What this covers
This episode has Aneal walking through Dexcom's business from the biology up. He separates type 1 diabetes, an autoimmune attack that affects about 1.6 million of the 34 million US diabetics, from type 2, where sustained high glucose erodes insulin sensitivity and patients drift toward insulin dependence. He explains why finger-sticking once or twice a day misses dietary swings and nighttime lows, and why a CGM is so hard to build: a reliable electrochemical signal from tiny glucose changes, then a calibrated algorithm. The product case rests on a roughly 1.5-point A1C reduction and on behavior, since about 90 percent of users report changing what they do after seeing a reading.
The middle of the discussion covers market structure and technology. Dexcom feeds pumps from Tandem and Omnipod, forming an Artificial Pancreas Closed Loop, and once rescued Tandem financially. It targets insulin-intensive patients, about 4 million in the US, while Abbott Libre dominates cheaper, less intensive type 2 use. Aneal contends that Wall Street's focus on competition is misplaced and that technology and manufacturing, not regulation, are the real barriers. He also covers the failed implantable route (Senseonics) and the contested threat of non-invasive sensing, including infrared spectroscopy.
The last stretch is about money. The US spends about 300 billion dollars a year on diabetes, yet insurers balk at a roughly 2,000 dollar annual device cost, partly because patients switch plans within four years. United's Level2 program is offered as an example of cost risk moving onto the payer. He also covers the low cost of persuading prescribers, hospital CGM use with steroid-treated COVID patients, and a more speculative case for glucose as a general wellness signal.
Aneal argues Dexcom's CGM wins on measurement accuracy and ecosystem integration within a stable duopoly, and that its long-term value depends on owning patient data and proving health-economic savings to insurers, which are the real gatekeepers of its market.
- Dexcom is the most accurate CGM, and clinical A1C data plus pump integration make it the preferred device for insulin-intensive patients
- Insurers gate how much of the large TAM is realized, so health-economic data showing savings above the device cost is the key risk and lever
- Value accrues to whoever holds the patient glucose data, which drives algorithm improvement and ecosystem lock-in
Hard manufacturing technology, not regulation or competition, keeps Dexcom's duopoly stable and protected.
- Dexcom charges about 40 to 50 dollars per 10-day sensor, which costs about 15 dollars to make, roughly 1.50 dollars per day of cost against about 5 dollars of daily revenue, a gross margin near 70 percent.
“It's about 70% gross margin business.”
- The barrier to entry in CGMs is mainly technology and manufacturing rather than regulatory approval, since the latest generation was cleared through the relatively easier 510(k) pathway.
“The latest generation was approved under a 510pathway is much easier. I view the barriers to entry as actually much more focused on technology in manufacturing component.”
- Only three firms have made CGMs (Medtronic, Abbott, and Dexcom), but Dexcom and Abbott are the only two technically able to serve the market accurately, making it a duopoly, because Medtronic needs calibrations and Abbott lacks a radio transmitter to a phone.
“There's only two players that actually are technologically able to service this market accurately.”
- The core technical difficulty is generating a reliable electrochemical signal from very small glucose changes under the skin and converting it with a calibrated algorithm into accurate readings during rapid fluctuations, which is much harder than it sounds and explains the graveyard of failed CGMs.
“This is actually a very difficult technology to master because there's wild fluctuations in glucose based on what you're doing and based on what you're eating.”
- Dexcom's brand targets insulin-intensive diabetics who take basal plus mealtime insulin, about 4 million in the US and 10 million across Dexcom's current markets, while Abbott Libre dominates less intensive type 2 patients because it is cheaper and needs less real-time monitoring.
“So that's a market where the Dexcom really own.”
- Consumer CGM startups like Nutrisense and January do not make their own sensors; they brand Abbott Libre patches, so the category is an awareness game rather than a zero-sum fight, and CGMs remain prescription-only in the US.
“Those companies are actually all supplied by the Abbott Libre. They're using the same Abbott Libre patch.”
