Amazon Now Deactivates the Failing FBM Offer Instead of Risking the Whole Account
Amazon changed its U.S. Account Health enforcement for Fulfilled by Merchant offers effective August 31, 2026. If an FBM offer puts Account Health at risk under one of four named performance policies, Amazon says it can temporarily deactivate that offer while leaving the seller’s other active listings and overall Account Health unaffected.
This is overdue analysis of an older development, not a new policy announced today. It is still important because the change reduces the threat of a whole-account interruption while creating unanswered questions about how one order problem becomes an offer-level penalty.
Four performance policies can trigger the action
The Seller Central announcement names Cancellation Rate, Late Shipment Rate, Order Defect Rate and On-Time Delivery Rate. It does not say that every breach produces an immediate deactivation.
Amazon promises a warning email before an offer goes down. The email should identify the issue and provide resolution steps. A deactivated offer appears in Account Health under Other Policy Violations, where the seller follows the reactivation instructions.
The old and new consequences are materially different. Under Amazon’s description of the prior system, listings that missed required metrics could put the entire account at risk. Under the new system, only the affected offer is deactivated and the rest of the catalog remains active.
That is real relief for a diversified seller. It is less reassuring when one SKU produces most of the business’s revenue.
Amazon did not explain the attribution method
The announcement is short. It does not define how Amazon selects the “affected offer” when the named metrics are normally monitored across orders and time windows. It also does not publish an offer-level trigger, the amount of advance warning or the duration of a temporary deactivation.
A detailed analysis of the rollout documents seller objections to treating a late shipment, carrier delay or customer claim as a property of the SKU. Those objections are anecdotal responses, not evidence that Amazon’s attribution is wrong in every case.
The low-volume problem is still worth testing. One defect on a product with few orders produces a very different percentage from one defect on a product with thousands. Amazon has not said whether it uses minimum-order requirements, confidence thresholds or another safeguard before taking an offer offline.
Do not assume the familiar account-level thresholds answer that question. The policy pages define the underlying metrics. They do not fully explain the new offer-selection method.
Your evidence file should begin before a warning
Create a weekly FBM exception report by SKU. Include seller cancellations, late confirmations, promised and actual delivery dates, A-to-z claims, chargebacks and negative feedback connected with fulfilled orders. Add carrier, service, warehouse or 3PL and the Amazon delivery promise shown at purchase.
EcomCrew’s guide to Amazon’s SKU-level handling-time rules for seller-fulfilled listings explains why a catalog-wide default can create promises that some products cannot meet. Align that setting with the SKU-level evidence file so you can distinguish a bad promise from a warehouse miss.
Save Amazon’s warning email, the Account Health entry and the order-level records used in your response. If the carrier scan or customer event is wrong, preserve the original tracking page and timestamps before the data changes. If the problem is operational, document the correction rather than arguing only about the metric.
The EcomCrew glossary of Account Health metrics is useful for orienting staff, but check the live Seller Central help page before relying on any threshold. Amazon has revised several fulfillment measures during 2026, and an older target can produce a false sense of safety.
Coordinate inventory and advertising when an offer goes dark
Offer-level enforcement changes the incident response. The rest of the account can keep selling, so the business may not experience an obvious sitewide outage. A high-value SKU can disappear while ad budgets, replenishment and customer-service work continue elsewhere.
Set an alert for the warning mailbox and give one person authority to open the Account Health case. Map every advertised FBM SKU to a backup fulfillment decision. That might be a corrected handling time, a different carrier, a temporary pause, a 3PL process change or an FBA offer if the economics support it.
Do not relist the same inventory under a new SKU to evade the action. The announcement gives a reactivation path under Other Policy Violations. Follow that process and preserve the case record.
Check campaign status after an offer is removed. It is reasonable to expect an unavailable offer to stop serving normally, but Amazon did not explain advertising behavior in the policy notice. Verify what Campaign Manager did instead of assuming spend paused cleanly or moved where you intended.
Treat the narrower penalty as relief, not immunity
The new system reduces the blast radius described by Amazon. It does not make the four performance policies optional, and it does not promise that every other type of violation will stay at offer level.
Review the Other Policy Violations section now, confirm who receives Account Health emails and export the last 60 days of FBM exceptions by SKU. The next important disclosure is Amazon’s definition of “temporarily,” followed by the trigger and attribution rules that decide which offer is removed.

