The Peptide Ad Review 2026: what gets research-peptide ads rejected

Last updated 15 September 2026 · Crawled 15 September 2026 · Method version 1.0 · 411 brands on the list, 334 readable, 77 unreachable

This is the first run of the Peptide Ad Review, a dated, reproducible check of what a platform reviewer meets on a research-peptide store's own public pages. We read the home page, up to three product pages and the cart page of 411 US brands on 15 September 2026 and scored nine public signals. Nothing was bought, no form was submitted and no account was created. Every count below is aggregate. No brand is named for a poor result, here or anywhere else we publish; the brands named on this page are named for a good one.

What this measures, and what it does not. The rejection-risk score counts how much of a store's own public copy matches the phrase families Meta's and Google's published advertising policies name. It is a measure of ad review exposure and of nothing else. It is not a statement about whether any product is legal, safe, pure or correctly sold, and it is not a statement about any company.

What did we find?

The nine signals side by side. Each bar is the whole list of 411 brands; the grey part could not be judged from public pages and is never counted against a brand.

The Peptide Ad Review, 411 US research-peptide brand websites, crawled 15 September 2026. Pass, fail and unknown. Our data

What do the platform policies actually say?

Both platforms publish the language. Neither says the word peptide. Both describe the wording that a research-peptide product page reaches for by default.

Meta, Drugs and Pharmaceuticals (Advertising Standards). The standard opens: "Advertisers can't run ads that promote the sale or use of illicit or recreational drugs, or other unsafe substances, products or supplements." Its "can't" list includes ads that "Solicit, buy, sell, trade, donate or gift (such as freebies) or illicit, recreational, or other potentially unsafe drugs, products or supplements" and ads that "Encourage the consumption of illicit, recreational, or other potentially unsafe drugs, products or supplements" (transparency.meta.com, read 15 September 2026). The operative phrase for this category is "encourage the consumption": that is a test of the words, not of the molecule.

Meta, Health and Wellness. The same standards bar ads that "employ clickbait tactics in a health, weight loss, or weight gain context, such as sensational language with exaggerated or extreme claims", and bar claims "to cure, heal, or eliminate" named incurable conditions (transparency.meta.com, read 15 September 2026). Before-and-after transformation imagery is permitted only in narrow, age-targeted cases, which is why it is a risk marker on a research catalogue rather than a neutral one.

Google Ads, Healthcare and medicines. "Google Ads is dedicated to following advertising regulations for healthcare and medicine, so ads and destinations are expected to follow appropriate laws and industry standards" (support.google.com/adspolicy/answer/176031, read 15 September 2026). Note the word destinations: the landing page is inside the policy, not outside it.

Google Ads, Unapproved substances. "Google doesn't allow the promotion of certain products, regardless of any claims of legality." The list that follows includes "Herbal and dietary supplements with active pharmaceutical or dangerous ingredients", "Products that imply they are as effective as prescription drugs or controlled substances", "Non-government approved products that are marketed in a way that implies that they're safe or effective", and "Products with names that are confusingly similar to an unapproved pharmaceutical or supplement" (support.google.com/adspolicy/answer/15595718, read 15 September 2026). The named-example list at answer/2423645 carries SARM and anabolic compound names directly.

Policy text read on 15 September 2026. Full quotes and links in Sources.

Which phrase families show up in the wild, and how often?

Each family is counted once per brand, on the pages we read, and only where the phrase is not inside a disclaimer line. A sentence such as "not intended to diagnose, treat, cure or prevent any disease" contains the word treat and is never counted as a claim.

Disease, treatment and outcome wording

Outcome wording by family, 334 readable brands, 15 September 2026. Our data

Dosing and human-use wording

Reconstitution wording (reconstitute, bacteriostatic water, a peptide calculator) appears on 180 of 334 sites. It is ordinary bench language in a research catalogue, so it is reported here and never scored as human-use language. Reviewers do not always draw that line, which is worth knowing before it turns up in an appeal.

Named substances

Before-and-after, testimonials and review widgets

Review text is usually loaded by script, so what the reviews say could not be read from the page source. The widget is counted, the words are not.