- Competition is not the main threat to Dexcom because manufacturing difficulty has created a graveyard of failed CGM attempts, so the market should remain a stable duopoly for the foreseeable future.
“competition is not one of them. This is a focus from Wall Street, but from studying the history of CGMs, there has been a graveyard of CGMs in development”
- The Wall Street focus on competition is misplaced for Dexcom, since the category is effectively a duopoly and the real constraints are manufacturing and insurance rather than rival device makers.
“competition is not one of them. This is a focus from Wall Street, but from studying the history of CGMs, there has been a graveyard of CGMs in development”
Payers drive CGM adoption and pricing, so Dexcom must prove health-economic savings to insurers.
- Of diabetes spending's three buckets (provider visits, pharmaceuticals, and end-stage complications), the complications bucket is the most important and scares insurers because it drives hospitalizations, amputations, and deaths.
“But the most important category is the third bucket, and this is what scares insurance companies. It's end stage diabetes where patients are progressing to hospitalizations and deaths.”
- UnitedHealth's Level2 program guarantees employers their diabetes cost will not rise over five years, moving cost inflation risk onto United, which then has an incentive to put CGMs in every diabetic's hands, since savings become profit.
“So now United is motivated to get CGMs into the hands of all their diabetics because ultimately they're on the hook for excess costs.”
- CGMs are paid through a mix of government and commercial insurance, with a commercial patient reimbursed a little under 2,000 dollars per year and a Medicare patient about 2,500 dollars per year, dispensed through pharmacies such as CVS.
“The product gets paid for a little under $2,000 per year for a commercial patient. And then for a Medicare patient, $2,500.”
- One third of Dexcom patients have a 0 percent copay versus about 10 percent for Abbott Libre, and Dexcom's effective price fell from about 3,000 dollars a year to just under 2,000 as it scaled, while Libre remains around 1,400 dollars.
“Dexcom gifts is one third of their patients have 0% copay. They say the Abbott Libre, only 10% of their patients have 0% copay”
- Insurers are the gatekeepers determining how much of Dexcom's theoretical TAM gets realized, so Dexcom must generate health-economic data showing benefits exceed the roughly 2,000 dollar cost for commercially insured patients.
“So insurance is really going to be the gatekeeper on establishing how much of that theoretical TAM is going to be realized.”
- Average non-diabetic healthcare spending is about 10,000 dollars per year (roughly 4 trillion dollars, 20 percent of GDP), but spending is skewed toward hospitalizations, surgeries, and ER visits, so one prevented hospitalization more than pays for a CGM.
“If a device can prevent one of those hospitalization events, it more than pays for itself.”
- Insurers fear negative media coverage of denied care, and the path to adoption runs through public awareness, employees telling HR they need CGMs for productivity, and employers weighing that in plan choices, so Dexcom must win over the public.
“The worst thing insurance companies are afraid of is someone goes to USA Today or Wall Street Journal or New York Times”
- Insurers focus on the immediate 2,000 dollar cost rather than long-run savings because the average patient switches insurance within four years, so a hospitalization event falls on another insurer's books.
“The problem is these insurance companies, they don't look at down the road savings. The average patient switches off insurance within four years.”
- Healthcare companies must benefit patients, doctors, and insurers together, since many firms fail by pricing a low-cost product for high margin without delivering the system savings insurers expect, while Dexcom lowers A1C, empowers patients, and generates health economic data.
“First, especially as a healthcare company, you have to benefit all your stakeholders.”
- To justify a low-cost product to insurers, Dexcom needs hard endpoints such as fewer hospitalizations and surgeries, and savings from reducing unnecessary drugs, since a CGM can reveal that a GLP-1 is ineffective and a cheaper drug is warranted.
“you have to show hard end points like decrease in hospitalizations and surgeries”
- Dexcom's device price is likely coming down to roughly $1,500 to $1,700, making it more palatable to payers.
“Dexcom price is likely coming down to probably closer to $1,500 to $1,700 price range”
The large, growing, underdiagnosed diabetes market leaves Dexcom with substantial headroom for growth.
