How to Protect Telehealth Margins When a Modality Disappears: Lessons From the GLP-1 Cliff

Direct answer: Protect your telehealth margins by never building your P&L on a single drug category, and by owning a fulfillment stack that can pivot pharmacies fast. When a modality gets restricted, operators with multi-pharmacy routing and a diversified formulary can reroute orders in days. Operators who can't, can't.

You probably watched it happen in slow motion. Compounded semaglutide and tirzepatide went from license-to-print-money to existential liability inside twelve months — and the brands that went all-in on GLP-1 felt every step of the slide.

This isn't a GLP-1 post. That cliff is well-documented, and you don't need another rundown of the FDA timeline. What you need is a framework for making sure this never happens to your business again — regardless of which drug category gets squeezed next.

---

Why the GLP-1 Cliff Is a Case Study in Stack Risk, Not Just Regulatory Risk

The compounded GLP-1 story looks like a regulatory story on the surface. FDA put semaglutide and tirzepatide on the shortage list, compounders filled the gap, then FDA removed them from shortage status and moved to permanently exclude them from the 503B bulks list. Comment period closes July 30, 2026. The writing is on the wall.

But regulatory risk was only half the problem. The real damage came from stack rigidity.

Operators who built on compounded GLP-1 often also built everything else around it: their Shopify catalog, their marketing, their pharmacy relationships, their entire unit economics model. When the modality started closing, they couldn't pivot fast enough — because the thing connecting their storefront to their pharmacy was either a one-drug integration or a locked-in platform that didn't give them the flexibility to switch.

The lesson isn't "avoid regulated drugs." Every drug category carries regulatory risk to some degree. TRT involves Schedule III, hair medications carry their own FDA scrutiny, peptides are in active regulatory review. If your business is in DTC Rx, risk is the operating environment. The lesson is that your fulfillment infrastructure should be able to route around a disrupted modality without requiring you to rebuild your entire stack.

---

What "Category Risk" Actually Looks Like on a P&L

Let's put some numbers on this.

Assume you're running a telehealth brand doing $150K/month in revenue. If 70% of that revenue comes from one compounded drug category — a concentration that was common among GLP-1-focused operators at the 2024–2025 peak — then a single regulatory restriction can eliminate $105K of monthly revenue before you've changed a single thing about how you operate.

Gross margins on compounded DTC are high, often 70%+. But net margins compress fast once you account for CAC, platform fees, and merchant processing. Hims & Hers — one of the best-capitalized operators in the space — ran net margins around 2–3% at scale on $2.35B in FY2025 revenue, even with adj. EBITDA in the hundreds of millions. The point: the engine is multi-script retention and cross-sell, not any single drug's margin. When you lose the anchor drug, you don't just lose its revenue. You lose the patient who was buying three other things.

---

The Three Compounding Factors That Made GLP-1 Concentration So Dangerous

1. Your pharmacy relationship was category-specific

Most compounders who scaled GLP-1 volume weren't fungible. Empower, Strive, Belmar, Revelation — they all compound, but their formularies differ, their licensing footprints differ, and their order intake systems differ. If you were integrated directly with one compounder via their intake portal, and that compounder had to wind down GLP-1 production, pivoting to a second pharmacy wasn't a flip of a switch. It was a new integration, a new set of credentials, new SKU mapping, and weeks of delay.

Multi-pharmacy routing isn't a nice-to-have. It's margin insurance. If your order can go to pharmacy B when pharmacy A can't fill it — whether because of a regulatory restriction, a formulation backlog, or a supply issue — you preserve revenue that would otherwise evaporate.

2. Your system of record was someone else's

Here's the other thing that happened to operators who relied on pharmacy portals or white-label telehealth platforms: when the disruption hit, they didn't own their patient data or their order history in a usable form. Their records were in a portal they logged into, not a database they controlled. Pivoting to a new pharmacy or a new platform meant potentially losing continuity on patient history, refill cadence, and subscription status.

Owning the system of record is a growth asset, not just a compliance posture. If every order flows through your stack — keyed to your order ID, your patient ID, your prescription history — you can port to a new pharmacy partner without rebuilding patient relationships from zero.

3. Your formulary was one-dimensional

Operators who built a brand around a single indication — "the GLP-1 clinic" — had no adjacent products to retain the patient when that indication dried up. Operators who had built across TRT, HRT, hair, ED, skincare, LDN, and oral weight-loss combinations had a different problem: which of my existing patients do I retain on an adjacent script?

