NSTG (2020 - Q3)

Ladies and gentlemen, thank you for standing by. And welcome to the NanoString Third Quarter 2020 Operating Results Conference Call. [Operator Instructions]. I would now like to hand the conference over to Mr. Doug Farrell, Vice President of Investor Relations. Douglas

Thank you, operator, and thanks for joining us today. On the call today is Brad Gray, our President and CEO; as well as Tom Bailey, our CFO. Earlier today, we released our financial results for the second quarter of fiscal year 2020. During this call, we may make statements that are forward-looking, including statements about financial projections, the impact of the COVID-19 pandemic, existing and future collaborations, future business growth, trends and related factors, prospects for expanding and penetrating our addressable markets, our strategic focus and objectives and the development status and anticipated success of recent and planned product offerings. Forward-looking statements are subject to risks and uncertainties, many of which are beyond our control, including risks and uncertainties described in our SEC filings. Our results may differ materially from those projected, and we undertake no obligation to update the forward-looking statements. Later in this call, Tom will be reviewing our financial results. Consistent with our most recent earnings release, we have prepared as a supplement to our results according to GAAP selected non-GAAP or adjusted measures that we believe makes it easier to interpret and compare our financial results. This afternoon's press release includes details regarding the definition and calculation of these non-GAAP measures, reconciliation of non-GAAP measures to nearest comparable GAAP measures as well as a discussion of the limitations and rationale for using these measures. Throughout this call, all financial measures will be GAAP, unless otherwise noted. To aid analysts and investors in building their models, we have posted exhibits out of the financial information tab of our Investor Relations page that include a presentation of our non-GAAP or adjusted measures and other selected financial data for each quarter and for the full year 2019. Later this month, we'll be participating in 2 virtual health care conferences, the Stifel Healthcare Conference and Credit Suisse. In addition, as you may have seen in our press release today, we're excited to announce our virtual Investor Day event, which will be held on Tuesday, December 1, where we'll provide updates as to the product road map, commercial initiatives and the development of the spatial biology market. We look forward to having the chance to speak with many of you there. With that, let me turn the call over to Brad.

