Accuray Inc (ARAY) 2018 Q3 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the Accuray Incorporated Third Quarter 2018 Earnings Conference Call. (Operator Instructions) And as a reminder, this conference is being recorded.

I would now like to turn the conference over to Mr. Doug Sherk of the EVC Group. You may begin.

Thank you, Amanda. Good afternoon, everyone. Welcome to Accuray's conference call to review financial results for the third quarter of fiscal 2018, which ended March 31, 2018, and recent corporate developments.

Joining us today are Josh Levine, Accuray's President and Chief Executive Officer; and Kevin Waters, Accuray's Senior Vice President and Chief Financial Officer.

Before we begin, I'd like to remind you that our call today includes forward-looking statements that involve risks and uncertainties including statements regarding our future business plans and strategies. There are a number of factors that could cause actual results to differ materially from our expectations, including, but not limited to, risks associated with the adoption of the CyberKnife, TomoTherapy and Radixact systems; commercial execution; future order growth; future revenue growth; and macroeconomic factors outside of the company's control. These and other risks are more fully described in the press release we issued after the market closed this afternoon as well as in our filings with the Securities and Exchange Commission. The forward-looking statements on this call are based on information available to us as of today's date, and we assume no obligation to update any forward-looking statements.

Two housekeeping items: first, (Operator Instructions); second, all references we make to a specific quarter in the prepared remarks are to our fiscal year quarters. For example, statements regarding our third quarter refer to our fiscal third quarter ended March 31, 2018.

With that out of the way, I'd like to turn the call over to Accuray's President and Chief Executive Officer, Josh Levine.

Thank you, Doug. Good afternoon, everyone, and thank you for joining us on today's call.

Accuray's third quarter results reflect the market momentum we've been building throughout the fiscal year. We generated $99.8 million in revenue during the quarter with Radixact and CyberKnife System sales as well as software upgrades, the main sources of year-over-year revenue growth. Year-to-date, total revenue was up 7%, while product revenue was up 9% compared to prior year. As a result of this performance and our outlook for the fourth quarter, today, we are raising the bottom end of our full year revenue guidance to reflect the progress we're making and more timely revenue conversion of our backlog. Kevin will provide more detail on both backlog growth and our revised revenue outlook in his prepared remarks.

Gross orders for the quarter were $74.9 million, which puts us slightly behind our internal targets through the first 9 months of the year. However, due to the number of high-potential multisystem order opportunities we're currently pursuing, we're still targeting approximately 5% order growth for the full fiscal year. In addition, we ended the quarter with backlog of $468 million, which is 4% higher than the year-ago level. Our third quarter gross orders were primarily driven by our TomoTherapy family of products, which represented 75% of the unit volume during the quarter. Radixact systems represented over 80% of all TomoTherapy order volume, which is up from 70% during the second quarter. To help illustrate the growing market interest in our Radixact System, third quarter orders for this product were the highest they've been for any quarterly period since our full commercial launch a year ago.

CyberKnife System orders in the third quarter were a bit behind the prior year due to a strong CyberKnife System order mix in the third quarter of fiscal 2017. Orders for system replacements to our existing customers were in line with our recent historical order contribution percentage of 15% to 20% of total global order mix. This means 80% to 85% of the remaining orders are coming from either greenfield opportunities or competitive wins. On a global basis, Accuray continues to be an effective competitor, and we are still capturing more vaults than we are giving up.

On a regional basis, our European and Japan regions continue to perform well and were strong contributors to the quarter's gross order activity. In our Americas region, we are beginning to see the benefits of our incremental investments in the U.S. sales and marketing organization that were made earlier in fiscal 2018. This week, we will be announcing that we won a large multisystem, multihospital order from Mercy in their St. Louis and Springfield, Missouri, area facilities. This order is the largest multisystem order in the U.S. in the past 5 years and was a mix of competitive replacements and replacements of our earlier generation installed-base devices. Mercy is a significant win for our company and is a major factor behind reiterating our gross order growth for the full fiscal year today.

