Dexcom 3Q15 – Record revenue of $104 million grows 54%; G5 launches to overwhelming demand; new inserter, transmitter, receiver in 2H16; enthusiasm for Google partnership – November 4, 2015

Dexcom reported 3Q15 financial results this afternoon in a call led by CEO Kevin Sayer. We enclose the top business and R&D highlights, followed by a pipeline summary and Q&A.

Financial and Business Highlights

1. Dexcom’s product revenue totaled $104 million, an all-time record and first quarter of sales exceeding $100 million. Revenue rose 54% year-over-year (YOY) and 12% sequentially. Sales will exceed the top-end of the full-year revenue guidance of $350-$375 million (36%-46% YOY growth). In a bit of conservatism (or uncertainty), the call did not provide a new range.

2. Cash-based net income of $20 million (19% of revenue) in 3Q15 increased 100% from 3Q14’s $10 million. Encouragingly, the business’ profitability is growing faster than top line revenue, implying impressive operating leverage.

3. Dexcom launched G5 Mobile in late September in both the US and Europe. The pipeline of demand is “exceptional” and the highest in the company’s history. Inventory challenges caused delays in shipping G5 transmitters, though these will be resolved with no impact on Q4 revenue. G5 is approved down to age two years and the European label permits insulin dosing without confirmatory fingersticks (as we learned at EASD).

4. Direct-to-consumer advertising campaigns for G5 are starting this month and will run into 2016. Increasing CGM awareness in the patient and HCP communities is a major priority. We believe integration with smart phones will be seen as a major advantage to previous generations.

5. Notably, over the next three years, Dexcom hopes to move 70% of its business to the pharmacy benefit channel. Management quelled Wall Street concerns that the move to the pharmacy channel will result in lower pricing. Dexcom’s goal is to roughly maintain current pricing.

6. Dexcom is investing in expanded manufacturing capacity, the Google Life Sciences partnership, and growing its international operations. R&D expense more than tripled to $65 million in 3Q15 vs. $18.5 million in 3Q14 compared to revenues that rose 54%. The R&D for 3Q15 included a one-time $36.5 million charge (Google Life Sciences, via stock). Subtracting out Google, R&D climbed 54% YOY to $28.5 million, exactly the same as the revenue increase.

R&D Pipeline

7. The call did not share any updates on the collaboration with Google (announced in August), but did reinforce Dexcom’s serious commitment to this partnership, and optimism for a simple, low-cost, disposable, 10-14 day sensor system to drive CGM use from intensive insulin users all the way to prediabetes. The August announcement called for the first product to launch in two to three years, with a follow-on within five years. We believe this deal is in direct response to Abbott’s Libre and the company’s decision to work with a much broader group of patients.

8. In new news, Dexcom will officially launch a new insertion system, a new smaller G5 transmitter (“much lower profile”), and a touchscreen receiver in 2H16. The lower cost hardware will offer a better patient experience.

9. FDA discussions “continue to evolve” on an insulin-dosing claim for G5. This was more conservative than in 2Q15, where such a claim was expected in 2016. There is still uncertainty over the necessary pre- and post-market data, also a departure from more firm commentary in 2Q15 (“in sync” with the FDA on the necessary data). But it sounds like things are getting close.

10. Dexcom says it still plans to launch an Android version of G5 in 2016. The G5 approval news slated Android for “early 2016,” and we’re not sure if that’s still the case – presumably it is less certain. Dexcom is making “significant investments in next-gen apps.”

11. Unlike 2Q15, there was no commentary on the timing or product specifics of the 10-day wear G6 sensor. In August, management expected to begin a G6 pre-pivotal study later this year, with a pivotal study to shortly follow, an FDA submission in early 2016, and launch in early 2017.

12. There was no meaningful commentary on pump partners or automated insulin delivery. J&J and Tandem are “pleased” with the G4 integrated products, but these will “not be major drivers” of Dexcom revenue.

Honorable Mention: In 2H16, Dexcom expects early data from the DiaMond study (n=338, 24 months, testing CGM in MDI users) and European reimbursement studies – this will be very important new data and could provide important marketing fodder for the company.

Business and Financial Highlights

1. Dexcom’s 3Q15 worldwide product revenue totaled $104 million, an all-time record and first quarter of sales exceeding $100 million. Revenue rose 54% year-over-year (YOY) and 12% sequentially (the previous all-time record came on 2Q15 revenue of $93 million). These are yet again stellar results, marking 12 straights quarters of 49%+ YOY growth for Dexcom since the launch of G4 in late 2012. Interestingly, 54% YOY growth is Dexcom’s “lowest” since 1Q13 (49%), as YOY growth over the previous nine quarters ranged from 56%-102%. This is to be expected given the massively larger base of sales. The late September G5 launch was obviously not a driver of sales in 3Q15.