Thanks, Doug. Good afternoon, and thank you for joining us today. NanoString is a leader in spatial biology, and we are addressing a large and growing market with huge potential across discovery, translation and clinical diagnostics. We have tremendous momentum, as evidenced by the strong orders for our GeoMx DSP instruments, increasing demand for GeoMx consumables and our rapidly expanding funnel of sales opportunities. During the third quarter, we generated over 20% pro forma revenue growth, despite ongoing academic-related headwinds. We preannounced these third quarter results on October 6, and successfully completed a $230 million equity financing. It gives us the resources to fund our current growth initiatives as well as to make future investments to help extend our leadership in spatial genomics. We have never been in a stronger position or more excited about our growth prospects. During the call today, I'll give you more specifics on our Q3 operational results and update you on progress towards our strategic objectives for the year. I'll then hand off to Tom, who will review the details of our financial results and outlook for the balance of 2020. During the third quarter, we maintained our commercial momentum in spatial genomics generating new orders for more than 25 GeoMx instruments, exceeding the guidance we provided during our August call. The leading indicators of future GeoMx demand are positive across the board, including a doubling in the volume of TAP projects over the prior year, a steady stream of GeoMx publications and a torrid pace of lead generation through virtual conferences and events. At the end of Q3, we had more than 150 cumulative GeoMx orders, with more than 120 instruments shipped, about 100 installed and about 90 GeoMx sites trained. The rates of GeoMx instrument shipments, installations and training all improved in Q3 as labs increased research activity levels compared to the second quarter. We're seeing GeoMx's consumer utilization ramp up as we install systems and train sites, and we recorded $1.4 million in GeoMx consumable sales in Q3, more than double the prior quarter. Our core nCounter business has been recovering nicely from pandemic-driven slowdown, growing sequentially by more than 50% compared to Q2. nCounter revenue was down less than 5% on a pro forma basis compared to the prior year versus the roughly 30% year-on-year dip we experienced during the second quarter. nCounter demand improved as research activity increased over the lows of Q2. Our sales team estimates that globally, about 25% of our customer labs are fully open, while the remaining 75% are operating at reduced capacity. Laboratories have not increased their activity levels meaningfully since August, and we expect most labs to continue operating at reduced capacity through the end of the year. We have not seen rising COVID-19 infection rates in the U.S., and Europe leads to a slowdown in research activities, but we are monitoring the situation closely and factoring that dynamic into a cautious Q4 outlook. Now I'd like to provide an update on our 4 strategic objectives that we laid out at the beginning of the year. Our first strategic objective is to accelerate the adoption of GeoMx DSP in translational research or nCounter as the preferred readout modality. Our lead in translational research remain large and accounts for the vast majority of 150-plus GeoMx DSP orders we've taken through September 30. More than 15 of our 25-plus GeoMx DSP system orders in Q3 were from translational researchers planning to use the nCounter-based readout. nCounter-based readout is well suited to the needs of researchers who are focused on running targeted protein panels, offering a simple workflow and lower cost per sample. We've recently expanded our market-leading menu of protein content, launching 3 new protein modules for readout using nCounter. This brings the total portfolio of genomic protein assays to over 300 validated antibodies. We continue to see strong demand for GeoMx nCounter bundles, demonstrating that we are penetrating the translational market beyond our historic installed base of nCounter systems. Our translational customers continue to publish at an impressive rate, and since our August call, there have been 6 new GeoMx papers for a total of 29 cumulative GeoMx publications. The body of peer-reviewed GeoMx data is growing on pace similar to the early days of single cell genomics. And we believe this is helping fuel demand. In September, we announced the formation of the GeoMx translational leadership network that is focused on pioneering the application of spatial biology and large translational research studies, developing and sharing standardized practices to facilitate collaboration between groups of researchers across the globe. Next week, we'll be participating in the Association for Molecular Pathology conference, where we are hosting a workshop focused on the diagnostic potential of GeoMx. While the clinical market is not expected to make a meaningful revenue contribution over the next year, it does provide an exciting long-term growth opportunity for spatial biology that we believe will arise naturally out of our success in translational search. Our second strategic objective is to expand the adoption of GeoMx into discovery research, leveraging the platform's next-generation sequencing readout capabilities. Interest in GeoMx amongst discovery researchers is being driven by the unique capabilities of this system, which include full automation, compatibility with formalin-fixed, paraffin-embedded samples and the ability to profile both RNA and proteins using the massive installed base of next-generation sequencers. In August, we launched the first NGS-based consumables for GeoMx DSP, the Cancer Transcriptome Atlas, or CTA, which profiles more than 1,800 RNAs and spatial context. In the third quarter, about 10 of our GeoMx DSP orders came for researchers who plan to make NGS their primary readout for GeoMx. As discovery researcher interest in NGS has increased, we are expanding opportunities for prospective customers to test drive GeoMx using their own samples, beginning with our Technology Access Program, or TAP. During the third quarter, more than half of the record 75-plus TAP projects ordered used NGS as the readout. In September, we expanded our TAP program to include our Whole Transcriptome Assay, or WTA, which is helping to build commercial demand and momentum ahead of WTA's commercial launch next year. This assay allows researchers to profile every gene in the human transcriptome in an unbiased manner, providing a powerful discovery tool that can be used to probe any aspect of spatial biology in human tissue. To complement our internal TAP, we formed an external service network called the GeoMx Premier Access Sites, consisting of 8 laboratories around the world that will provide customers with easy access to run their own samples, allowing them to use that data to write grants and budget requests to purchase their own GeoMx system. Finally, as we announced this morning, we have expanded our NGS readout capabilities to include our industry-leading portfolio of protein panels. Customers receiving our Cancer Transcriptome Atlas will now be able to add complementary protein content, including more than 50 protein targets that are focused on immuno-oncology applications. With the addition of protein capability to our RNA offerings, the ability to analyze markers in virtually any tissue type, we now offer the most versatile and diverse portfolio of content for discovery research in the emerging field of spatial biology. Our third strategic objective is to maintain the momentum of our nCounter business by expanding our installed base while maintaining consumable pull-through. In the third quarter, we generated healthy instrument demand and grew our installed base to approximately 915 nCounter systems, an increase of 12% compared to a year ago. While about 20% of our nCounter instrument sales were driven by GeoMx bundles, the majority of placements were motivated by traditional gene expression in the field of oncology, immunology and neuroscience. Immunology is an area of particularly strong interest, driven in part by new nCounter panels. For example, we're seeing strong interest in our new host response panel that allows scientists to study the immune system response to SARS COV 2 or any other pathogen. This panel can be used in combination with our COVID-19 gene spiking, which has been ordered by more than 75 customers to date. nCounter gene expression studies of COVID-19 now have been published in both New England Journal of Medicine, and in science, reaffirming the unique position of nCounter to generate key translational results on hard-to-analyze FFPE samples. Our fourth strategic objective for 2020 is to select the key applications for our Hyb & Seq platform. As we mentioned in August, our team has been exploring potential applications for this unique sequencing chemistry. While for the time being, these remain confidential, we plan to provide an update on the direction of this program and our overall product development pipeline at the virtual Investor Day that we'll be hosting on Tuesday, December 1. Doug will provide details shortly after the call. Now I'd like to turn the call over to Tom to review the details of our operating results for the third quarter.

