Good day, ladies and gentlemen, and welcome to your Q3 2017 Gogo Earnings Conference Call. [Operator Instructions] As a reminder -- I would now like to introduce your host for today's conference, Ms. Alva, Treasurer and VP of Investor Relations.

Thank you. Good morning, everyone. Welcome to Gogo's third quarter earnings conference call. Joining me today to talk about our results are Michael Small, President and CEO; John Wade, Executive Vice President and COO; and Barry Rowan, Executive Vice President and CFO. Before we get started, I would like to take this opportunity to remind you that during the course of this call we may make forward-looking statements regarding future events and the future financial performance of the company. We caution you to consider the risk factors that could cause actual results to differ materially from those in the forward-looking statements on this conference call. These risk factors are described in our earnings press release and are more fully detailed under the caption Risk Factors in our annual report on Form 10-K and other documents we file with the SEC. In addition, please note that the date of this conference call is November 2, 2017. Any forward-looking statements that we may make today are based on assumptions as of this date. We undertake no obligation to update these statements as a result of new information or future events. During this call, we will present both GAAP and non-GAAP financial measures. We include an explanation of adjustments and other reconciliations of our non-GAAP measures to the most comparable GAAP measure in our third quarter earnings release. This call is being broadcast on the internet and is available on the Investor Relations section of Gogo's website, at ir.gogoair.com. The earnings press release is also available on our website. After management's remarks, we'll host a Q&A session. And now it's my great pleasure to turn it over to Michael.

Thanks, Varvara. Good morning, everyone. Before I get into the details for the quarter and the progress we are making against our plan, I want to take a step back and provide some perspective on our business. Our rapid deployment of 2Ku is transformative. We will have much more to say about this at our upcoming Investor Day on November 17, but at a high level, the increase in bandwidth transforms our offerings in several ways. First, it gives us the ability to vastly improve the customer experience, including, very importantly, across our North American networks. Second, it allows us to accelerate take rate growth through Gogo, airline and third-party programs. And third, it positions us to begin offering new products and services. We are confident this transformation will accelerate revenue growth and drive long-term profitability. In fact, we are already seeing evidence of this in the form of increased take rates and improved ARPA on our satellite systems. We recognize that having great technology is not enough. We need to get that technology on planes quickly. To that end, we have made significant progress in improving our installation times and reducing installation costs. There is much work to do, but we have the right plan in place and our progress to date is substantial. With that, let me turn to our results. Revenue grew 17%, to $173 million, fueled by strong service revenue growth in ROW and BA segments. Adjusted EBITDA for the quarter of $13 million was up $3 million sequentially and was impacted by $4.5 million in charges that Barry will discuss. We expect a significant ramp in adjusted EBITDA in Q4. Also, we added $110 million of cash to the balance sheet in anticipation of more airline wins as demand for 2Ku remains high. In Commercial Aviation, we continue to expand our market-leading position with key wins for 2Ku in North America and Rest of World. In North America, we announced a deal to install 2Ku on all of Alaska's aircraft, including some win-backs on the Virgin American fleet. More recently, in ROW, we announced a deal with LATAM, the largest airline in South America. We now have more than 1,900 aircraft awards for 2Ku, making it the most rapidly adopted broadband satellite in-flight connectivity solution in aviation history. 2Ku has been selected by 4 of the world's largest airlines and 14, over all. We are rapidly building critical mass around the globe. Operationally, we continue to set the industry standard in terms of speed for broadband satellite installations. We installed 2Ku on 76 aircraft in October, bringing year-to-date installs to over 320. We are on target to hit our installation guidance of 450 to 550 2Ku installs for the year. I am also pleased to report that our bandwidth per aircraft grew 47% in CA and 20% in BA year-over-year in the quarter, even with only 9% of our CA-NA flights equipped with satellite connectivity. Bandwidth growth will accelerate in 2018. By this time next year, not only will approximately a third of our CA-NA fleet have satellite connectivity, but our 100-megabit per second next-gen ATG network will be commercially available for the regional jet and BA market. This rapid migration to higher bandwidth for all aircraft types will result in a significant improvement in customer experience, a crucial factor in driving take rates in ARPA. We are already seeing this happen. Take rates increased to 7.5% in CA-NA, up from 6.5% in Q3 of last year. Drivers of this take rate increase include the recent fleet-wide launch of Messaging Pass with Delta, an expanded relationship with T-Mobile and continued adoption of Alaska's messaging product. In September, we announced Gogo Vision Touch. Gogo Vision Touch brings internet-enabled entertainment to the aircraft by leveraging our existing connectivity system to deliver content wirelessly to seatback screens. It is a cost effective, lighter-weight option than what is available in the marketplace today. We are launching this product with Delta and have seen a very high level of interest from other airlines. We see Gogo Vision Touch redefining and disrupting IFE. To sum it up, the transformation brought by high bandwidth is real and incredibly exciting. It's still early days, but we are already seeing satellite-connected aircraft in North America generate far more ARPA than ATG aircraft. Annualized ARPA for our satellite-connected aircraft across both NA and ROW is about $220,000. Turning to Business Aviation, we had another outstanding quarter. Service revenue grew another 30%, to reach $43 million for the quarter. We also launched a new connectivity platform through BA called AVANCE. From now on, you'll hear us talk about AVANCE L3 and L5 connectivity systems that were previously called Gogo Biz and Gogo Biz 4G. AVANCE is a unified platform for in-flight experiences from streaming video and audio to on-demand movies, personal smartphone use, real-time data apps for cockpit use and remote diagnostics and support. Before I turn it over to John Wade, I want to leave you with a few takeaways. Our dedicated aviation networks and aircraft connectivity systems are delivering industry-leading and rapidly improving performance in terms of bandwidth coverage and availability in both BA and CA. This performance drives aircraft wins at an industry-leading pace. Operationally, we are and will continue setting records for getting more aircraft online. Finally, we are engaging more passengers in CA and more customers in BA with more bandwidth and cutting-edge products and services. This combination of factors is already driving increased take rates and positioning us to deliver improved profitability and long-term shareholder value. Now I'd like to turn it over to our Chief Operating Officer, John Wade.