That's a solvable problem. An empty formulary isn't.

---

What a Resilient Formulary Actually Looks Like

Category diversification doesn't mean stocking fifty SKUs on day one. It means designing your formulary so that no single modality represents more than 30–40% of your revenue, and so that your top categories sit in different regulatory risk buckets.

Here's a working framework, from lower to higher regulatory complexity:

Lower complexity (fewer moving parts, high refill rates):

• Tretinoin / skincare — easy attach, low CAC, strong refill
• Hair (finasteride + minoxidil) — high demand, well-established safety profile, minor FDA scrutiny on topical compounding formulations to monitor
• LDN (low-dose naltrexone) — DTC telehealth market is meaningful; ~$30–60/month, high patient loyalty

Medium complexity (real volume, requires clinical diligence):

• ED (compounded tadalafil / sildenafil troches) — strong attach, high attach-rate from men's health stack; note FDA scrutiny on combination troches, verify with your pharmacy
• HRT (estradiol / progesterone) — steady, cash-pay-heavy, large addressable market; patient retention is excellent once initiated

Higher complexity (volume is there, compliance overhead is real):

• TRT / men's hormones — testosterone is Schedule III, so Ryan Haight / EPCS compliance applies; DEA telemedicine flexibility extensions run through December 31, 2026 with no confirmed permanent resolution yet; get legal guidance before launching
• Oral weight-loss combinations (metformin + naltrexone + topiramate) — the post-GLP-1 oral pivot that brands like Shed are building on; no shortage-list exposure, cleaner regulatory profile than compounded injectables

GLP-1 as optionality, not thesis: If you already have GLP-1 revenue, treat it as upside until clarity emerges. Do not anchor new patient acquisition or a new brand around it. If you're starting fresh, build the formulary above and add GLP-1 as a category you can switch on if the landscape changes — not the load-bearing beam.

---

How Your Fulfillment Stack Either Enables or Blocks the Pivot

This is where infrastructure becomes a margins conversation.

If you're on a single-pharmacy portal — even a well-run one — your ability to respond to category disruption is constrained by that pharmacy's formulary and its operational status. One restriction, one supply backlog, one compliance action against your pharmacy, and you have nowhere to reroute.

Multi-pharmacy routing changes the math. The same order request hitting your system can be evaluated against multiple licensed, credentialed pharmacy partners: which pharmacy holds the state license for the patient's ship-to state? Which has stock? Which has the best turnaround? If pharmacy A can't fill, pharmacy B fills automatically, and neither you nor the patient notices.

This is structurally impossible if your "integration" is a portal login. It requires an abstraction layer that sits above the pharmacies — normalizing their different intake formats, credential sets, SKU identifiers, and webhook schemas — so your storefront can speak once and the routing layer decides where the order goes.

The other side of this: every order that flows through your own stack is a record you own. Patient demographics, prescription history, refill cadence, which pharmacy filled which SKU, what the tracking number was — all of it lives in your database, keyed to your order ID. Not in a portal you can be locked out of. Not in a platform you can't export from. Yours.

When a category disruption hits, operators who own their data can answer the right questions fast:

• Which patients are on the affected category?
• What adjacent scripts have those patients filled before?
• Which of my pharmacy partners can take overflow on the adjacent modality?

Operators who don't own their data can't answer those questions at all.

---

The Expand-Fast Playbook When a Category Shifts

A category restriction gives you roughly three windows:

Weeks 1–2 — Stop the bleed. Audit current active subscriptions and upcoming refills in the affected category. Identify which patients can transition to an adjacent indication. Communicate proactively — patients who are already seeing a licensed provider through your platform are more likely to stay than churn if you give them a clear path.

Weeks 3–6 — Reroute. If you have multi-pharmacy routing, confirm which of your pharmacy partners can still fill the adjacent categories at volume. Activate SKUs that were in your catalog but not actively marketed. Verify state-by-state licensure for the categories you're expanding into — your compliance gate should be enforcing this automatically on every order, not manually.

Months 2–4 — Rebuild the CAC engine. The formulary you expand into needs a paid acquisition strategy. LegitScript certification should already be in place before you touch ads — Visa, Mastercard, Google, and Meta all require it for telehealth and Rx categories. If you applied on day one, you have it. If you didn't, the certification process takes 2–4 months and you're running paid acquisition blind until it's done.

