How to start a research-use-only (RUO) peptide company in 2026: an operator’s playbook.

Who this is for, and what an RUO company actually is

I run a research-use-only peptide store, and I built Torva OS because running one took far more than a website. This guide is the plan I would hand a friend starting today: what to decide first, what to spend on, what reviewers and customers look at, and the mistakes that cost the most. It is written for founders starting a store, a brand or a small supply company in 2026.

An RUO company sells materials for laboratory research, labelled and presented that way. It is not a pharmacy, a clinic or a supplement brand, and the moment its pages, emails or ads describe what a product does in a person, it stops being presented as research supply. Everything in this playbook follows from that one line: the catalog, the copy, the payments, the ads you can buy and the partners you can work with.

The practical consequence is that this is a documentation business. The companies that last can show, for every lot they sell, where it came from, what an independent lab found and when. This is general information, not legal advice; talk to an attorney who works with RUO businesses before you rely on any of it.

Choose a business model and a position

There are three common shapes, and they need different amounts of cash and skill. Pick one deliberately; most early mistakes come from drifting between them.

Entity, banking and bookkeeping

Form a legal entity before you buy inventory, with an accountant’s advice on which type suits your state and plans. Get a tax ID, open a business bank account in the entity’s name and run every dollar through it. Processors, banks, carriers and suppliers all expect a real business, and underwriting reads your bank statements.

Expect some banks to ask what you sell and to decline the category. Tell them accurately. An account opened on a vague description tends to be closed later, usually at a worse moment. Keep a second business account at a different institution once you are trading, so one closure does not stop payroll, suppliers and refunds at the same time.

Set up bookkeeping on day one, not at tax time. The detail that matters most in this business is cost per lot: what you paid for the product, the freight, the duties and the tests, divided by the units that arrived. Without it you cannot price, and you cannot tell which products actually make money. This is general information, not legal advice; talk to an attorney who works with RUO businesses before you rely on any of it.

• Legal entity, registered agent if you use one, and a tax ID.
• Business bank account in the entity’s name, and a second one once trading.
• Domain and business email on that domain; free webmail reads as temporary to everyone who matters.
• Bookkeeping software with a chart of accounts that separates product, freight, testing, packaging, postage, payment fees and commissions.
• A business address you can publish on policies and applications.
• Sales tax: ask your accountant where you have to register as you grow.

Suppliers and landed cost

Finding suppliers is easy; verifying them is the job. Shortlist three to five, ask each the same written questions, buy a small first order from the two best and send a sample from each to a lab you chose and paid. A supplier who discourages independent testing has answered the question. The full checklist is in how to find and vet a peptide supplier.

Price everything as landed cost, not unit price. A cheap unit price with expensive freight, duties, a payment surcharge and a 20% short shipment is not cheap. Work it out per lot: everything you paid to get the lot onto your shelf, divided by the units that arrived in sellable condition.

Independent testing and a COA library by lot

Handle every lot as its own product record. When a lot arrives, give it a lot number that matches the vial label, send a sample to an independent lab, and only list it for sale when the certificate is back and matches. A certificate of analysis describes one sample from one lot on one date; it is not a statement about the product in general.

Publish the certificate against the lot a customer is actually buying, reachable without an account and readable on a phone. Keep older lots’ certificates available and clearly marked, because customers holding an older vial will look for theirs. Never edit a lab’s document, and never show one lot’s certificate on another lot. How to read and check one is in how to verify a certificate of analysis; how to present them is in a COA page that builds trust.

• Lot register fields: product and size, lot number as printed, supplier, date received, units received, landed cost, lab, test date, result as printed, certificate file, status.
• Statuses: received, testing, listed, sold out, withdrawn. Nothing moves to listed without a matched certificate.
• Cadence: test the first lot from any new supplier before it reaches the catalog, then every lot you publish a certificate for.
• Ask what the lab actually runs. Purity by HPLC is the base test; identity confirmed by mass spectrometry is more specific than identity by retention time alone and usually costs more. Published US prices we checked in September 2026 ran from about $100 to roughly $350 a sample depending on the panel, and standard turnaround at one lab was fifteen business days. Plan the timeline around it.
• If a result disagrees: stop listing the lot, tell the supplier in writing, keep every record.

Site essentials reviewers look at

Payment underwriters, banks, ad reviewers and careful customers all read the same public pages. They are looking for a store that is plainly research supply, run by a real business that can be contacted and that has thought about returns, shipping and privacy. None of this is decoration, and none of it makes any provider’s decision for them.

Run your site through the free processor check or the Processor Risk Score once it is live: they read your public pages and list what a reviewer would find, item by item.

Payments: plan more than one rail from day one

Card processing is the part of this business most founders underestimate. Mainstream processors commonly list this category as restricted, high-risk processors review it closely and often hold a reserve, and an account that is open today can be reviewed again later. A store with one way to get paid has a single point of failure for the whole company. Plan at least two rails before launch and make sure both are tested with a real order.

The common combination is pay-by-app (Zelle, Cash App or Venmo business accounts) or bank transfer from the first day, with card processing applied for once the finished site is live on your own domain. Some stores add other rails; whatever you use, describe your business accurately on every application and never route payments through another merchant’s account.

