You’ve probably heard the rumors. Maybe in a podcast, over coffee with a colleague, or in a LinkedIn post with three exclamation points. “Google axed middle management!” “Managers are obsolete!” It sounds like Silicon Valley sci-fi—lean, algorithm-driven, flat organizations with no need for suits or org charts. But is any of it true? Let’s dig in.
The Origin of the 35% Myth: What Actually Happened
Project Oxygen—that’s where the myth starts. Launched in 2007, Google wanted to answer a deceptively simple question: do managers matter? In a company built on code and data, the instinct was skepticism. Engineers joked that managers were “people who don’t code but get in the way.” Internally, some teams operated with near-total autonomy. The hierarchy was flat by design. But Google, being Google, didn’t assume. They measured.
They analyzed performance reviews, retention rates, team productivity, and employee feedback across thousands of employees. They used machine learning to isolate variables. And after two years, the results shocked even the skeptics: good managers significantly improved team performance. Teams with high-scoring managers were more productive, innovative, and less likely to quit. The data didn’t just validate management—it redefined it. But here’s where it gets messy. Google didn’t fire 35% of managers then. They didn’t fire any, not en masse. What they did was more surgical: they identified the lowest performers—those scoring in the bottom quartile on employee surveys—and encouraged them to improve or transition out.
Over several years, some managers left voluntarily. Others were reassigned. A few were let go. No single announcement. No press release titled “Google Cuts Middle Management.” Just quiet attrition. And somewhere along the way, someone—probably in a newsletter or a blog—rounded that attrition rate to 35%. The number stuck. But it wasn’t a purge. It was a filter.
Project Oxygen’s Seven Deadly Sins of Management
Google didn’t just collect data; they reverse-engineered bad management. Their research identified specific behaviors that tanked team morale. One manager didn’t give feedback—ever. Another hoarded information like a dragon guarding gold. A third failed to advocate for their team during budget fights. These weren’t firing offenses on paper, but collectively, they created toxic pockets. Google distilled these into eight behaviors (later refined to ten) that predicted managerial success. Things like “empowers the team” and “is a good coach.” Managers scoring low were offered training. Those who didn’t improve? They faded out.
And that’s exactly where the 35% figure likely came from—not a layoff, but a multi-year attrition rate among underperforming managers. But even that number is fuzzy. Google has never confirmed it. Internal documents suggest the actual figure was closer to 25–30%, and it wasn’t a single event. More like a slow leak in a balloon.
Why Google Doesn’t Talk About This Publicly
Because it’s awkward. Google prides itself on transparency. But admitting that a third of your managers weren’t cutting it? That’s a PR landmine. It implies past failure. So instead, they rebranded. Project Oxygen became a “leadership development initiative.” The narrative shifted from “we fired bad managers” to “we’re helping all managers grow.” Smarter. Safer. But less honest. And we’re far from it when it comes to transparency in corporate restructuring.
How Google Actually Measures Manager Performance Today
Forget annual reviews. Google’s system is continuous, data-driven, and mercilessly transparent. Every manager receives an Annual Employee Survey—called the Googlegeist—that includes specific questions about leadership. Employees rate their managers on a 1–10 scale across ten behaviors, from “provides career development” to “has a clear vision.”
Results are public within the team. Managers see their scores. So do their reports. So does their boss. A manager scoring below the 25th percentile is flagged. They get coaching. If scores don’t improve in 12–18 months? They’re unlikely to stay. No dramatic firing ceremony. Just a quiet exit. The system works because it’s not about punishment—it’s about accountability. And because it’s based on peer data, not top-down opinion, it’s harder to game.
But—and this is key—not all low-scoring managers are bad. Some lead high-stress teams (like incident response or legal compliance) where satisfaction scores are naturally lower. Google adjusts for that. They use statistical controls. Yet, the pressure remains. Because in a culture obsessed with metrics, a low score is a stigma. It follows you. And that’s the point.
The Role of 360-Degree Feedback in Tech Giants
Google didn’t invent 360 feedback, but they weaponized it. Today, companies like Meta, Netflix, and Microsoft use similar systems. Employees rate managers. Peers rate peers. Even interns can flag leadership issues. The data feeds into promotion decisions, bonuses, and retention. One engineer at Meta told me they once saw a director passed over for promotion because their team’s feedback showed “poor listening skills.” That’s power. But it’s also risky. Feedback can be biased. A charismatic but ineffective manager might score high. A quiet, analytical one might be misunderstood.