Thanks, Brad, and thanks all for joining us today. I'm excited to walk through our results from this quarter, which are strong top bottom, despite a challenging operating environment. For the third quarter of 2020, product and service revenue was $30.1 million, representing pro forma year-over-year growth of 22% and sequential growth of 42%. Recall that our pro forma metrics reflect the December 2019 transaction with Veracyte, is that that transaction occurred at the beginning of the comparable periods. Pursuant to the terms of the Veracyte transaction, we now recognize about 1/3 of the previous Prosigna revenue over the same units sold. As Brad noted, GeoMx demand was strong in Q3. Genomics revenue recognized was $8.9 million. $7.5 million was derived from approximately 30 instruments shipped, and $1.4 million was derived from consumables sales. nCounter instrument revenue was $5.4 million, and are 8% lower compared to Q3 '19, but showing a strong recovery with sequential growth of about 50% as compared to Q2. NCounter consumables revenue has been more impacted by lower lab activity. However, in Q3, we began seeing substantive recovery. Q3 nCounter consumables revenue was $12.3 million, 9% lower on a pro forma basis compared to Q3 '19, but sequential growth of 55% compared to Q2. Our Q3 nCounter consumable sales imply annualized pull-through of about $55,000 per system, about 90% of our pre-pandemic pull-through guidance and a substantial improvement over the $37,000 annualized pull-through we experienced during Q2. Service revenue derived from both nCounter and GeoMx-related services was about $3.6 million for the quarter, 16% year-over-year growth and 20% sequential growth. Driven primarily by increasing GeoMx DSP TAP projects and increased service contract revenue due to our growing instrument installed bases. Turning now to margins and expenses, I'll be providing results on a non-GAAP or adjusted basis, removing the impact of stock-based compensation, depreciation and certain onetime items. Please refer to our press release with all of the exhibits we have posted to our Investor Relations web page for detailed information on how our non-GAAP or adjusted measures, including adjusted EBITDA, are prepared. Q3 adjusted gross margin on product and service revenue was 55%, about 600 basis points lower than Q3 last year. About half of that change was driven by the lower Prosigna pricing we received from the Veracyte transaction. Pro forma for the Veracyte transaction, adjusted gross margin was about 300 basis points lower, primarily due to greater instrument revenue as a percentage of our total sales mix with the acceleration of GeoMx sales. We reduced operating expenses compared to the quarter a year ago. Adjusted R&D expense was $12.7 million, a decrease of 16% year-over-year. The decrease was driven primarily by reductions related to the Veracyte transaction and the conclusion of certain collaboration agreements, offset in part by expenses incurred for our continued GeoMx product development efforts. Adjusted SG&A expense was $16.6 million, a decrease of 16% year-over-year. The Q3 SG&A expense decline was driven primarily by reductions relating to the Veracyte transaction as well as expense savings realized from reductions in travel and trade show activities. These savings were partially offset by our continued investment in GeoMx-related commercial initiatives, in particular, investments made in our service and customer support group and in certain digital marketing initiatives. Adjusted EBITDA loss was $12.7 million, an improvement of 33% as compared to the prior year and reflecting the combined impact of our revenue growth and lower operating expenses. Turning to the balance sheet. We are pleased to have the strongest balance sheet in NanoString's history after completion of a common stock offering last month, which we upsized due to strong investor demand. The offering generated net proceeds of about $216 million, including these proceeds and cash and equivalents on hand at the end of Q3. Our cash, cash equivalents and short-term investments balance is currently over $445 million. Transitioning to guidance. For Q4, we currently expect product and service revenue of about $31 million to $34 million, which assumes lab activity levels that are approximately consistent with Q3. This range assumes $21 million to $23 million of nCounter revenue and $10 million to $11 million of GeoMx revenue. Our nCounter guidance represents just over 3% sequential growth from Q3 at the midpoint of the range, consistent with last year's Q3 to Q4 nCounter sequential growth. Our Q4 guidance range is wider than we typically provide, reflecting current pandemic-related recent uncertainty, with infection rates had a new records in North America and new lockdowns in EMEA. While our sales funnel remains strong and October purchasing activity remained at levels similar to those observed in Q3, we are aware of a few specific examples of purchasing processes slowing down in Europe in recent weeks. As has been true throughout the pandemic, revenue in Q4 may be negatively impacted if customers are unable to place orders or receive shipments due to site closures. GeoMx revenue recognized might also be impacted by the pace of our service and support team's ability to install systems and train customers, which is dictated by our own safety and operational procedures as well as those of our customers. Against this backdrop, we are guiding for 25-plus new GeoMx instrument orders during Q4, approximately flat to our strong Q3. Regarding expenses, we expect gross margins and operating expenses to be about consistent with what we recorded in Q3. Now I'll turn the call back over to Brad for our closing comments.

