For much of the past decade, automated insulin delivery has been viewed primarily through the lens of Type 1 diabetes. Insulin pumps integrated with continuous glucose monitoring have created one of the most significant advances in diabetes care: the ability to continuously monitor glucose levels, automate insulin adjustments, and reduce the daily decision-making burden on both patients and physicians. At the same time, companies such as DreaMed have been building software-driven models designed to bring similar continuous decision support to the much larger population of insulin-treated Type 2 patients who remain on injections.
Today, major diabetes technology companies such as Insulet and Tandem Diabetes have invested heavily in bringing automated insulin delivery to patients with insulin-treated Type 2 diabetes. Clinical studies published in leading journals continue to show improved glycemic control, lower management burden, and better patient outcomes. By late 2025, FDA indication expansions made clear that this is no longer a niche opportunity; it is becoming a major commercial priority.
The Opportunity Between Injections and Pumps
Clinically, the logic is straightforward. If automation improves outcomes for Type 1 diabetes, extending similar benefits to millions of insulin-treated Type 2 patients makes sense.
But healthcare markets are not shaped by clinical logic alone. Type 2 diabetes introduces a fundamentally different reimbursement equation because of its sheer scale. Millions of patients rely on multiple daily injections rather than pumps, and broad reimbursement of pump-based automated insulin delivery would create enormous cost pressure for insurers.
Even conservative estimates place annual pump therapy costs at several thousand dollars per patient. When applied across a meaningful portion of the Type 2 insulin population, the numbers quickly become difficult for payers to justify, suggesting that what works clinically may not work economically.
This is where the real market gap begins. Most insulin-treated Type 2 patients still rely on injections, and for many, that is unlikely to change. Pumps require higher costs, more training, physician support, and meaningful workflow changes. For providers and patients alike, injections often remain the simpler and more practical option.
Hence, a new category of companies is emerging around a simpler premise: continuous insulin optimization does not necessarily require full pump adoption. The opportunity lies less in replacing hardware and more in adding intelligence to existing treatment workflows. By combining continuous glucose monitoring data, dosing algorithms, and connected injection-based care, software-driven systems can help physicians and patients make better insulin decisions while keeping patients on injection therapy.
When paired with connected injection devices, this approach begins to deliver many of the advantages of continuous care without the operational complexity, infrastructure requirements, or full cost structure associated with pump therapy. Companies like DreaMed are building around this approach, focusing on algorithm-based insulin titration and patient-guided dosing support rather than full pump replacement. The goal is not necessarily to clinically outperform pumps in every scenario, but to create a model that can scale to the much larger Type 2 population and deliver meaningful clinical impact.
Why the Economics Are Changing Now
A few years ago, continuous insulin optimization without a pump would have been far harder to justify commercially because the supporting infrastructure was missing. Today, that is changing rapidly as continuous glucose monitoring becomes more widely reimbursed in the care of Type 2 diabetes.
In many cases, the sensor is already covered. Patients are already generating continuous glucose data, and providers have access to the information needed to make better insulin decisions. Once that foundation exists, the economic logic changes: the next layer of value may not come from adding another expensive device, but from improving how existing data is translated into treatment decisions.
That makes software fundamentally different from pump adoption. A pump requires new hardware, training, operational support, and significantly higher reimbursement commitments. A software-driven dosing model with a connected insulin pen device can sit on top of existing infrastructure, creating a much lower barrier to adoption for both providers and payers.
In other words, pump companies are helping prove the clinical value of continuous insulin optimization, but they are also exposing the limits of scaling that model across the full Type 2 population. The companies that may capture the broader market are not necessarily those building the most advanced hardware, but those creating the most efficient path between glucose data and better dosing decisions.
