Expedia Backtracks on Outage Cause: ‘Software Issue’ Took Down Global Operations

What caused the widespread outages at many of Expedia Group properties this weekend? The company initially attributed them to “maintenance” but provided a different explanation Monday:

“Expedia Group performs regular updates and changes to its systems, which is a complex and involved process,” the company told Skift. “Yesterday, we did experience an outage caused by a backend software issue. We were able to bring our systems back online quickly and we are continuing to monitor the situation. We also have measures in place to minimize the risk of something like this happening again.”

Skift reported Sunday that many of Expedia Group websites around the world went down, including Expedia.com, Hotels.com, Vrbo, Orbitz, Travelocity, WotIf, and eBookers. All of those sites are on a common backend technology stack.

Sister brand Hotwire, which is on a different tech platform, remained up and running.

A source familiar with what went wrong confirmed it was an “an expired root certificate in the tech stack.”

Expedia also confirmed Monday that the problem went beyond just the consumer-facing websites and included some internal operations.

The updated cause of the outage from Expedia Group came after a tipster told Skift that an “expired root certificate in the tech stack” was the culprit, and claimed that layoffs were a contributor to the problem. Expedia acknowledged that was the cause, and disputed that layoffs were a factor.

The Expedia Group Layoffs

In February, Expedia Group announced that 1,500 employees, or about 9% of its workforce, would be fired. Many of the affected employees had worked on the multiyear tech platform migration that had consolidated Group brands Expedia.com, Hotels.com, and Vrbo onto a common backend tech platform.

Expedia Group disputed the fact Monday that the layoffs had anything to do with the outages, or that the technology platform migration contributed to the outages. It said some maintenance was taking place over the weekend, and that it responded quickly to to outage, considering it took place on the weekend.