What exactly does the Amazon A-to-z Guarantee cover?

It covers three situations: an item not delivered, an item arrived damaged or defective, and an item not as described in the listing. If Amazon upholds the buyer's claim under any of those grounds, the seller loses the sale value and return shipping, and the case counts against your Order Defect Rate.

Does the A-to-z Guarantee apply to FBA sellers or only marketplace FBM?

It applies to both. FBA sellers still receive A-to-z claims when the product is defective, mislabeled, or not as described, because fulfillment by Amazon does not override your listing accuracy obligation. FBM sellers face additional exposure from tracking and delivery disputes that compound the claim.

How many A-to-z claims can I absorb before losing selling privileges?

It depends on your order volume because the threshold is a percentage, not a count. Your Order Defect Rate must stay below one percent over a trailing ninety-day window. At ten thousand orders per month, that means no more than roughly ninety defective outcomes; at one hundred thousand, about nine hundred. The absolute number that triggers suspension scales with your sales.

Can I dispute an A-to-z claim that Amazon has already upheld?

Yes, you can appeal through Seller Performance or by contacting Seller Support within a limited window after the ruling. Your best chance is specific evidence: tracking confirmation, pick-and-pack records, photos of the item at fulfillment, and a written explanation tying the listing data to what was actually shipped. Generic apologies do not overturn a ruling.

Amazon Seller Operations

How Amazon's A-to-z Guarantee Shapes Your Seller Account Health

By VisibleProducts · July 27, 2026 · 6 min read

A-to-z Guaranteeaccount healthseller metricscatalog accuracydispute resolution

A long fluorescent-lit warehouse aisle stretching into the distance, towering industrial shelving units on both sides stacked with sealed cardboard boxes in various sizes and shades of brown, a yellow forklift parked mid-aisle with its forks raised slightly off the concrete floor, overhead LED tube lights casting even cool light across the scene, the perspective looking straight down the corridor with the far end receding to a vanishing point, no people visible

What the Guarantee Actually Covers

The A-to-z Guarantee is Amazon's buyer-protection policy that covers three specific situations: an item was not delivered, it arrived damaged or defective, or it was not as described in the listing. If a buyer files under any of those three grounds and Amazon sides with them, the seller absorbs the refund cost plus any return shipping. It is not a general return policy; it is a guarantee that what the listing promised will be what the buyer receives, intact, on time.

A common misconception is that A-to-z claims are rare or only apply to marketplace sellers. They are filed every day, and they hit both FBA and FBM accounts. FBA sellers still get A-to-z claims when the product itself is defective, mislabeled, or not as described, because Amazon's warehouse fulfillment does not override the listing accuracy obligation. FBM sellers face an additional layer: if your tracking shows delivered but the buyer says the item never arrived, or it arrived in a different condition than listed, the claim lands on you.

What makes A-to-z distinct from a standard return is the financial and reputational penalty. A regular return costs you shipping and restocking. An A-to-z claim that Amazon upholds counts against your Order Defect Rate, which is one of the three pillars of account health alongside late shipment rate and valid tracking rate. Cross the ODR threshold of one percent for thirty consecutive days, and Amazon can suspend your selling privileges regardless of how many units you move.

How Claims Hit Your Account Metrics

The Order Defect Rate is calculated as the sum of A-to-z claims, chargebacks, and negative feedback divided by total orders over a trailing ninety-day window. At a volume of fifty thousand orders per month, one percent means five hundred defective outcomes. In practice, a seller moving two to three thousand units a week needs fewer than twenty problematic orders in that window to cross the line. The math is unforgiving at scale: you cannot absorb sloppy listing data or inconsistent fulfillment and stay under threshold.

The second metric, late shipment rate, interacts with A-to-z directly. If your handling time promises two days but your warehouse ships on day four, and the item arrives in a state that triggers a not-as-described complaint, Amazon may classify it as both a late-shipment event and an A-to-z claim. One operational failure, two metric hits. We have seen accounts where a single week of misconfigured shipping templates produced enough compounded penalties to trigger a Performance Improvement Plan within the following thirty-day window.

The third pillar, valid tracking rate, matters most for FBM sellers. If your carrier scan shows the package in a different city than the buyer's address, or the tracking number is reused across multiple orders, Amazon flags those as invalid. When the buyer then files an A-to-z claim because the item never arrived, the invalid tracking record makes it nearly impossible to prove delivery. The combination of an invalid tracking event plus an upheld A-to-z claim is the fastest path to a suspension that I have seen in account reviews.

