Restricted Product Policy Violation You Did Not Deserve: How To Appeal

Twenty two ASINs came down in a single morning. The policy cited was Restricted Products, and the specific reason given was pesticides. The brand makes toys.

We understand how that lands, because there is no pesticide anywhere in that business, no chemical, no EPA registration to produce and nothing to confess to. The notice does not care. It arrives with the same wording it would use for an unregistered insecticide, the ASINs stop selling, and the Account Health rating drops while somebody tries to work out what Amazon thinks it found.

The turn in this story, and the reason the appeal eventually worked, is that Amazon had not made a judgement about the products at all. It had made a judgement about a few words in the listings. Once that is clear, the appeal stops being an apology and becomes an argument about classification, which is a much better argument to be having.

What Amazon actually flagged

Restricted Products is a wide policy, and pesticides sit inside it. The US definition comes from federal law rather than from Amazon, which is why the enforcement feels so literal. The US EPA (government primary source) defines pesticides as substances "intended to prevent, destroy, repel, or mitigate pests", and a device as "any instrument or contrivance (other than a firearm)" intended for trapping, destroying, repelling or mitigating any pest.

The word doing the damage in both definitions is intended, and intent is read from what the listing says. So a product becomes a pesticide or a device, for enforcement purposes, at the moment its listing claims to do one of those things, whatever is physically in the box. Microorganisms count as pests, which is how an antimicrobial claim on a fabric lands in the same policy bucket as an insecticide.

Amazon's own pest control products and pesticides policy and its wider restricted products policy (both Amazon primary) sit on that legal definition and add a documentation requirement: a pesticide marking attribute, an EPA registration or establishment number where one applies, and label images matching the physical label. A listing that trips the classification without those fields completed is a listing with missing paperwork, and that is what the takedown is technically for.

Two independent write-ups describe the same trigger list and the same fix, one from Blue Wheel, published 2022, and one from Goat Consulting, updated 2026. Both are third party, both list wording such as antimicrobial, antibacterial, mildew resistant and sanitises as the language that flags an ordinary product, and both note that images carry claims exactly as text does.

Why well-run catalogues get caught

There is an assumption that this only happens to sellers cutting corners on chemicals. That is not what we see, because three ordinary situations produce it.

The first is inherited copy. A phrase like "antibacterial coating" was written by a supplier or a freelancer years ago, sat harmlessly through several enforcement cycles, and then a sweep picked it up. Nobody in the current team wrote it and nobody remembers it being there.

The second is the A plus content and the image stack, the places people forget. A claim in a lifestyle image or an infographic module counts, and so does a claim printed on packaging that appears in a photograph, which catches sellers who cleaned their bullets carefully and thought they were finished.

The third is scale. Enforcement runs across a catalogue rather than a listing, so one phrase reused in a template across twenty two ASINs produces twenty two notices on the same morning. That is why these arrive in clusters, and why the fix is a catalogue exercise rather than a listing one.

None of this is careless. It is what happens when copy is written once and lives for four years while the enforcement around it tightens. The annoying part is that the phrase was fine when it was written.

Reading the Account Health dashboard properly

The dashboard holds more detail than the email that brought you there. The path is Performance, then Account Health, and the section that matters is Policy Compliance, which plenty of sellers still call the Policy Center from an older layout.

Inside it, restricted product violations appear in their own group, each naming the ASIN, the policy cited and the date. That naming is the point of going there. The email says something is wrong. The dashboard says which ASINs, under which policy, which is the difference between guessing and diagnosing. Our page on Amazon policy warnings covers how the different violation types read.

Above that sits the Account Health rating, a score from 0 to 1000. Two independent sources describe the same bands: green and healthy from 200 to 1000, yellow from 100 to 199, red below 100. Those sources are Riverbend Consulting, last updated January 2026, and RetailDogma, both third party, and they agree on the numbers.