The research-use line, and where it sits

Where the line appears matters more than whether it exists. A reviewer opening a product page from an ad sees the product page, not the terms page.

A site can carry more than one of these.

How exposed is the category as a whole?

The score is the sum of the weights in the method below, capped at 10. The median measured store scores 4 of 10 and the mean is 3.79. Nothing here is a pass or fail verdict on a business; it is how much of the copy matches policy language.

Rejection-risk distribution, 334 readable brands, 15 September 2026. Our data

Are the brands that already advertise any different?

165 of 334 readable stores carry at least one advertising tag in their page source, which is a proxy for "this brand buys ads" and not proof of live spend. A harder cut: ads for 42 brands were visible in the public Meta Ad Library on 15 September 2026, of which 36 are in this crawl. Absence from a keyword search is not evidence that a brand runs no ads, so this is a floor, never a ceiling.

Median rejection-risk score in this group: 3 of 10, against 4 of 10 across all 334 readable brands.

A site can carry several. Google Tag Manager is present on 129 sites and can load a tag out of sight, which is why a tag manager alone is scored unknown.

What do the brands that look review-safe do differently?

73 of 334 readable stores are clean on all four copy signals. Reading their pages against the rest, the difference is not subtlety or clever wording. It is subtraction. Four habits separate them.

1. The product page describes the compound, not a person. Sequence, form, quantity, purity, storage, paperwork. No sentence about what happens to a body.
2. The research-use line sits on the product page, in the buying flow, not only in the footer or the terms. 208 of 334 stores do this (62%); 62 carry the line somewhere else on the site but not on a product page we read. In the lowest-risk decile, 14 of 34 carry it on a product page.
3. Substantiation replaces persuasion. A certificate of analysis, a test method, a batch number. 260 of 334 stores publish testing or a certificate (78%); in the lowest-risk decile it is 24 of 34 (71%).
4. A research purpose is confirmed before checkout. 140 of 334 stores ask for an age or a researcher attestation anywhere a visitor meets it.

34 stores sit in the lowest-risk decile of this crawl, all of them scoring 0 of 10. We name the 12 of them whose good result rests on pages we actually read: the research-use line appears on a product page, testing is published, and at least one product page was in the crawl. Naming here describes public copy on one date. It is not an endorsement of any company or product, and it is not advice about anything except advertising copy.

One thing the numbers show that is not a recommendation

Several of the lowest-scoring stores list compounds under internal codes rather than compound names. The score falls because the score reads words. That is not a technique to copy: Google's Unapproved substances policy separately names "Products with names that are confusingly similar to an unapproved pharmaceutical or supplement", and anything whose purpose is to read differently to a reviewer than to a buyer is the fastest way to lose an account. A catalogue behind a sign-in scores lower here for the same mechanical reason: less was visible to read. The method section says what that does and does not prove.

How do you run this check on your own store?

Six steps, about twenty minutes, no tools beyond a browser.

1. Open your store the way a reviewer does. Your home page, then three product pages, then the cart. Not the terms page; a reviewer rarely gets there.
2. Search each page for the five families in the tables above. A dose amount, an injection word, a cycle or protocol, a how-to-use line; heal, repair, recovery, fat, muscle, anti-aging, sleep, libido, immune; any GLP-1 or prescription-drug name; before and after; a first-person account of taking it.
3. Find your research-use line and note where it is. If it is only in the footer, a reviewer opening a product page from your ad never sees it.
4. Read your product titles alone, with no page around them. That is what appears in a shopping feed and in an ad headline, with nothing to soften it.
5. Check what a first-time visitor is asked to confirm before a cart exists.
6. Score yourself with the weights in the method below and fix the two heaviest lines first.

If you would rather paste text than read pages, the free rejection-risk check at wealthmarketing.ai/peptides runs the same phrase families in your browser. Nothing you paste is sent anywhere. The longer write-up of the policy text, with the FDA letters that explain why a research-use disclaimer does not save a page, is at can you advertise research peptides in 2026.

