On March 25, Meta notified about 700 employees to leave, involving five departments including Reality Labs, Facebook social media, recruitment, sales, and others. On the same day, the SEC disclosed an executive options plan, which will provide stock options tied to a $9 trillion market value for six core executives. This is the first time Meta has issued options to executives since its IPO in 2012.

On one hand, laying off employees while rolling out the most aggressive executive incentive plan in Silicon Valley history. The two actions Meta took on the same day are not contradictory; they are two sides of the same strategy. The AI race does not need more people; it needs more expensive people and more machines.

Fewer people, each person is more "valuable"

2022 was the peak year for Meta's employees, with a total of 86,482 people. That year, Zuckerberg bet on the metaverse, hiring aggressively, but the annual revenue dropped from $117.9 billion the previous year to $116.6 billion. Revenue per employee fell to a low of $1.35 million.

What happened next is known to all. In November 2022, 11,000 people were laid off, and in 2023 another 21,000 were cut, resulting in a total reduction of one-quarter of the company's workforce. Zuckerberg declared 2023 as the "Year of Efficiency."

The results of efficiency are reflected in the numbers. According to Meta's Q4 2025 financial report, by the end of 2025, the company will have 78,865 employees, nearly 8,000 fewer than the peak. But at the same time, annual revenue increased from $116.6 billion to $201.0 billion, a 72% increase. Revenue per employee soared from $1.35 million to $2.55 million, an increase of 89%.

The meaning of these numbers is very straightforward. Meta earned more money with fewer people. In 2022, the marginal revenue brought by adding an employee was declining, whereas in 2024 and 2025, the revenue increase corresponding to each employee reduction expanded. This is a typical scale effect of technology companies, but Meta accelerated this process through layoffs.

This is the background of the recent round of layoffs involving 700 people in March 2026. According to The Register, this is already Meta's second round of layoffs this year, having just cut about 1,000 people in Reality Labs in January. NBC News cited insiders stating that there could be larger reductions ahead, affecting up to 20% of the total workforce, or about 15,000 people, which would bring Meta's total employees back to 2021 levels.

Zuckerberg's original words in the January earnings call were to "flatten teams" and allow excellent individual contributors to complete projects that previously required large teams. A spokesperson for Meta also provided a template response, stating, "Teams undergo regular restructuring or adjustments to ensure they are in the best position to achieve their goals."

Continuing to bet on the AI arms race

Where the money saved from layoffs is going can be seen by looking at capital expenditures.

According to the Q4 2025 financial reports and public guidance of various companies, combined capital expenditures for Amazon, Google, Microsoft, and Meta will reach about $650 billion in 2026, a year-on-year increase of about 130%. Among them, Amazon's expenditure is approximately $200 billion (an increase of 167%), Google's about $175 to $185 billion (an increase of 140%), Microsoft's approximately $145 billion on an annualized basis (an increase of 127%), and Meta's between $115 and $135 billion (an increase of 73%).

According to CNBC, this is the largest single-year capital expenditure in the history of the tech industry. The investment in AI infrastructure by these four companies in one year exceeds the GDP of Sweden for a year.

Meta's absolute value ranks fourth, but relative to its own size, the density of investment is astonishing. Based on a median of $125 billion, Meta's AI infrastructure investment per employee is approximately $1.59 million, close to 62% of the revenue per employee ($2.55 million). In other words, for every $100 Meta earns, $62 is invested in data centers.

The cost of this money is also very direct. According to CNBC citing Barclays analysts, Meta's free cash flow in 2026 is expected to drop by nearly 90%. Amazon is even more aggressive, with Morgan Stanley estimating that Amazon will have approximately negative $17 billion in free cash flow in 2026. All four giants are doing the same thing: trading today's cash flow for tomorrow's AI infrastructure.

A $9 trillion bet

Next, let's look at the options plan. According to documents disclosed by the SEC and analysis from Motley Fool, this plan covers six executives, including Chief Technology Officer Bosworth, Chief Product Officer Cox, Chief Operating Officer Olivan, Chief Financial Officer Susan Li, Chief Legal Officer Maoni, and Vice Chairman McCormick. Zuckerberg is not on the list, as his super voting stock already makes him exempt from additional incentives.

The conditions for exercising the options are designed with a tiered price threshold. According to Motley Fool, the minimum exercise price is $1,116 per share, requiring the stock price to rise 88% from the current approximately $615. The highest tier is $3,727 per share, corresponding to a market value of about $9 trillion, which is six times the current $1.5 trillion. There is a five-year window, and they must be fulfilled before 2031. If Meta really achieves a $9 trillion market value, according to Motley Fool's calculations, the potential earnings for the top four executives (Bosworth, Cox, Olivan, Susan Li) would be about $2.7 billion each.

The signal of this plan is clear. Meta is not giving executives bonuses; it is binding the core team to an extremely aggressive growth target with options. Current market value is $1.5 trillion, with a target of $9 trillion, a difference of $7.5 trillion—Meta is betting that AI can create this value.

In comparison of scale, $9 trillion is roughly equivalent to the combined current market value of Apple and Nvidia. No company in the world has ever reached this market value. Meta has given its core executives five years to touch a number that has never existed in the history of human business.

A formula

When viewed altogether, Meta's logic is a simple resource allocation formula. Total employee compensation (including equity incentives) is basically flat between 2022 and 2026, around $26 to $28 billion. However, AI capital expenditures skyrocketed from $32 billion to $125 billion, growing about threefold in four years. At the same time, a completely new executive options pool has appeared, locking the six most core individuals in for the next five years.

According to Benzinga, Meta's equity incentive expense in 2025 is about $42 billion, which has consumed most of its free cash flow. Signing bonuses for AI researchers have reached nine figures, with a researcher hired away from OpenAI reportedly receiving a signing package worth $100 million. These numbers, in contrast to the 700 laid-off employees, clearly illustrate Meta's pricing logic regarding "people."

The money saved from laying off 700 people is roughly equivalent to Meta's AI infrastructure spending for one and a half days.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。