HOW IT WORKS

Every word ruled against the same 109 published rules. Signed by a licensed pharmacist.

Keep the copy that sells. Lose the lines that can shut it off. We read every public claim your brand makes three times, reword each failing line to the strongest wording the rules allow so it still sells, and a licensed pharmacist signs the file in five business days.

A suppressed listing earns nothing while it is down. Set $3,500 against one week of sales on yours. Your URLs are enough to start.

Request an audit See a real finding first
The audit$3,500 flat
Up to 10 products · five business days · $1,000 to start, the rest on delivery
If a platform flags a line the pharmacist signed, we fix it and back your appeal, free, for 12 months. The limits in our terms apply.
01 The fileEvery claim, 3 passes
02 The standards109, drawn from nineteen rulebooks
03 The AI sweep244,000+ phrase checks · minutes, not weeks
04 The sign-offEvery verdict reviewed before delivery
Verdict readyDelivered day 5
The five days

URLs on day 0. A signed file on day 5.

DAY 0
Order placed

Scope confirmed in writing. We start from your URLs; anything else comes through a private upload link. No scoping call required, and $1,000 starts it.

DAYS 1 TO 2
The sweep and the standards

The engine reads every claim across your listing, A+ content, site, ads and emails, and rules each against all 109 standards.

DAYS 3 TO 4
Pharmacist review and rewrites

The engine proposes the severity and cites the statute or policy each one breaches. A licensed pharmacist reviews and signs every finding, and can change the severity or send it back. Each gets a paste-ready rewrite at the strongest wording the standard allows, so it keeps the promise customers respond to.

DAY 5
Signed report delivered

The complete findings file: severity-rated, statute-cited, with the pharmacist’s signature. If we find nothing Critical, you get that in writing. The balance is due on delivery.

DAY 30
Re-check, included

Once your rewrites go live, we re-run the scan against the same rules and issue the updated score, included.

What the rewrite keeps

Say more, not less.The rules allow more than most passing listings use.

The “keeps the benefit” line is from our scan of the 2,243 best sellers we ruled, 14 to 27 Sep 2026. The audit rewrites every line that fails to the strongest wording the rules allow, with the benefit the line reached for still in it, and rules it again before you see it. The listing keeps its sales argument and loses only the words a platform can act on.

Sleep
Reaches too far · failsTreats insomnia.Critical under our published standard
✓ Strongest allowed · passesPromotes restful sleep.*19 best-selling sleep brands use it, no finding
Says little · passesMelatonin, 5 mg.
Joint
Reaches too far · failsRelieves joint pain.Critical under our published standard
✓ Strongest allowed · passesSupports joint comfort.*33 best-selling joint brands use it, no finding
Says little · passesGlucosamine and chondroitin.
Immune
Reaches too far · failsImmune booster.High under our published standard
✓ Strongest allowed · passesSupports healthy immune function.*9 best-selling immune brands use it, no finding
Says little · passesVitamin C, 1,000 mg.

*With the FDA disclaimer. The top and bottom lines are constructed examples; the green line is counted from our scan, 14 to 27 Sep 2026. A structure/function claim still needs the brand’s own evidence. Amazon says Basic A+ Content can increase sales by up to 8%, and well-implemented Premium A+ Content by up to 20% (Amazon internal data, sell.amazon.com). We rule the A+ copy too, so it says the most the rules allow.

Pull it apart

Every claim, on every surface, ruled before the platforms do.

Watch one brand's lines get read, flagged, reworded to the strongest the standard allows, signed and recorded, then watched weekly on monitoring. See what each step protects: the listing, the ads, the payouts and your rank.

An illustrative sleep supplement bottle surrounded by the six places its claims appear (label, product page, marketplace listing, A+ content, social ad and email), every line marked green after review, and a category rank bar moved from among the riskiest to among the cleanest in its category.
Illustrative · fictional brand
Your rank against the best-sellers in your category, before and after.

Ruled. Signed. Still selling.

We review the marketing claims, not the formulation.

Check one listing free
0:35 / 0:35

Drag to turn it. Illustrative brand.

  1. 1CriticalCV-NC-12 · treats insomnia
    Treats insomnia, naturally.

    Insomnia is a named sleep disorder, so the line reads as a drug claim. Source: Marketplace listing, bullet 2.

    Who can act on itAmazon can suppress the listing. It earns nothing until fixed.
    Strongest wording the standard allowsSupports restful sleep.
  2. 2HighCV-FR-04 · clinically proven
    Clinically proven deep sleep.

    Proven is an absolute. It needs strong evidence on the finished product. Source: Product page, headline.

