How Much Do Telehealth Companies Make Per Patient?

The formula, and why LTV claims mislead

Gross revenue per patient is price times tenure: what a patient pays per month, times how many months they stay. Contribution per patient subtracts the per-patient cost stack (medication or platform rates, visits, processing, support) and then acquisition cost once. Every honest analysis in this category is those three lines; every misleading one hides in a variable, usually tenure (assumed heroic) or the cost stack (assumed away).

The reason to distrust quoted LTVs: tenure distributions in this category are extremely skewed. The persistence research shows a large early cliff (historically, most weight-loss GLP-1 patients stopped before twelve weeks) with a long tail of durable patients; an average built across that skew flatters the number while describing almost nobody. Median months retained, and survival at fixed checkpoints (week 4, month 3, month 12), are the honest tenure statistics, and they are what the sourced studies actually report.

The whole business in one line: (monthly price minus monthly cost stack) times months retained, minus acquisition cost. Every strategy debate in telehealth is secretly an argument about which term it improves.

Anchoring the price variable (sourced)

What patients actually pay as of September 2026, from sellers' own publications: the membership-plus-medication programs charge around $149 per month in program fees (Ro and LifeMD both publish that number, each with a $39 first month), with medication billed near the manufacturer cash floors ($349 per month for Wegovy through Novo's channel, $299 to $449 for Zepbound vials through Lilly's, an oral option entering at $149). All-in, a branded-medication patient runs roughly $450 to $600 a month across those published components; compounded-lane offers advertise below that with the regulatory caveats our compounded guides cover.

Independent spend data lands consistently: Earnest Analytics' credit-card panel reported average monthly tickets rising to about $134 at Ro and $158 at Hims & Hers in 2024 as GLP-1 programs scaled, blended across their cheaper non-weight-loss product lines. For modeling, the useful anchors are: non-Rx-heavy telehealth products blend to low-hundreds tickets; GLP-1 programs run mid-hundreds all-in when branded medication is inside the number.

Anchoring the tenure variable (sourced)

The persistence studies give the range. Historical, shortage-era: 58 percent of commercially insured weight-loss GLP-1 patients stopped before completing twelve weeks (Blue Health Intelligence, 2024), and about a third of 2021-2022 initiators remained at one year (Prime Therapeutics). Current era: one-year persistence around 63 percent for early-2024 initiators (peer-reviewed in JMCP, March 2026). Those are insurance-population reference curves rather than cash-pay telehealth cohorts, but they bracket the honest modeling range: a poorly run program lives near the bad curve, a well run one near the good curve.

For public-company corroboration that long tenure exists at scale: Hims & Hers' shareholder reporting tracks revenue retention of multi-year-tenure subscriptions as a headline KPI, which tells you durable multi-year cohorts are real and material at the category's largest operator, whatever any individual program achieves. Model with checkpoints, not averages: what fraction of your starts survive week 4, month 3, month 6, month 12.

Worked models (illustrative arithmetic, not projections)

Model A, a branded-medication program on the membership architecture: a patient paying a $149 membership while buying medication direct from the manufacturer channel generates membership revenue only. At five months' tenure, that is roughly $745 gross per patient before the program's visit, support, and processing costs and before acquisition. The membership architecture lives or dies on operating lean and acquiring cheap, which is why its operators emphasize brand and organic audience.

Model B, an all-in program: a patient paying $399 a month all-in, on a stack whose true monthly cost per patient (medication, visits, operations) is $250, contributes $149 a month. At three months' tenure (a bad, cliff-shaped outcome) that is about $447 of contribution against acquisition cost; at nine months (a good outcome by the current persistence data) about $1,341. Same program, same price, tripled economics: tenure is the multiplier on everything.

Model C, the sensitivity that matters: take Model B and move one variable at a time. Cutting the cost stack by $50 a month adds $450 over nine months; raising price $50 does the same arithmetic but risks the retention curve that generates it; adding two months of tenure adds about $298 without touching price or costs. This is why the retention levers in our benchmarks guide are worth more than pricing cleverness, and why medication economics (the biggest line in the stack) decide which programs can afford to be good. All three models are illustrations built from the sourced anchors above; they are not projections, promises, or typical results, and your inputs will differ.

• Model in contribution per month, then multiply by honest tenure checkpoints
• Use median and survival checkpoints, never blended-average LTVs
• The cost stack's biggest line is medication economics: know yours in writing
• Tenure is the multiplier: two extra months usually beats any price increase
• Label every model illustrative, to yourself first: believing your own deck is the classic failure

What this means for choosing your stack

The formula explains the platform-selection debates this site covers elsewhere. Revenue share prices your upside (it taxes the price term as it grows). Medication markup prices your biggest cost line opaquely. Flat rates make the contribution line knowable in advance, which is the entire point of the arithmetic above: a program that knows its monthly contribution per patient can compute what it can afford to spend on acquisition and on retention, and a program that cannot is guessing with a spreadsheet costume.

EmbedCare's model is built for the knowable version: flat product rates fixed in a signed partner agreement, medication included on GLP-1 programs, your retail above the rate, your spread per patient computable before launch. Our earnings calculator runs the arithmetic with your own assumptions, and a demo prices the real thing; both frame results the way this page does, as models rather than promises.