- Dexcom has about one million CGM customers, roughly 80 percent in the US and 20 percent abroad, while Abbott has about 3.5 million, so about 4.5 million people worldwide use a CGM.
“Dexcom has about a million customers right now. Abbott has about three and a half million customers”
- Extending CGMs to the pre-diabetic population would quadruple the total addressable market, and hospital and pregnancy use cases sit outside even that expanded estimate.
“you have the pre-diabetic market, which would quadruple that TAM.”
- The core market of about 10 million insulin-intensive patients at roughly 2,000 dollars each is a 20 billion dollar opportunity, and treating every diabetic would imply about a 120 billion dollar worldwide market.
“So if you look at that core market, that's 10 million people worldwide at $2,000. So you're talking about a $20 billion market right there.”
- Dexcom is about 40 percent penetrated in type 1, about 25 percent in type 2 insulin-intensive patients, and only low single digits in non-insulin-intensive type 2, so the growth runway is significant within a duopoly.
“They're only 40% penetrated in the type 1 market. They're 25% penetrated in the type 2 insulin intensive market.”
- The US spends about 300 billion dollars a year on diabetes and related complications, globally about 700 billion, and the average diabetic costs the system about 17,000 dollars per year, projected to rise toward 30,000 over the next decade.
“the average diabetic cost to system is about $17,000 per year. And it's estimated in the next 10 years that that number is going to go up closer to $30,000 per patient.”
- Roughly one third of the 340 million Americans are diabetic or pre-diabetic, about 10 percent diagnosed diabetics plus about 90 million pre-diabetics.
“So in total, one third of Americans are pre-diabetic or diabetic.”
- Diabetes is fundamentally a disease of the blood and blood vessels, producing cardiovascular events like heart attack and stroke, nerve damage through damaged small vessels, and foot ulcers that lead to amputation and death.
“Diabetes at its core ends up being a disease of the blood and ends up becoming a disease of blood vessels.”
- Diabetes is becoming an epidemic because of worsening diets, more sedentary lifestyles, and broader behavioral shifts across society, alongside rising awareness and testing that uncovers more cases.
“It's worsening diet, more sedentary lifestyle, just the large scale behavioral changes that we see the way society is moving towards that's largely driving it.”
- Of the 34 million US diabetics, only 27 million have been diagnosed, and diagnosis rates internationally are substantially lower than in the US.
“Even though there's 34 million diabetics in the US, only 27 million of those have actually been diagnosed.”
- Type 1 diabetes is an autoimmune attack on insulin-producing cells, about 1.6 million of the 34 million US diabetics, while type 2 arises when sustained high glucose makes cells lose insulin sensitivity, so the pancreas overproduces insulin until it fails and type 2 patients converge toward type 1-like insulin dependence.
“So it's the loss of insulin sensitivity and ultimately loss of insulin production which actually makes type 2 patients at the end similar to type 1 patients.”
Patient-level glucose data is the moat powering Dexcom's algorithm, pricing, and later applications.
- Insulin manufacturers are not interested in buying CGM makers because they are pharmaceutical businesses, and Dexcom is staying in its lane because value accrues to whoever captures patient glucose data first.
“I think they realize that the value actually accrues to the party that gets the patient data.”
- More patient glucose data improves Dexcom's algorithm, which makes the device more accurate and allows more features like predictive alerts and personalized recommendations, so patient-level data is the power behind each later application.
“As you get more patient data, you develop a better algorithm.”
- The latest CGM generation alerts users when glucose is about to enter a hypoglycemic or hyperglycemic range, letting them eat a snack or walk to intervene before the event, and these predictive analytics are becoming central to Dexcom's story.
“they'll alert you when you're reaching hyperglycemic area”
- Dexcom's unique data on glucose responses to behaviors could be licensed to service providers that deliver diet and exercise guidance, or to insurers to price employer policies more accurately, creating a data-provider business model.
“Those are three alternative business models. They're all really interesting because they do expand the TAM”
Dexcom can grow through software, new business models, direct-to-consumer demand, and pump integration.