    Who can act on itA processor can hold payouts; the FDA can send a letter.
    Strongest wording the standard allowsFormulated with clinically studied ingredients.
  3. 3CriticalCV-NC-01 · treats anxiety / for anxiety
    Calms anxiety so you can drift off.

    Anxiety is a recognised disease. Naming it creates a drug claim. Source: Social ad, primary text.

    Who can act on itMeta and Google can reject the ad and restrict the account.
    Strongest wording the standard allowsPromotes relaxation before bedtime.
  4. 4HighCV-SB-03 · works in [X] days
    A week later, caught by monitoring
    Deep sleep that works in 3 days.

    A time-to-result claim needs trial data on the finished product. Source: A+ content block, module 2.

    Who can act on itAmazon can pull the A+ content and suppress the listing.
    Strongest wording the standard allowsHelps you fall asleep faster.

Ruled. Signed. Still selling.

We review the marketing claims, not the formulation.

Check one listing free
What we read

One audit, every channel your revenue runs through.

The review is external and read-only. You hand us URLs. We cover every claim a regulator, platform or acquirer would see.

01Amazon listing and A+

Title, bullets, A+ modules and backend keywords, every line ruled and rewritten to the strongest wording Amazon’s policy allows.

02Product pages

Claims that hold on your own site, judged as a regulator or plaintiff would.

03Meta and Google ads

Creative checked against Meta and Google health policies before it goes to review.

04Email flows

Welcome and cart emails too. The FTC treats them as advertising.

05Search and SEO copy

Keyword and meta copy, disease claims out.

06What AI tells shoppers

When available: ChatGPT, Claude, Gemini, Perplexity and Grok, asked as a shopper would, plus the strongest allowed wording worth quoting. Each answer is ruled against the same standards and each bad line is traced to its source. It is included, not sold on its own, with one re-test after your rewrites to the standard go live.

07Packaging

Label and carton copy checked before the print run, with placement for our seal ready if you pass.

08Influencer and UGC briefs

A clear do-say list under FTC endorsement rules, so creators start from wording the rules allow.

What we read, and what we never touch.

The audit is built to be safe to commission. It runs entirely on what the public can already see, and it is reported to you alone. Every audit runs under our confidentiality terms, and a mutual NDA is available to sign before you send us anything.

✓ We read only what the public can see
  • Your live listing, A+ content and product pages
  • Published ad creative and landing pages
  • Email flows you forward to us
  • What AI assistants answer when asked about your brand, when available
✕ We never touch
  • Seller Central or any account login
  • Customer data, order data or analytics
  • Your formulation or supplier documents
  • Disclosure to any platform or regulator, unless the law requires it
The three passes

The AI does the reading. That is why the audit is $3,500 flat.

A fixed rulebook and a machine sweep do the reading. A pharmacist signs the result. Nothing is sampled, and no finding ships until that signature is on it.

PASS 1
The standards
109, drawn from nineteen rulebooks

Every claim is checked against all 109 standards, drawn from FDA structure/function rules (21 CFR 101.93 · 21 U.S.C. §343(r)(6)), FTC Act §5 substantiation, Amazon seller policy, and Meta and Google ad policies.

PASS 2
The AI sweep
The engine behind 244,000+ phrase checks

Reads your pages in minutes, not weeks. It flags named conditions, drug and certification claims, and superlatives, and cites the published source behind every one.

PASS 3
The sign-off
Pharmacist-in-the-loop

Every finding is reviewed and signed by a licensed pharmacist, who can change it, and Sign stays locked until every finding has the pharmacist’s decision. That signature is what you hand a retailer, a payment processor or an acquirer.

What the engine actually doesSix steps, every audit
01Reads every word

Everything you publish on every channel, hidden backend keywords included, and, when available, what AI assistants tell shoppers about you.

02Splits copy into claims

Each sentence is judged on its own, because the law is enforced sentence by sentence. Explicit and implied, testimonials included.

03Judges each claim

Every one is checked against the full set. It passes, or it fails with a severity, the exact rule it breaks and a rewrite at the strongest wording the standard allows.

04Quotes the source

Every finding quotes your published copy word for word, with the field it came from, so you can check it against the source.

05Scores the risk

The same violation on three channels merges into one finding. Your file gets a risk score benchmarked against the best sellers we ruled in your category.

06Hands it to the pharmacist

A licensed pharmacist reviews the results, can change any of them or send them back, and signs every Claims Review Report.

The rulebook is written down, so the same phrase gets the same verdict for every brand. One sentence can fail four different rules at once. The categories, sources and report format are on the standards page. The full text ships inside every report.

The severity scale

Know which line to fix first to keep the money coming.