A single packing station seen from a three-quarter elevated angle, a half-assembled brown corrugated shipping box open on a steel workbench with crumpled bubble wrap spilling out and a small rectangular product wrapped in white tissue paper resting inside, a compact thermal label printer sitting to the left of the box, a neat stack of five identical unopened inventory units leaning against the bench edge, warm task lighting from above creating soft shadows on the metal surface, no text visible anywhere

The Catalog Accuracy Connection You Miss

Most A-to-z claims that reach my desk trace back to a listing data problem, not a fulfillment failure. A buyer orders a 12-pack of a product and receives a 6-pack because the quantity attribute was wrong in the feed. A buyer expects a specific color variant and gets a different one because the variation relationship in Merchant Center was misconfigured. A product is listed as waterproof when it is only water-resistant, and the buyer files not-as-described after the first rain. In every case, the listing told the truth to no one.

This is where treating your catalog as a legal document rather than a marketing asset changes everything. Every attribute you submit to Amazon's feed, including item name, product description, bullet points, variation attributes, quantity, dimensions, weight, and condition, becomes part of the contract between you and the buyer. When an AI shopping assistant pulls your listing data to answer a question like Is this item available in small or Will it fit a 24-inch frame, that same attribute set is what the buyer sees before purchasing. If the data is wrong, the expectation is wrong, and the A-to-z claim is inevitable.

At our scale, we run automated attribute validation across every SKU before it goes live on any channel. Dimension mismatches between the physical product and the feed value get flagged. Quantity fields are cross-checked against the actual pack count in inventory. Variation hierarchies are verified so that a color swatch does not silently map to a different size than intended. The cost of catching these errors in a pre-publish check is zero; the cost of an upheld A-to-z claim on a high-velocity SKU is a refund, a shipping charge, a data point against your ODR, and potentially a suspension review.

Responding to Claims Without Losing Money

When an A-to-z claim hits, Amazon gives you a limited window to respond. The buyer files, the system opens a case, and you get roughly forty-eight hours to provide evidence or offer a resolution before Amazon makes a determination. In that window, your response format matters as much as its content. A generic We apologize for the inconvenience is treated as an admission of fault. What Amazon wants to see is specific documentation: your shipping label showing the correct address and tracking, your inventory record confirming the right SKU was picked, photos of the item at the packing station if you have them, and a clear statement of what the buyer received versus what the listing specified.

If the claim is legitimate and the product was genuinely not as described, fight the framing rather than the fact. Offer a partial refund that covers the discrepancy without conceding the entire order. A buyer who ordered a 12-pack and got 6 can be offered a fifty percent refund plus a replacement for the missing units. This resolves the claim, keeps your ODR clean, and often earns a neutral or positive feedback update. The alternative, a full refund with return shipping on a defective item, costs you the product value, the outbound shipping, the return shipping, and the restocking labor.

For claims where you believe the buyer is in error, do not ignore the case. Silence is interpreted as non-response, and Amazon will rule in the buyer's favor by default. Provide your tracking confirmation, your pick-and-pack record, and a polite request for photos of what the buyer received. In most disputes we have reviewed over the years, the majority of upheld claims involved a genuine listing or fulfillment error on the seller side. The minority that were buyer error still required a documented, specific response to overturn.

Preventing Claims Before They Ever Happen

The single highest-leverage prevention step is getting your listing data right and keeping it right. Audit your top two hundred SKUs for attribute accuracy: does the quantity field match the physical pack? Are variation relationships mapped correctly so that selecting red small actually ships red small? Do your dimensions and weight values match what the carrier will measure? Run this audit quarterly, or more often if you onboard new products. A fifteen-minute spreadsheet check on your highest-velocity items will prevent more A-to-z claims than any amount of post-hoc dispute fighting.

Second, tighten your handling time and shipping templates to reflect what your warehouse can actually do. If your average pick-and-pack cycle is thirty-six hours but your template promises two days, you are building a late-shipment pipeline that feeds directly into A-to-z claims when the item arrives stressed or damaged. Set your handling time at ninety percent of your actual cycle, not the theoretical best case. The one-day buffer you lose in perceived speed is cheap insurance against an ODR penalty.

Third, monitor your account health dashboard weekly, not monthly. Amazon's metrics update continuously, and a single bad week can push your trailing ninety-day average toward threshold. If you see your ODR tick from zero point three to zero point seven, investigate immediately. Identify which SKUs generated the claims, find the root cause in listing data or fulfillment process, fix it, and document the fix. When Amazon does open a performance review, that documentation is what separates a quick resolution from a suspension hearing. The sellers who get suspended are almost always the ones who wait until the threshold is crossed to start looking.