, how should we think about the trend of that? Obviously, the start is probably back half-loaded; safe to assume the fiber and ground network investment as well will follow a similar trend. And with that, as well as the somewhat elevated R&D, I think, which should wane in the back half of the year. I'm just trying to think about what the full year, fiscal year EBITDA trend should be with all those factors.

Sure, sure. So this is Shawn. So I think, obviously, one thing keeping in mind government can be a bit lumpy. But this is always probably one of the toughest quarters for government and it was a fantastic quarter. But on the backdrop of that, obviously just all the dynamics that you talked about is we're going to have -- our R&D is going to be a little bit heavier in the first half than the second half overall, but it's going to stay elevated for the next couple of quarters and then we'll have strong growth with ViaSat-2. So it's probably good to think about it as a little bit pressure in the next couple of quarters, and then we'll have strong growth as we look to the second half.

If you just think government business and profile it out, we don't think there's material changes in our trajectory there. And then as Shawn showed you that chart that showed what happened to EBITDA as we've started up on ViaSat-1, and I think the same effects we see occurring on ViaSat-2, somewhat offset by R&D reductions. So it's...

And our...

We don't really give industry guidance, so kind of give you guys the pieces.

That's helpful. And then just one other question just on the IFC segment. I think last quarter, you mentioned that as you get into next early calendar year, installs should hit about 100 per quarter. Obviously, should we expect that to ramp as we go through the remainder of the year, somewhat obviously, in the September quarter as maybe aircraft availability for installs increases a little bit there? And just what is the, I guess, expectations to get to that 100 per quarter? Is it a calendar 1Q kind of thing or 2Q? How should we think about that? And just another question for IFC is, the ARPA trend I guess, overall, IFC revenues were up nicely in this first quarter. But as you move from a wholesale, I guess, capacity provider model to more of a direct service provider, how should we think about ARPA trending as well as maybe some of the IPTV and in-flight entertainment initiatives you've been working on, I guess, maybe post Arconics? I'm just trying to think of the revenue trend in that segment of your business.

Okay. I'll talk about the second part first. So the second part, we think ARPA is generally trending higher and for the reasons that -- well, one of the reasons we described was that there's more passengers using it and the passengers are using more bandwidth, so that's a factor. We are adding additional services, including in-flight entertainment. That's one of the things we talked about. For some aircraft -- for some airlines that are also interested in TV, although one of the things we're excited about is delivering TV to over-the-top services as opposed to broadcast which plays a little bit into the overall theme that we talked about in the first part which is the bandwidth getting to more passengers and having passengers use more bandwidth. We're also trying to work with our airline customers to help them offload those bandwidth usage costs as well in multiple ways. But the general trend, we think, is going to be up in terms of ARPA. And then the issue on the installs, I would say well, first or second quarter of calendar '18 is when we'll hit that, where we're targeting to hit that 100 a quarter type of range.

And we're hoping it continues to go get through other wins.

And the next question will come from the line of Chris Quilty with Quilty Analytics.

A question on the government mobility. Is that growth coming from the growth in the number of aircraft platforms or the amount of bandwidth you're using? And then the second part of the question is, given that, that was historically not a requirement, where's the money coming from?

Okay. So first, it is coming from -- basically, the 2 things, both of the things that you said which is getting on more platforms and those platforms using more bandwidth, both. And some bandwidth usages depend on the missions for the platform and the requirements, the amount of bandwidth that they want us to hold to assure that they get the services that they need for their missions. The point that I made about requirements would be, as an example, you won't see a program requirement for -- in certain types of command and control, in-flight connectivity for a number of new platforms that we're trying -- that we're actually not trying to get on that we're being tested on. And so some of those include rotary wing aircraft, as an example, that we talked about in the past, where operational elements are finding that if they do have in-flight connectivity, that they can perform their mission a lot better. So rotary wings are really an interesting market because there are lots and lots of them that would be a really big growth area for us. That would be an example of one. There's multiple applications that we don't really or generally talk about, where by putting connectivity on certain types of platforms, they can either enhance their mission or add missions that they don't currently perform.

Or in many cases, reduce flight hours required for certain missions.

Bet you the pilots love that. Two questions on consumer. One is, I think you've been running around 40,000, 50,000 gross adds. Is it still fair to assume that if you're warehousing some capacity, that the gross adds were down pretty significantly in the quarter? And second question, if you're migrating customers now to 25 MG plans, what happens with the introduction of ViaSat-3 and the new set of plans with higher tiers?

Yes. So okay, with ViaSat-3 or ViaSat-2 on your question?

Let's stick with 2 for now.