Method, in short

• The list. 411 US research-peptide brand domains, compiled from vendor directories and review roundups, search results for the category, and the brands seen advertising in the public Meta Ad Library. Sources of the domain list: adlib_active (42), seeds (377), send_list (228), send_list_held (142).
• How they were read. Plain GET requests with a browser user agent, a twelve second timeout, eight sites at a time and one request at a time per site with a pause between them. At most six pages per brand: the home page, a collection page when product links were not on the home page, up to three product pages, and the cart page.
• What was not done. Nothing bought. No form submitted. No account created. No checkout started. No login anywhere. No message, no call, no email to any brand as part of this crawl. The Meta Ad Library was read through its public search, never behind a login.
• Scoring. Each signal is pass, fail or unknown with the evidence quote, the URL and the date. Unknown is never counted as a fail. The rejection-risk score adds 2 for dosing or human-use language, 2 for disease or outcome claims, 2 for a named GLP-1 or prescription drug, 1 for before-and-after or use testimonials, 1 for no age gate or researcher attestation, and 2 when no research-use line appears anywhere (1 when it appears but not on a product page), capped at 10.
• What it does not do. It does not open pages with a real browser, so copy drawn entirely by script is missed. It does not read review text inside a widget, which is why the before-and-after signal is unknown for 175 brands rather than a pass. It does not see the ads themselves, only the destination. It does not judge products, companies, legality, safety or purity, and it is not legal advice.
• A low score can also mean less was visible. Six pages per brand is a sample, catalogues behind a sign-in show less, and stores that list compounds under internal codes carry fewer matchable words. The score measures the copy a reviewer can read on those pages on that date, which is the same thing a reviewer does, with the same limit.
• Names. A brand is named on this page only when it is in the lowest-risk decile and the research-use line was on a product page we read and it publishes testing. Nobody is named for a poor result. Everything else is a count.
• Unreachable. 77 of 411 sites could not be read with an ordinary browser request and are left out of every share on this page: botwall (28), empty or script-rendered (26), dns (9), http:403 (5), timeout (3), http:503 (2), http:404 (2), http:523 (1).
• Reproducibility. The engine is a single stdlib Python file and the six steps above reproduce any brand's result by hand. Re-run monthly; the date on this page is the date of the crawl.

Want your own store's result?

Every brand on the list has a nine-line scorecard in this dataset, with the quote, the URL and the date beside each line. We do not publish it. Three lines on what we do with it:

1. We build the compliance-first creative system per SKU, the landing pages built for review, the account hygiene and the tracking, and we run the campaigns. Flat $5,000 a month, month to month.
2. You fund ad spend in accounts you own, and everything we build stays with you.
3. Nothing whose purpose is to get around review, ever. The reviewer and the buyer see the same page. GLP-1 compounds and PT-141 are off limits in everything we produce.

Frequently asked questions

Sources

1. Ours Wealth Marketing AI, The Peptide Ad Review, 411 US research-peptide brand websites, crawled 15 September 2026; 334 readable, 77 unreachable. Public pages only, GET requests only. Data: data.csv.
2. Ours Wealth Marketing AI, public Meta Ad Library search for this category, read 15 September 2026: 42 brands with ads visible (search terms: BPC-157; research peptides). The Ad Library Graph API refused this application ("Application does not have permission for this action"), so the public Ad Library search was used instead.
3. Primary Meta, Advertising Standards, Drugs and Pharmaceuticals, read 15 September 2026. transparency.meta.com
4. Primary Meta, Advertising Standards, Health and Wellness, read 15 September 2026. transparency.meta.com
5. Primary Google, Advertising Policies, Healthcare and medicines, read 15 September 2026. support.google.com/adspolicy/answer/176031
6. Primary Google, Advertising Policies, Unapproved substances, read 15 September 2026. support.google.com/adspolicy/answer/15595718
7. Primary Google, Advertising Policies, Unapproved pharmaceuticals and supplements (the named-example list), read 15 September 2026. support.google.com/adspolicy/answer/2423645

Compliant paid acquisition for research peptide brands  ·  Can you advertise research peptides in 2026  ·  The 9pm Test, Los Angeles