Thanks, Tom. NanoString is an early leader in spatial biology, a rapidly developing field that will revolutionize biology. We generated tremendous momentum with our initial GeoMx offerings among translational researchers, and we are now leveraging NGS readout to rapidly expand the capabilities of this platform to accumulate an even broader set of researchers. Following our recent financing, we believe we have the capital needed to both grow the sustainability and to invest in strategic growth opportunities. We look forward to our upcoming Investor Day, when we'll share additional details on how we plan to put this capital to work organically within our innovative pipeline, helping to further extend our leadership in spatial biology. With that, I'd like to open up the line for your questions.

[Operator Instructions]. And our first question comes from Dan Arias with Stifel.

Brad, can you just kind of expand on the comment that Tom made a few minutes ago just in terms of lab access for installation and for training on your instruments? We've kind of gotten some varying comments from the different academically focused companies so I'm just curious what your experience in shaping up like in the U.S. versus Europe and the ability to get in there and kind of step people up?

Thanks for the question, Dan. It's a very dynamic situation. I'd say beginning in the last few weeks, especially as infection rates increased in Europe, we've started to see some travel restrictions and complications begin to reemerge in Europe. The -- we don't have service and support personnel, for instance, in every single country in Europe. We have people that have to cross borders to do installations and trainings and preventative maintenance-type work. So when we have quarantine procedures that culminated in places like Europe, for NanoString that becomes complicating. It becomes hard to send a service engineer from one country to another and then have to take that person out of the field for 7 or 14 days into quarantine post installation. So those are the kinds of complications that we're dealing with. It is largely in Europe at this time. And I think we are -- our guidance range of $10 million to $11 million for GeoMx, basically, this is at the top end of that range that we continue to be able to install our -- and recognize revenue for GeoMx systems at a reasonably undisruptive manner over the next, say, 8 weeks. And low end of that range really implies that things get more complicated, we're not in a position to recognize revenue quite at the pace that we had hoped. Lab activity itself in terms of utilization of consumables has not slowed down. We have not -- our October activity was consistent with Q3 and remain -- it remained stable. But we are watching closely our indicators to see if some of the same lockdown procedures that you're beginning to hear about in Europe keep people working from home rather than in the lab and whether that plays through on consumables. To be clear, we have not seen that today, but we are cautious as we look at the rising infection rates globally.

Yes. Okay. Very helpful. And then maybe on the nCounter side, I imagine a lot of the folks that are doing gene expression work on the instruments also have a sequencer. So what are you finding at this point are the factors that have labs running the nCounter rather than doing an RNA-seq experiment? Yes. Said another way, where are you winning? And where are you seeing people choose other approaches? And then maybe along those lines, on the pull-through that you're seeing with the nCounter, obviously, this year, a bunch of factors playing into the growth equation with Veracyte and COVID. But any reason why holding steady with double-digit consumables growth wouldn't still be in the picture if we just sort of think about normalized trends over the longer period of time?

Great. So nCounter has always served the need of a set of researchers who value a very simple workflow, a very high level of throughput and specific content that's tuned to the biology of their interest. So for instance, cancer is a field where NanoString has always had a distinct capability with nCounter based on the FFPE compatibility, the ability to process 48 or to really drive at 96 samples a day and on the basis of our very popular PanCancer panels. We've replicated that success in the field of immunology and neurology. And really, that's where you see most of the nCounter uptake. And despite the fact that RNA sequencing has grown less expensive over time, it really hasn't gotten much simpler to either view it to a workflow perspective or from a data analysis perspective. And our customers really want to have independence from their core lab to be able to run of that nCounter system themselves and analyze their data themselves to be highly productive. So that's really why nCounter continues to add demand in the order of, call it, 20, depending on the pandemic dynamic, sort of 20 to 30 new instruments per quarter. And we see that continuing, but we don't see that slowing down in the next year or 2. In terms of the consumable demand, the nCounter system has a razor-razor blade model. Every year, we try to replace about 120 new systems, ex pandemic, growing our installed base at a low double-digit rate and then translating that into a low double-digit consumable revenue growth. So our -- I said during my prepared remarks, for instance, that our installed base is up 12% year-on-year. At constant pull-through, you'd expect that to translate into consumables up about 12% year-on-year. So I think you're thinking about the model the right way, Dan. I think there's no reason to believe we're going to see a slowdown in nCounter instrument placement. That can drive approximately 10% order of magnitude consumable growth in our model going forward.