Thanks, Michael. During the quarter, we continued to ramp our 2Ku install pace in Commercial Aviation. In October, we installed 76 aircraft, up from 50 in September, bringing total 2017 installed aircraft to 322. Our installation time has dropped to about 36 hours. We are well positioned to hit our install guidance and ramp from here. On the OEM side, we have now delivered 3 A350 aircraft to Delta with our 2Ku system already installed. We have continued to advance our Airbus OEM programs and expect to announce our first Commercial Aviation order through Airbus some time early in 2018. For Boeing, we expect the first B787 installation to take place around year-end. We will have our first Bombardier production installation delivery in May 2018 for Delta. We are also progressing production installations for other Airbus and Boeing aircraft types and are targeting delivery in 2019. I would also like to add more color to the Gogo Vision Touch product Michael mentioned earlier. Gogo Vision Touch evolved out of an airline need. Delta wanted a seatback solution without the weight, cost and complexity of legacy solutions. For Gogo, this amounted to a software-centric add-on to our existing connectivity service. It allows us to offer Delta a product which will engage even more passengers. This is how it works. Gogo Vision Touch puts a commercial-grade tablet supplied by Delta in the seatbacks of Delta's new Bombardier C-series aircraft. The tablet wirelessly connects to the Gogo in-cabin network and can be pulled in and out without the complexity of a wired system. The simplicity and potential cost savings are why we are already seeing significant interest from other carriers. Turning now to Business Aviation, we have begun to ship our AVANCE L5 connectivity system, which we formerly referred to as Gogo Biz 4G. The market reception has been strong, and we have now sold 86 units. The AVANCE L5 system was recently selected as a factory option on Embraer Legacy 450 and Legacy 500 business jets, as well as several key Dassault Aviation Falcon aircraft. Recently, we also announced the addition of a global Ku band solution that will utilize a small tail-mounted antenna for larger business aircraft and leverage our existing global Ku satellite network. We believe this product will give us access to a new market in excess of 2,000 aircraft. This solution, combined with our 2Ku solution, gives Business Aviation 2 satellite connectivity products to serve the entire global business aviation market. Before I turn it over to Barry, I would like to highlight a couple of points. Our operational execution is industry leading, as evidenced by our aftermarket track record. We have made considerable progress with OEM delivery aircraft and will continue to accelerate our efforts. And our Business Aviation position and competitive strengths are formidable advantages. In sum, we see significant growth opportunities ahead, and we are already driving the execution necessary to capitalize on those opportunities to deliver value for our customers and shareholders. With that, I'd like to turn it over to Barry.