With that said, the Mercy order illustrates the potential impact on order timing that these large multisystem orders can have on our quarterly results. Because these opportunities involve a larger number of systems across individual sites within a targeted IDN or strategic account, they involve a broader range of participants in the final decision-making process and greater variability in order timing. As a result, completion of these multisystem deals can often move from one quarter to the next.

As mentioned previously, gross orders for our Radixact System set a high watermark this quarter. Radixact's improved performance characteristics related to treatment speed and overall throughput in both direct and helical delivery mode have allowed us to position Radixact as a highly efficient and virtual treatment platform, exceedingly capable of treating routine cases with great speed and efficiency, while allowing for great complex case mix treatment capability. Radixact offers users a highly abbreviated 5-step workflow and has a small dimensional footprint that can easily fit into the smallest of radiation vaults, including those that may have designed for legacy cobalt systems. Across active Radixact customer locations currently treating patients, treatment volume is averaging in the high-30 to low-40 patients per day, depending on case mix.

Further, with the enhanced performance characteristics of our Radixact System, we continued to improve our position as a viable competitor in the single and dual vault space, which expands our overall market opportunity by an additional 2,000-plus vaults in the U.S. alone.

Turning to our growth opportunities in China, the recent announcement from China's National Health Commission on April 9, provided the definitions for and classification of medical equipment imported to China. This release provides the clarity we've been waiting for regarding our Onrad system and the base system TomoTherapy H product, which have both been categorized as class B, meaning these devices will not be subject to the class A quota. Our CyberKnife and Radixact systems, as well as our TomoTherapy HD and HDA Series products, remain in the class A category.

The next step in this process will be the confirmation from the China National Health Commission regarding the number of class A licenses that will be issued, which will enable us to begin the conversion of our existing backlog of class A orders to revenue. While we await the class A license announcement, we're moving forward with the selling process at the provincial level with our Onrad and TomoH devices, which we now know are definitively class B products. The sales activity will take several additional months to secure orders due to the class B provincial quota process, but thanks to the efforts of our commercial team and distribution partners in China, we believe, we're well positioned to drive order activity with our Onrad and TomoTherapy H systems in the value segment of the market early in fiscal 2019.

On another key element of our China growth strategy, we continue to make progress towards establishing a joint venture partnership in China during the quarter, and we believe we will be in a position to provide a more definitive update by the time we report our full fiscal year 2018 results.

As we highlighted on our previous calls, we made the decision as we began fiscal 2018 to incrementally invest in both R&D and our U.S. commercial infrastructure, which has resulted in higher operating expenses during the past several quarters, and that was the case as well in the third quarter. The increase in R&D spend will ensure accelerated development timing and speed to market for enhanced functionality and innovative upgrades for both CyberKnife and Radixact systems. The increased sales and marketing spend reflects the additional commercial infrastructure investments we've made during the first half of the fiscal year in our U.S. sales and marketing team. We believe that both the incremental R&D and U.S. infrastructure investments are essential in driving increased revenue growth going forward, but require that we must balance these investments with our goal of reaching a level of sustainable cash flow and profitability. These increases in our spending, together with lower-than-anticipated margins, are the primary factors that impacted our lowering of full year EBITDA guidance. Kevin will further expand upon our revised EBITDA range in his prepared remarks.

Lastly, I'd like to note the developments that came out of our participation in the annual ASTRO meeting, which was held in Barcelona this year. We had strong customer interest in our recent software upgrades that have been focused on system connectivity, imaging enhancement and faster planning capabilities as well as some of the software enhancements on our road map design to further enhance our CyberKnife treatment planning efficiency. In fact, at the meeting, we released an enhanced CT imaging capability for Radixact that our early customer sites have confirmed offer a significant improvement in soft tissue resolution. We've branded this image guidance software as CTrue iterative reconstruction, and it represents the first of several imaging improvements we expect to make on the Radixact platform. In addition, we continue to build the clinical evidence for radiation therapy procedures performed by Accuray devices.

At ASTRO, Dr. Donald Fuller from Genesis Healthcare in San Diego presented critical quality-of-life data from his 5-year study on prostate SBRT procedures performed with CyberKnife. The study showed the key areas of the patient's quality of life, including urinary, bowel and sexual health returning to normal within 1 year following treatment. This approach to treating prostate cancer in low and intermediate-risk patients only requires an average of 4 treatment visits and provides very high rates of local disease control and allows patients to regain their health without major interruptions in the quality of their life.

