How many workers would Big Tech’s revenue predict?
This chart compares reported Big Tech headcount in 2022–2025 with an out-of-sample benchmark predicted from each company’s revenue and the pre-2022 revenue–headcount relationship. It is a descriptive counterfactual, not an estimate of optimal staffing.
What does it show?
Big Tech reported 27% fewer workers in 2025 than its pre-2022 revenue–headcount relationship would have predicted. The gap appears at every company, suggesting revenue has recently grown with much less labor.
Methodology
Fits a log-linear panel model to 30 company-year observations from 2016–2021: log reported fiscal year-end headcount equals a company fixed effect plus a common coefficient on log whole-company revenue. It then holds that pre-2022 relationship fixed and uses each company’s observed 2022–2025 revenue to generate an out-of-sample headcount benchmark. The left panel aggregates predictions and reported headcount across Apple, Microsoft, Alphabet, Amazon, and Meta; the right panel shows each company’s 2025 percentage gap. This is a descriptive counterfactual, not an estimate of optimal staffing or proof that layoffs were justified. Revenue can separate from labor because of prices, product and business mix, capital intensity, outsourcing, acquisitions, automation, or reporting definitions. The 2025 aggregate gap remains negative when adding net-income margin (-26%), allowing company-specific revenue slopes (-35%), or benchmarking against each company’s median 2018–2021 revenue per employee (-29%).