• Management pointed to several driving factors in the strong performance: “viral” demand from parents following G4 Platinum with Share, the Nick Jonas ad campaign (launched in mid-summer – this has been a big hit in our office), CGM becoming more readily accepted as standard of care, and the beginning of business shipping through the pharmacy channel (starting in 3Q15 with United and Anthem).
• International sales of $14 million accounted for 13% of revenue, a 33% YOY rise and an 11% sequential gain. The business is obviously still US-dominated, which provided 91% of the quarter’s share of growth. This may change in the coming years as international efforts expand and more EU reimbursement studies support the benefits of CGM. We believe the slower growth outside the US (33% vs. 54% in the US in 3Q15) also may reflect competition from Abbott’s FreeStyle Libre, although we doubt this is a major factor since Libre is still so capacity constrained.
• Management is “comfortable” that sales will exceed the top-end of the full-year revenue guidance of $350-$375 million (36%-46% YOY growth). In a bit of conservatism (or uncertainty), the call did not provide a new range. For context, Dexcom has done $270 million in sales in 2015 thus far. Assuming the fourth quarter represents a similar 32% of full-year sales (as it did in 2014), Dexcom could expect full-year sales of $396 million, or 54% YOY growth from 2014. That might even be a conservative estimate given what should be a boost from a full quarter of shipping G5. As a reminder, the $350-$375 million guidance was already increased in 2Q15 from the original guidance for $340 million-$360 million (32%-40% YOY growth) in full-year sales. The company may certainly try to exceed $400 million, though we would guess this would be a stretch goal.

Figure 1: Dexcom’s Quarterly Worldwide Product Revenue (2Q12-3Q15)

• Dexcom pre-released revenue two weeks ago in what appears to be an effort to alleviate Wall Street worry over payer relationships. Apparently, investors grew concerned that United and Anthem had trouble processing patients through the pharmacy channel. Management honestly called payer relationships “challenging,” but admitted such a pre-release will not be a regular policy – this was specifically intended to counter some of the Street’s rumors. The pharmacy channel is clearly a major company priority – the goal is 70% of the business in three years (see below) – though the new distribution channel obviously comes with a learning curve (e.g., prior authorizations, updating payer policies, etc.). Clearly that did not impact 3Q15 results as much as the Street feared.

2. Cash-based net income of $20 million (19% of revenue) in 3Q15 increased 100% from 3Q14’s $10 million. Overall, the company reported a substantial $42.5 million net loss, which included: (i) a one-time $36.5 million charge as part of the Google Life Sciences transaction (via stock payment); and (ii) $22.6 million in share-based compensation on the company’s rising stock price. The major positive is that the bottom line is growing faster than the topline: through the first nine months of 2015, sales have increased 55% YOY while cash-based net income has increased 155%. This is impressive operating leverage. Said management, “Our business model works.” Indeed, 3Q15 marks two straight quarters of 16% sequential cash increases (now at $113 million). 

3. Dexcom launched G5 Mobile in late September in both the US and Europe. The pipeline of demand is “exceptional” and the highest in the company’s history. Since the approval came earlier than expected in August, Dexcom launched G5 earlier than originally expected and didn’t have time to build up transmitter inventory. This resulted in some inventory challenges in shipping transmitters to new patients, though management is confident these will be resolved in the coming weeks with no impact on Q4 revenue. The upgrade program for in-warranty users (free-$299) has seen a much greater response than previous programs; however, the company only deferred $800,000 in revenue for transmitters that did not ship prior to the quarter’s end. As a reminder, Dexcom users out-of-warranty (1+ years since the last receiver) must get the full G5 starter kit (transmitter + optional receiver). 

• Dexcom is shipping two G5 transmitters to each patient, as the labeled (and actual) useful life is three months. The price per transmitter is consequently lower than with G4. Management did not give a specific price range, but said patients will truly need four G5 transmitters per year; by contrast, the G4 required only 1-2 transmitters per year (labeled for six months, but really lasted ~9-11 months). The result of lower pricing is lower margins on transmitters with G5. This was clearly a tough design choice, since a larger transmitter would last longer but take up more space on the body. We think it was the right move.
  
