Virtual clinics are very good at getting GLP-1s into patients’ hands fast. That chapter is behind us, and the margins are compressing to prove it.
Leaders like Ro, hims & hers, Calibrate and Found have started building more holistic health solutions, layering in coaching, CGMs, and behavioral support. That is a decent first step. But breadth without depth isn’t a moat; it’s a feature set. And it’s only a matter of time before feature sets get copied.
Moreover “holistic” by itself begs the question, “holistic for whom?” The honest answer is for the “average” patient, which is to say for the “lowest common denominator” or “for no one in particular.”
Personalization is the missing link, and even more specifically, biology-based personalization.
The platforms that will matter in the long run are the ones that can look at a patient’s phenotype, their biomarkers, their physiology and build a regimen that addresses their unique needs. Not a formulaic program but an individualized one.
The science to do this exists. The tools exist. What most of the landscape is missing are the vision and near term incentive to move beyond the current “land grab” environment. But for better or worse, economic reality is about to force the issue.
A GLP-1 Tailored Nutrition Wraparound
A landmark joint advisory from the American College of Lifestyle Medicine, the American Society for Nutrition, the Obesity Medicine Association, and The Obesity Society put it plainly: GLP-1s achieve meaningful weight reduction, but they come with a set of challenges — gastrointestinal side effects, nutritional deficiencies, and significant loss of muscle and bone mass.
Patients on GLP-1 therapy who don’t maintain adequate protein intake risk sarcopenic obesity – losing muscle without proportionally losing fat. During major weight reduction, experts recommend protein consumption of 1.2 to 2.0 grams per kilogram of adjusted body weight per day. Unfortunately for most GLP-1 patients, achieving even baseline protein targets is difficult because the drug suppresses appetite so aggressively.
Specifically, over 20% of GLP-1 patients are nutritionally deficient within 12 months of initiating therapy. Loss of lean mass, if left unaddressed, increases long-term risk and erodes the metabolic health outcomes that made the treatment worthwhile in the first place.
GI side effects compound the problem. Nausea, vomiting, and general discomfort are key drivers of discontinuation. The implications are clear: maximizing the benefits of GLP-1 therapy requires a multidisciplinary approach integrating evidence-based nutrition, physical activity, and proactive management of gastrointestinal side effects.
A high-protein, nutrient-dense dietary regimen — one that goes beyond dietitian guidance and is built around the physiological realities of GLP-1 therapy — is a clinical necessity, not an upsell. Virtual clinics that fail to provide or partner to develop such a wraparound leave their patients at genuine risk of medical harm, and their businesses at risk of poor outcomes data, high churn, and reputational exposure.
The Moat Is Knowing Your Patients, Not Medicating Them
The first generation of virtual clinic evolution — adding coaching, CGM, behavioral support — was about breadth. The critical next move is about depth.
A coaching module or a CGM add-on doesn’t differentiate you if every competitor can bolt on the same features. What does differentiate you is the depth and intelligence of the patient relationship, which requires personalization at a level that most platforms haven’t yet reached.
Consider what precision-based obesity care might entail. Dr. Andres Acosta and colleagues at the Mayo Clinic have demonstrated that obesity is not a single disease but a heterogeneous one, driven by distinct biological and behavioral phenotypes. Their research identifies four core phenotypes: hungry brain (impaired satiation), hungry gut (impaired satiety), emotional hunger (reward-driven eating), and slow burn (reduced metabolic rate). Their research concluded that a phenotype-guided approach to treatment resulted in 1.75 times greater weight loss at one year, with 79% of patients losing more than 10% of body weight compared to only 34% in the non-phenotyped group.
This science also reframes one of the most contested questions in GLP-1 care: how long should a patient stay on the drug? The answer is that it depends. Some patients will need to remain on GLP-1 therapy indefinitely to maintain their metabolic health. Others, as they adopt sustained dietary changes, build lean muscle, and address the root behavioral drivers of their weight, may be candidates for tapering or microdosing as their metabolic resilience rebuilds. And those who make the most significant lifestyle transformations eventually may be able to go off the medication entirely if their metabolic health is sufficiently restored.
The critical word is may. None of these paths can be determined by a generic protocol or a standard titration schedule. They require continuous, biologically grounded assessment: blood biomarkers, body composition tracking, physiological data from wearables, and phenotype-informed clinical judgment. When you combine that data infrastructure with AI-driven personalization and the right human expertise in the loop, you have something that actually reflects how complex and individual obesity truly is. Not just “take this medication.” But: here is your medication, here is the protein intake target calibrated to your lean body mass, here is why your phenotype makes you particularly susceptible to GI side effects and how your diet should account for it, here is how we’ll assess over time whether your regimen should evolve, and in which direction.
The virtual clinics that build toward that model will have something competitors cannot easily replicate. The clinics relying on a rigid protocol will have little more than a customer list ready to migrate at the first price change.