Now I'll turn the call over to Kevin to provide more details on the fiscal third quarter financial highlights and full fiscal year guidance metrics.

Thank you, Josh, and good afternoon, everyone.

As Josh highlighted, we had $74.9 million of gross orders in the quarter, with a substantial contribution from our Radixact System. On a net order basis, we've generated $40.9 million in our fiscal third quarter, which included adjustments of $26 million for age-outs and $8 million for cancellations and other adjustments. Age-outs were slightly higher than expected due to the timing of certain orders converting to revenue. While cancellations can fluctuate quarter-to-quarter, on a year-to-date basis we've had $10 million of cancellations, which is consistent with prior year and a smaller percentage of our revenue as compared with last year. On a year-to-date basis, orders represent a positive book-to-bill ratio of 1.12 and increases our ending backlog, which drive future revenues.

We anticipate Q4 net age-outs to be at a comparable level to our prior year Q4, which is approximately $20 million. We continue to work on converting these orders to revenue and age-in as we've seen happen in previous quarters.

Turning now to our income statement. Total revenue for the third quarter was $99.8 million, representing a 3% increase over prior year. Revenue growth in the third quarter was driven by more than 30% growth in Europe and high single-digit growth in both the Americas and Asia Pacific. This was offset by a decline in revenues in Japan due to a very tough comparable in the third quarter of prior year.

Product revenue for the quarter was $43.2 million, a decrease of 10% over the prior year period. Although total product revenue declined, we recorded growth in every region in the world outside of Japan. Japan product revenues decreased approximately $10 million over prior year. However, Japan is performing very near our internal plan on a full year basis. Product revenue was evenly mixed between our CyberKnife and TomoTherapy Systems in the third quarter. On the TomoTherapy platform, approximately 80% of the units to revenue were from Radixact.

Service revenue for the quarter was $56.6 million, an increase of 15% over the prior year and a sequential increase of 6%. The growth rate in service revenue was well above both our historical averages and our expectations and can be attributed to increased upgrades purchased on service contracts, the timing of which could vary from quarter to quarter. These upgrades were primarily from our recent software releases around both treatment planning and connectivity. While we are pleased about the increase in service revenue growth and related upgrade sales, we believe that going forward, service revenue should grow at or around our installed-base growth rate.

On a year-to-date basis, total revenue grew 7% compared to the first 3 quarters of the prior year. The year-to-date revenue growth was driven by product revenue growth of 9% and service revenue growth of 6%.

Product revenue growth is being driven by a variety of factors. First, Radixact revenue has more than doubled its year-over-year contribution as initial orders of Radixact are converting to revenue at a faster rate than planned. Second, our CyberKnife product line has increased revenue year-over-year primarily in the U.S. and European regions. And lastly, our European region continues to show positive traction in backlog-to-revenue conversion and has outperformed our internal plan and is our fastest-growing region.

Overall gross margin for the third quarter was 36.3%, which is consistent with the prior year period gross margin of 36.4%. Product gross margin increased to 41.4% in the quarter compared to 38.4% in the third quarter of prior year. As discussed during our previous calls, prior financial periods included a quarterly $2 million intangible amortization charge from the TomoTherapy acquisition, which was fully amortized in Q4 of the prior year. Normalizing to exclude this charge, product margins are more in line with the prior period.

The volume decline in product revenue and associated product margin impact was partially offset by a larger percentage of our overall revenue attributable to CyberKnife systems and increased Radixact systems as compared to the prior year. The Radixact System continues to generate favorable pricing and margin as compared to our older generation TomoTherapy systems.

Service gross margins in the third quarter were 32.4% compared to 34.4% in the prior year. The primary driver of lower service margins was higher installed-base part consumption of approximately $3 million above our expectations.

We believe the part consumption in our third quarter was an isolated incident related to specific high-dollar parts, and we should see service margins return to more historical levels in the fourth quarter. However, the third quarter service margin shortfall will impact our full year EBITDA performance that I will touch on later. On a full year basis, our year-to-date service gross margins were 36.3% compared to 36.1% in the prior year.