  • An advantage of a shorter transmitter life is lower-hassle upgrade cycles – when Dexcom comes out with a smaller, lower-cost transmitter next year (see below), most patients on G5 will move over to it within three months. Early G5 users’ transmitters will start running out of battery towards the end of this year, and we look forward to hearing patient reactions to the shorter useful life.
• Management reminded listeners that G5 is approved down to two years, and the European label permits insulin dosing without confirmatory fingersticks (as we learned at EASD): “The G5 Mobile System is designed to replace fingerstick blood glucose testing for diabetes treatment decisions.” Management said the latter is the “first ever” for a CGM, presumably excluding Abbott’s FreeStyle Libre since it is technically not “CGM.” We’d note that the G5 European label is pretty strong in that it does not ask patients to confirm sensor readings with a fingerstick in certain scenarios (rapid change, hypoglycemia). Of course, G5 still requires two fingersticks per day, so it is not on par with FreeStyle Libre’s compelling “no fingersticks” marketing.
• Management said the patient response to G5 “has been exceptional.” Read Adam and Kelly’s diaTribe’s test drive here. CEO Kevin Sayer noted that patients have shared “plenty” of advice on how to make G5 better. Some of the common complaints include glanceability of the app (must unlock the phone), Apple Watch integration, and Android compatibility. At DTM a couple weeks ago, Senior R&D Director Apurv Kamath said all were in progress. These high expectations are of course a sign of the times we live in!
• “With G5 mobile, we are now receiving millions of real-time data points each day.” Dexcom is continuing to invest in data, software, and apps, and G5 brought the launch of Clarity, an outstanding new web-based data management software. Clarity gives an intuitive retrospective view of CGM data without any need to plug in the receiver (assuming a patient is using the G5 app) – this was a highlight of Adam and Kelly’s diaTribe test drive.
  
  • Mr. Sayer said Clarity is geared to patients for now, and future versions will integrate more advanced HCP tools. We’ve heard anecdotally that some key opinion leaders want more functionality. Ultimately, Dexcom even plans to provide a Clarity payer platform – that is truly the power of data flowing to the cloud, though we wonder about the implications: “Oh, it looks like you’re actually doing well, so we’ll be taking your CGM away.” Conversely, it could help payers identify patients doing well and give them lower premiums, or proactively reach out to those doing poorly.
• “This is the first ever class III medical device without a printed user guide. The user guide is on the phone.” It was a compelling testament to the world medical devices are moving in – the G5 guided quick start guide simply comes right on screen when you open the app for the first time. Instructional videos are embedded right in the app too (e.g., sensor insertion). Management called it a very different user experience – our take too – particularly for kids who didn’t want to wear G4 but now tell their parents they want to wear G5. Said CEO Kevin Sayer, “It really starts us with a whole new future.”

4. Direct-to-consumer advertising campaigns for G5 are starting this month and will run into 2016. Increasing CGM awareness in the patient and HCP communities is a major priority. CEO Kevin Sayer said Dexcom is in a “much stronger position with HCPs than ever before.” But several recent days in the field have reminded him that there is lots of work to do. “CGM first rings true for many, but not enough,” he said. Shifting to pharmacy distribution could also make prescribing CGM easier and streamline the new-patient experience.

• With the planned product iterations, management is confident that 24/7 CGM will be viable for all US type 1s and many type 2s on MDI. Q&A elaborated that non-insulin type 2s could be a viable markets even with intermittent wear (e.g., eight sensors per year might be clinically more valuable than two fingersticks per day). Dexcom will of course have to do the studies to prove this, and talks are ongoing with Google to define what these might look like. Of course, if millions of people are wearing even a few sensors per year intermittently, that is indeed a big business (and certainly the route Abbott is taking with the blinded FreeStyle Libre Pro, currently under FDA review).   

5. “Pharmacy benefit is the future for us.” Over the next three years, Dexcom hopes to move 70% of its business to the pharmacy benefit channel. This was more specific than the “vast majority” expectation we heard in 2Q15. Management did not say what percentage runs through the pharmacy now, though it is still very early – the first deals with Anthem and United began processing claims in 3Q15, and we assume others will take time to get in place.

• As a reminder, the pharmacy channel offers serious advantages: (i) patients can pick up Dexcom sensors at retail outlets like CVS and Walgreens, offering lower out-of-pocket expense (similar to a drug co-pay) and faster time from the initial prescription to sensor pickup (important for new patients); (ii) less expense for Dexcom (e.g., no need to have a huge call center for reorders like in the DME world); and (iii) often lower prices for payers (since DME distributors mark up the price). It is rare to see a business decision that benefits all three stakeholders – the only losers in this deal seem to be the distributors.
• Wall Street is concerned that the move to the pharmacy channel will result in lower pricing. Dexcom’s fervent goal is to roughly maintain current pricing, and management painstakingly explained how that’s possible. In some cases, DME distributors are currently gouging payers, charging $90 per sensor when Dexcom actually gets $70-$75 per sensor. In those cases, Dexcom could negotiate a price higher than $70-$75 per sensor, but lower than what the payer is giving the distributor – a win-win! In other cases, Dexcom will have to give up some pricing to move things over to the pharmacy channel. Management’s ultimate goal is a neutral impact on pricing, and with higher potential volumes and less admin hassle, this seems like a business no-brainer to us. 
• There have been some small early hiccups as Anthem and United have moved to the pharmacy channels. Dexcom has seen some inconsistent application of prior authorizations and the failure of some plans to actually move from DME to the pharmacy (in spite of the new contracts). Management said the company is learning in this initial foray, and there were no concerns about the impact on new patient growth or reorders (certainly, the 3Q15 order numbers back that up, and Anthem and United are two large payers).
• The one area where pricing may change in the next few years is with the 10-day wear G6 sensor. Management said G6 pricing is still to be decided, and we could see it going either up or down. On the one hand, Dexcom might command higher pricing per sensor – 10-day wear only requires three sensors per month, so even a 15% increase in sensor pricing would still reduce total patient costs of wearing CGM (subtracting out one $70-$75 sensor per month). On the other hand, maybe Dexcom will keep the same $70-$75 pricing, since the marginal cost per sensor is presumably similar, and many patients currently use only three sensors per month anyways (i.e., they extend the wear beyond seven days).