Severity of condition, tech fluency, level of intrinsic motivation all matter here too. Some consumers can be successful with the efficiency of an AI-curated plan. Others will want the accountability and support of a human coach or registered dietician. Best-in-class platforms will offer both, and will know which patients benefit most from which approach.
The Coverage Crisis Is a Direct Consequence of One-Size-Fits-All Care
Here is the part of the story that is changing fastest, and that virtual clinics cannot afford to ignore.
The way GLP-1s are being prescribed today — broadly, reactively, and largely without individualized clinical criteria or structured wraparound programs — is not a sustainable model for the health system.
Employers are walking away from GLP-1 coverage at an accelerating rate. According to the Mercer National Survey of Employer-Sponsored Health Plans and the Business Group on Health, 67% of large employers currently cover GLP-1s for weight management. Of those, only 72% said they were likely to maintain that coverage into 2027. Blue Cross Blue Shield of Massachusetts stopped covering GLP-1s for obesity in employer plans, projecting that doing nothing would cost close to $1 billion this year, nearly double its 2025 spend and seven times its 2023 costs. HCA Healthcare, which employs more than 300,000 workers, cut weight-loss GLP-1 coverage entirely after use surged 90% in a single year. The data and diagrams below paint a clear story about where this is heading.
The problem isn’t simply the cost of the drugs; it’s that coverage decisions are being made in the absence of the outcomes data that would justify them. When GLP-1s are prescribed without individualized clinical criteria, the ROI case for payers becomes difficult. The result is blunt, across-the-board coverage cuts that hurt the patients who genuinely need and would benefit most from these medications.
The employers and payers that are continuing to cover GLP-1s are attaching conditions: participation in weight management programs, verification of clinical eligibility through biometric data, restrictions on prescribing to specific providers. In other words, they are beginning to demand evidence that the investment will produce outcomes. That is exactly the right instinct, although the implementation is still blunt.
This trajectory points toward an inevitable reckoning. Payers will increasingly want to reimburse for holistic, outcomes-driven obesity care, individualized programs that can demonstrate adherence, biomarker improvement, lean mass preservation, and a rational, biology-guided approach to dosing over time. The virtual clinics that have built that infrastructure will be positioned to make the case for their patients. The rest will be fighting a price war on commodity prescriptions and programming with eroding coverage support from employers and insurers.
The case for the full regimen is also backed by outcomes data. Research tracking 192,000 GLP-1 users found that individuals who maintained therapy for 18 months showed slower medical cost growth and a reduction in hospitalizations. The drugs, when supported by the right behavioral and nutritional infrastructure, do drive long-term cost reduction.
What an Individualized GLP-1 Program Looks Like
The model I am describing is not theoretical. The components exist. What’s missing is integration and intentionality.
A virtual clinic serious about building for the long term should be assembling:
A personalized nutritional plan built around the patient’s protein requirements, GI tolerance, and dietary preferences, with a specific emphasis on protein-enriched, nutrient-dense food that supports muscle preservation. This is not generic healthy eating advice. It is a clinically informed regimen that accounts for the physiological realities of GLP-1 therapy.
Biomarker-driven personalization that incorporates both blood-based data (metabolic panels, inflammatory markers, hormonal profiles) and physiological data from wearables (activity levels, sleep quality, HRV, glucose trends where applicable). This data should inform both the treatment plan (including dosing decisions over time) and the ongoing measurement of progress.
Phenotype-informed decision-making — drawing on the growing body of research that shows different patients respond very differently to both the medication and to specific behavioral and dietary interventions. Understanding which phenotype is driving a patient’s obesity changes the treatment calculus significantly, including how long they are likely to need pharmacological support.
A dynamic dosing framework — rather than a static prescription, a structured approach to evaluating whether a patient should stay on their current dose, taper, microdose, or potentially transition off the medication, based on continuously updated biological and behavioral data. This is the layer that transforms a prescription business into a genuine clinical outcomes business.
Human-AI optionality — an AI-driven experience for patients who want efficiency and self-direction, with pathways to registered dietitians and health coaches for those who benefit from human support and accountability.
Outcomes tracking that goes beyond the scale — capturing lean muscle mass preservation, biomarker improvement, and adherence metrics that can be shared with employers and payers to build the evidence base for holistic reimbursement.
The Strategic Imperative
The GLP-1 market is here to stay. Even as employer coverage contracts, the patient population seeking these therapies continues to grow, and the clinical evidence for their efficacy remains strong. Newer oral formulations will expand access further.
But the virtual clinic category is about to bifurcate. The Losers: commodity prescription platforms that win on price and convenience until they don’t, squeezed by coverage pullbacks, high churn, and undifferentiated positioning. The Winners: platforms that have invested in the full-spectrum patient relationship — personalized nutrition, biomarker intelligence, phenotype-informed care, dynamic dosing, and demonstrable outcomes.