Every finding is rated, cited and rewritten to the strongest wording the standard allows. Critical is the class that can draw warning letters and takedowns. Fix those first.

Critical
Named condition or drug claim

“Cures anxiety,” “treats insomnia.” The lines that turn a supplement into an unapproved drug and can draw FDA warning letters.

High
Unsubstantiated or certification claim

“Clinically proven,” “#1 doctor recommended.” Prove it on demand or remove it: FTC substantiation territory.

Medium
Hedged or unverifiable language

“GMP certified” with no certificate on file, softened wording that still implies outcomes. Fix in the next content cycle.

One finding, in full

Drop the risk. Keep the sale.

CriticalNamed conditionSource: Amazon A+ content
As published

“…has also been shown to soothe anxiety.”

Rewrite the standard allows

“Promotes a sense of calm and supports restful sleep.”

Ruling. “Anxiety” is a named disease state. Treating it turns the product into an unapproved drug under 21 U.S.C. §343(r)(6), a pattern that appears throughout published FDA warning letters to supplement brands. In your report, the rating and rewrite carry a pharmacist’s signature, not a model’s guess.

Pulled from a live scan of a stress-category listing. Every finding in your report is formatted exactly like this.
One claim, start to finish, in 25 seconds: the line our AI flags, the rule and its citation, who can act on it, and three rewrites checked again that keep the keyword and the benefit. A constructed example, with no brand. Every real ruling on a listing we scan is dated in our daily public record.
The report, page by page

The file you hand over when a platform, processor or buyer asks.

A findings file built to be handed onward: to your copywriter to fix, to counsel to rely on, to a platform appeal or an acquirer’s data room as-is.

P.1
Cover and scope of review

What was read, when, under which rule sets, and the terms it was ruled under.

P.2
Risk score and peer position

Your risk score and where you sit against every ruled listing in your category.

P.3
Findings index

Every finding in one table: severity, type of claim, where it appears, and the statute. Your triage list.

P.4 on
The findings

Each quoted verbatim, cited to the rule it breaches, rated, and explained in plain language.

+
Paste-ready rewrites that keep the sale

The strongest line the standard allows for every finding, written to keep the promise customers respond to.

+
AI-answer test

When available: what AI assistants say about your brand, flagged where they repeat your risky claims.

End
Sign-off block

The reviewing pharmacist’s signature, the licence on file, the seal verdict, and a verification code a platform, processor or certifier can check.

D.30
Re-check report

After your rewrites go live we re-run the scan, included. Your claims agent then answers from your own file and rules new drafts line by line.

Read a complete sample report →See the client dashboard, with sample data →
Forty seconds: every risky line found in minutes, the strongest wording the standard allows with the keyword kept, the signed report, your rank in your category, what AI assistants tell shoppers about you, and what changed since you last looked. No sound. Sample data. Every ruling on a listing we scan is dated in our daily public record. Click around the demo yourself.
Ask before you publish, in 15 seconds: type a line, get a verdict and three rewrites from safest to strongest allowed. An AI answer in minutes, in your client dashboard. Sample data.
The software, stage by stage

Two of these stages are AI agents. A licensed pharmacist signs every Claims Review Report.

Each stage is named as what it is: an agent, a check or an automated step. We call a stage an agent only if it works from your own file toward a goal you set, and acts only when you approve or hands the work to the pharmacist.

Weekly readCheck

Reads each monitored Amazon listing and page and compares its text with the last reading on file.

RulingAutomation

Rules each line of the text that was read against the published standards, quoting the line and citing the rule.

RewordingAutomation

Writes up to three rewordings of a flagged line and rules each one again. A rewording that draws a finding is dropped.

Watch agentAgent

Follows your monitored listings and pages from week to week: when the weekly read finds changed text, it has that text ruled again, sends a new Critical or High finding to the pharmacist, and drafts a rewording for each new finding for you to send to the pharmacist or dismiss.

It never edits or publishes anything, and it never signs.

Public recordAutomation

Publishes one fingerprint of every listing ruling we hold, so that no ruling can be backdated.

Your claims agentAgent

Answers your questions from your own file, rules a line you want to publish and proposes rewordings, showing the finding or the rule each answer relied on.

It never approves, signs or certifies an answer, and it never edits a listing. An answer is an AI draft until the pharmacist signs it.

No agent and no automated step approves, signs or certifies anything. A licensed pharmacist signs every Claims Review Report. The five-point test, and each stage with its limit →

Why brands buy the auditWhat one flag costs, who decides whether you get paid, and how much more you are allowed to say.
What a claim can cost

What one bad line can cost you.

