Amazon Outage Sparks Seller Concerns Over Uncharged Ad Fees and Widespread Service Disruptions

Amazon experienced a significant widespread outage on Thursday, impacting a broad spectrum of its services and leaving customers and sellers frustrated. The disruption, which began around 2 pm ET and was largely resolved by 8 pm ET, prevented shoppers from browsing, adding items to their carts, and completing purchases. The issue manifested for many as Amazon’s familiar "UH-Oh. Something went wrong on our end" error page, often accompanied by images of dogs, a playful but unwelcome sight for those trying to shop. One particularly exasperated customer humorously listed a multitude of dog names they encountered, highlighting the pervasiveness of the error.

Amazon acknowledged the problems in a statement to media outlets, attributing the difficulties to an issue related to a software code deployment. The company apologized for the inconvenience, stating that "some customers may have temporarily experienced issues while shopping." However, the impact extended far beyond simple browsing difficulties.

Customers reported being unable to retrieve pre-purchased items from Amazon lockers, with one individual describing their item as being "held hostage." Their experience detailed a spinning wheel on the app when attempting to access the locker, followed by an extended period of unresponsiveness despite reassurances from customer service that the issue would be resolved quickly.

The outage also affected other Amazon services. Members of the Amazon Vine Review program were unable to post their feedback, and the Kindle ebook platform also suffered disruptions. Customers reported an inability to view ebook descriptions or download new titles, leaving them unable to access their digital libraries.

For sellers operating on the Amazon marketplace, the outage translated into significant financial losses and operational challenges. Beyond the direct loss of sales due to customers’ inability to complete transactions, sellers faced difficulties accessing their own product listings. Reports on Amazon’s seller forums described listings showing 404 errors, missing buy boxes, appearing as "currently unavailable," or encountering errors when items were added to the cart. One seller noted that many listings lacked images, buy boxes, seller information, and even product details, speculating that Amazon Web Services (AWS), which underpins much of Amazon’s infrastructure, was severely impacted.

Some shoppers were unable to see item prices, with one user describing a "Suspenseful Music" message appearing when they attempted to view pricing options. This, combined with numerous "Product Unavailable" and "404 Dogs" errors, painted a picture of a crippled platform. For a seller accustomed to daily sales exceeding $2,000, the outage meant a complete halt in revenue for a five-hour period.

Communication and customer service were also compromised. Sellers reported that when attempting to reply to customer messages, the system indicated the messages were being processed, only for it to appear as if no reply had been sent. Furthermore, sellers found it challenging to reach Amazon seller support, with one seller asking on a forum if others were in the same predicament.

A primary concern for advertisers and sellers who utilize Amazon’s advertising services emerged: the potential for being charged advertising fees during the outage. These sellers argued that while their ad campaigns continued to run, generating clicks, customers were unable to complete purchases due to the site’s malfunction. This led to a significant increase in ad spend with virtually no corresponding sales or conversions.

In a detailed post on Amazon’s discussion boards, one seller outlined these concerns, highlighting a substantial increase in ad spend during the outage window, hundreds of clicks with minimal to no conversions, and the appearance of customers frantically clicking and retrying pages amidst broken functionality. The seller argued that these clicks did not represent genuine shopping intent as purchases were impossible.

The implications for advertisers were multifaceted: advertising spend accumulated during a period when the marketplace was non-functional, conversion rates were artificially suppressed, and campaign performance metrics were distorted. This distortion, they posited, could negatively impact optimization algorithms, bid efficiency, and even organic ranking signals. The core of the concern was that advertisers were paying for traffic during a time when the marketplace itself was not functioning properly, a situation entirely outside of their control.

Consequently, advertisers sought clarification on several critical points: whether Amazon would automatically refund ad spend generated during the outage, if performance metrics or algorithmic learning would be adjusted for affected campaigns, and if Amazon had identified the precise timeframe of the outage for accurate impact assessment. The seller emphasized that while acknowledging the scale of Amazon’s operations and the possibility of technical incidents, charging for clicks when the marketplace was not functioning raised serious fairness concerns. They respectfully requested a review of ad traffic during the outage and appropriate reimbursement or adjustments for affected advertisers.

EcommerceBytes reached out to Amazon on Friday evening for comment on whether it would credit seller ad fees incurred during the outage and pledged to update the article upon receiving a response.

CNBC reported that user reports on DownDetector surged around 2 pm ET, indicating the widespread nature of the issue.

In an update on March 12, 2026, an Amazon spokesperson confirmed to EcommerceBytes that the company is indeed reimbursing sellers for affected ad campaigns. This confirmation addresses the primary concern raised by advertisers and sellers regarding the financial implications of the outage.