Insulet Corporation (PODD) Earnings Call Transcript & Summary

Larry Biegelsen, the medical device analyst at Wells Fargo. And it's my pleasure to host this fireside chat with the management team from Insulet. With us, we have Ashley McEvoy, President and CEO; and Flavia Pease, Executive Vice President and CFO; and Claire Trachtman, Head of Investor Relations. Ashley and Flavia, thanks so much for being here.

Thanks, Larry. Pleasure.

Thanks, Larry.

So Ashley, it's hard not to start with the Q2 update. So -- or update. So let's start there. You took down guidance because of 3 issues: lower retention and utilization in type 2. Lower new starts due to softness in Q1 and less pricing. 2/3 of the change was due to the type 2 issue. And probably the #1 question we've been getting, and my guess is you, why was this such a surprise?

Yes. I mean, Larry, first of all, it's great to be here in Boston, our hometown. Before I get into the timing and the data, let me just first share our perspective on the type 2 market, which we see as a tremendous growth opportunity for Omnipod as well as the AID category. It's the largest TAM in the category with the lowest penetration. If you look at the 2.5 million people who are using basal/bolus insulin, AID has around a 5% penetration and CGM is well north of 50% plus in that area. In the basal-only population, we have the indication. We haven't yet promoted that and CGM is around 30% penetration. So collectively, that represents about a $28 billion TAM. The second is the ADA guidelines recommend AID for people with type 2 diabetes because it has a superior clinical and economic value proposition to MDI patients. We pioneered this indication on behalf of the industry. We have tens of thousands of very satisfied podders with type 2 diabetes. And so we are very committed to nailing the first 90 days, so more people can benefit. And as you know, we enjoy a reoccurring revenue business model. So it's really about the lifetime value of the patient. Now when it comes to timing and the data, Larry, we saw late in quarter 2, some of the trends start to elevate versus our expectations. We did a very strong forensic. We identified with the root cause. We've implemented actions on those. One of the areas that we've innovated on is really the transparency and the frequency of some of our KPIs. Attrition, as one example, is a bit of a lagging indicator. It allows for people to have a pump holiday, if you will. And so we've addressed that of looking at some of a more real lifetime data of our KPIs. I will tell you, it's early on. It's been less than 30 days, and some of these programs are going to have time before they have a durable effect. So in the headline, I'd say this is an execution challenge, nothing to do with the structural advantage or attraction of the market, and we were acting with urgency.

That's helpful. I mean why was there an uptick? You just mentioned a minute ago, an uptick at the end of Q2. Why do you think there was an uptick?

Yes. We -- I would say it was -- it happened. We saw it trending ahead of our expectations at an elevated rate in quarter 2. And we've identified several root causes of why we saw some elevated attrition levels, and we're taking actions. Those 5 actions I spoke about on the quarter 2 call, I'm happy to elaborate more.

That's helpful. We definitely want to hear about the actions and the progress. Just with first, the math, historically, we've had about a 10% attrition rate in our model for Insulet when you were mostly a type 1 company. You also on the call confirmed that type 2 attrition runs at least double type 1. So it looks like the type 2 attrition is, by our math, close to 40%, which is pretty high. Are we in the right ballpark? And again, maybe now talk about the progress you're making trying to addressing that.

Yes. I mean we -- again, we've been at type 1 for 25 years. We're newer to type 2. When we were setting our expectations, we studied the CGM analogs. Again, they've been at serving the type 2 community, both basal/bolus as well as basal-only for more years than we have. We looked at what good would look like. And we -- again, as I said, we saw an elevation versus that. They're not near the levels that you mentioned, Larry. But importantly, we're getting after that. And again, we have a recurring revenue model, so it's worth it for us to nail the first 90 days. As I shared, the demand is really strong, and the data will show that once we get them past 90 days, retention rates really stabilize. So there are several things that we're doing. The first thing is sampling. So we deployed a sample program to, kind of, demystify the fear of trying AID therapy. The vast majority of the volume in AID is coming from the MDI category. And so for some, there's a fear factor of wearing insulin on your body. So we deployed samples. We uncovered when we did a territory analysis that in certain territories that we had a prescription that accompanied the sample. And when we adjudicated the health insurance, we had much higher retention rates. So we've tightened up that best practice and are now deploying that across the country. The second thing we did is related to our sales force incentives. Again, going from a type 1 company to a type 2, we needed to evolve. We are still ensuring that we have good quality new customer starts, but we're also going to hold them accountable for durable starts, that first month, that first 30 days. I was in the field last week. They're very enthusiastic to win with the company on patient success. We've augmented our customer service reps to make sure that we have the right, kind of, resourcing and staffing. We've also added clinical retention specialists that we're now deploying in the field. And then we've launched a new data platform, which enables really strong patient engagement and clinician engagement, and that's called Omnipod Discover. And in our limited market review that we were doing in quarter 2, we had around 12,000 people with diabetes in the program and around 1,600 clinicians. We saw not only just stronger customer satisfaction, but we saw elevated retention levels. So it's worth it for us to nail the first 90 days. We see a tremendous opportunity of growth ahead of us.