Our next question from Doug Schenkel with Cowen.

This might be kind of a similar question to what Dan just asked, although I just -- my brain may not be working mathematically quite as quickly as Dan's is tonight. But I was wondering, on nCounter consumable pull-through, you rebounded back to levels last seen in Q1 and Q3. I'm just wondering if you could say a little bit more about pacing by month, whether you think there's any stocking activity. And just given all you've talked about in your prepared remarks on just kind of the changes in lab activity or in certain instances, the lack of change in lab activity over the course of the quarter. Do you think that this is kind of the new normal for a while, meaning the $55,000 annualized per nCounter, do you think that's what we should be modeling moving forward?

Yes. Thanks for the question, Doug. Yes, the pacing over -- let me go all the way back to Q2 and tell you what we saw a little bit about the pacing over the last couple of quarters. So you'll remember, in the depth of the COVID-19 walk down, we saw just abysmal demand in lab activity happening during the months of April and May during Q2. We saw a substantial increase in June. And so we exited Q2 at a pretty decent run rate of consumables, much better than the average of Q2 overall. The June run rate was much better than the overall quarterly average. We saw that continue to improve in July and August as labs moved into, really, out of a situation where some labs will close. We have virtually no labs that are totally closed at this stage. Most labs operate in this partially open scenario, really since the end of August, where maybe the staffing is not quite as high as it normally would be in terms of the number of actual people in the lab, and therefore, the amount of research activity is not as high as it would be, but it's a lot more than 0. So really since the end of August, we haven't seen that change much. We've seen about the same level of lab activity and the same level of, therefore, consumable order pacing. I think this is the state we're going to be in at least through the end of Q4. Labs are figuring out how to get as much work done as possible, spreading people out physically in their labs, spreading people over multiple shifts, maybe into the weekends. But I don't think that we will see us get all the way back to pre-pandemic pull-through until 2021. So our Q4 guidance, I think, a good way to think about it is applying that $55,000 per system per year consumable pull-through on our modestly larger installed base on a sequential basis, and that's totally consistent with Tom's guidance.

Got it. And just to be real clear on that, and that was super helpful, Brad. But the good thing is there is no stocking or catch-up dynamic in Q3. So just...

No, I really don't know.

Okay.

No, we don't believe we saw any stocking dynamic. Our consumables are sort of -- you've got to know what experiment you're going to do before you know what panels to order or what content to order. So we very rarely see stocking dynamics. That was not what was in account for the sharp rebound in Q3.

Okay. All right. That's great. And then I know we kind of ask some form of this every quarter. But recognizing, it seems like it's really, really early days for spatial profiling in the sense that there's a lot of interest, but you and your competitors are just getting going. So that's great. I do want to ask if you've seen any changes in competitive dynamics amongst the competitors that you've noticed as activity has picked up. And then to the extent there is any change, if you could talk about whether there's things that you could point to, whether it's application or end market. And then kind of last part of this. Is the ReadCoor-10x combination changing the tone of discussions at all with key customers?

Yes. So I'd say in the last quarter, there really have been no material change in competitive dynamics. NanoString continues to be incredibly strong, where we have been strong with GeoMx, which is in the translational markets. We feel very secure in our leadership. We really do not encounter head-to-head competition in that market. When we watch at the latest scientific meetings to see if there's any kind of sea change in the representation of GeoMx relative to other spatial profiling platforms, we don't see anybody catching up with us. For instance, this morning, the Society of Immunotherapy and Cancer abstracts published. We were glad to see that GeoMx appears to have 3x as many abstracts as the competition. And we feel really great about our momentum, with GeoMx showing up in immuno-oncology translational research. So really no change in the competitive dynamic. The acquisition by 10x Genomics of ReadCoor has also, I'd say, not meaningfully changed the dialogue we're having with our customers. That's a product that we will all look forward to seeing what 10x plans to do with it and when it will be coming to market. I think it sits in a category of imaging-based products that I think will be interesting over time that will probably not be commercially impactful for another couple of years. And certainly, it's not top of mind for the customers right now who are buying GeoMx. And when it does become top of mind, I'd say that imaging-based product category is complementary to profiling if the imaging-based category provides high resolution in a single cell or subcellular level. Its plex is going to likely be not quite the whole transcriptome, but hundreds or potentially thousands of genes over time, and it's going to be much lower throughput. It's going to occupy a complementary place to where the profilers sit, which tend to be higher throughput, higher plex all the way up the whole transcriptome and lower resolution. So I'd say it remains for us to see how the imaging category develops over time. But right now, the competitive dynamic is just as it was when we talked to you in August with NanoString feeling very secure in its leadership in translational, and as we talked about in the prepared remarks, a lot of movement towards discovery.