Which means a rating of 200 is not comfortable. It is the bottom edge of the green band, where a handful of unresolved violations have eaten everything above it, and it is the number the toy brand was sitting on. A score can read as technically healthy and still be one bad week from something else.

How to tell whether this is a false flag or a real one

Worth settling honestly first, because the two situations need different submissions and a wrong guess costs a rejection cycle.

1. Read what the product physically does, not what the listing says it does. If nothing in the box is intended to prevent, destroy, repel or mitigate a pest, the product is not a pesticide and the claim is the only thing that made it one. That distinction is the spine of the appeal.
2. Search the full listing for the trigger vocabulary. Title, bullets, description, backend search terms, A plus modules and every image, including packaging visible in photographs. The phrase is almost always there, and it is almost always somewhere nobody looked.
3. Check whether the same phrase appears on other ASINs. If it does, those listings are on the same list and have not been swept yet, which turns the job from an appeal into a catalogue correction with an appeal attached.
4. Open the compliance fields on the listing. If a pesticide marking attribute has been set to a pesticide value at some point, whether by you, a supplier or a bulk upload, the classification is coming from your own data and no amount of copy editing will move it.
5. Compare when the copy was written to when the notice arrived. A gap of years points to an enforcement sweep rather than anything you changed, and knowing that keeps the appeal factual instead of apologetic.

What to check before you appeal

The instinct is to write to Amazon the same day, worth resisting for a few hours, because an appeal submitted before the listing is clean gets rejected on its own terms and the second attempt is always read more sceptically.

1. Confirm every affected ASIN is on the list. Working from the dashboard rather than the emails catches the ones flagged quietly, and a partial fix produces a second wave a fortnight later.
2. Remove the claim everywhere before the appeal goes in, not after. An appeal that promises to remove language reads as a promise. An appeal that describes language already removed reads as a completed correction, and those are treated differently.
3. Save evidence of the before and after. Dated screenshots of the old and new copy turn an assertion into something a reviewer can check in ten seconds.
4. Get the compliance attribute right. Where the product genuinely is not a pesticide, the compliance fields need to say so, because a listing arguing one thing in the appeal and another in its own data does not survive review.
5. Decide what you are actually asking for. Reinstatement of the ASINs and removal of the violation record are two separate outcomes, and an appeal that only asks for one tends to get only one.

Building an appeal that proves root cause rather than apologising

Most rejected appeals we are asked to look at fail the same way. They are polite, they accept responsibility, they promise to do better, and they never say what caused the flag. Reviewers are not looking for contrition. They are looking for evidence that the seller understands the mechanism, because that is the only thing that predicts whether it happens again.

The structure Amazon's process expects is consistent across violation types: root cause, corrective actions taken, preventive measures. Sources: Emplicit and Goat Consulting, both third party, both describing the same three headings.

1. State the classification plainly in the first two lines. The product is not a pesticide or a pesticide device under FIFRA, and the listing contained wording that read as a pesticidal claim. A reviewer who understands the case in ten seconds is a reviewer who reads the rest.
2. Name the exact phrase and where it sat. "The bullet point read antibacterial finish" is a root cause. "Our listing contained non-compliant content" is not, because it could describe anything and proves nothing.
3. Explain how the phrase got there. Supplier copy carried over at launch, a translation, a template reused across a product family. This is the part that separates a root cause from a description, and it is the part almost every rejected appeal is missing.
4. List the corrective actions in the past tense with dates. Wording removed from titles, bullets, description, backend terms, A plus modules and images, on a stated date, across a stated set of ASINs. Past tense matters, because it is the difference between work done and work promised.
5. Attach structured evidence rather than a paragraph about it. Before and after screenshots, the ASIN list, the compliance field showing the product declared as not a pesticide, and the label or specification showing no pesticidal substance or mechanism.
6. Give preventive measures specific enough to be checkable. A named banned-phrase list applied at copy sign-off, a quarterly catalogue scan for that vocabulary, and a rule that supplier copy is rewritten rather than pasted. Vague promises about "enhanced review processes" are what a reviewer skims.
7. Follow up when the first reply is a template. It often is, and it is the thing that irritates us most about this process, because the reply frequently answers a question nobody asked. The answer is to resubmit the same evidence with the reference numbers rather than to rewrite the argument into something weaker.