Yes, okay, okay. So yes, we expect to introduce higher speed plans. One of the things we expect when we introduce higher speed plans is -- and this is exactly what happened when we introduced ViaSat-1, is that customers with lower speed plans will want to migrate to those services. So with ViaSat-1 and 2, for speeds better than the 25 megabit-per-second range, we actually have the option of migrating them on either satellite. We may add infrastructure to ViaSat-1 that would allow us to migrate customers on that satellite to plans that are higher. What we expect and we talked about this multiple times, is that as we migrate customers to higher speeds and plans that use more bandwidth, ViaSat-1 will have fewer customers on it, and we're starting to see that effect, but ViaSat-2 will have a lot more. And that net effect will be to have more subscribers than we had with ViaSat-1 only. But especially, one of the effects that we've seen with ViaSat-1 is that the marginal earnings that come from adding additional subscribers are much higher than our average earnings. So by adding marginal subscribers, we can grow our earnings a lot faster than we can grow revenues. That's basically the approach that we will use which will add, particularly, the ViaSat-1 like amount of subscribers on ViaSat-2 and that the actual ViaSat-1 subscribers would go down. Does that answer your question?

I think so. But I think with WildBlue, I think you went from like a peak of 415,000 subs down to whatever the number is, 100,000, 150,000. Do you expect to see that order of magnitude of decline on ViaSat-1 or is it a little more gradual?

No, it's more gradual. So...

Well, we went from 500 kilobits on ViaSat-1 (inaudible) WildBlue-1 to (inaudible)

Speeds went up by a factor of 20, bandwidth consumption also went up. So with ViaSat-2, there were -- we're really looking at speeds that would go up in range from 2 to maybe as high as 8, right? That was -- or given that we already have 25-megabit subscribers, something like 2 to 4 type of blended range. So I think what we've -- also, we talked about bandwidth economics for ViaSat-2 being kind of double what they were for ViaSat-1 because we underran our budget, the economic threshold a little bit better than that. So those are kind of the ways to think about it. Well, what we've said multiple times is that we want to go out and find in the market the way that we can best optimize performance by this combination of speed and volume.

But not just subscribers, were able to -- we've been able to reuse WildBlue-1 and Anik F2 for other purpose, mobility is one, for example. And so having that fleet allows us to -- adjusting subscriber count isn't really the key.

It's not the only way to measure it, yes. We're really focused on revenue and earnings growth.

And then the gross adds in the quarter?

I think we've told you guys which would -- Mark told you what the key drivers were in the quarter and we said over half of the net has came from the 0 gross adds source which was our wholesale source.

And the next question comes from the line of Louie DiPalma with William Blair.

Government bookings growth this quarter ramped to the strongest it's been in at least 7 years. Was that mainly due to demand for your traditional data encryption appliances and tactical data links? Or were there new products, like the BATS-D, contributing to that?

It's mostly -- for that quarter, it's mostly more traditional existing products. I think we've got good prospects ahead for some of the newer products. But right now, it's more existing.

Okay. And when do you think the BATS-D and the small tactical terminal can become programs on record?

So the small tactical terminal has already happened. So the history there was, originally, there was going to be a variant of a JTRS radio, the airmobile version, [A and F] version, that was going to get used on small platforms and that program basically didn't compete successfully. So we -- one of the things we talked about in the last few months is that we were selected for the Apache helicopter which is one of the biggest applications of that small form factor radio. So that's already happening and I think that the prospects for getting more and more of the platforms that are sort of orphaned by the JTRS program going away is really good for the small tactical radio. But the BATS-D handheld Link 16 application is something that's really not envisioned at all in any of the DoD radio programs, but has been shown to provide great value. And we think it's a really good example of the type of market that we're going after. So that one has been in some evaluations, has gotten really favorable reviews. And we think over the next year or so, that we'll start to see some meaningful revenue from that product.

Okay. And lastly, is the timeline for the ViaSat-3 Americas launch still in late 2019? And what are the future milestones that we should monitor ahead of that launch?

Yes, so we've been talking about late '19, early '20. I don't want to mislead people, there's still -- this is really hard stuff. There's still uncertainty associated with those dates. The kind of the main milestones along the way, one is we've got a critical design review for the payload coming up.

This month.

Yes, this month. So that's where we would freeze the design. That's a big milestone for us. Another milestones will be commencement of the actual in-flight payload hardware. And one of the artifacts you'll see associated with that is, actually, we'll start capitalizing more of the satellite and decreasing the amount of expenses, R&D expenses associated with it. So you should see that happening shortly after the CDR. I think those would probably be the next near-term milestones.

Just to reiterate, we haven't announced any kind of slip in that schedule. That's where it's been for a while. I think that's it, Mark. Last call, last question.

Okay, good. And thanks a lot. I know we've covered a lot of material. Thanks a lot for your patience and everybody dialing in. And we'll be back with you next quarter.