Moving down the income statement. Operating expenses for the quarter were $40.1 million, an increase of 9% from $36.7 million in the year-ago period. The majority of the operating expense growth is coming from the R&D and sales and marketing functions, as Josh highlighted in his previous remarks.

Third quarter operating loss was $3.8 million compared to an operating loss of $1.3 million from a year ago. Adjusted EBITDA for the fiscal third quarter was $1.4 million compared to $7.1 million in the prior year.

Turning now to our balance sheet. We had approximately $73 million of cash, cash equivalents, short-term restricted cash and investments as of March 31. In February 2018, we used cash on our balance sheet to retire $40 million of convertible notes upon maturity. This strategy minimized shareholder dilution and avoided issuance of approximately 7 million shares. After retirement of the February notes, all debt has a maturity that is now greater than 4 years from today. This consists of the 2022 convertible notes of $85 million and the term loan and asset-backed revolver of up to $72 million. Of the $72 million available, we have $67 million outstanding as of March 31.

Turning now to our guidance for fiscal 2018. Today, we reaffirmed our full order growth guidance of approximately 5%. On revenue, we're raising the bottom end of our range from the previous guidance of $390 million to $400 million to a range of $395 million to $400 million. Stronger-than-anticipated service revenues and overall year-to-date revenue performance are the primary factors in narrowing this range to the high end of our original guidance.

We are increasing our annual operating expense guidance to be approximately 7% above prior year from the previous guidance range of 3% to 5%. Increased expenses relate to higher spending in research and development, allowing us to bring our product road map activities to market sooner, adding additional sales infrastructure in the U.S, and lastly, accelerating our spending around business development activities. As Josh highlighted, we believe these investments will yield long-term revenue growth.

For EBITDA, we are lowering our guidance from the previous range of $25 million to $30 million to approximately $18 million to $20 million. The lower EBITDA range is due to the expense increases we have mentioned coupled with lower-than-anticipated margins.

Service margins are now expected to be relatively flat to 2017 given the higher-than-anticipated part consumption I discussed previously. Additionally, product margins for the full year are expected to be approximately 43%, which is below plan, primarily due both to lower volumes and products and regional mix.

And with that, I would now like to hand the call back to Josh.

Thanks, Kevin. Before we open up the call for your questions, I'd like to thank the entire Accuray team for their continued focus, commitment and improving execution, supporting our important work that's making a difference in patients' lives.

Operator, we're now ready to open the line for questions.

(Operator Instructions) Our first question is from the line of Josh Jennings of Cowen.

I was hoping to start off and ask about new orders in fiscal third quarter and also what guidance implies for fiscal fourth quarter. And I guess I was just needing help just thinking about the fiscal third quarter performance. You guys did have a very challenging comp from fiscal Q3 last year, and typically, at least the prior 2 years, in fiscal '15 and '16 the fiscal third quarter, you saw on a dollar basis new orders drop sequentially. Can you just help us -- remind us about fiscal third quarter last year, why that was so strong and how you're thinking about setup for fiscal fourth quarter? Was it just -- Josh, in your prepared remarks, and I apologize for the lengthy question here, but you talked about multisystem orders and it sounds like a multisystem may have fallen out of fiscal third quarter and into fiscal fourth quarter. So I'd love to just hear your thoughts on fiscal third quarter, the comp and then also how we should be thinking about the strong growth in (inaudible) fiscal quarter new order guidance? And I just have one follow-up.

Yes. Okay, Josh. So let me take the third quarter just as a starting point. We highlighted in the prepared remarks that we see a growing number in terms of funnel opportunities in multisystem multifacility orders. That's the good news. The challenge is because the complexity of those transactions, i.e., the number of devices spread across a number of facilities, spread across the number of participating clinical personnel and administrative personnel that have input into the decision-making processes, it creates longer timelines and quite frankly, some variability in order timing to be predictive about close time. And I think that the reference that we made to the announcement that we'll have this week about Mercy reflects that. That was an order that we had expected in Q3, that slipped to Q4. But it -- quite frankly, it's a significant opportunity for us. As we indicated, it's the largest order that we had in the U.S. market in 5 years. And we feel good about that win. But again -- so the opportunity said that we're facing relative to more multisystem orders is a positive. It's just a little bit more difficult to predict when that will happen. And so if you have one of those big ones that slips from Q3 to Q4, it can create some disruption relative to the expectations that exists out in the market. As far as prior year comps, I'm going to let Kevin give you his response on that one.