Thank you, John, and good morning, everyone. We delivered another strong quarter of revenue growth. Service revenue was up 19%, to $153 million, and total revenue of $173 million was up 17% from the prior year. Service revenue grew across all 3 segments and was particularly strong in CA-ROW and BA, which grew 117% and 30%, respectively. Adjusted EBITDA of $13 million was up $3 million sequentially, but lower than the prior year by $2 million. As expected, expenses related to airline launches and OEM programs did decline sequentially. In addition, this quarter we incurred $4.5 million in charges related to write-downs of legacy product lines and retirement of Gogo test aircraft. Excluding these charges, adjusted EBITDA was $17.5 million. The charges associated with legacy product lines included previously capitalized software and related inventory expenses totaling about $3 million. We also decided to retire our test fleet: Jimmy Ray, our CA test aircraft, and Challenger 600, our BA test aircraft. With hundreds of 2Ku aircraft now flying, we don't need to own a plane to test this technology. For next-gen ATG testing, it will be far more cost effective to charter aircraft as needed. The $4.5 million in charges impacted CA-North America and BA segment profit by $2.4 million and $2.1 million, respectively. We also closed an add-on bond financing that raised $110 million in net proceeds during the quarter. With the market success of 2Ku and increased visibility into more customer wins, we felt it was prudent to cushion our balance sheet to fund the success-based capital investment in aircraft installations. Let me now turn to the performance of our business segments. For Commercial Aviation-North America, service revenue increased 7% over the prior year, to $94 million. This was driven by an increased number of aircraft online, which now stand at more than 2,800. We installed 114 aircraft in the quarter, which included 70 2Ku retrofits from ATG-4 aircraft. Annualized ARPA was flat year-over-year, at $133,000. With increased bandwidth now beginning to arrive through our global satellite network, we believe it's important to track ARPA by these 2 primary network delivery systems: satellite and air-to-ground. For the quarter, satellite ARPA was $220,000 annualized, as compared to ATG ARPA of $125,000 annualized. We see higher levels of ARPA on satellite-connected commercial aircraft due to the increased bandwidth provided through this network, and it is installed on larger planes. Annualized CA-NA ARPA declined sequentially, largely due to seasonality and, to a lesser extent, due to lower average revenue per session and travel slowdowns associated with the recent hurricanes. As we increase the percent of satellite-equipped aircraft in North America, improve the customer experience and increase connectivity options available to our airline passengers, we expect ARPA to expand further over time, driven by improved take rates. In CA-NA, our take rate increased to 7.5%, up from 6.5% last year, led by increased passenger engagement from airline-paid and third-party offerings. With increasing bandwidth, we are in a position to expand our service offerings through more partnerships like the messaging program we announced a few weeks ago with Delta. We expect this initiative to favorably affect take rates beginning in Q4. CA-North America segment profit of $16 million included the $2.4 million in charges I mentioned earlier. Excluding these charges, segment profit was $18.4 million, or 19% of revenue. In the quarter, we also recognized approximately $3 million in expenses related to our next-generation ATG development, which remains on track for commercial availability in 2018. Turning to CA-Rest of World, quarterly service revenue more than doubled for the third quarter in a row, to $15.7 million. This continued strong performance is a result of growing ARPA, which was up 30% to $226,000 annualized, and a 41% increase in equivalent aircraft on line. The take rate for satellite-connected ROW aircraft was 13.5% for the quarter, 80% higher than the 7.5% take rate for North America, which includes a blend of satellite and ATG aircraft. Over all, ARPA was flat compared to Q2, reflecting the expected dilution from newly launched airlines which represented 25% of aircraft online for Q3. We continue to expect strong double-digit growth from our existing fleet, although consolidated CA-ROW ARPA is likely to be lower in Q4 and during 2018 due to new aircraft additions throughout that period. Aircraft online increased to 352, up 96 versus the prior year. Following the LATAM announcement, our awarded CA-ROW backlog now stands at approximately 680, most of which will be installed by the end of 2019, approximately tripling the number of ROW aircraft flying today. Rest of World segment loss for the quarter increased to $24 million, from $20 million in the prior year. However, segment loss improved $7 million sequentially. Approximately half of this improvement came from lower spending on OEM programs, with the other half resulting from increased utilization of our global satellite network. As we add more aircraft, we expect to see continuing improvements in capacity utilization, paving the way to accelerating EBITDA performance. Let's now turn to our Business Aviation segment. BA continues to outperform. Service revenue was up 30%, to a record $43 million, with 15% growth in ATG aircraft online, to over 4,500 planes, and a 13% increase in ATG service ARPU, to nearly $2,900 per month. Total revenue was $61 million for the quarter, up 24%. BA equipment revenue was $17 million for the quarter, up 11% from the prior year, driven by an increase in shipments of L5, formerly Gogo Biz 4G. We sold 66 L5s in Q3, representing over 30% of our ATG shipments. As we look farther down the road, we are enthusiastic about the long-term growth prospects for Business Aviation. We see opportunity to expand the business along 2 dimensions. First is further penetration of the lower end of the market in light jets and turbo props, as we hone our product offerings and pricing to meet customer needs. Second is the opportunity to serve business aircraft flying internationally or domiciled outside the US through our recently announced satellite-based solution. With the launch of our AVANCE platform, as well as the introduction of satellite-based connectivity systems for our BA customers, our sales momentum and long-term prospects are strong. BA segment profit for the quarter was $21 million, reflecting a 35% segment profit margin, which was down from 42% in the prior year. There were 2 main factors affecting segment profit margin. First, as I mentioned, we wrote off approximately $2 million in expenses related to our legacy One Phone product line, a cord-connected phone no longer desired by business jet customers. Excluding this charge, BA segment profit margin would have been 39%. Secondly, we increased ED&D and sales and marketing investments to further drive market penetration. We expect this level of spending to continue in Q4, but we expect segment profit margin to return to 40-plus-percent in 2018. Turning to cash and CapEx, we ended Q3 with $411 million of cash on hand. Excluding interest payments on debt, which occur during the first and third quarters, our free cash flow, which we define as cash from operating activity less capital expenditures, was a negative $35 million, the lowest level in 2017. Consolidated cash CapEx of $53 million was $17 million higher than the prior year, reflecting our investment in bringing significant numbers of 2Ku aircraft online. This increase in CapEx is largely related to our success-based co-investment with our airline partners in 2Ku airborne equipment. As we approach the end of 2017, let me comment on our guidance for the year. Based on the continued strength of revenue in CA-ROW and our Business Aviation segments, we expect revenue to come in at the high end of the $670 million to $695 million range we provided. We expect adjusted EBITDA to approximately double from the first half of 2017 to the second half of the year, putting us at the low end of the $60 million to $75 million range, excluding the $4.5 million in charges we incurred this quarter. We expect adjusted EBITDA to be significantly higher in 2018 than in 2017. We expect to come in at the low end of our 2017 cash CapEx guidance of $230 million to $260 million, of which approximately 70% is related to success-based airborne equipment CapEx. Operator, we're now ready for our first question.

