The message on the screen said it plainly: “Workday is currently unavailable.” Behind it, for the customers it hit, sat the pay, the leave and the job applications.
That list comes from Workday’s own service agreement, in the section for its most severe kind of failure. Severity 1, the document says, is when the service is unavailable or when a problem prevents timely payroll processing, entry into time tracking, or the processing of candidate applications. “No workaround exists.” The company told KRON4 that some customers were experiencing a service disruption and it was working urgently to investigate and fix it. “We sincerely apologize for the challenges this is causing impacted customers,” it said. It did not say when service would be fully back.
The cause, as Workday gave it, was found while testing the backup systems. The company placed some of its systems into maintenance mode and said it would post updates every 30 minutes. Workday said the problem had left some information in an inconsistent state, which could cause errors when users tried to sign in or finish a task. The same service agreement that defines Severity 1 also says Workday will test its disaster-recovery plan once every six months, and that it promises to bring a production tenant back within 12 hours. Counted from the first user reports, the outage was about nine hours into that 12-hour promise by Friday night.
Workday posted no official incident, and its own status page still read operational while the reports piled up. The outage tracker Instatus, which checks that page every minute, reported that the page said operational at the same time it counted 2,051 reports in the last hour. Georgia Tech’s IT page carried Workday’s own wording: the service was unavailable “due to an unexpected issue requiring the service to be placed into maintenance mode,” and the downtime had begun ahead of the vendor’s regular Friday maintenance window, which starts at 10 p.m. Eastern. Iowa State’s page said its Workday tenant was down for emergency maintenance across all five of its modules, finance and payroll among them.
The people locked out were not talking about maintenance windows. Meredith Johnstone, replying on Threads to the San Francisco Chronicle’s post about the outage, wrote: “That’s fine I can’t even use workday when it’s not experiencing an outage.” Erinoanna, replying to the same post, had the sharper problem. “No big deal, last day of open enrollment for us. Let’s hope no one procrastinated or no insurance for them.” Her employer, not every employer, had set Friday as the last day to pick health insurance.
Walmart, Target, Chipotle and Skechers were reported as affected on the word of a single unnamed source, relayed by the outlet The Newsground and repeated across the coverage. None of the four companies has confirmed it, and Workday has named no one. One of the four, Target, is a customer Workday itself names on its own site. The rest is an unconfirmed list. What is confirmed is the scale behind the word “some”: Workday says it serves more than 11,500 customers, with more than 80 million users under contract.
Bryan Lee, posting on X, put the payday question in his own words: “Roughly tens of millions of people rely on Workday for payroll, so a full outage could leave a large number without a paycheck on a major payday—though not every Workday Payroll customer is paid on the same day.” The tens of millions is his estimate, and his own caveat follows it in the same post. Another user, Blue- FeedJeremiahSmith, supplied a different theory in the gap the company left: “Yeah, you cant tell me this isnt a cyber attack at this rste.” It is speculation, not evidence. Workday’s stated cause is the backup-test issue, and no official has suggested an attack.
The Workday outage was the second workplace system to go down in two days. On Thursday, Microsoft’s Outlook, Teams and Copilot stumbled, with reports of problems with Outlook passing 130,000 around 1:27 p.m. Eastern. The disruption reached government systems in Japan, and Microsoft blamed unusually high load and servers that had failed health checks. Users answered that outage with memes, asking the engineers to take the weekend. Nobody asked that about Workday. Microsoft’s outage locked people out of email. Workday’s locked them out of pay, leave and applications.
Instatus lists 15 user-reported Workday incidents in the last 90 days, including one on November 9 that lasted about an hour, one on November 7 that lasted two, and a seven-hour one on August 29. Every entry on the list carries the same note: no update from Workday. Friday was the fourth in eight days. It was also the longest-running of them. Another tracker, Technobezz, claims there were no outage days at all in the last 90 days. The two trackers count different things, and the article follows Instatus because it publishes the list.
The reports kept coming. Downdetector’s 24-hour peak stood at 2,835, and the latest report StatusGator logged came from Texas: “Service down.” Instatus counted 2,051 reports in the last hour, 48 times normal.