6. Aside from the standard near-term pipeline investment, management highlighted three major upcoming spending areas: expanded manufacturing capacity, the Google Life Sciences partnership, and growing international operations. Comments particularly called out the Google partnership as a “chunky” investment in 2016. As a reminder, this deal included a one-time upfront $36.5 million payment (paid this quarter), milestone payments to Google of $65 million (payable in cash or shares of Dexcom stock), and incremental R&D expense going forward (not defined). On manufacturing, management is comfortable with current capacity, but wants to start investing in additional facilities now, particularly if foreign demand expands with better reimbursement (“in two to three years, the way demand is growing, we could butting up into capacity”). The next earnings call in January will give detailed operating expense guidance for 2016 for the first time.     

• R&D expense more than tripled to $65 million in 3Q15 vs. $18.5 million in 3Q14. This included the one-time $36.5 million charge pursuant to the Google Life Sciences agreement (via stock), additional payroll costs, and $3.3 million in additional non-cash share-based compensation. Subtracting out the Google payment, R&D still climbed to $28.5 million, a 54% gain. There is no question that Dexcom is making serious investments in the future pipeline. We also imagine that the company’s growing suite of apps will require constant updates and maintenance (new phone operating systems), increasing R&D expense on an ongoing basis.

R&D Pipeline

7. The call did not share any updates on the collaboration with Google (announced in August), but did reinforce Dexcom’s serious commitment to this partnership. Management called the objectives “far-reaching”: develop a simple, low-cost, disposable, 10-14 day sensor system the size of dime integrated into an advanced data analytics platform to drive CGM use from intensive insulin users all the way to prediabetes. The call’s comments centered most on the lower-cost advantage rather than the smaller size or disposable components. Management did not share any updated launch timing from the August announcement, which called for the first product to be commercialized in two to three years, with a follow-on product to be commercialized within five years. The partnership does not affect the near-term pipeline: G5 (just launched), the new insertion system and new transmitter (launching in 2016), and G6 (launch expected in 2017).

• The progress is still early (the teams are “just getting to know each other”), but we caught a glimpse of the potential product at EASD. We were pretty surprised to see Senior VP of R&D Jake Leach actually show the below picture in Dexcom’s corporate symposium:

• There are still many questions on what the ultimate product design and cost profile will look like. Will it be factory calibrated? Will it be accurate enough for dosing insulin? Will it give traditional CGM data (number, trend arrow, alarms)? Will it continuously send real-time data to a nearby phone? How much will it cost? How will it compare to Abbott’s FreeStyle Libre? Will the Google/Novartis contact lens be available by that time?
• The deal gives Dexcom access to Google’s IP and R&D resources, including engineers that they would be hard-pressed to find (much less hire directly), tremendous capabilities on data analytics (if Google doesn’t have it, who does?), Android expertise, and a clear product improvement pathway to expand beyond intensively managed type 1s and into type 2 diabetes and effectively all diabetes (including hospital, gestational, and prediabetes). Dexcom also gets a “cool factor” from working with Google, which certainly doesn’t hurt. Our August report on the deal discusses the partnership goals, deal terms, respective advantages, barriers to CGM use, competitive implications, and Q&A.
  
  • Dexcom will manage commercialization and retains all sales and distribution rights for the products developed under this agreement. Dexcom has an exclusive license to use certain Google intellectual property related to these products; management said in August that the “primary use” of Google’s IP is in the transmitter.
  • Dexcom paid Google $36.5 million upfront (in Dexcom stock), $65 million in milestone payments (cash or stock), and royalties (mid/high single digits) once annual product sales exceed $750 million). Overall, we see the deal terms as highly favorable to Dexcom (purchasing R&D with valuable currency [its own stock!], rapid product improvement, back loaded $) and Google (monetization of IP assets, low financial risk, high upside).
• We believe Dexcom saw the major value Abbott is building through FreeStyle Libre, and it will be fascinating to watch Dexcom emerge with a very competitive product (especially one that might compete on cost and price). The deal is a major victory for CEO Kevin Sayer and the Dexcom team, who remain focused on building great sensors and partnering to make them better and more widely adopted. We admire Dexcom for identifying external expertise in hardware/electronics, miniaturization, and data management.