[Operator Instructions] Our first question comes from the line of Robert Gutman, of Guggenheim.

In terms of the ROW [ ARPA ] dilution that you're expecting, going forward, from the new installations, what would be the magnitude there of the impact? And if you could provide us a little more detail on this tail-mounted, high-throughput satellite antenna going on the business jets? How does that compare to the 2Ku antenna and how do you provide that level of throughput on a smaller antenna tail mounted?

Barry will take the first question, and I'll take the second question.

I think it is worth reiterating that the ARPA for ROW is very robust, at $226,000. And in this quarter is about a quarter of the ROW aircraft, including Virgin Atlantic and GOL. The dilution that we expect, going forward, is totally related to the new aircraft airlines coming online. So it will have a relatively meaningful impact, over all, but I think it's important to look at the fact that we expect the ARPA for the existing aircraft to grow in a continuing double-digit rate throughout that period.

And the reason it will grow next year is we're going to accelerate the rate of aircraft additions in Rest of World next year. It will be substantially higher than it was this year. And the second, on the tail-mount antenna, we are getting once again to do in BA what we did in the past with the ATG network. We built the ATG network for the commercial business and then had a very profitable opportunity to sell it into BA. And we have now built a global high-throughput satellite network for CA, and we've also built a very sophisticate modem that's going into the marketplace right now. And so bringing that high-throughput satellite network and that modem to the BA network, combined with a more standard tail-mount antenna, produces dramatically higher performance than anything that has yet been available in the business aviation market. We expect peak speeds to consistently be at the 25-megabit per second level, or better, and the business aviation market has not seen that kind of performance, historically. The quality of our coverage globally is tremendously strong since we already serve 98%-plus of the global commercial aviation flight hours. So this is going to be a powerful, powerful offering in the business aviation market.

Our next question comes from the line of Philip Cusick, of J.P. Morgan.

You mentioned that the new $110 million is for anticipated wins, and we've seen fewer EMEA and APAC deals in the last year than I would have expected, given your comments at the Analyst Day last year. Is this the area of new potential? Or do you think there are more widespread deals?