- Dexcom's software apps are currently free, and with the next-generation G7 the company believes it has optimized the hardware, so innovation shifts to software and predictive analytics.
“They believe they've optimized the hardware component and they're kind of reaching late innings of innovation there.”
- CGM is becoming more consumer-driven as patients ask doctors for a Dexcom or Abbott by name, prompting both companies to run direct-to-consumer advertising, with Dexcom's Super Bowl ad featuring Nick Jonas.
“Some patients are going into their doctor's office and they're actually asking for a Dexcom or they're asking for an Abbott.”
- Dexcom could move to enterprise-to-enterprise sales to insurers, sell its glucose data to service providers and insurers, or shift to a software subscription, each expanding the addressable market and changing the margin profile.
“Those are three alternative business models. They're all really interesting because they do expand the TAM”
- In healthcare the relevant cost is doctor acquisition rather than customer acquisition, and a clinical study presented at a major medical meeting can influence most prescribers in the country within a 30-minute presentation, making it very cost efficient.
“And in a 30 minute presentation, you have influenced most of the prescribers in the country. So it's very cost efficient.”
- Dexcom is not pursuing acquisitions and instead invests internally in the G7 launch and software, with the strategy focused on building an ecosystem of pumps, smart pens, and digital health partners.
“Right now they're not interested in doing M&A. The best use of their capital in their opinion is to invest internally”
- Enterprise and software subscription models would have lower gross margins than the hardware model, but higher operating margins because they require less selling and SG&A effort.
“While gross margins would be lower, in most of these other business models there would be much higher operating margins because it doesn't require any more selling”
- When Dexcom sends glucose signals to insulin pumps such as Tandem and Omnipod, the system automatically releases insulin as glucose rises, creating an artificial pancreas outside the body that reduces patient effort for type 1 diabetics.
“So essentially this device in the system creates an artificial pancreas that sits outside of the body.”
Ecosystem integrations make Dexcom's device increasingly integral to care and trusted by patients.
- As competing ecosystems grow, Dexcom's integrations with insulin pens, pumps, and digital health services make its device more integral to care, and the company is in early innings of building that ecosystem.
“I think in terms of building the ecosystem, they're in very early innings but that's where they need to be spending a lot of their money”
- Dexcom provided a lifeboat funding to Tandem, a near-bankrupt insulin pump maker, which became a key ecosystem partner, and Tandem's t:slim pump integration is a top reason type 1 patients prefer the G6.
“Dexcom actually decided to step in and provide a lifeboat to this company”
- Dexcom's partnerships with Apple on the Apple Watch, with Google on a small microchip, and with service providers like Onduo extend its ecosystem beyond the device.
“They've also made important partnership decisions with Apple on the Apple watch, with Google on creating the small microchip for their device”
- Dexcom CEO Kevin Sayer has over 20 years in diabetes technology, starting at MiniMed, the company that made the first CGM, and has been at Dexcom for several years.
“The CEO of the company, Kevin Sayer, he's been in this field for over 20 years. He originally started at MiniMed”
- In healthcare, reputation is everything: Dexcom built trust through clinical data showing it is the most accurate CGM and through an ecosystem where patients can rely on its device communicating with other devices.
“in healthcare, reputation is everything.”
- Through integration with smart insulin pens, a Dexcom reading can automatically calculate and display the insulin dose needed based on carbohydrate intake, reducing the manual calculation burden for patients on mealtime insulin.
“the Dexcom device can speak directly to the insulin pen, which is now a smart pen, and automatically tell you how much insulin you need”
CGMs outperform finger-sticks by catching glucose swings and changing patient behavior.
- A continuous glucose monitor is a subcutaneous patch with a small filament that sits under the skin, measures glucose roughly every five minutes, and transmits the signal to a phone or handheld receiver.
“It has a follicle like needle that goes underneath the skin and it will measure your glucose levels every five minutes and transmit that signal to a smartphone or a receiver”
- Most diabetics finger-stick only once or twice a day, often just fasting in the morning, which misses glucose swings driven by diet, exercise, sleep, and cortisol and fails to catch nighttime hypoglycemia.