8. In new news, Dexcom will launch a new insertion system, a new smaller G5 transmitter (“much lower profile”), and a touchscreen receiver in 2H16. This was the most specific timing on any of these pipeline projects, and it’s a major positive that the transmitter and receiver are lower cost AND offer a better patient experience. Management called the new insertion system “one of most complicated tasks Dexcom has ever taken on.”

• We first saw pictures of all these projects at JPM 2015 (see below), though aren’t sure if they changed in look and feel since then. Management does not see the new receiver as transformative (in the context of G5 mobile), though it will improve the experience for users that prefer a receiver over the phone. We wonder how much smaller the transmitter component can get prior to the Google product launching – the commentary from JPM suggested it will decline in height and volume, though it does not look dramatic.

9. FDA discussions “continue to evolve” on an insulin-dosing claim for G5. This was more conservative and uncertain than in 2Q15, where such a claim was expected in 2016. There is still uncertainty over the necessary pre- and post-market data, which is also a departure from more firm commentary in 2Q15 (“in sync” with the FDA on the necessary data). Management said things are getting close, and we assume a clearer update will come in 4Q15.

• Comments implied the timing and calibration of G6 is very wrapped into this discussion – management stated it does not want to obtain a dosing claim for G4/G5, only to go back to adjunctive use with G6. For instance, if G5 obtains a dosing claim on two fingerstick calibrations per day, but then G6 launches with one fingerstick calibration per day (as implied in 2Q15), the dosing claim may not transfer over. In other words, Dexcom seems to be juggling two factors: (i) the necessary data for a dosing claim; and (ii) the calibration scheme it should launch with G6. That is a very tough call, though getting the dosing claim as soon as possible seems like the bigger business upside. The 2Q15 call talked about different labeled G6 products as Dexcom operates within the labeling for insulin dosing off of CGM – perhaps G6 would launch with two fingersticks calibrations and then go down over time.
• The T1D Exchange’s REPLACE BG study (n=250, six months) is testing insulin dosing off the G4 Platinum alone vs. the G4 Platinum + BGM, with estimated completion in May 2016. The study started in March 2015, though we’ve never been clear on whether this trial would bring the necessary pre-market data to support FDA approval of a dosing claim. It does seem like the study is using the updated Software 505 (G4AP) algorithm (two fingerstick calibrations per day), meaning the data would apply to G5 too. The primary endpoint is time-in-range, though secondary endpoints include severe hypoglycemia and change in A1c. It’s great to see the study is long enough and potentially large enough to capture meaningful changes in events meaningful to payers. We give the study good odds of  showing that CGM alone is non-inferior to CGM + confirmatory fingersticks – it would be incredible to show superiority, and that’s not unfathomable given the added impact of hand washing and meter inaccuracy.

10. Dexcom still plans to launch an Android version of G5 in 2016. The G5 approval news slated Android for “early 2016,” and we’re not sure if that’s still the case. As a reminder, the initial G5 app is only available for iPhone and iPod touch. The Follower app for remote monitoring is still available for both Apple and Android.

• Dexcom is making “significant investments in next-gen apps.” Management did not say anything further, but we learned at EASD that this may include real-time therapy support “like an insulin dosing calculator.” Dexcom has really pressed forward on the data ecosystem and we have high hopes for some meaningful apps from the company and its growing list of partners.     

11. Unlike 2Q15, there was no commentary on the timing or product specifics of the 10-day wear G6 sensor. In August, management expected to begin a G6 pre-pivotal study later this year, with a pivotal study to shortly follow, an FDA submission in early 2016, and launch in early 2017. The previous expectation was to launch G6 with one fingerstick calibration per day, and a gradual reduction over time. Today’s call still suggested G6 will not be factory calibrated initially, but management still believes this is possible with this future sensor and algorithm. What is less clear is whether the incremental approach stems more from FDA caution, cost-effective manufacturing a factory calibrated sensor, or both.

12. There was no meaningful commentary on pump partners or automated insulin delivery. J&J and Tandem are “pleased” with the G4 integrated products, but these will “not be major drivers” of Dexcom revenue. This continues several quarters of muted management commentary on this front. Dexcom has obviously committed to working with pump/artificial pancreas partners (e.g., J&J, Tandem, Insulet, Bigfoot, Bionic Pancreas), though the muted expectations are probably an indicator of the company’s ambitions – make CGM far bigger than pumps (our speculation).