Yes. So, Josh, first off, good afternoon. If you're looking at the year-over-year comps, I think it's important to point out that we had double-digit order growth in this most recently completed third quarter in all of our regions except Asia Pacific, led primarily by our European region. So year-over-year basis, you're really looking at the decline in APAC, which was expected and in our plan due to our tempered deal on the class A quota in fiscal 2018.

Okay. And just on fiscal fourth quarter, the guidance implies north of 20% new order growth. You do have a very (inaudible). But is there a multisystem baked into that guidance? I just with the commentary, Josh, you just provided, again, just in terms of the complexity of the timing of those orders, I just wanted to get a sense of how that fiscal fourth quarter is shaping up? And whether we have -- are you relying on another multisystem order with that kind of complexity of timing issue surrounding it?

I think we feel pretty good, Josh, about where we sit in Q4. It is a big step-up. But given kind of what's already in the hopper, I think we're feeling good about where we should land there. So again, the multisystem orders -- the funnel opportunities that we're chasing and that we have in front of us are spread across all regions. And we have those opportunities kind of presenting themselves through Q4 and into the beginning of fiscal '19. So it gives us some confidence that we should be able to make that step-up in Q4.

Okay, great. And just one last one on the -- just looking out into 2019, I know you're not going to give out your guidance, but just considering order growth you're expecting to experience for the full year '18 where by the backlog stands, can you help us just think through your confidence level in terms of the ability to grow the revenue base off of the fiscal '18 level? Were you accelerating revenue growth? I think your guidance implies kind of a 3% to low-4% range (inaudible) modest acceleration to the higher kind of 4% range. I just wanted to kind of get -- test your temperature there on the ability to either maintain this level of growth or accelerate it in '19.

Yes. So without giving guidance, maybe I will just go -- attack your question in a different way. If you take our implied gross order guidance in the fourth quarter along with the implied revenue range, you're looking at backlog growth at the end of the year that should be up 7% to 9%, which historically would be a good reflection of what we believe the future product revenue growth of the business should be moving forward on the product side. And then in my prepared remarks, I had said that you should expect service revenue to grow more in line with our installed base, which is in that 4% to 5% range. So I definitely think, Josh, we should be able to maintain our growth rates moving forward with our increases in backlog.

Our next question is from the line of Brooks O'Neil of Lake Street Capital.

I know you talked a little bit about where you're at with the U.S. commercial execution. But could you just amplify a little bit about how you feel about your U.S. commercial organization and what the outlook is in this country?

Yes, Brooks. You know that we basically made what, for us, has been really probably the most substantial commercial infrastructure investment we've made in the time that I've been here in the U.S. market in the first part of this fiscal year. In absolute headcount, it was probably something in the low- to mid-teens in terms of total personnel. That's not all salespeople, but it's a combination of salespeople, folks across clinical applications, marketing, et cetera. So pretty significant incremental investment for us from a commercial standpoint. We feel good about the people we brought onboard. The vast majority of them are competitors -- folks from competitors and people with proven track records. And -- but again, just given the nature of the length of the sales cycle in this business model, you're going to see some ramp -- time to ramp as you might expect. And I think we're starting to see -- as I mentioned in my prepared remarks, we're starting to see some light at the end of the tunnel with regards to indicators both in terms of funnel activity and funnel strength and some of the activity that we're starting to see that would lead us to believe that, that incremental spend is going to start to kick in here. I think we saw a little bit of it in this order that we highlighted in our prepared remarks. And I feel good about the opportunities that exist in the U.S. market for us going forward. There is more -- there are more multisystem orders. There are more opportunities for us in single and dual vaults that we really haven't had a chance to play in historically. So I think we feel pretty good about where we should be. But I mean on a comparative basis, we still have work to do in the U.S. market. The U.S. is not operating at the level of consistent performance that we see in our EMEA region or the Japan region by comparison. So getting better. Funnel looks good. Feel good about kind of the ramp we're starting to see, but more work to do. It's too early to declare victory.