That is an area of potential, APAC and the Middle East, but globally there's opportunities everywhere, including North America. So we -- 2Ku is being extraordinarily well received as the leading solution in the marketplace, and our ability to install planes quickly is being recognized. So we think we'll see -- the funnel will be global. The reason we focused on Asia is we built a global network and we want a lot of aircraft flying on it everywhere because it helps the utilization of the network. So there are a lot of planes in Asia, and it's beneficial to us to get planes in that region and I expect us still to be successful in signing in that region.

If I can follow up, can you talk about where your satellite capacity utilization is today? Where are we now versus 2016 and how does that track over the next year?

We saw and explained one of the reasons for the improvement in Rest of World's EBITDA sequentially from second quarter to third quarter was the increased utilization of the network. We didn't spend much more on satellite capacity, but we had some more incremental revenue. That trend will continue. The utilization is going to be best initially in North America and in Japan, where we have good regional fleets under the satellites. It will improve materially in the rest of the world as we install substantially more rest of the world aircraft next year than this year. So there will be significant improvement during the course of next year.

Our next question comes from the line of Simon Flannery, of Morgan Stanley.

This is Landon Park, on for Simon. Just talking about the CA-NA ARPA, are you able to provide a more specific apples-to-apples revenue uplift for your converted 2Ku planes?

So we'd make a couple of comments. We improved our reporting by breaking out ARPA between ATG and satellite in North America. Those are our 2 fundamental networks, and we think that's the right way to report. And you see that the ARPA is $220,000 for satellite and $125,000 annualized for ATG. That's a function of 2 things. One is larger aircraft get the satellite solution. That just makes sense that planes that fly over water and have more people get the satellite. The planes that stay over land and fewer people get ATG. But there's also the benefit of the additional bandwidth that is helping the ARPA on the satellite aircraft. There will be mix issues as we do these conversions, which ones we put 2Ku on and when. So that number will have that factor in it for the next year, or so. But more bandwidth translates into more revenue. We see it in the rest of the world. We see in the satellite aircraft in North America.

And Landon, I would just add one point to what Michael said, which is that when you see the economics as we've now broken them out, with ARPA of $220,000 for satellite, the unit economics are very attractive at these levels. And if you think about that across each of the segments, for the BA segment, for the regional jets, for example, that are largely on ATG, and for 2Ku, that the unit economics are very attractive even at these ARPA levels, which is an important foundational element of our overall company economics.

But are you able to provide a breakout between those 2 drivers or a specific uplift that 2Ku is providing on new installs or conversions?

We are not providing that data. There is no good ongoing way to report that. We're reporting it in a sensible way by the 2 networks. We will continue to make technological innovations that will drive bandwidth. On the ATG side, we have ATG and ATG-4. To break that out, we don't. We have Ku and we have 2Ku. We are introducing the [ Glot ] modem, has a major impact. We're introducing high-throughput satellites that has a major impact. This is -- we improved our reporting, and we think we improved it in the way that is going to be most helpful to investors over the long term, because there will be rapid technological innovation that drives bandwidth improvements.

Our next question comes from the line of John Hodulik, of UBS.

It's Lisa, for John. Just a couple of questions. As far as the expanded T-Mobile relationship, can you tell us a little bit more about what that involves, how it's different than the prior relationship, and whether you are still able to potentially strike partnerships with other wireless carriers? And then as far as the Gogo Vision Touch with Delta, does that have incremental impact on your revenues from Delta or was it already contemplated in your last renegotiation of your relationship with them?

So the expansion of the relationship is across several fronts. It's time, technology and service offerings. And we've, over time, increased the service offerings and who gets what among their customer base, and they now have a more premium solution where their best customers get a full flight solution. We're working on ways to make it easier to access the service, just a simpler sign-on, and we've expanded in terms of time, the duration of the contract. Now on the Gogo Touch, the first deployment of that happens in the spring of next year as the first C-series is delivered. And so it will be deployed. But, yes, there is a commercial relationship with some economics for Gogo that will begin to appear next year. It won't be a large number next year, simply due to the number of aircraft that have it. But our ability to revolutionize in-flight entertainment, make it lower cost per seat, get more modern screens in there, make it more upgradable over time and make it more uniform from one aircraft type to the next, which airlines like, is going to be a pretty big deal for the in-flight entertainment business, and it will result in revenue for Gogo over the long term.

Okay. Thanks. Just one more, if I could. I think there's been some press coverage of Netflix working with a number of airlines that are your partners and using new compression algorithms to help pass through [indiscernible] air access. Netflix has a better experience. And just wondering your involvement in that partnership, and potentially do we see revenue uplift there as that increases [ bandwidth ]?