The issue remains: no system is perfect. But Google’s bet is that aggregated data beats individual judgment. And for the most part, it does.
Flat Organizations vs. Managerial Necessity: The Great Tech Paradox
Startups preach flat structures. No managers. Just “leads” or “coordinators.” It feels democratic. Agile. But scale kills that dream. When you have 500 engineers, someone has to align priorities, resolve conflicts, allocate resources. You can’t run a company by consensus. Not unless you want endless meetings and zero shipped code.
Google learned this the hard way. In the early 2000s, some teams tried manager-free models. Results were mixed. Brilliant on paper—autonomy! innovation!—but in practice, coordination collapsed. Projects stalled. People burned out. The ones who thrived were also the loudest. Quiet contributors were overlooked. So Google reintroduced managers—but redefined them. Not bosses. Coaches. Enablers. The title changed. So did the job description.
But here’s the irony: by optimizing managers so aggressively, Google may have made them more powerful, not less. A high-scoring Google manager has influence, data, and trust. They’re not just administrators. They’re culture carriers. And that’s exactly where the myth of the “managerless company” falls apart.
Spotify’s “Squads” vs. Google’s “Manager Coaches”
Spotify famously ditched traditional hierarchies for “squads,” “tribes,” and “chapters.” No managers—just “chapter leads” for technical guidance and “product owners” for direction. In theory, it’s agile heaven. In practice? A 2020 internal review found squads struggled with long-term planning and career development. Chapter leads were overloaded. Accountability blurred. By 2022, Spotify quietly reintroduced formal leadership roles. Not managers, they called them “team leads.” But functionally? Same job.
Google’s model is different. They never fully abandoned management. They just made it earn its keep. Managers must prove value, every year. And that’s the real difference: not structure, but accountability.
Frequently Asked Questions
Did Google officially announce a 35% manager layoff?
No. There was never an official announcement. No press release, no internal memo titled “Manager Reduction Initiative.” The 35% figure appears to be an estimate—possibly inflated—about the attrition rate among underperforming managers over several years. Google’s approach has always been incremental, not revolutionary. They improve systems quietly. Then declare victory later.
Are managers still relevant in tech companies?
Yes—but not in the old sense. The micromanaging boss who assigns tasks and monitors time sheets? Obsolete. The coach who unblocks talent, aligns vision, and advocates for their team? More critical than ever. The role has shifted from control to enablement. And that’s where conventional wisdom gets it wrong: it’s not that managers are disappearing. It’s that the definition is evolving.
Can a company operate without managers entirely?
Theoretically, yes. Practically? Only at small scale. Valve, the game company, famously has no managers. Employees choose projects. Pay is peer-determined. Sounds ideal. But ex-employees report chaos during major decisions. Hiring? Slow. Conflict resolution? Messy. Without structure, power vacuums form. Informal hierarchies emerge—often less fair than formal ones. So while manager-free models work in niche environments, they don’t scale. And that’s exactly where the idealism of flat orgs crashes into reality.
The Bottom Line
Did Google fire 35% of managers? Not in the way the internet claims. There was no grand purge. No sweeping memo. What happened was slower, smarter, and more sustainable: they built a system where poor managers couldn’t survive—and good ones thrived. They turned management into a performance-based role, not a default promotion. That’s not anti-management. It’s pro-meritocracy.
I find this overrated—the idea that companies like Google are “killing management.” What they’re doing is holding it to a higher standard. And honestly, it is unclear whether this model works outside elite tech firms. It requires cultural buy-in, data infrastructure, and a tolerance for transparency most companies lack.
But here’s my take: the future isn’t managerless. It’s better managed. The 35% myth distracts from the real lesson. Google didn’t eliminate managers. They redefined them. And that changes everything. So if you’re a leader, ask yourself: would your team rate you highly on coaching, vision, and empowerment? If not, you might not get fired—but you won’t last. Because in the new era of work, relevance isn’t granted. It’s earned. Every single day. And that’s not a rumor. That’s the truth.