Honorable Mention: In 2H16, Dexcom expects early data from the DiaMond study (n=338, 24 months, testing CGM in MDI users) and European reimbursement studies. These trials are critical to show the therapeutic and cost benefits of full-time CGM use regardless of insulin delivery method. Dexcom has always fought an uphill battle with some HCPs, as many prescribe pumps first or suggest CGM to those only on pumps.

Sensor and Software Pipeline

Pump Partner Pipeline

Questions and Answers

Q: A question on operating expense trajectory. You've been growing $10-$11 million sequentially each quarter this year. Do you expect to continue at the pace for the next several quarters or should things taper?

A: One of the big growth factors obviously has been our non-cash expenses, particularly the stock compensation because the stock has performed so well. Those numbers were much bigger than we'd anticipated and certainly much bigger than we'd had a year ago. As far as the increases in our quarter-to-quarter spending, again, there are moving pieces and we continue to make investments. A perfect example is the investments we made in the DTC campaign this quarter, where we spent in excess of $2 million just getting geared up for the start. And there are other things that we've spent on the R&D side. Next year we're going to be a little more transparent on expenses and a little more firm as we give our guidance at the start of the year. As I've said in my remarks earlier, we've certainly increased our cash based net income at a more rapid rate than we've grown. But we've also turned around and spent what we can to position us for future periods and that's where we sit today.

Q: What has been the early user experience of the G5 device? I know Bluetooth can be a little trippy. What are you hearing from customers? And what are you seeing in terms of pipeline and demand as the launch progresses?

A: Well, pipeline and demand has been exceptional. Our opportunities coming into our system every day are much higher than we've ever experienced before, so demand has been great. As we've gone through and looked at the data – other than when it appears that people have walked away from their phones – our data capture rate is pretty much equivalent to what we had with the old receiver. We’ve learned a few other things. For example, a lot of patients don't experience the 30-foot range that they had with the G4 receiver with Bluetooth. That Bluetooth radio is being used for many things on the phone, so it depends on how people use their phones. So we've had to really gear up our tech services people for calls that are different than what they've had to deal with before. We've learned a lot. And we have software upgrades coming in the future that we think can address any concerns we have.

The experience with the patients – as far as just looking at the phone, and the interaction – has been fabulous. I mean, this is the first ever Class 3 medical device without a printed user guide. The user guide is on the phone. If somebody has a question, they can touch the screen. They can get the user guide, they can watch videos on their phone on how to insert the product. It has been a great experience from an actual interface perspective, and very well accepted, particularly by our young patients.

Q: Can you give us more color on what drove third quarter trends? It came in well above where The Street was thinking, and this was almost entirely without the G5 impact. And in terms of the G5 launch, this is the first time that pediatrics will have access to the AP algorithm. What are you seeing there so far in the early launch and how do you expect it to play out over time?

A: What we’ve seen since the pediatric launch of G4 Platinum and G4 Platinum with Share was a shift from endocrinologists pushing patients to get the product to it going viral – patients and parents of patients becoming our biggest advocates. The Nick Jonas campaign is hard to measure, but I think that had some impact as well. CGM is quickly becoming standard of care.

We do know that the physicians are happy. And at the beginning of the quarter we started shipping much more of our business through the pharmacy channel with United and Anthem. That's starting to have an impact as well.

Q: It’s been several months since the Google partnership was signed. How much incremental R&D spending might we see per quarter going forward from it?

A: We don't really have an incremental number to give you for Q4. We will give you very firm guidance on that the first of the year. It hasn't been that many months and we're really starting to get going now.

Q: On the PBM side, 3Q15 is the first quarter moving to Anthem and UNH. What’s the percentage of patients now that are currently through that channel and where do you see that going in the next 12-18 months?

A: We've hit our internal goal of covered lives and we have seen an increase in the number of patients that have gone over, but we're not going to give a whole lot more color than that today. Our goal is to move 70% of our business to pharmacy benefits as the primarily reimbursement source over a three-year period.

Q: It’s still early obviously with the G5 launch, but you highlighted looking to put a new receiver on the market in the second half of next year. How should we be modeling a new receiver launch?

A: The new receiver launch is strictly a function of coming up with something that is more modern than what we have - touch screen, a better experience for patients. It will be lower cost. The revenues aren't going to change. There are some patients, as we've learned, that would rather look at the data on their receiver than on their phone, so this will provide a much better experience for them, but it's not going to change much of anything.

Q: What drives your confidence that manufacturing and inventory constraints will not have an impact on the fourth quarter from a new patient starter revenue perspective?