Okay. That's very helpful. I really appreciate it. Would you say that in the U.S. you are seeing a real manifestation of the replacement cycle we've been expecting? Or are you seeing more competitive takeaways or opening new bunker opportunities? How would you characterize it?

I wouldn't characterize it from a contribution standpoint. New bunker opportunities are not part of that mix. I'd say it's a combination of replacement sale opportunities. So bunker retention of our legacy bunkers and competitive takeaways in the context of us moving into single and dual vaults that we haven't been able to compete in really effectively, if you look back over the last several years.

Great. Just one more from me. I'm pretty excited about the success you're having with Radixact. That's tremendous. Are you seeing -- would you say you're seeing any cannibalization of your CyberKnife opportunities because of that? Or is it pretty much incremental in your mind?

No. There's no -- the success and the ramp we're seeing with Radixact really doesn't impact the CyberKnife discussion. We'd -- as you've heard us talk about in prepared remarks, we had a little bit softer quarter in CyberKnife than we would have liked. But at the end of the day, it came against a prior year comp that was very strong, in prior year third quarter. And, if you look historically, CyberKnife sales for us have been somewhat cyclical. It will run kind of in a -- it will oscillate, if you will, between quarters. We'll have a strong one, we'll have a softer one, just the kind of nature, I think, of a sales process with that device because it is more of a specialty device. It's less of a replacement -- a pure replacement opportunity, although we have a fair number of replacement opportunities with CyberKnife in the U.S. and Western Europe as well.

Our next question is from the line of Brandon Henry of RBC Capital Markets.

With more clarity in China now, can you spend some time discussing the opportunity for Onrad in China? And how quickly you think this product could ramp? And then a couple of follow-ups.

So, Brandon, we have -- our distributors there and our team there have really been working hard to position Onrad in the value segment over the course of probably the last 15 to 18 months. As we've talked about in the past, prior to real clarity around where those products were going to land in terms of their classification, I think that we characterize it as a little bit of a headwind in that people are kind of taking the wait-and-see approach rather than being more aggressive in jumping in with Onrad decisions. With the decision now behind us and clarity around where it falls, I think we feel like we should be able to see some impact -- order impact within the next quarter or so. And I would say, you'll start to see some -- we'll start to see some in Q4 and it should improve from there. So I think we sense several months. But I think that means the beginning of which would be next quarter, and moving forward from there into early fiscal '19.

Okay. And then any clarity you can provide on the size of the Mercy order? Would you guys have hit your gross order guidance with this order in fiscal 3Q?

Yes.

Okay. And can you also provide some thoughts in terms of your confidence in reaffirming the guidance -- the gross order guidance for the year? And how much flexibility do you have in that fiscal 4Q guidance should several of these multi order -- system, system orders slip into fiscal year '19?

So we -- again, the funnel opportunity that we have, improving in the U.S., still strong in Western Europe. The 1 big area, quite frankly, from a regional contribution standpoint that really now should be also kicking in is the Asia Pacific discussion because of China, which we just discussed. So I think when you look at these things in the aggregate, it leaves us feeling pretty good -- feeling pretty good about where we are and our ability to deliver what we recognize is a pretty good size step-up or step change in Q4 order activity.

Okay. And then just last question for me. Obviously, fiscal year '18 EBITDA guidance was lowered. I think you talked about strategic investments you're making in gross margins. Can you just provide some additional insight to both of those issues? And on the investment side, where you making the investments? And what types of technology do you think you can accelerate?