And then sorry, last one, Brad, and that was great. On Whole Transcriptome Atlas, and I think the planned launch for 2021. I apologize if I missed this in your prepared remarks. But at least in my notes I have -- really accessing 8 global sites. Is there any plan to expand this before the full commercial launch next year?

So Doug, if I heard your question properly, you're asking about the Whole Transcriptome Atlas availability as a service. So to be clear, the Whole Transcriptome Atlas, right now, is available only through our Technology Access Program. So -- and we have a lot of interest in it. People who are sending their samples to us here in Seattle, and we're running the Whole Transcriptome Atlas and shipping results back to them. The GeoMx priority site program is for people interested in everything but the Whole Transcriptome Atlas. There's the Cancer Transcriptome Atlas, the nCounter-based readouts, the protein readouts, those will be done by the 8 global sites we've added. We haven't yet enabled those to hold the Whole Transcriptome Atlas, and we really don't plan to ahead of launch. We'll probably just keep that product run out of Seattle until its global availability as an assay that we can ship to customers.

And our next question comes from Tycho Peterson with JPMorgan.

Brad, you previously talked about capacity to install 25 systems a quarter on GeoMx. You did 35 this quarter. So can you maybe just talk to what's led to the step-up in installation capacity and then where do you think that can go?

Yes. Thanks for the question, Tycho. We've been investing and hiring additional support and field service engineers in anticipation of growing demand for an order flow from GeoMx. So for the last, I'd say, for Q2 and Q3, our ability to recognize revenue on GeoMx had been more limited by our customers' ability to accept the systems and invite our people in to install and train, then it has been our own capacity in terms of manpower to do that. Really, with COVID-19 and all the employees' health and safety precautions around that, it's been less about our manpower and more about lab access. But we are at a state now where I think we are prepared to install GeoMx systems at whatever pace is required to more or less match order flow and revenue recognition. Yes. By at least by the end of the fourth quarter, we will have kind of caught up and worked through most of our backlog that we had accumulated during 2019. We'll exit our guidance -- in terms of our guidance, we'll exit the year with slight GeoMx systems in backlog. So think of that as puts us on pace to basically install systems at the same pace that we accept orders next year. So I don't think about that as a meaningful constraint anymore. We've made the hiring to reduce that constraint, or eliminate it, rather.

And then maybe you could provide us with some updated thoughts on the NGS readout scale up. You talked about 10 orders this quarter and half of the technology access orders being tied to NGS readout. How quickly do you think you can kind of leverage some of those installed base and scale that up?

Well, I mean, I see that -- I would expect next year, I think about 2021, I would expect the majority of our GeoMx orders to be coming from people who want to use NGS readout. That will be true in 2021 and every year beyond. And that's simply based on the tremendous installed base of the limited systems that are already out there, the expanded capabilities, including the addition of proteins as we announced this morning, to the NGS capability set and of course, the addition of a whole transcriptome assay, which will be an incredibly exciting consumable launch next year. So I think demand -- while demand will remain robust with the nCounter readout, most incremental growth will come on the NGS readout side. And we really look forward to seeing what customers are able to do in terms of sample processing and pull-through using NGS readout. As we've discussed many times in the past, the consumable pull-through potential of those customers is probably larger and even more exciting than the nCounter readout customers. It's too early for us to declare what a number is, but I'll say to you were very pleased with Cancer Transcriptome Atlas orders that came in the third quarter. It was a pretty strong start from some of those first customers. Another interesting phenomenon we'll see that we didn't talk about in our prepared remarks is, many of our customers who acquired GeoMx initially for the nCounter readout also have access to a sequencer in the same lab. Now that we've enabled NGS readout, many of those customers are asking us to come and train them on how to take advantage of the NGS readout, even though they may use nCounter in some experiments, they want to be able to drive the higher plex on NGS readout than in others. And so I think that just -- it shows a strong demand for the Cancer Transcriptome Atlas even in the translational markets, where our early installed base was on. So I'm very bullish overall on what the NGS readout is going to be for us, but it's hard for me to be very quantitative about that potential at this time.

