previous thread: https://www.thelayoff.com/t/1kqgha9tv
Limitations: to answer a common question what exactly is counted here and how.
Between once a day up to every few hours I manually record Slack total participants count in #general Slack channel of Oracle One workspace - that's where all Oracle employees onboard into corporate Slack account auto-join.
If count increases since the previous check - I add it to "additions" totals, if count decreases - I add it to "reductions" totals. Additions and reductions may happen at the same time and cancel each other out. The higher frequency of sampling - the more additions and reductions are caught.
"Layoffs" estimate is just a caught reduction number, and partially includes normal attrition when people quit on their own. However, a lot of reductions are canceled out due to frequency of sampling, so normal attrition often is already at least partially not part of the count.
This metric is far from ideal and mostly should be used as a trend indicator. It's especially less accurate (has more delay) in EU and Pacific regions, where employees are disconnected from Slack with long delay. However, it reflects the same day layoffs in NA and IDC usually.
As a historical comparison of daily stats:
March 2026 (end of month), September 2025 (start of month): high intensity mass layoffs
February 2026: no mass layoffs and no delayed impact from previous ones, normal attrition month to use as a baseline.