So Brandon, I'll start and I think Kevin can amplify on this. But one of the things that we recognize very clearly is, we're at a point in our evolution that we don't have the luxury of being able to just grow from a top line standpoint and not care about what the bottom line looks like. Our balance sheet, our capital structure, the improvements we've made in all of those areas have been directly related to our ability to generate more consistently free cash flow and generate more EBITDA. And so this is a balancing act for us that we take very seriously. Our focus right now is to continue to be able to show that we can accelerate and ramp revenue growth, but not give up the impact to the bottom line. Again, we made some conscious decisions coming into fiscal '18 that have been, I think, both Kevin and I feel very certain that they have been the right decisions for longer-term revenue momentum. But given some of that spend and some of the one-time or margin impacts that we've seen and really been surprised by, quite frankly, in some aspects of the service business. It's put this quarter certainly and then on a full year basis, given how late in the year it's been, it's put our ability to deliver the committed number from the beginning of that FY '18, it's put it in jeopardy. And so -- but our view going forward is, we have to do both. We have to continue to accelerate revenue growth, and we have to be able to show that we can generate incrementally more free cash flow going forward and get this business to a sustainable level of profitability in terms of GAAP operating income or GAAP net income.

Our next question comes from the line of Tycho Peterson of JPMorgan.

It's Tejas on Tycho. Just a couple of ones here from me on APAC. So first on China. You said you'd announce a JV by your next earnings call. Can you just elaborate on what you're looking for in terms of an ideal partner there? And what segments of the market you hope to target via the JV? And then on India, recently you highlighted the sale of 2 Radixacts by Apollo Hospitals. So what does your installed base look like today in the country? And how big could the opportunity potentially be for you down the road?

Yes. So let me take the China piece first. Basically, the ability for us to be able to ensure that we have good market access longer term in China through the presence of having a local branded product solution with the assistance of a local manufacturing partner, I think, is an important part of our growth strategy going forward. So that's kind of the background on that. And I think our view is that someone that has from either an existing manufacturing capability or someone that has the experience in the regulatory processes that will allow us to bring product to market, whether it's from a prior experience in the linac business or not is less important probably. But those are the primary drivers in terms of what I would describe as the characteristics of what we're looking for. I would also point out that this is not a pure R&D and start-from-scratch product development activity cycle. This is really more of a product branding and local regulatory and assembly type of presence in that context. So something that would look very different than essentially a start-from-scratch build it entirely ourselves kind of a scenario. On the India opportunity, our installed base, I think, right now is probably somewhere in the low- to mid-20s kind of installed base -- device base size and growing nicely, quite frankly, and one that we feel good about going forward.

This is Kevin. What's been, I don't want to say surprising, but I think it's nice to see is we've actually seen some receptivity to our next-generation Radixact platform in India. So I know a lot of folks may view India as more of an emerging market, lower price point type of market. But we've been successful. We have a direct operation there. We've been able to take Radixact orders. And I think what we're hearing from customers in India, the throughput of Radixact and the reliability does make it an ideal solution for the hospitals in India that many of those devices can be running 18 to 22 hours a day. So we see some nice uptake there.

Got it.. That's helpful. And one quick one here from me on the U.S. So I know in last quarter you said your commercial buildout and the reorg in the States was largely complete. And how should we think about sort of visibility overall in the market given that you've had some success there in terms of your commercial organization, but on the other hand you have these large lumpy orders? Is there anything you can do in terms of these large multifacility orders in terms of just getting a better sense of order timing and flow and getting people to sort of place orders before the quarter closes? Or is it largely driven by factors that are beyond your control? And then secondly, what's the split like in terms of competitive wins at the single, dual vault unit level versus replacing your own instruments? Is it 50-50?

We're still replacing our own bunkers or replacing our own installed base situations at a level of about 15% to 20% of total order volume on a global basis. The biggest geographic representation or areas of concentration for replacement sale opportunities continues to be the U.S. and Western Europe. So I'd say that's the primary driver of the replacement sale discussion and where those opportunities are.

And then if you look at just replacing conventional linear accelerators and competitive bunkers, those obviously vary from quarter to quarter. But we've been running in the 15% to 25% of all of our sales are competitive replacements, which would leave roughly 50% of our remaining sales to more greenfield-type opportunities, which you'd see in APAC and Eastern Europe.

Got it. And then on the commercial organization in terms of getting better visibility on the larger orders any thoughts there?

