Advertising Peptides on Meta: What Actually Gets Approved

the single most useful thing to understand about advertising peptides on meta is that the review system is not looking for a product list. it is looking for claims, implied personal attributes, and destination-page consistency. that's why two advertisers can run the exact same SKU and one runs 294 live ads while the other gets a permanent account restriction in week two. the difference is almost never the product. it's the language, the creative, and the page.

this is an operator guide to what actually clears review, written from running spend on med spa and supplement-adjacent accounts. it is not legal advice and it is not medical advice. peptides sold as research chemicals are for research use only, are not for human consumption, and are not evaluated by the FDA. talk to a healthcare attorney about your specific product, your claims, and your state before you spend a dollar.

how meta actually reviews a peptide ad

meta's review is layered. an automated classifier scores the ad text, the image or video frames, the landing page, and the advertiser's history. most ads never touch a human. the classifiers are trained on the policies that matter here: health and wellness, personal health, unsafe substances, prescription drugs, and misleading claims. each of those triggers differently.

• —personal attributes — the classifier flags copy that implies it knows something about the viewer's body, health, or condition. 'struggling with stubborn belly fat?' is the textbook violation. it is not about whether it's true, it's about the implication that meta targeted someone because of a health condition.
• —health and wellness / weight loss — before-and-after imagery, scale shots, body-part close-ups, and 'lose X lbs' claims are restricted or prohibited depending on framing. weight loss ads also carry a minimum age restriction of 18+.
• —prescription drugs — semaglutide, tirzepatide, and other Rx compounds sit under a policy that requires meta's online pharmacy certification for most advertisers. an uncertified account running the compound name in ad copy is a fast disapproval and a fast strike.
• —unsafe substances — vague or unapproved 'research chemical' framing that reads as recreational or self-administered gets caught here.
• —misleading claims — guarantees, 'clinically proven', 'FDA approved', percentages without substantiation.

each disapproval is logged against the ad account, the business manager, the page, the domain, and the payment method. this is the part most operators miss. rejections compound. an account with fifteen policy hits in thirty days gets reviewed differently than a clean one — same ad, different outcome.

what gets approved: the four compliant angles

there are four categories of peptide ad that consistently clear review. everything that works is a variation of one of these.

1. 1.brand and category ads — the ad sells the company, the sourcing, the testing, the catalog. 'third-party tested research peptides. COAs on every lot. ships in 48 hours.' no outcome claim, no body reference. these are the workhorses.
2. 2.education and authority ads — the creative explains what a peptide is, how research use only classification works, what a certificate of analysis means. these build a warm audience you can retarget with offers later, and they almost never get rejected.
3. 3.offer and logistics ads — free shipping thresholds, bundle pricing, restock announcements, subscription discounts. commerce language, not health language.
4. 4.provider-facing ads — for med spas selling to practitioners or to their own patient list, ads about consultations, memberships, and appointment booking rather than about compounds. 'book a wellness consultation' clears where 'get on GLP-1 today' does not.

that number matters because it is the real proof of a compliant system. you do not get to 294 simultaneously active ads on a med spa account by finding a loophole. you get there by building a creative library where the baseline version of every concept is approvable, then iterating on hooks inside that safe frame. livv well scaled over 1,200% in six months on that structure, with top creatives running between 6.79 and 14.96 ROAS.

the copy rules that decide approval

write ad copy about the product and the company, never about the reader's body. this one rule prevents the majority of disapprovals.

• —no second-person body language. 'your belly fat', 'your low energy', 'tired of being overweight' — all violations. replace with third-person or product-first framing.
• —no outcome promises. 'lose 20 lbs in 30 days' is out. 'research-grade purity, verified by third-party HPLC' is in.
• —no before/after implication in text either — 'transform', 'finally', 'get your body back' read as outcome claims even without an image.
• —no medical framing. 'treat', 'cure', 'heal', 'therapy', 'dose', 'protocol' push you into drug-claim territory for a product classified research use only.
• —no Rx compound names in copy unless you hold the relevant meta certification and the appropriate licensure. do not use them decoratively.
• —avoid urgency built on health fear. scarcity on inventory is fine. scarcity on 'your health is running out' is not.

creative: what the classifier sees in your video

meta samples frames from video and runs them through image classification. it reads on-screen text. so a clean voiceover with a 'lose 30 lbs' text overlay on frame four gets caught. treat every frame as ad copy.

• —no body close-ups, no scale shots, no shirt-lift, no side-by-side comparison composition — the composition alone reads as before/after even without labels.
• —vials, labs, packaging, COA documents, shipping, and founder-to-camera all clear cleanly and, in practice, perform.
• —keep on-screen text as compliant as spoken text. captions get read.
• —no syringes pointed at skin or injection demonstrations for a research-use-only product. it directly contradicts 'not for human consumption' and it's a fast rejection.
• —if you use a person on camera, have them talk about the business, the sourcing, or the category — not about what happened to their body.

the landing page is half the review

meta crawls the destination. a perfectly compliant ad pointing at a page with a testimonial that says 'i dropped 40 lbs in six weeks' gets the ad rejected and the domain flagged. the page is where most operators lose accounts, because they clean up the ad and never touch the site.

1. 1.put the research use only statement above the fold on product pages, not buried in the footer. 'for research use only. not for human consumption. not for diagnostic or therapeutic use.'
2. 2.remove every dosing instruction, protocol chart, and reconstitution guide that reads as human-use direction on a public page.
3. 3.audit testimonials and reviews. user-generated reviews making outcome claims count against you.
4. 4.publish real policy pages: shipping, returns, terms, privacy, contact with a working phone and physical address. thin sites get flagged as low-quality destinations independent of health policy.
5. 5.make sure the page you send traffic to actually matches the ad. mismatched destination is its own violation category.
6. 6.if you sell both compliant retail products and restricted ones, do not put them on the same landing page.

the ad, the page, and the checkout are one artifact as far as review is concerned. fix one and leave the other two broken and you've fixed nothing.

account structure that survives

structure is the difference between a bad week and a dead business. build it before you need it.

• —one verified business manager, fully verified with real legal entity documents. verification is not optional at scale.
• —separate ad accounts by risk profile. never run your most aggressive test creative in the account that carries your scaled winners.
• —warm new accounts. small budgets, clean brand-level creative, a week of clean delivery before you push volume.
• —keep a separate page for each brand. page-level violations follow the page.
• —assign real admins with two-factor enabled. security holes look like compromise, and compromise looks like fraud.
• —document every disapproval and appeal. when you talk to support you need a record, not a memory.

what to do when an ad gets rejected

appeal once, quickly, and only if you genuinely believe the ad is compliant. don't mass-appeal. spamming appeals on ads that are actually in violation trains the account history against you. if the ad was borderline, rewrite it instead — a clean rewrite that gets approved is worth more to your account health than a won appeal. if the account gets restricted, request review through business support and stop launching new ads in the meantime.

and never respond to enforcement by creating new entities, buying accounts, or hiding the destination page. those are the actions that turn a recoverable restriction into a permanent ban across every asset connected to you. the operators who run for years run clean.