That's helpful. Okay. I have to go back to the 40%, which you said was too high because we've used 10% for type 1. And when we talked after the Q2 call, you said that the blend of U.S. attrition is likely higher than mid-teens. So we just tried to back into the 40%. So just -- so help me understand where my assumptions are wrong.

No, I think we said that it's more than double. It's trending more than double that. And we don't disclose what our retention rates are now. It's something that we're looking at maniacally, so I have confidence in that. We're going to talk more about how these programs are progressing on the quarter 3 call. But we -- importantly, we're looking, Larry, at what our active customer base. All of our data goes to the cloud. So we can see lifetime the health of what our active customer base is and then see the flows in and out. So you'll see us continue to start to drive the business more on real lifetime data sources.

Blended attrition rate, mid-teens, is that a good number? Because that's the number people have in their -- we use in our model.

I think what I would say is the 10% on type 1 is what we've talked about before. And as we said, type 2, we had already assumed that it would be higher than that. It was elevated versus our assumption, and it's not as high as the 40% that you disclosed.

And one on -- basically, your competitors are saying they're not seeing the same issue with type 2 retention. You've probably heard that. Any -- what do you think the disconnect is?

Yes. I mean we were the first -- I think they're very early on. We have a little bit of a head start from them. We were the first to pioneer the indication. Again, we have tens of thousands of very satisfied people with type 2 diabetes using OmniPod. We have very strong demand. We did some of this when we didn't have really tight governance on our sampling. So we're addressing that. And we're -- I would say we're ahead of the curve of making sure that we deliver delightful customer service the first 90 days and all of our commercial leaders line up to what company success looks like.

Okay. And your guidance for 2026 assumes no improvement from the Q2 trends. Do you think you can start to see a benefit in 2026 from the actions you're taking and therefore, the updated guidance is fairly derisked?

Larry, I think the way we think about guidance is we took a prudent approach. As we said, we are not assuming, as you just stated, any impact of these measures. They will occur over time at different speeds. Some of the things that Ashley mentioned, the change in the commercial incentives and the change in sampling are already being executed, but we expect that they will have an impact by the end of the year. Discover will go into full market release next month. So the impact is probably more next year. And similarly, with the customer care and retention specialists, we are ramping up their onboarding. So it takes some time to get them onboarded, to get them productive. So the impact will be more next year. So that's why the approach to the guidance is to assume that we are executing on these strategies, but that the impact won't be felt this year and is not contemplated on the guidance.

Got it. More of a maybe check the box question, but GLP-1s continue to be a hot topic, especially for the type 2 population. What gives you confidence that it's this execution issue that you talked about, not GLP-1 usage showing up in your base once patients are a few months in?

I mean, we saw very healthy demand, as I talked about on quarter 2 and year-to-date. And again, they're staying post 90 days, retention stabilizes. So we don't see the use of GLP-1s having an impact on the first 90-day experience. I mean we view GLP-1s as very complementary and not a substitute to insulin. The data and the evidence would show that it can slow the progression by about a year to when insulin is initiated. But we've seen very stable rates of insulin initiation over the past 2 decades where GLP-1s have been indicated for diabetes. And the unmet need is that the vast majority of people with type 2 on insulin, they are using a GLP-1, but predominantly, their primary source of insulin delivery is MDI. And 2/3 of those are not achieving glycemic control. And so that's why the ADA guidelines recommend AID as the standard of care. And we think that there's a tremendous opportunity ahead for the category as well as Omnipod as a market leader to take advantage of that opportunity for the 5.5 million people on insulin and the vast majority of them not benefiting from AID therapy.

All right. Helpful. So we'll move on.

Sounds good.

So just next topic is quality and manufacturing. Manufacturing scale has been a key advantage for Insulet, but there have been concerns about quality as you scale. You've had 2 Class I pod corrections this year, and investors are taking a closer look at public adverse event data, the MAUDE database, which we know is flawed.

You should say that again, but...

But look, we're getting the question, and I'd be surprised if you're not getting the question. So has anything changed actually operationally? And why are you confident these issues are isolated rather than indicative of a broader quality trend?

Listen, as I mentioned, we weren't proud of those 2 medical device corrections. We did identify root cause analysis. We've implemented interventions to address that. We have a very rigorous post-market surveillance program. I see no new signals or no new quality events. There were no reported fatalities related to those 2 MDCs. I'm very familiar with the MAUDE database. It's helpful for a surveillance tool. It does not establish causality in adverse events nor does it establish increased -- a change in rates. And any kind of MDR that was associated with fatality, we've assessed it, and we haven't identified any kind of issue related to Omnipod 5 safety.