Okay. That's helpful. And then last one for me. Just curious if you could talk to use of proceeds from the recent offering. I think there was a view when you did the convert back in March that you had what you needed. So I'm just curious if you could talk about where you're going to use the incremental cash flow, is it all for commercial scale up? Or is there an eye to look externally, potentially do some M&A?

Yes. The convertible debt financing that we did back in the spring really did provide enough capital for us to become -- to grow into being a breakeven company. We've always said we're at current, $200 million-plus in product and service revenue. But what it didn't provide was incremental strategic capital to make investments in what we now feel is an enormous and underappreciated spatial biology opportunity. So we really think of this incremental capital as being put to work in 3 potential areas. The first is commercial investment. Capitalizing on the momentum we have in the discovery markets, for instance, by adding additional sales reps there or by finally choosing to go direct in major markets where we still work with distributors, such as in China and Japan, where we can probably do better if we went direct on our own. So that's category 1, commercial investment. Category 2 would be pipeline investments. We haven't yet revealed everything that we have cooked up here organically, and we will talk more about that on December 1, and I think people will be excited by some of the innovation that's continued to happen here in a road map that leads beyond GeoMx. And we're excited to have the capital to fully pursue those options. And then, of course, as the leader in spatial biology, we do track a host of external innovations that are happening in companies large and small. And with the increased capital on our balance sheet, we're in a position to pursue those as strategic licenses or acquisitions or partnerships should one of those technology platform or products resonate strongly with us. So to be clear, and we were clear about this at the time we did the financing, we're not in a hurry to add inorganic growth but certainly, we want to build on and extend our leadership position here in spatial biology, and we'll do that inorganically if the right opportunity arises.

And our next question comes from Dan Brennan with UBS.

Great. Maybe, Brad, just on the kind of the funnel, if you will, you talked about a torrid pace during the prepared remarks. Any more color beyond kind of the number of customers in TAP? I know you just had a publication, but what else could go into that torrid pace that might give us some insight towards kind of the opportunities we look out beyond 2020?

Well, we continue, Dan, to have really great success with our virtual marketing approaches, both at major conferences like the recently completed ASHG, some spatial biology conferences that we've really pulled together on our own, like one we held in September, and other small events that we've done both by ourselves and more recently, in collaboration with Illumina. Just taking the Illumina example, we've now held about 10 events with them jointly, generating about 1,500 marketing qualified leads. Those events have turned out to be really popular. They allow us to talk jointly about the unique capabilities that the GeoMx and the sequencing lineup from Illumina bring. And we just continue to have an unprecedented flow of new leads into our organization that can be qualified and added to our sales funnel, and it grows every single week. So we're going to -- and I think, as I said, back in the springtime, while the pandemic had certainly slowed down the utilization of systems in people's labs, our customers are still looking to the future, and spatial biology is on the top of a lot of people's shopping list when they get back and as a capability they want to acquire and apply their search going forward. So we have really a lot of growth every single week and month in our genomic sales model.

Okay. And then maybe as you look ahead in terms of the outlook in '21 in terms of funding. Can you just discuss a little bit about where the recent new orders came from by customer type? And then as you think about '21 from an academic funding standpoint, any sense about the prioritization or not for spatial? Is that something that -- any color you can give us on grants or from a top-down basis, how many NIH is thinking about spatial as a potential funding opportunity?

Yes. So in terms of where our orders have been coming from recently, it's been more of the same, Dan. It's been largely cancer researchers on either the translational side or the discovery side because of our first discovery assay for NGS is a Cancer Transcriptome Atlas. It's probably been 60% academic, 40% biopharma and CRO, and a good geographic split that's representative of what our nCounter business looks like. So in terms of the incremental demand next year, I think spatial is going to be getting more and more attention as a modality that people want to add to get incremental insights on top of their bolt or their single cell-based genomics. I think it will be things that editors are going to be -- or pure viewers are going to be looking for in papers. I think the tremendous success that our early customers have had in getting high-profile publications is going to be inspiring to a new wave of additional customers. So I feel really good about how spatial will be positioned as a 2021 investment area for anybody who's in biology. Look, I'm also encouraged by our President-elect, who is by far the most connected -- will be the most connected President to the field of cancer that we've ever had, right? I mean don't forget Joe Biden started -- lost his son to cancer, started a Moonshot program, spent a tremendous amount of effort over the last several years on that field. And while biological funding in science has always had bipartisan support, I think we could be coming up on a really positive era of investment in public health and science in the next several years. So I'm encouraged by that.