“The problem is diabetes is a 24 hour disease. So your glucose levels will increase or decrease through the day based on your diet, your exercise, your sleep, your cortisol levels”
- Clinical trials show CGM use lowers A1C by about 1.5 percentage points, a difference that moves a patient from severe diabetes on a path to hospitalization to well controlled, delaying insulin treatment or hospitalization by years.
“The use of CGMs has resulted in a decrease of A1C by about 1.5%. That means the difference of being a severe diabetic that is definitely on the path to hospitalizations to being a well controlled diabetic”
- The biggest reason CGMs work is behavioral: when patients see glucose spike after a meal, they drink water, exercise, or eat more carefully, and about 90 percent of diabetics report changing their behavior in response to CGM readings.
“So this product actually works and the biggest reason it works is behavioral changes.”
- During COVID, hospitals used CGMs on patients given steroids, which cause glucose dysregulation, because nurses lacked time to finger-stick every patient to titrate steroid doses.
“So during COVID actually, a lot of patients in the hospital setting wore CGMs because the nurse didn't have the time to finger stick each patient”
A non-invasive glucose monitor is the second major risk that could displace current CGMs.
- The second major risk is technological: a non-invasive glucose monitor that avoids penetrating the skin could displace current CGMs, and prior attempts such as Novartis's contact lens and GlucoWatch failed on accuracy and skin irritation.
“The second big risk is technology. Dexcom and Abbott and the MiniMed device all go underneath the skin.”
- Dexcom originally aimed for a long-term implantable, but implants risk inflammation and immune detection that blocks the electrochemical signal, which is why Senseonics' implant has seen little uptake despite Dexcom and Abbott's strength.
“The biggest thing you worry about when you go underneath the skin is you don't want inflammation.”
- Infrared spectroscopy, which uses lasers to measure glucose below the skin, is a fairly nascent non-invasive approach that Dexcom is monitoring, and management must stay patient-centric and watch for disruption.
“One new technology that's being studied is infrared spectroscopy, which uses lasers to measure how much glucose is sitting below the skin”
- The host speculated the future is a combination of an Apple wearable and implanted devices pinging phones with glucose and other data, and Aneal countered that the Dexcom origin was implantable but the GlucoWatch-style non-invasive wearable would be the more compelling path.
“That would be awesome if like the Apple watch was able to pull glucose through.”
In choice-driven markets, patients shift to whichever patient-centric device serves them best.
- Medtronic, the former market leader, restricted its CGM to work only with its own pump, but its CGM required several finger sticks a day for calibration and patients wanted to mix pumps such as Omnipod and Tandem, so patients shifted to Libre and G6.
“Medtronic was previously the market leader in both CGMs and insulin pumps and they made the business decision that they wanted to create a closed ecosystem”
- Companies that prioritize business over patients eventually suffer, because patients and doctors choose the most patient-centric product when a category offers choices.
“If you're prioritizing business over patients, that ultimately does show up. Patients are very empowered.”
- Telling patients to reach specific glucose targets (for example from 140 to 110) rather than prescribing diet and exercise rules gives them choice and makes them far more likely to comply, because they can choose behaviors that suit them.
“the patient becomes much more empowered because they feel like they have choice”
CGM value extends beyond diabetes into pregnancy care and general health and wellbeing.
- Hormonal changes in pregnancy can push women with glucose dysregulation to dangerous levels, so Aneal expects OBGYNs may eventually recommend wearing a CGM for the full pregnancy.
“I can very easily see in the future where every OBGYN says, 'Just wear the CGM for the nine months you're pregnant”
- Glucose is a marker of other behaviors such as sleep quality, stress, and exercise intensity, so glucose monitoring has value for overall health and wellbeing beyond diabetes and may reduce healthcare utilization over time.
“glucose is a very important marker of other behavioral functions. It is telling you how well you are sleeping at night.”