Okay. Fair enough. I'd be surprised if you're not getting the question, too.

Absolutely.

So I had to ask.

Of course.

You gave some helpful color on 2027 on the Q2 call. I think you said you expect growth next year to be similar or better than the Q4 exit rate, which is mid-teens. And so we're assuming U.S. grows about 10%, international grows about 20%. Is that generally how you're thinking about the base case for next year?

Yes. I think the way you took our framework for 2027 is the correct approach, Larry. We talked about for both total company U.S. and international using the exit rates of the fourth quarter as a consistent way to think of 2027. We're not suggesting any acceleration of growth next year. If you look at the updated guidance that we provided in the second quarter, it would suggest an exit rate for the total company between 12% and 17% on a constant currency basis. You can apply the same math for international and the U.S. Let me provide a little bit more color on the U.S. and international, what it means on those ranges. The U.S. at the low end of the guidance would be at a 9% exit in 2026 that would contemplate essentially flat-to-down NCS growth next year, a mid-single-digit share loss of NCS and attrition and retention rates consistent with where we are today and price to be stable versus where we are today. At the top end of the range for the U.S. would suggest an exit of 14%. Same assumptions around NCS, but it would contemplate attrition and utilization improving slightly and price to be slightly positive. So those are the building blocks as you think about the range for the U.S. When I turn to international, we see continued momentum and opportunity internationally, but at a more, I would say, moderate rate. And the reason for that, we've been talking about we are anniversarying the launch of Omnipod 5 in most of our large markets in Europe. In addition to that, we have been benefiting from price/mix realization as we convert from DASH into Omnipod 5. We're about 75% converted, Larry, on that installed base. So that price/mix realization will slow down a bit. Having said that, we still have a tremendous opportunity internationally. It will happen over time. We just launched in Spain, as you know, in July. We're also continuing to grow in Saudi and the Middle East. We have launches planned for Libre 3 in a couple of our large markets in Germany and in Canada. So all of those will continue to fuel momentum in our international business, but it will be at a more moderate pace. So to summarize it, we see our framework for 2027 as a jump-off point, consistent with the growth rates that we will exit 2026.

That's helpful. What's the range for international that exit rate you...

19% to 24%.

19% to 24%. And going back to the U.S., the mid-single-digit share loss for both scenarios. Why is that the right -- what's the thought process behind that? What are the assumptions? There's 3 potential competitors?

So obviously, as you just said, we know the market will become more crowded. We continue to have confidence in the differentiation of Omnipod, and we believe that competitors are going to continue to enhance the attractiveness of AID as a source of growth from MDI. What we did model is, obviously, we have the benefits of our product launches this year. We did the enhancement on the algorithm of Omnipod 5 as well as our sales force expansion. And we essentially are assuming that those give us no benefit and that it will be offset by the more crowded market next year. So we'll continue to have NCS in actual numbers. It's just that the growth rate of that NCS will be flat to down.

Okay. And the attrition, basically the low end, no improvement in attrition.

Attrition and utilization. And then on the high end would be attrition and utilization, both improved from where they are.

And price stable, I think.

Stable on the low end and slightly positive in the high end.

Got it. Okay. Super helpful. And despite the reduction in the top line outlook, you maintain your operating margin goal of 100 basis points per year. So what's giving you the confidence in that? And why is that the right decision for the business?

We have a really strong financial profile, Larry. We are delivering top-tier growth. We continue to have gross margin expansion behind manufacturing efficiencies. We are disciplined in cost management, and we are investing in the activities that drive long-term growth, like innovation, like commercial capabilities, like the investments in type 2 that Ashley described. So we are in this envious position where we actually can do both. We can continue to invest in the business to support that top-tier growth that I talked about and we can deliver 100 basis points of margin expansion. In fact, as you know, we actually also increased our adjusted EPS growth target to be better than 30%, showing the earnings power of the business that we have.

Okay. And Dexcom recently committed to a 10% revenue growth floor through 2030, 100 to 150 basis points of margin -- annual margin expansion and pledged at least 50% of annual free cash flow back to -- free cash flow to buybacks. So I guess my question for you, Ashley, is would you consider a similar kind of model algorithm that Dexcom has put out there?

Yes. I mean I would just build off of what Flavia said, which is we are committed to -- again, we're not guiding for '27, but we think that we can deliver like top-tier growth for the industry. We have a very compelling, I think, quality of earnings profile around really strong gross margin on efficiencies in manufacturing, really strong G&A efficiencies that allow us to continue to invest in the business. And we have a very healthy EBITDA profile with a good attractive runway ahead of us that we are calibrating based upon the investments that we want to keep plowing in to drive category expansion and keep innovating on the S-curve. While investing in capacity, we're standing up our fourth plant. And so we enjoy a reoccurring revenue business model. And we're on the first 90 days, I would tell you, in T2, but that is going to be a meaningful growth opportunity for the industry as well as disproportionately for Insulet.