The audit takes five days. Brands buy it to avoid everything below, and none of it arrives as a fine.

The listing goes down

A suppressed ASIN earns nothing while it is down, and the way back is an appeal built from evidence you either have or do not.

The money you earned is held

A processor review puts a reserve on the account. The revenue is yours and you still cannot spend it, and the reviewer wants documentation, not an explanation.

Acquisition stops

A restricted ad account can end paid traffic the day it happens. Google’s strikes apply to the account, not the ad: an account on hold cannot run any ads.

The exit is priced down

Claim exposure is a diligence line. An acquirer prices what it cannot verify, and a signed file is something it can verify.

What you get back
Say more, legally

Every failing line comes back rewritten to keep the promise that sells. The point is rarely to say less.

A list of passing claims to write from

The claims that passed, in writing, so the next campaign does not bring back a line the audit removed.

Your rank in your category

One number that tells you whether fixing your copy is urgent or can wait.

Remove what takes revenue away. Keep every line that earns it.

Who asks

Four gatekeepers decide whether you get paid.

The FDA, the FTC and Health Canada write the rules. Every gatekeeper reads them. Enforcement arrives late, and in public. We apply the same rules first, in private, on your side.

The platforms now read claims with AI, and the machines keep getting better. Amazon’s 2025 Trustworthy Shopping report describes AI scanning billions of attempted changes to product detail pages daily. Google says it stopped over 99% of policy-violating ads before they ran in 2025, with Gemini-powered tools. Amazon’s shopping assistant, renamed from Rufus to Alexa for Shopping in May 2026, answers shoppers from your listing details. Never flagged doesn’t mean cleared. It means no machine has acted yet.

FDA, FTC, Health Canada and platform policy→ One signed file
MarketplacesAmazon · Walmart · TikTok Shop
Ad platformsMeta · Google · TikTok. Google strikes apply to the account: an account on hold cannot run any ads.
Payment processorsStripe · your acquiring bank
Retail buyersVendor questionnaires · category buyers
Claim territory

The strongest thing you are legally allowed to say is usually not what you are saying.

Most brands hedge into language so soft it barely names the benefit. A minority overreach into claims that draw enforcement. The permitted middle is wider than either group assumes, and it is where the revenue is.

What you say now
“Supports wellness.”
“Made with quality ingredients.”
“May help you relax.”
“A gentle formula.”
Says little · no enforcement risk
What you may say · permitted
“Helps maintain healthy sleep cycles and supports a normal stress response.”
“Third-party tested for identity and potency in an ISO 17025 laboratory.”
“Promotes relaxation before bedtime.”
“Formulated without habit-forming ingredients.”
Keeps the benefit and passes · where we put you
Competitors say · draws enforcement
“Clinically proven to cure insomnia and reduce anxiety.”
“Pharmaceutical grade. Doctor approved.”
“Balances hormones and eliminates stress.”
“Non-habit forming.”
Sells until a platform reads it · warning letter, suppression

The middle column is not playing safe. These are stronger claims you are already allowed to make and are not using, written into the copy that does your selling.

$1,000 starts it. The rest on delivery.

Find the lines a platform can act on, and the wording that lets them say more. Send your URLs. No scoping call required. Not ready? Try the free listing check first and see a real finding on your own copy.

Request an audit Check a listing free
Claims Verified

The independent standard for supplement and telehealth claims. Our AI rules a brand’s public supplement claims against 109 published standards. Each failing line is rewritten to the rules. A licensed pharmacist signs every Claims Review Report.

audit@claimsverified.org
Toronto · 1200 Bay St, Suite 1201
New York · 228 Park Ave S, NY 10003
Services How it works Free listing check Listing and Launch Clean Monitoring Listing Rescue and appeals Warning-letter response Telehealth: FDA Letter Match M&A diligence Proof on demand Request an audit Book a call
Research Research and data Claims by category Claim rulings Enforcement tracker Enforcement wire Telehealth letter index Daily record
Check our work Standards and method Sample report Corrections What this does not tell you How to verify us Our AI agents and their limits
Company About us Verify a signed file Partner program Client login Common questions Contact
Claims Verified is an independent business, not a law firm, and is not affiliated with the FDA, the FTC, Health Canada, Amazon or any platform. It provides supplement claims review. We do not test product, verify ingredient identity or potency, provide legal advice, or represent clients before any agency. Nothing on this site constitutes a guarantee of regulatory outcome.
© 2026 Claims Verified Privacy Terms Unsubscribe Every Claims Review Report signed by a licensed pharmacist Published from Toronto · 109 current standards · every figure recomputed from the scan databases before each deploy