There is a predictable moment in the life of a modern direct-to-consumer health or wagering platform. Marketing finally lands, orders triple in a fortnight, and then an automated email arrives: your account is under review, payouts are on hold, and a reserve has been applied. Nothing illegal happened. The business simply outgrew the risk tolerance of the payment tool it started with — and nobody warned the founder that the tool had one. This is not a rare event in telemedicine, peptide retail, supplements, nicotine alternatives or prediction markets. It is closer to a rite of passage. Understanding why it happens — and what underwriters actually want to see — is the difference between a two-day interruption and a quarter spent rebuilding your revenue stack while refunds pile up.
Aggregators are built for the average merchant, and you are not average
Mainstream platforms like Stripe, Square and PayPal onboard merchants in minutes because they place thousands of businesses under shared master accounts. That speed comes with a condition: the aggregator, not the merchant, carries the loss if something goes wrong. So the acceptable-use policy does the underwriting that a human would otherwise do, and it draws hard lines around prescription products, compounded medications, research chemicals, peptides, event-based wagering and anything that looks like medical advice delivered at scale. Crucially, the line is enforced retroactively. An account can process happily for months at low volume, then trip a review when velocity spikes, when a chargeback ratio crosses a card-brand threshold, or when a risk analyst finally reads the landing page and sees a dosage chart. Industry commentary on the mechanics of sudden holds — including this American Banker opinion piece on how merchants can stay unfrozen during the e-commerce wave — keeps returning to the same point: the freeze is usually a documentation failure, not a verdict on the business model.
The three triggers that do most of the damage
- Mismatch between the website and the application. You applied as “wellness e-commerce” and your checkout sells a physician-prescribed GLP-1 refill plan.
- Chargeback ratio drift. Subscription refills and free-trial funnels generate disputes at a far higher rate than one-off retail, and card networks monitor the ratio monthly, not annually.
- Volume outrunning the approved profile. An account approved for $25,000 a month suddenly submitting $300,000 looks identical, to an automated system, to a laundering pattern.
What dedicated underwriting actually asks for
A direct merchant account — your own MID with an acquiring bank — replaces a policy document with a conversation. That conversation is more work up front and vastly more durable afterwards. Expect to assemble:
- Three to six months of processing statements, or bank statements if you are pre-revenue, plus a projected monthly volume and average ticket you can defend.
- Corporate documents, beneficial ownership, and a business bank account in the legal entity’s name.
- Clinical governance for telehealth: state medical licences for the prescribing physicians, the asynchronous or synchronous consult workflow, and how identity and age are verified.
- Pharmacy relationships in writing — 503A or 503B partners for compounded products — and evidence of physician oversight rather than a self-serve order form.
- DEA registration where controlled substances are genuinely in scope; if they are not, say so plainly, because ambiguity reads as risk.
- Product-level substantiation for peptides and supplements: labelling, claims language, and whether items are sold for research use or human consumption.
- Refund, cancellation and shipping policies visible at checkout, and a descriptor that customers will recognise on a statement.
HIPAA adds a parallel track. If payment data travels alongside, or can be linked to, protected health information — appointment notes, prescription SKUs, diagnosis codes — the processor and gateway become part of your compliance perimeter. That means a signed Business Associate Agreement, tokenisation so raw card data never touches your servers, and restraint about what you push into metadata fields. Naming a medication in a transaction descriptor is a small convenience that creates a genuine disclosure problem.
Recurring billing, FSA/HSA and the unglamorous plumbing
Refill revenue is the reason these businesses are valuable, and the reason they are risky. A tokenised card vault lets you re-bill without storing card numbers, and an Account Updater service quietly refreshes expired or reissued credentials so that a reissued card does not silently end a two-year patient relationship. Soft declines should be retried on a schedule, not hammered — repeated retries on the same failing card inflate your decline ratio and attract attention. FSA and HSA acceptance is its own capability. Those cards route against merchant category codes and, for substantiation, against item-level data. If your MCC says “general merchandise”, eligible consultations and devices will decline at checkout and your support queue will fill with confused patients. Getting the category right during underwriting is cheaper than appealing it later.
Chargeback defence is an operations discipline, not a plugin
Most disputes in this sector are friendly fraud or confusion: an unrecognised descriptor, a forgotten subscription, a refill shipped two days before the customer meant to pause. The defensible response is evidence assembled before you need it — consult timestamps, IP and device records, delivery confirmation, the exact cancellation terms the customer accepted, and the support thread. Pre-dispute alerts can resolve some cases before they become formal chargebacks. None of this eliminates disputes; it keeps the ratio below the thresholds that end accounts. Security hygiene belongs in the same conversation. Underwriters and card brands both care whether your admin tools, staff laptops and support inboxes are hardened, because a compromised back office produces fraudulent orders that look like your fraud. Basic endpoint discipline — patching, credential rotation, and sensible use of protective software, such as the anti-malware features worth enabling on business machines — is cheap insurance for a company whose entire revenue depends on remaining in good standing with a bank.
Where a specialist provider earns its keep
At some point the answer is not a better workaround but a provider whose business is underwriting exactly this risk. 2Accept works in that lane, connecting merchants to a network of acquiring banks and handling the parts founders tend to underestimate: the underwriting file, gateway or EHR-native integration, BAA-backed HIPAA-compliant setups, tokenised recurring refill billing with Account Updater, FSA/HSA acceptance and ongoing chargeback defence. Its focus on telemedicine merchant accounts extends to adjacent categories that aggregators restrict — peptide sellers and prediction market operators among them — with approvals conditioned on the things that actually matter, such as documented physician oversight, 503A or 503B pharmacy partners and DEA registration where controlled substances are involved. The company reports a 98% approval rate across 40+ acquiring banks and more than $2B processed yearly, with a 48-hour approval average and go-live typically inside 48 to 72 hours.
Build for the account you will need in a year
Three habits prevent most of the pain. Apply as the business you are, with the clinical and product documentation attached, rather than the simplest category that gets you live today. Hold a secondary MID so a single review cannot take revenue to zero — load balancing across acquirers is standard practice in high-risk verticals, not paranoia. And treat compliance artefacts as living files: licences expire, pharmacy partners change, state rules on asynchronous prescribing shift. Payment access is infrastructure. Founders who treat it that way from the first thousand dollars of volume rarely meet the frozen-payout email at all — and if they do, they have the file ready to answer it.


