Overview

Error 90146 is a listing validation error that Amazon triggers when a seller attempts to update a product attribute with a value that drops too sharply compared to the existing entry. This safeguard is designed to prevent accidental or unauthorized changes to critical listing data — such as price, weight, or item dimensions — that could mislead customers or disrupt catalog integrity.

Key Points / What Sellers Need to Know

• Dual-threshold trigger — This error fires only when a proposed change exceeds both an absolute decrease limit and a percentage decrease limit simultaneously, meaning small corrections typically pass through without issue.
• Attribute-specific limits — Amazon sets different thresholds for different attributes; a weight field may tolerate a different percentage drop than a price or quantity field.
• Template format — The raw error reads: "Value for '{attribute_name}' must not decrease by {delta_limit} or more and decrease by {percentage_limit}% or more," where the placeholders are replaced with the specific attribute name and threshold values relevant to your submission.
• Feed and flat file submissions — This error most commonly appears when uploading inventory files, flat files, or making updates through the SP-API, rather than during manual edits in Seller Central's listing interface.
• Not a permanent block — The error does not lock your listing; it simply rejects the specific update attempt, allowing you to correct the value and resubmit.

How This Error Works

Amazon's catalog system applies a set of business rules to protect listing accuracy. When you submit an update to an attribute — for example, changing a product's package weight from 10 pounds to 3 pounds — the system calculates both the absolute decrease (7 pounds) and the percentage decrease (70%). If both figures exceed the predefined limits for that attribute, the update is rejected with Error 90146. This dual-condition approach means that a modest absolute drop on a high-value attribute or a large percentage drop on a very small value will usually still be accepted. The error is specifically targeting changes that are large in both absolute and relative terms, which Amazon considers likely to be mistakes.