Amazon Outage Disrupts Shopping, Sellers Concerned About Ad Fees Amidst System Failures

On Thursday, Amazon experienced a significant widespread outage that prevented customers from shopping and completing purchases, leading to widespread frustration and a flurry of reports on industry forums. Shoppers encountered Amazon’s familiar "UH-Oh. Something went wrong on our end" error page, often accompanied by images of dogs, a signature element of Amazon’s error displays. One particularly detailed account listed numerous dog names, highlighting the frequency of the errors encountered by a single user.

Amazon acknowledged the disruption in a statement to media outlets, attributing the issues to a "software code deployment" and stating that "some customers may have temporarily experienced issues while shopping." The company issued an apology for the inconvenience.

However, the impact of the outage extended far beyond simple shopping cart failures. Customers reported being unable to retrieve items from Amazon lockers, with one individual describing their purchased item being held "hostage" due to the inability to access the locker system. The user detailed a frustrating experience of the app showing a spinning wheel, followed by a promise of a quick fix that failed to materialize, leaving them unable to collect their order for hours.

Members of Amazon’s Vine Review program were also affected, unable to post their reviews. The outage even impacted the Kindle ebook platform, with users reporting an inability to view book descriptions or download new titles, raising concerns about the accessibility of digital content.

For Amazon sellers, the consequences were multifaceted. Beyond the direct loss of sales due to customers being unable to complete purchases, sellers faced difficulties accessing their own product listings. Reports on Amazon’s seller discussion boards described listings showing 404 errors, missing buy boxes, or appearing as "currently unavailable." Some listings even generated errors when added to a cart. One seller noted that the issue seemed to extend to Amazon Web Services (AWS), suggesting a broader infrastructure problem.

Further compounding the problem, some shoppers reported that item prices were not displaying correctly. Instead of prices, they encountered an "Options" button that, when clicked, led to a "Suspenseful Music" message and returned them to the item page without any additional information. Many listings were also marked as "Product Unavailable" or displayed 404 errors, often accompanied by the now-infamous dog images.

The financial implications for sellers were significant. One seller, accustomed to daily sales exceeding $2,000, reported zero sales during a five-hour period of the outage. The disruptions also extended to seller-customer communication. One seller detailed an issue where replying to customer messages resulted in a "being processed" notification, followed by the message appearing as if it was never sent. Seller support also seemed to be impacted, with one seller questioning if others were experiencing similar difficulties in reaching assistance.

A primary concern for sellers who utilize Amazon’s advertising services was the continued charging of ad fees during the outage. Advertisers reported that their Amazon Advertising campaigns continued to run, generating clicks even though customers were unable to complete purchases due to the platform’s inoperability. One seller detailed their experience in a post on the Amazon discussion boards, highlighting a significant increase in ad spend during the outage window, coupled with hundreds of clicks but virtually no conversions.

The seller articulated several serious concerns for advertisers:

• Accumulated Ad Spend Without Sales: Advertising costs were incurred during a period when purchases were impossible.
• Artificially Suppressed Conversion Rates: The platform outage skewed conversion rates, making campaign performance appear worse than it was under normal operating conditions.
• Distorted Campaign Performance Metrics: The outage threatened to negatively impact optimization algorithms, bid efficiency, and potentially organic ranking signals due to abnormal traffic patterns.

The argument was made that advertisers were essentially paying for traffic during a time when the marketplace was non-functional, raising questions of fairness. The seller respectfully requested clarification on several key points:

• Whether Amazon would automatically refund ad spend generated during the outage window.
• Whether Amazon would adjust performance metrics or algorithmic learning for campaigns affected by the outage.
• Whether Amazon had identified the exact timeframe of the outage to allow advertisers to assess the impact on their accounts.

The seller emphasized that while technical incidents are understood given Amazon’s scale, charging advertisers for clicks when the marketplace was not functioning properly raised fairness concerns. They requested that Amazon review ad traffic during the outage and provide appropriate reimbursement or adjustments for affected advertisers.

EcommerceBytes reached out to Amazon on Friday evening to inquire whether the company would be crediting seller ad fees incurred during the outage and stated that the article would be updated upon receiving a response.

CNBC reported that the number of user-reported issues on DownDetector spiked around 2 p.m. ET, with the majority of problems appearing to be resolved by 8 p.m. ET.

In an update on March 12, 2026, an Amazon spokesperson informed EcommerceBytes that the company is indeed reimbursing sellers for affected ad campaigns.