And then maybe just one more. I know it's too early to discuss kind of pull-through on GeoMx and kind of where that can go to. But since you brought up some of the strength, I guess, on some of the CTA customers. Is it possible to just describe like maybe what a range of heavy users are kind of using GeoMx in terms of a pull-through? Is there anything that would give us a flavor about kind of the range of outcomes that you're seeing?

There's too few of those customers to even give you a range, but I will say we have at least 1 6-figure order in terms of Cancer Transcriptome Atlas during the third quarter to a large study that was being initiated by one of our first customers to get access. And that's a really good positive side about the scale of biology that people are thinking about with this assay. I don't think -- I would not represent that, that's by any means typical. But it does represent a good, enthusiastic early adopter, and is encouraging.

[Operator Instructions]. And our next question comes from Tejas Savant with Morgan Stanley.

So Brad, just one quick question for you on instrument installs this quarter. Obviously, both nCounter and GeoMx look good. Can you just help us think through how much of this was a catch-up in terms of orders which you weren't able to ship in 2Q? Or was it just all sort of organic demand, and we should think of this as a baseline number, if you will, going forward?

Yes, Tejas. I'll tell you the Q3 instrument revenue was not a bunch of backlog shipping that had been accumulated during the second quarter. So we did -- it was -- if your question is, did we receive a bunch of orders in Q2 that we failed to ship for a whole host of reasons and then recognized in Q3, the answer is no. I'd say backlog entering and exiting the quarter was probably about the same. So what you're seeing in Q3 is current demand. So it is the best measure we have of what the current natural run rate for the business is in terms of instruments.

Got it. That's helpful. And then, just following up on your AMP comment. Can you share some color, Brad, on how you think about the clinical opportunity for GeoMx down the road? And to make it a little bit more specific, I mean, even with your launch of the protein assays this morning, you highlighted sort of a use case in prostate cancer at the Hutch. And so clearly, I mean GeoMx would make sense for those sorts of applications. But what remains to be done before the platform is ready for clinical prime time? And when that time comes, should we expect you to partner with someone like you did for the nCounter? Or will you go after that segment yourself?

Tejas, it's very early days in knowing exactly where GeoMx will fit in the clinical diagnostic realm or whether we'll choose to go alone or partner. What we're doing today is really focused on allowing some of our translational customers to do what they do: to translate their early protein-based assays into clinical assays and to build the case for why that would be more useful than traditional methods like immunohistochemistry. The most obvious use case of GeoMx would be to allow a pathologist to run a large number of immunohistochemistry assays that they normally either run sequentially on different slices or because they were tissue-limited, would not run at all in a single -- they run all those in a single assay. And I think that will be the first area we see our diagnostic users try. Not run a -- making a 10-plex immunohistochemistry assay in the clinic. That's the most obvious thing. But there are also potential for novel assays. Some of the first papers that came out on GeoMx, for instance, in the field of immuno-oncology showed that incremental predictive benefit could come from knowing exactly what cell type certain biomarkers were expressed on. So for instance, David Rimm from Yale showed PD-L1 expression on macrophages was more predictive than PD-L1 expression overall in terms of who responded to certain immunotherapies. So there's a possibility for novel assays to come out as well. And I expect the next several years as we build our installed base at academic medical centers and biopharmaceutical companies for there to be a whole host of innovation and discovery about diagnostic content. And we'll start by working with those centers to disturb -- to try to develop, rather, a laboratory developed test. And then we'll take our strategy from there, whether we want to ever have an in vitro diagnostic version of the system, and what role we would play in that remains to be seen.

Got it. That's helpful. And then apologies if I missed this earlier, but could you just give us a quick update on your informatics collaboration with Illumina? Is that in Early Access now? Or is that still ahead of you? And if it's already out in Early Access, any feedback from customers who are using DRAGEN for spatial biology?

Yes, that partnership continues to go well. I believe it's going to be going into Early Access in the weeks and months ahead. And we'll have more to say on that as we move through Q4. So it's a little too early for me to give you actual customer feedback on that, Tejas. But in terms of the collaboration between the 2 companies, we're confident this is going to help create a streamlined workflow, yes, we continue to feel enthusiastic.

Thanks, everybody, for joining us today. There should be -- the replay will be up on our website in the next couple of hours. In the meantime, if you want to dial-in for a replay, if you missed any part of the call, domestic dialers can dial 800-585-8367. International callers can dial 416-621-4642. The conference ID for both is the same: 5172128. With that, thank you very much for your time, and have a great day.

And that does conclude today's conference call. Thank you for your participation. You may now disconnect.