A: We're catching up on the manufacturing side. This was largely an issue of getting enough components in place. We're building at full capacity right now. The only way we'd have a problem by the end of the quarter, barring something unforeseen today, would be if it was just a true mega-quarter beyond anything we could comprehend, so we feel pretty good about it. This isn't anything other than we launched G5 earlier than we'd planned and we took the month that we were supposed to be building up inventory and shipped everything instead. So we put ourselves a little bit ahead of schedule. But we'll catch up.

Q: On operating expenses, you walked through a fairly long list of investment priorities. Would it make sense if there were just one or two or maybe three of those to call out that are going to be chunky pieces of incremental spend into 2016?

A: We’ve always continued to invest and continued to spend on the core technology development. In 2016, Google could be chunky for example. We are looking strategically to make some additional investments in our foreign operation. We have to start investing in Class III regulated manufacturing facilities now in anticipation of growing demand in the next two to three years. Those are the kinds of things that will be chunky. And as Kevin mentioned, we're going to do our best to give you some much better OpEx guidance for 2016.

Q: You talked a little bit about the pricing dynamics in the shift to the pharmacy, and that is a common question from investors. Can you give us a little bit more on the mechanics of how that changes? And how you might be actually realizing a higher price as you shift into that channel?

A: As an example, we have seen contract pricing between payers and our independent third party distributers to the tune of $90 a sensor when we've set our ASP at $70 to $75. When you look at a contract like that and go to a payer and say, look, from a cost perspective we can get you this as a pharmacy benefit at a price much lower than $90, it's pretty easy to show how that's better for the payer. On our side we're selling the stuff to third party distributer at a discount. When all is said and done, net to Dexcom in some cases can be better, in some cases it will be lower. But we believe at the end we're trying to keep this as even as we possibly can, and we evaluate all the pricing opportunities very carefully.

Q: So should we be thinking about the potential for some price benefit in addition to the underlying unit volume that you are generating as we think about the next couple of years?

A: Our goal is to remain even on the price perspective as we push out the distribution channel to the pharmacy benefit, and that's how we look at the contracts. Where the price benefit may change in future years and future models is with an extended-wear sensor like we're planning for G6 when we go to 10 days, but that has yet to be determined. We expect pricing to come down in the next five or ten years. One of the key driving factors behind the Google deal was to develop products that are much, much lower costs for us to enable pricing to come down to make the technology more accessible to people. I don't see that happening in the next two to three years.

Q: It looks like you have a lot of leverage available in the model. As we look at the next year or two, should we think about the increased spending that you're talking about as maybe taking away some of that leverage? How should we be thinking about matching the incremental expense with the revenue growth that you're generating?

A: As we're in the middle of our planning cycle for 2016, that's our great debate here internally. Steve identified a couple of chunky areas as far as investing in additional manufacturing capacity, our Google relationship, and data analytics. What we're going to do as we start next year is really break out those areas that are going to be above and beyond the spend in the core business. There is a lot of leverage in the model. There's no question that our business model works. We want to get big, and we want to serve more patients. And so we're going to have to make some investments. And as you look at the people who are coming to our markets, A, they're really big. And B, they don't have to report this level of granularity as far as their spend into what they're doing and their activities. So we'll give you more guidance. But as we continue to do some of these other things, as we outlined, we're going to have to spend some money that's outside and above the norm and that's okay because that's preparing us for the future.

Q: Circling back to the investment comments from a sales and marketing perspective, is it still fair to assume you're going to add maybe 20 or so reps in 2016? Could it be a more significant expansion? Or do you think at this point you're getting more bang for your buck spending it on DTC or R&D?

A: On the commercial side, they're going through their plans right now. Our more recent investments have been made on the channel side with respect to the distribution channel, with respect to payers, with respect to programs to go out to pharmacies to prepare pharmacists to market and to sell this product. So we know there's going to be an investment there for next year.

We’re getting tremendous bang for the buck out of our team. They're doing a fabulous job. We are investing in a lot of new tools with respect to IT infrastructure and with respect to things we can do to make that process go faster with our online store.

The R&D spend for us is almost completely independent of the commercial spend. There's really, really near-term products that are going to come as we talk about the applicator and the new receiver for example. And the applicator also has a new transmitter that is smaller and lower cost, so you see there's a new hardware configuration and that's a very short-term 2016 event. Then you get into more near-term things like clinical trials. And then we've got longer-term sensor R&D investments. So we will spend money in a very controlled, structured spend to drive the commercial business.

Q: Looking at the shifting hardware model with G5, can you give us an idea on where ASPs are heading for G5 transmitters?

A: I'm not going to give you a specific ASP, but the transmitter now carries a three-month warranty unlike the G4 transmitters because of the increased power requirements of Bluetooth. Patients are going to have to go from two or maybe less than two transmitters per year to close to four transmitters per year.