What this looked like on the toy brand account

Back to the twenty two ASINs. The products were toys, the flag was pesticides, and the cause was wording in the listings that read as a pesticidal claim rather than anything in the products.

Our team worked from the catalogue rather than the notices, so the sweep covered every ASIN carrying the same phrasing and not only the flagged ones.

The copy was corrected across titles, bullets, descriptions and imagery, the compliance fields were set to reflect what the products actually are, and the appeal was built around the classification argument with before and after evidence for each ASIN. Then it was followed up, repeatedly, which is the unglamorous part that decides these cases.

All twenty two ASINs were reinstated within seven days, and the Account Health rating came back from 200 to 1000.

We are deliberately not presenting that as a timeline you can expect. It was a clean false positive with one identifiable cause, evidence that was straightforward to assemble, and a catalogue small enough to sweep in a day.

Cases with genuine compliance gaps move differently, and anyone quoting you seven days before looking at your account is quoting a marketing number. If policy work is a recurring cost on your account rather than a one-off, our account health management page explains how our team handles it.

How to keep the next sweep from reaching you

Enforcement on this policy tightens rather than relaxes, so the useful goal is a catalogue that does not depend on a particular sweep being lenient.

1. Keep a banned-phrase list and apply it at copy sign-off. Antimicrobial, antibacterial, antifungal, sanitises, disinfects, kills germs, mould and mildew resistant, repels. Twenty minutes to write, and it prevents the entire situation described above.
2. Scan the catalogue for that vocabulary once a quarter. A search across an exported listing report finds in one pass what an enforcement sweep would, and finds it before the ASINs come down.
3. Treat images and A plus content as copy. They carry claims exactly as text does, and they are where the leftover phrases survive longest because nobody re-reads an infographic from 2021.
4. Rewrite supplier copy rather than pasting it. Manufacturer marketing language is written for a catalogue with different rules, and it is the single most common origin of these phrases.
5. Describe the benefit rather than the mechanism. A shopper cares that a fabric dries fast and washes clean. Saying that instead of claiming it kills bacteria keeps the argument out of federal pesticide law entirely, and it usually converts better because it answers what the buyer was actually worried about.

That last point is where compliance and conversion agree. Our view, and it is not the popular one in this industry, is that a listing is a narrative rather than a keyword container. The phrases that trigger this policy are usually there because someone was chasing a keyword, not describing the product. Writing the use case, the material and the situation instead tends to be both safer and more persuasive.

What we would do first if this were our account

Open the Policy Compliance section and write down every affected ASIN with the policy cited and the date, because the emails are an incomplete record and the dashboard is not.

Then find the phrase. It is nearly always one phrase, reused, and locating it settles whether this is a false positive or something needing a different conversation. Only then is it worth touching the listings, because the sweep has to cover the unflagged ASINs carrying the same wording, otherwise the notice returns in a fortnight and the second appeal starts from a worse position.

Then the appeal, built around root cause, written in the past tense, with evidence attached rather than described. If ASINs are already off the page, our note on what happens when a detail page has been removed covers the parallel catalogue work.

We cannot promise reinstatement, and we would not trust anyone who does, because some of these cases turn out to have a genuine compliance requirement underneath them and a few need a lawyer rather than a consultant.

What we can tell you, before you spend anything, is which category yours is in, which ASINs are exposed and what a realistic appeal looks like. That is what the free, no-obligation audit covers, and if the honest answer is that your listing genuinely does need an EPA registration, we will say that too.

Related guides

• Submitting a compliance document that passes
• Handling a Used Sold as New complaint
• The ODR, LSR and VTR metrics behind most warnings

Common questions about restricted product policy violations