So we love it when over-the-top content providers make it more efficient to stream their services. Netflix has always been a leader in adaptive compression techniques to deliver a good-quality video in a variable-bandwidth environment, and they're taking it to the next level here and we think that's great. With 2Ku, streaming is available for everybody on the plane, and this only makes it better, just great technological advancement on their part.

Our next question comes from the line of Louie DiPalma, of William Blair.

Thanks for the satellite ARPA disclosure, but on the flip side to ARPA I have a question regarding satellite bandwidth pricing and, specifically, Gogo's average satellite bandwidth cost per aircraft. Michael, in your prepared remarks you noted that bandwidth per aircraft in North America has increased by 46%. And over the past 2 years for Gogo, bandwidth cost per aircraft has fallen from around $230,000 to around $160,000, as the satellite aircraft online has more or less quadrupled. So in the context of falling pricing per gigahertz and the increased efficiency or the increased utilization that Phil Cusick just referenced, I was wondering if you could provide a long-term target range for satellite bandwidth cost per aircraft with the context of most of your 3,000 commercial fleet going to the satellite network.

It's a good question, and that ultimately Gogo's mission/vision is to get the highest quality bandwidth to the most aircraft everywhere at the lowest possible cost. And our open architecture to capture all the innovation that's happening, our open architecture to buy the bandwidth where we need it when we need it, not being trapped into a [indiscernible] is all the way we drive our costs down over time. We are seeing rapid declines. We forecast rapid declines. And then we repeatedly revise our forecasts to lower and lower costs per bit over time. Those trends are powerful. And some of the things that are happening to make that in very real terms what's driving it is the 2Ku acceleration. So we're getting those on planes at a more rapid rate, and there's more efficient costs per bit. We also -- in January, SES-15 is scheduled to come online. That's our high-throughput satellite over North America. It will address a very high percentage of our megabytes. That comes on in January, goes live. We'll optimize it during the first quarter, but that's going to change. We're deploying the [ Glot ] modem, our new modem, and at a pretty rapid rate. It's on about 50 planes today. That's a much more efficient modem, particularly on the reverse link. We're driving down our costs there. I would also say one of the ways to drive down costs is, quite simply, scale. The more planes you have flying, the better you can utilize your capacity, and that's happening at a rapid rate. So this is what we're all about, is getting more bandwidth at lower cost to aircraft, and all the trends are going in the right direction.

Great. And my second question concerns CapEx. Barry, can you provide more color as to what is in the non-success-based CapEx bucket? And is any of the non-success-based CapEx one-time in nature?

The primary components of the non-success-based CapEx have to do with the network cost, over all, for ATG and satellite network upgrades. So those are the primary drivers. We also invest in software. So those are really the primary elements of that. In terms of how we see that flowing over time, as we've mentioned, we expect to be at the low end of the cash CapEx guidance this year, and it's mostly due to those non-airborne equipment-related expenses, but we're also seeing some improvement in the co-investment per aircraft, as well.

Okay. And you previously earmarked approximately $50 million in CapEx for the development of the next-generation ATG network. How much of that $50 million is expected to be incurred in 2017 versus 2018?

The lion's share of that will be incurred in 2018. As you saw from the release about next-gen ATG, we are now testing that in the embedded test sites during this year. It will be rolled out pretty extensively in 2018 to the point that it will be commercially available there. So most of the expenditure will be in 2018, with some carryover into 2019.

Okay. And lastly, when you do the rough math regarding success-based CapEx and the number of2Ku installations for this year, it comes out to roughly $340,000 per aircraft. And I just wanted to know if that amount of co-investment is still expected to decline in 2018 and 2019? And what are the factors behind the decline in the CapEx, if that is actually happening?

First, Louie, your number is very high. And I think probably the major reason for that is that the CapEx this year includes the equipment that's actually installed on the aircraft as well as the inventory build ahead of the installations for next year. So that's a meaningful percentage of that. So you're off pretty substantially, on the high end. In terms of what brings that down is that we our co-investment is declining. So there was more investment with some of the early airlines that we expect that to be lower, going forward. We also see the ability to lower installation cost and the cost of the shipped sets over time. So it's a combination of those factors that has it starting substantially lower than that number and then continuing to reduce over the next several years.

Our next question comes from the line of Paul Penney, of Northland Capital.

Could we get a refresh on where you expect take rates and ARPS to trend into next year? And specifically, how much drag on price per session will come from the rollout of texting on Delta?