But we're not anticipating that the payers are going to pay us; the old ASP was probably around $400, $450 for a transmitter. We're not expecting that we're going to receive that four times a year. So the pricing is going to come down now, and that will be somewhat offset as we ramp up the G5 launch by increased volumes. So the margins will end up being better over time because of the increased volumes.

Q: Is there a difference that you've seen this quarter in terms of sensor utilization?

A: No, I think it continues to remain very strong. Everywhere I go I hear “people get on this and stay.” I mean anecdotally we hear all sorts of things. Out there we heard one patient say the data wasn't very good, and when we asked her how long she had worn the sensor she was on week six. Well, for a 7-day labeled product week six is pretty good. And so we're going to learn a lot more and have a lot more understanding of those patterns but no, utilization is good.

Q: If you are shipping G4 transmitters two at a time, then are patients still ordering just twice a year, or getting scripts twice a year? Is that how it works, and is that how the insurance companies understand it?

A: Insurance systems are never really up to date, and we're always experiencing challenges. But no, the fact that we're sending two means that the patient's probably not going to have to come back for another six-plus months.

Q: You said you now expect revenue to exceed the top end of your previous range, but you haven’t provided a new top end. Do you expect flat sequential results or a decline in Q4? You've targeted a long- term growth rate of 40%, but all your commentary today – as well as your results – suggest your business is actually accelerating. Is there any reason you believe growth will decelerate below that 40% in Q4?

A: No, we're sticking with our guidance, what we said. We updated our guidance last quarter to a top end of 375, which was north of our original stated guidance of 35% to 40%. We didn't say that we expect a sequential decline, we didn't say that we expect sequential flat. We said we are comfortable that we're going to exceed the top end of our 375, and we're going to leave it at that.

Q: Can you give us a sense of how patients are responding to the shorter transmitters? Can you provide more color on the new transmitter that will launch, along with the sensor applicator, in the second half of next year? You said, smaller and lower costs, should we also be looking for another shortened useful life?

A: No, the useful life will remain the same. And they really haven't had time to respond to a three-month life, because we've only been shipping for a month and a half, so we haven't heard anything on that front yet. There are advantages and disadvantages to a three-month transmitter life. Obviously, the disadvantage is that you do have to buy more of them. The advantage is, when we get something new, that upgrade cycle is going to be really short because no matter when you bought your last transmitter, within 90 days you can move to the new one.

Q: In the past you guys have talked about 40% sustainable growth going forward and obviously you've been growing much better than that this year and last year. How should we be thinking about that longer-term?

A: We continue to plan for that sustainable 41% number. These past several quarters of 50% plus growth have been great, but just by the law of increasing numbers some of those percentages will have to come down. There’s still a whole bunch of market to go penetrate and get in our type 1 market. We still don't have the penetration that we need. There's a large international market for us to go capture. Growing 40% on last year's number is certainly much tougher than growing 40% on the number from the year before. So the law of big numbers eventually gets to you a little bit. But we're very optimistic about our business future and we're not slowing down at all.

Q: As you think about the total addressable market, what is a realistic penetration rate for type 1 patients assuming you're successful in continuing to evolve the technology? How big do you think this can get for type 2s?

A: For type 1 in the US, if we're successful with the product iteration that we have, there's no reason that every type 1 in the U.S. who has some form of insurance wouldn't use CGM. Why would you prick your finger six, seven, eight times a day if you're a type 1? If you have to monitor, wear a sensor.

The type 2 market is still very much evolving. You have to segment the type 2 market. There are intensive insulin using patients who are taking mealtime injections of insulin. I think the penetration there could be quite extensive.

If we look to the broader type 2 market, or the pre diabetes market, the obesity market, it's still to be determined. I think we can have meaningful impact there. I don't necessarily think that these are going to be patients wearing sensors in real time all the time like our type 1s or intensively managed type 2s, but as the standard of care becomes monitoring, we might see them wearing several sensors a year.

A: We've got to develop the clinical evidence that says you're better off as a type 2 patient wearing four to eight sensors a year than you are sticking your finger twice a day. We believe that's very doable and very true.

Q: Last quarter I thought I remember you saying you expect the dosing claim by next year. Is there a change there? Can you clarify that?

A: I'm just being a little conservative with our expectations. We are getting close to defining exactly what we have to do here.

Q: Is it the new receiver that's coming next year? Or is it a new transmitter as well?

A: With the new insertion system there will be a new transmitter that has a much lower profile, and a new receiver. So it'll be a big hardware year for us next year. Hardware and new apps and things of that nature will be our big product launches next year. We’ll give you more color on that first of the year when we provide our 2016 guidance.

-- by Adam Brown, Ava Runge, and Kelly Close