Fulfilment

Customers judge a new store on the first parcel. Pick a handling time you can meet every business day, ship in rigid boxes or padded mailers that protect vials, add cold packs where your storage information calls for them, and put a packing slip with the order and lot numbers in every box. Buy labels through a carrier or postage account in your business name so rates, tracking and claims are yours.

Decide your policies for the three problems that will happen in the first month: a parcel marked delivered that the customer says never arrived, a damaged vial, and an address typed wrong. Write the answers into the shipping policy before launch. Start with domestic shipping; international orders add customs rules that vary by country and that you need to read yourself before accepting them. Packaging and carrier detail is in shipping and fulfilment.

• Test one complete order end to end before launch: order, payment, label, tracking email, delivery.
• Record the lot number shipped on every order, so a question about a vial can be answered in one lookup.
• Reships and refunds are a real cost per order; the unit economics section counts them.

Email flows that do the selling for you

Email is the channel you own. Send from your own domain with sending authentication set up, ask for marketing consent at checkout, and email only people who gave it. Hold email to the same copy rules as product pages: no outcomes, no directions for use, no customer stories about their own use.

Transactional flows come first because every order needs them. Marketing flows come second and should be triggered by something the customer did rather than blasted to the whole list. Templates and timing are in abandoned checkout and receipt emails.

Organic growth before paid ads

Most major ad platforms restrict or prohibit promoting products in this category, and accounts that try are often disabled. Think of paid ads as something to test carefully later, not as the launch plan; what the policies actually say is in can you advertise research peptides. The channels that work early cost time rather than cash.

Unit economics: know what one order leaves

Revenue is not the number to watch; contribution per order is. Take a worked example with illustrative inputs, not market data: an average order of $150.00, landed product cost of $30.00, testing spread at $2.50 an order, packaging $2.50, shipping you pay $6.00, packing time $2.50, blended payment fees of 2.5% plus $0.30, 15% commission on the 25% of orders that come through affiliates, and 3% of revenue lost to refunds, reships and disputes.

That order leaves $92.33, or 61.6% of the order value, before acquisition and fixed costs. So $92.33 is the most you could pay to win that order without losing money on it. At 120 orders a month and $1,500 of fixed costs, the month contributes $11,079 and $9,579 is left before tax and your own pay; add $3,000 of ads and that falls to $6,579.

The per-order margin in this category can look wide. What sinks stores is rarely the margin; it is too few orders, cash tied up in inventory and reserves, and a payment rail that stops. Run your own numbers in the unit economics calculator and read are research peptide businesses profitable for the full worked example.

What it costs to start

Startup cash is mostly inventory and testing, and both scale with how many products you list. As planning estimates, not quotes: a lean launch with about 5 products tends to land around $5,000–$12,000, a standard launch with about 12 products around $18,000–$40,000, and a brand-heavy launch with about 25 products and custom branding around $70,000–$150,000. The example inputs for those three come to $7,188, $24,495 and $95,760.

Change the inputs to your quotes in the startup cost calculator, and read what it costs to start a research peptide business for the line-by-line budget and the costs that arrive after launch, such as a payment reserve.

A 12-week launch timeline

Twelve weeks is realistic for a small catalog if the supplier ships on time and the first tests come back clean. Each phase has a finish line; do not start the next one on a phase that has not reached it.

The mistakes that cost the most

Almost every expensive mistake in this category is one of these. Most are cheap to avoid if you decide early.

• Launching with too many products. Every product is a minimum order and a test per lot. Twenty products is twenty of each.
• Publishing the supplier’s certificate as your own. Test it yourself; once it is on your page, the claim is yours.
• Letting claims in through the side door. Reviews, emails, affiliate posts and image text count as much as product pages.
• One payment rail. When it stops, the business stops. Have a second rail running before you need it.
• Asking customers to mislabel payments. It breaks the payment apps’ terms and risks everyone’s account.
• Applying for cards with an unfinished site. Underwriting reads the live site; a placeholder wastes the application.
• Buying ads before the basics work. Restricted categories get accounts disabled, and paid traffic into a store without documentation converts poorly anyway.
• Not knowing landed cost per lot. Without it, pricing and quantity breaks are guesses.
• Ignoring the reserve. Held revenue is cash you cannot spend on the next restock.
• Skipping the attorney. An hour of review by someone who knows RUO businesses costs little next to what it prevents.

How Torva fits, if you want it to

Torva OS is store software built for research-use-only suppliers. Every storefront opens with a 21+ research-use gate, product copy is checked for claims language before it goes live, certificates attach to each lot and the product page shows the newest lot first, and pay-by-app orders wait in one list until you mark them paid, with reminders to customers who have not paid. Card checkout runs through your own Stripe account, which you set up inside Torva, and Stripe makes its own decision about your business. Torva does not underwrite, does not give legal advice and cannot promise any provider’s outcome.

You can see your store built from your catalog before you pay anything: paste your product list or your current site and Torva builds a free preview, no card needed. Plans and what each includes are on the pricing page.

Sources

• U.S. Small Business Administration — choose a business structure ↗
• IRS — apply for an employer identification number (EIN) ↗
• 21 CFR 201.128 — meaning of “intended uses” (eCFR) ↗
• FTC — Advertising FAQ’s: a guide for small business ↗