So a couple of comments on that, Paul. We introduced texting at Delta right at the start of the fourth quarter, and you will see a significant increase in take rates in the fourth quarter as a result of that. You will also see significant reduction in average revenue per session. If we've done this right, it shouldn't have a major impact on normal revenue trends. We'll have to see exactly how it rolls out, but it was fairly revenue-neutral in the short run. We have switched strategy from mining the current customer base, that business traveler, to maximize yield. That's basically been our strategy for the last few years, given the limited capacity. We are now trying to increase engagement. Given the 2Ku rollout, we're only partially through it, the only airline-wide solutions we can put in place today, because some planes are ATG and some planes are 2Ku, are the low-bandwidth solutions, like messaging. As the 2Ku rollout progresses, we will have more flexibility to introduce higher ARPS offerings on a fleet-wide basis. So we, quite simply, we're switching from maximizing yield on the business traveler segment to a strategy that engages the entire aircraft, and I'm very pleased with how that's going.

Our next question comes from the line of Andrew Spinola, of Wells Fargo.

I was wondering if you could just add a little additional color to the ARPA in North America this quarter. You mentioned before the hurricane, and I was wondering how big of an impact that was. And also just any of the sort of other puts and takes as to sort of why that number was relatively flat this quarter.

So, first, Andrew, there was -- seasonality is definitely a factor, and actually last year it was down, ARPA and revenue, from the second to the third quarter. That's been true most years I've been here, but greater magnitude this year. About $1 million, or so, of the greater magnitude is explained by the hurricanes. The rest of it is the reason I just explained. Instead of looking for price increases where we could take them, which is what we've been doing in prior years, this year we focused on increasing engagement and introduced the messaging passes and, in fact, if anything, took a few price decreases rather than increases against the business traveler base. So that explains it. The trend, going forward: lots more bandwidth, lots more engagement, and when you add a lot more people engaged in paying you, there's going to be plenty of opportunity to create revenue and shareholder return out of that.

Understood. And then on the cost of services, wondering if you had any cost from the AMC-4 satellite that you procured in Q3.

I will -- I can't answer that question off the top of my head for you. We'll have to get back to you on that one.

Fair enough. And then just last one, just wanted to understand, the third quarter looked a little unusual in BA, and you pointed out that it's going to bounce back to over 40% next year. And I just wanted to drill down on, specifically, the equipment margin in that segment. It looked like it was artificially low this quarter. I just wonder if you can confirm that and just help us understand what that margin is going to look like, going forward, as you go after the turbo props and the international jets with the satellite antennas.

So the equipment margin was affected by the One Phone write-offs that we had this quarter that we talked about. So that had a meaningful impact. As we look at it, going forward, the equipment margin we expect to kind of be in the general ZIP code where it has been. But as we are also introducing the lower-end products to go after the turbo prop and light jet markets, part of that strategy is to be able to offer lower upfront cost to those aircraft, with very attractive ongoing revenue streams. So that will unfold over the next couple of years as we move into that market. But very attractive unit economics in that market, as well.

So just to confirm, the entire $2.1 million flowed through cost of equipment in the quarter?

Yes, that's correct.

And our next question comes from the line of Brian Dennis, of Cowen.

A couple of things. If I'm looking at this correctly, it looks like the year-over-year increase in take rates decelerated a little bit in the quarter versus 2Q and 1Q. Are there any one-time items driving this? Or how would you explain that?

There are no material one-time, other than the seasonality. Take rates, yes -- I don't think there's a big explanation. You will start seeing a material acceleration of take rates starting next quarter, and that will continue, going forward. We are working to engage the whole plane, and simply with the Delta Messaging Pass hitting a huge number of aircraft -- by the way, that's available on every single one of their planes, a fleet of well over 1,100 aircraft globally. So that will make a big difference.

Got it. Thanks. Just one more. The stock has sold off since you guys did the tack-on. We thought it was a positive move, but can you elaborate a little bit on the rationale behind taking down the extra $100 million?

