How the AWS outage caused retailers to stumble
March 1, 2017 12:17 PM
Retailers and a big chunk of the internet flailed for several hours Tuesday when Amazon.com Inc.’s Amazon Web Services reported "high error rates" with its S3 (simple storage solution) cloud-computing service.
The websites for Express Inc., No. 99 in the Internet Retailer 2016 Top 500 Guide; Lululemon Athletica Inc. (No. 96) and One Kings Lane (No. 104) went down completely when AWS S3 experienced problems at a northern Virginia data center for about four hours, according to an analysis by Apica, a website monitoring and optimization firm. Amazon worked on the problem throughout Tuesday and said on its website at 2:08 p.m. Pacific Tuesday that S3 was operating normally. No AWS issues were reported Wednesday.
Almost 150,000 websites use AWS S3, which can be used to house data, manage web applications and host software that can be accessed and downloaded via the internet. Among the Internet Retailer 2016 Top 1000, only eMusic.com Inc. (No. 234) lists AWS as its web hosting provider, but many of the vendor services used by retailers rely on AWS, and an outage has a ripple effect. EMusic.com could not immediately be reached for comment.
Among the top 100 retailers monitored Tuesday by Apica, in the Top 500, 54 experienced a decrease of 20% or greater in average website performance because of the outage, Apica says. Average website load time jumped to more than 30 seconds from just a few seconds, Apica says. Retailers logging the most severe slowdowns were as follows:
- Disney Store USA LLC (No. 78 in the Top 500), 1165% longer to load
- Target (No. 22), 991% longer to load
- Nike Inc. (No. 47), 642% longer to load
- Nordstrom (No. 18), 592% longer to load
Nike declined to comment on the impact of the AWS S3 problem; the other retailers could not be reached immediately. The outage did not affect Amazon (No. 1) or Amazon-owned Zappos.com, Apica says. Data from web analytics companies SimilarWeb and SimilarTech show that Amazon is the leading shopping site that uses AWS S3, with about 2.47 billion global monthly visits.
“For marketers, a degradation in website performance means lost revenue. It’s as simple as that,” Apica CEO Carmen Carey says. “Consumers are very sensitive to how web pages perform. 34% of consumers say they would abandon a website that takes over 10 seconds to load. During the Amazon outage, some of the world’s largest [sites] took over 30 seconds to fully load their home pages,” she says.
Some retailers were able to spot-check and evaluate their site performance while they were down, and this helped them find a way to work around the problem, such as using images from a local server so the site would load faster, according to Apica says. “One of the dangers of moving to the cloud is an over reliance on a single cloud vendor,” Carey says. “Companies can deploy a multi-cloud strategy to mitigate issues when one or more of their vendors has an outage.”
AWS is Amazon’s fastest-growing and most profitable division, generating $3.5 billion in revenue in the fourth quarter. It’s the biggest cloud-services provider, with data centers around the world that handle the computing power for many large companies and other large organizations, such as government agencies.
Bloomberg News contributed.