Yes. Lots of thoughts. We're working hard at it. The reality is, we -- I think some of this is -- it's a lumpy business to begin with. I think our lack of scale creates actually higher peaks and valleys than our competitors. And I'm not complaining about that. The onus is on us to figure out how to be forward-looking and dialing in expectations from a forecast accuracy standpoint to the greatest degree we can and greatest of accuracy we can. And we're working hard at that. I think it's -- maybe complicated a little bit in the U.S. market by, again, how new the resources are that we've got in place. We've got probably 65%, 70% of our territories or people that have been in place for less than a year. And so just, again, given the length of the sales cycle and the complexity of the deal flow that we're looking at with more multisystem orders, it becomes a little bit more challenging with the folks that have been in place on a shorter time frame, quite frankly, than would be ideal. But I would expect this would be improving over time. I don't know that we'll ever be -- until we grow our way, quite frankly, to a level of scale that we -- an order, even a single order, if it's a multisystem deal, not having the ability to move or to impact us in terms of we can make the commitment or not make the commitment, until we get to that level of scale, I think that we'll probably be subject to some continuation of this. But I think you should be confident in knowing that we're working hard to minimize the quarter-to-quarter variability in order activity.

Our next question is from the line of Amit Hazan of Citigroup.

This is Bhalla on for Amit. So the first one, on the increased R&D spend, I think you guys noted future imaging enhancements that are going to be coming. Can you just touch briefly on what that pipeline may entail? And then any updated thoughts that you have around adaptive therapy offering and what that might look like for Accuray?

Yes. So we're excited about the things, and we've got our biggest investments being made in terms of the road map. The specific road map projects, I'd say, from the key programs right now, the big 3 would be diagnostic CT quality capability on Radixact that would be in the form of kVCT imaging capability; the next generation optimization and volumetric imaging capability on the CyberKnife, which is the programs we've talked about in conjunction or collaboration with medPhoton; and then true adaptive therapy capability on Radixact. We'd also talked about in prepared remarks, we just launched an MVCT enhancement on the existing Radixact platform. That's the CTrue iterative reconstruction, which is really focused on soft tissue resolution enhancement and it's gotten some very good feedback, at least in early-stage discussions with early users of the product and the software. So again, it's probably very more heavy to the imaging -- upgrade imaging enhancement end of the spectrum in terms of the topical areas. And we're excited about the things we've got going on. We will get to something approaching diagnostic CT quality capability on the next turn of image upgrades for Radixact and CyberKnife both.

That's all great. Any timing around that next turn of image upgrades?

I think we've talked about something on the Radixact side, maybe, something in the next 18 roughly months or so. And CyberKnife, probably a little bit beyond that.

That's great. Second one, I certainly appreciate the increased revenue range, I guess, tightening higher range. But you noted expectation for service revenue to return back towards installed-base growth. Does that -- that essentially means that you exhausted the opportunity there with the software upgrades. Why isn't there more of a runway there?

I actually think there is more of a runway. I was pointing to more our full year service revenue growth given where we started off the year kind of in the low single digits. So if you go back to Q1 and Q2, we're only growing at 2%. So the Q4 guidance would imply that there's runway left in our software upgrades. And I actually do think that's an area, as a company, we're putting some focus into installed-base upgrade sales because not only does that help the top line, those upgrades are typically more margin-rich than product revenue sales. So it also helps the bottom line.

One of the areas we've added resources when we talked about the build or the investment incrementally in U.S. commercial infrastructure, one of the areas, specifically, was installed base -- dedicated installed-base personnel, people that we're focused on mining the installed base, specifically, for things just like the topic you talked about, software upgrades, other upgrades from a hardware standpoint and so, I think we're getting better focus and better -- more feet on the street, better focus, better intensity, if you will, on those types of products specifically, which I think should continue our trends going forward.

Thank you. And at this time, there are no further questions. I'd like to turn the conference back over to Josh Levine, Accuray's President and CEO.

Thank you, operator, and thank you, everyone, for participation this afternoon. We look forward to speaking with you on our fiscal fourth quarter earnings call. Thank you, and have a good evening.

Ladies and gentlemen, thank you for your participation in today's conference. This does conclude the program. You may now disconnect. Everyone, have a great day.