To state the obvious, it's been more than frustrating to see what's happened there when there's been no negative news from the company during that period of time. Let me talk about the reason for the $100 million and then some observations about that. The reason that we took on the additional $100 million was it was an opportunity to do that in response to the strength of the sales pipeline which is trending toward the high end of our expectations and to provide some additional cushion to our balance sheet. So we were able to do that at a cost effective way, as the yield to worst was less than 8%. I think what's happened is that we did that. We announced that additional financing before we were able to announce the additional airline wins. So we had obviously announced Alaska prior to that. Since then, we have announced LATAM. So we knew that the pipeline was strong, as Michael had talked about. So I think that was a big part, I think, of what had happened. And I think second observation is that the transformative nature of 2Ku bandwidth is really not obvious yet in the market. It is coming on now, and we're going to see the benefit of that for the reasons that we've pointed out on this call. And I guess the final point I would make is that this really is an execution story, to a large degree, at this point that has to do with getting more planes online. So the installations that you saw are materializing as we had set the expectation they would and hoped they would. And even if you look at the October rate of 76, that's over 900 per year in terms of installation. So given that you can't do that at every month, with seasonal considerations, that positions us well to provide the 750 in installation capacity on an annualized basis. So at this stage, it really is about executing on getting those 2Ku planes in the air and getting the benefits from them. That's not yet been revealed -- it's not yet been demonstrated, obviously. That's all happening in the coming quarters.

Our next question comes from the line of Andrew DeGasperi, of Macquarie.

I joined the call late, so I'm not sure if you already answered this, but how should we be thinking about in terms of aircraft count for 2018, when you include the American Airlines potential de-installs? And then, secondly, at your Analyst Day last year, you mentioned that you saw demand in terms of throughput of satellite bandwidth at around 70 gigabit per second. Just curious to know, given the new wins at Alaska and others, as well as the 2Ku uptake, do you feel that number has changed or do you think that's potentially a little stale at this point?

On aircraft count, we have no way of knowing for sure the pace of deinstallations that could occur at American. But at this stage, we would forecast a modest decrease in aircraft count in North America due to those deinstallations next year, in 2018. But we expect to far more than offset that in Rest of the World, a significant acceleration there. And so Gogo's overall aircraft count will increase next year at an appreciable rate. So we feel good about the net aircraft count. The bandwidth, we're still -- our aircraft counts haven't really changed materially since we were looking at them a year ago. What has changed more is our optimism, and reality, that we can keep driving down the cost of bandwidth and [indiscernible] the availability of bandwidth. So our bandwidth picture is substantially better than it was a year ago, and I think it's substantially better than we believe the competitive options are in the marketplace. Gogo will have the most bandwidth, the highest-quality bandwidth, to the most planes at the best economics.

And we have a follow-up question from the line of Louie DiPalma, of William Blair.

Given all the controversy over the ATG network in Europe, do you believe that validates how big of a competitive advantage your ATG network here is in the United States and, particularly, how that relates to your ability to serve the commercial regional jets?

So the ATG network is a huge competitive advantage for Gogo, and it's a much bigger advantage in North America. The geography is just inherently more friendly. It's more contiguous, uniform. More aircraft stay over land in North America than they do about anywhere else. It also happens to be true that two-thirds of the business jets in the world are in North America and a very high percentage of the regional jets in the world are over North America. In fact, today we serve over 1,000 regional jets, and that's a really good business for us. And they never fly over water, and the lower cost overnight install of the ATG is perfect for them. So yes, ATG has a huge advantage for us, and it's much more applicable in North America than it is any place else in the world.

And do you think that most or all of the regional jets that you currently serve with your current ATG network will be interested in upgrading to your next-gen network? Or is that mostly going to be a Business Aviation product?

I think they should be and I think they will be. Our current ATG network will get better as we offload to 2Ku, but it's still a 10-megabit per second peak speed network. With next-gen ATG, we'll take it to 100-megabit per second peak speed network. And it also has the benefit of further offloading ATG with next-gen ATG. So this is all good news. It's a low-cost overnight upgrade. So I'd be surprised if the RJ fleet in North America doesn't upgrade to next-gen. I do think it is applicable to the larger business jets, and I think the biggest competition to next-gen ATG on those business jets will be our global Ku solution. Right now, most big global planes invest in a global solution and then add our air-to-ground when they're in the US because it's higher performing than the global solutions. But our new Ku solution for business jets is going to be pretty high performing. So we'll have to see whether those jets vote for both or are satisfied just with the global solution.

And I'm showing no further questions in the queue at this time. I'd like to turn the call back to Michael Small, CEO, for closing remarks.

Thank you, Operator. So I'd like to close, as I started, by saying that the bandwidth is arriving and it will be transformational. I view this, and we are focusing as a company, as this is an executional story: we've got to get those planes installed with the solutions we have. And that will result in what we've all been waiting for, is EBITDA to start growing, and I expect that to happen even in the fourth quarter and into next year and beyond. So thank you all.

Ladies and gentlemen, this does conclude the conference. You may disconnect. Everyone, have a wonderful day.