Meta Forces 20% "Below Expectations" Ratings While Spending $65 Billion on AI Development

By
CTOL Editors - Dafydd
5 min read

Inside Meta's Corporate Battleground: Performance Quotas Clash With Record AI Ambitions

In the gleaming corridors of Meta's Menlo Park headquarters, two parallel realities have taken hold. Behind closed doors, managers frantically strategize to designate which team members will fall into the expanded "Below Expectations" category, while just floors away, elite AI researchers are being courted with compensation packages reaching astronomical heights of $300 million. This jarring contrast reveals the high-stakes gamble Meta is making on its future.

Meta (wikimedia.org)
Meta (wikimedia.org)

"The Quota Machine": How Meta's Performance System Turned Ruthless

Meta has dramatically intensified its performance management process in 2025, implementing what multiple insiders describe as "engineered attrition." According to internal documents and sources familiar with the matter, the company has mandated that managers label 15-20% of employees as "Below Expectations" during the current Performance Summary Calibration cycle—up from previous years' 12-15% threshold. Employees from several orgs, such as Ads and Infra, confirmed to us that the actual percentage of "Below Expectations" is 20%.

"The numbers are decided before anyone's performance is even evaluated," confided one mid-level manager who requested anonymity due to fear of retaliation. "We received explicit instructions to fill the quota, regardless of actual team performance."

The policy applies uniformly across Meta's organizations—Ads, Infrastructure, and Product—creating an environment where survival, not innovation, has become the primary focus for many.

What makes this round particularly troubling is the emergence of pre-selection practices. Multiple sources confirm that some directors and senior managers identified potential BE targets before performance reviews began. New hires, who haven't had sufficient time to demonstrate their capabilities, have become particularly vulnerable.

"People who joined just months ago are being sacrificed to protect more established team members," explained a senior engineer with knowledge of the process. "Even those who resigned earlier this year are being retroactively counted toward the quota, which tells you everything about how arbitrary this has become."

The Weaponization of Attrition Labels

Meta's internal classification system for departing employees has evolved into a controversial tool within this high-pressure environment. When employees leave—whether through resignation or termination—managers assign one of two designations: "Regrettable Attrition" or "Non-Regrettable Attrition" .

These seemingly administrative labels carry significant consequences. Only those designated as NRA count toward a manager's BE quota, creating a perverse incentive to mark departures as "non-regrettable" regardless of actual performance or contribution.

"It's an open secret that nearly everyone leaving gets the NRA label now," said a veteran recruiter with visibility across multiple teams. "Even top performers who resign by choice. It's purely about hitting numbers."

The implications for future employment are substantial. Those labeled RA can return to Meta within a year through a simplified rehiring process, while NRA-designated former employees face the full interview gauntlet should they attempt to return—effectively burning bridges with one of tech's most powerful employers.

"Survival Politics": The Deterioration of Collaboration

Perhaps most concerning is how this system has transformed Meta's once-collaborative culture into what one product manager described as "corporate Survivor."

Cross-functional collaboration has deteriorated into territorial disputes. In one particularly revealing incident, six different teams engaged in what witnesses describe as a "shouting match" over ownership of a relatively minor project. The dispute escalated through multiple management layers, eventually requiring director-level intervention.

"TLs [Technical Leads] routinely run to their director's office to 'snitch' on each other," said one engineering manager. "Basic recognition posts become battlegrounds for credit. People who contributed nothing demand co-ownership, simply to pad their performance metrics."

This environment has cultivated a culture where undermining colleagues is viewed as a survival strategy. "Everyone's desperate to make someone else take the BE mark, thinking it'll keep themselves safe," observed a human resources professional with knowledge of multiple team dynamics.

The $65 Billion Contradiction: Meta's All-In AI Bet

The ruthless efficiency drive stands in stark contrast to Meta's unprecedented investment in artificial intelligence. The company is channeling up to $65 billion into AI development in 2025 alone—one of the largest technology investments worldwide this year.

This includes acquiring over 1.3 million GPUs, constructing massive specialized data centers, and assembling what industry observers call an "AI dream team" under the newly formed Meta Superintelligence Labs. High-profile recruits include former Scale AI CEO Alexandr Wang and ex-GitHub CEO Nat Friedman, who have been lured with compensation packages that dwarf even traditional executive pay scales.

CEO Mark Zuckerberg has framed this investment as a "historic turning point" for both the company and society, publicly declaring Meta's ambition to develop "personal superintelligence for everyone."

The Strategic Calculus: Why Meta Is Playing Both Games

For investors and market analysts, this dual approach represents a fascinating case study in resource allocation. Meta appears to be aggressively pruning what it considers non-essential functions while simultaneously placing an existential bet on AI dominance.

"What we're witnessing is a company-wide reallocation of human and financial capital," explains an investment strategist at a major Wall Street firm. "Meta believes the AI race will determine the next decade of tech leadership, and they're restructuring everything around that conviction."

The strategy carries significant risks. Employee morale has deteriorated, with internal surveys showing concerning drops in engagement metrics. Institutional knowledge is being lost through both engineered and voluntary departures. Yet leadership appears convinced that these trade-offs are necessary to secure Meta's position in what they view as technology's most consequential race.

Investment Perspective: Reading Meta's Mixed Signals

For investors navigating Meta's contradictory signals, the key may lie in understanding the timeline of expected returns. The company's aggressive performance management approach could deliver short-term cost efficiencies and margin improvements, potentially boosting quarterly results through 2025-2026.

However, the AI investment represents a longer-term thesis that may not yield significant returns for 3-5 years. This creates a potential valley between immediate productivity impacts and future AI-driven growth—a period where Meta could face innovation challenges as surviving employees focus on political safety rather than breakthrough thinking.

Market analysts suggest monitoring several key metrics: voluntary departure rates among high performers, product launch velocity compared to competitors, and early commercial adoption indicators for Meta's AI initiatives. These may provide early signals about whether the high-risk strategy is yielding the intended balance of efficiency and innovation.

As with any significant investment thesis, past performance does not guarantee future results, and investors should consult financial advisors for personalized guidance tailored to their specific circumstances and risk tolerance.

As Meta navigates this complex organizational transformation, the question remains whether it can successfully transform into both a lean operator and an AI powerhouse—or whether one ambition will ultimately undermine the other.

You May Also Like

This article is submitted by our user under the News Submission Rules and Guidelines. The cover photo is computer generated art for illustrative purposes only; not indicative of factual content. If you believe this article infringes upon copyright rights, please do not hesitate to report it by sending an email to us. Your vigilance and cooperation are invaluable in helping us maintain a respectful and legally compliant community.

Subscribe to our Newsletter

Get the latest in enterprise business and tech with exclusive peeks at our new offerings

We use cookies on our website to enable certain functions, to provide more relevant information to you and to optimize your experience on our website. Further information can be found in our Privacy Policy and our Terms of Service . Mandatory information can be found in the legal notice