The Brutal Math of Talent Rotation

The narrative dominating the business press for the last three years has been one of collapse. We are told that the tech sector is shrinking, that the boom times are over, and that austerity is the new religion in Silicon Valley. If you strictly read the layoff announcements, you would be forgiven for thinking these companies are dismantling themselves brick by brick.

But if you ignore the PR releases and look at the 10-K filings, a completely different reality emerges. The giants aren’t shrinking. They are rotating.

Despite years of high-profile termination rounds, headcount at Alphabet and Amazon is effectively flat or rising. This is not a recession; it is an aggressive inventory management strategy applied to human capital.

The Hiring Binge Hangovers

To understand the current maneuvering, you have to look at the sheer insanity of the 2020-2021 period. This wasn’t just growth; it was hoarding. Capital was cheap, demand for digital services was artificially spiked by lockdowns, and companies hired bodies simply to keep them away from competitors.

Amazon’s numbers from that period are almost difficult to comprehend. In just two years (2020 and 2021), Amazon increased its staff by 810,000 people. They more than doubled their workforce (+102%) in twenty-four months. When an organization adds nearly a million people in that timeframe, operational discipline evaporates. You cannot integrate that many people efficiently. You end up with layers of management managing nothing but other managers, and remote workers holding down multiple full-time jobs because oversight mechanisms have broken down.

Alphabet saw a similar, albeit smaller scale, surge of 60% during the same window. This was the era of excess. And naturally, when the cost of capital rose and the pandemic distortion field faded, the bill came due.

The Clean Up Phase

Since late 2022, we have been in the “sorting out” phase. This is what the market interprets as a crash, but it is actually just operational hygiene. Companies are shedding the deadwood, the redundant layers, and the projects that no longer map to revenue.

Take Google (Alphabet). By the end of 2025, after years of “serial layoffs,” their global headcount actually rose by 7,497 year-over-year. They are currently sitting just a hair below their all-time peak headcount. If this is a crisis, it is a very strange one.

Google VP Brian Welle noted that the cleanup included sweeping out 35% of managers overseeing small teams. This is the key metric. They aren’t just cutting costs; they are de-layering. A manager with three direct reports is an expensive bottleneck. By removing that layer, the company doesn’t just save salary; it speeds up decision-making.

Amazon shows the same pattern. After shedding 16,000 and 67,000 roles in 2022 and 2023 respectively, their headcount climbed back up by 20,000 in 2025. They are sitting at 1.576 million employees—down only 2% from their chaotic peak. They trimmed the fat, but the muscle mass remains enormous.

The Zero-Sum Talent Swap

The most telling example of the current strategy comes from Intuit in July 2024. They announced 1,800 layoffs—roughly 10% of their workforce. Simultaneously, they announced plans to hire 1,800 people to accelerate their push into AI.

This is the playbook. It is a one-for-one swap. They are firing employees whose skills are depreciating (legacy maintenance, redundant middle management, underperformance) and using that exact budget to hire employees whose skills are appreciating (AI, machine learning, specialized engineering).

Layoffs are instantaneous. Hiring is slow. This creates a temporary lag that looks like a reduction in force, but the end goal is to keep the headcount stable while upgrading the capability per head.

Follow the Value

When you see a headline about 10,000 people being let go, do not assume the company is in trouble. Assume the company has realized those 10,000 people are no longer generating a return on investment that exceeds their cost.

These corporations are planning hundreds of billions in CapEx for AI infrastructure in 2026. They need cash flow to fund that. They are generating it by swapping low-velocity legacy workers for high-velocity technical talent.

The headcount isn’t dropping. The bar for staying employed is just getting higher.