---

What This Has to Do With Owning Your Stack

The operators who navigated the GLP-1 disruption best shared two traits. First, they had a diversified formulary. Second, they had a fulfillment stack they actually controlled — where the order system of record was theirs, where multi-pharmacy routing was operational, and where adding a new pharmacy partner didn't require re-integrating their entire front end.

The operators who got hurt had the inverse: a great brand, loyal patients, and a single modality running through a single pharmacy via a portal they had no control over.

This is fundamentally an infrastructure conversation. The GLP-1 cliff is a cautionary example. The next cliff — whatever the modality — will test the same thing: can your fulfillment rail adapt, or does it break?

A provider approves every order in either scenario. That part doesn't change. But whether that approval routes to one pharmacy or five, whether your order history is yours or someone else's, whether a new category goes live in a week or takes a quarter — that's decided by your stack, not your prescribers.

---

Key Takeaways

• Never let a single drug category exceed ~35–40% of revenue. The GLP-1 cliff proved that concentration risk in regulated DTC is existential.
• Formulary diversity is CAC amortization. A patient retained on three modalities is far more valuable than three single-modality patients — and far more resilient to any one category disruption.
• Multi-pharmacy routing is the operational hedge. When one pharmacy can't fill, another can. This requires an abstraction layer above the pharmacies, not a direct portal integration.
• Own your system of record. Every order that flows through your stack should live in your database, keyed to your IDs. When a disruption hits, the operators who can query their patient data pivot fast. The ones who can't, wait.
• LegitScript certification is not optional and not fast. Apply day one. It gates both payments and paid acquisition across every major platform, and takes 2–4 months. Running ads without it means running on borrowed time.
• GLP-1 as optionality, not thesis. If you already have it, manage the wind-down. If you're starting fresh, build a formulary that doesn't require it.

---

FAQ

What categories are the safest bets for a new telehealth brand post-GLP-1?

No category is without regulatory risk, but lower-complexity starting points include tretinoin/skincare, hair (topical minoxidil + finasteride), and LDN — high refill rates, well-established safety profiles, and less acute regulatory pressure than GLP-1 or controlled substances. Verify specifics with your pharmacy and legal counsel before launching; the landscape shifts.

How much does single-pharmacy dependence actually cost?

The direct cost is hard to model in the abstract, but the structure is clear: if your top category gets restricted and you have no routing alternative, you have no continuity on the revenue from that category until you build or buy a new integration. For a brand doing $100K/month with 65% concentration in one category, that's potentially $65K/month in revenue that can't move to a backup pharmacy because the backup pharmacy isn't integrated. Multi-pharmacy routing eliminates that structural gap.

Is multi-pharmacy routing something I can build myself?

You can build the abstraction layer yourself, but the complexity is real: each pharmacy has different credential requirements, different intake formats (LifeFile, Empower's EIP, Honeybee, etc.), different SKU identifiers, different webhook schemas, and different state licensure footprints. The routing logic — picking the right pharmacy for each order based on state license, drug availability, turnaround, and failover — adds another layer. Most operators find it's better to sit on top of a fulfillment rail that has already abstracted these differences rather than build N pharmacy integrations in-house.

Does provider approval change when I route to a different pharmacy?

No, and it shouldn't. The provider approval gate sits above the routing layer — a licensed provider approves every prescription before any routing decision is made. The pharmacy that fills the order changes; the requirement that a human provider approves the order does not. If a platform tells you otherwise, that's a compliance problem.

When should I start thinking about owning a pharmacy outright?

When monthly order volume makes the margin capture worth the operational overhead. A 503A gets you started but is state-limited (the 5% out-of-state rule creates real ceiling risk, even if enforcement has been soft). A 503B is the real moat — full interstate capability, office-use dispensing — but it requires cGMP infrastructure and FDA inspection exposure. Neither is the right first step. Build the orchestration layer first; let volume justify the pharmacy acquisition later.

---

Nothing in this post is legal or medical advice. Drug categories, compliance requirements, and regulatory positions change — verify specifics with your pharmacy, LegitScript, and qualified legal counsel before making formulary or infrastructure decisions.

---

neolife is the fulfillment rail that connects your Shopify store to your pharmacy. Orders out in under 60 seconds. A licensed provider approves every one. And because neolife becomes your system of record — not your pharmacy's portal — your patient data and your order history stay yours, no matter what happens to any individual modality or pharmacy partner. See how multi-pharmacy routing works → · Talk to us about your stack →