Satya Nadella thinks Microsoft’s biggest advantage has quietly turned into its biggest handicap. And that's why, on his weekends, the CEO of one of the world's most powerful tech companies is doing something unexpected: studying startups.
In a recent conversation with Mathias Döpfner, CEO of Axel Springer (Business Insider’s parent company), Nadella revealed that he spends his own weekends examining how small, fast-moving companies develop products. His reasoning? Microsoft’s sheer size, he admits, has become “a massive disadvantage” in the ever-accelerating race to build AI.
He shared on Döpfner’s MD MEETS podcast that he “spent the entire weekend trying to understand how new companies are building products.” Startups, he noted, often sit everyone around the same table — engineers, product designers, scientists, and infrastructure experts — allowing decisions to happen immediately. “They can iterate in real time,” Nadella explained. “Meanwhile, at Microsoft, I have three division heads who manage those three areas separately.”
That one observation exposes the challenge of scale: large organizations move slower, not because they lack talent or resources, but because their structure naturally resists agility. And Nadella isn’t alone in recognizing this problem. Across Silicon Valley, giants like Meta, Google, and Amazon are flattening their management hierarchies, cutting through layers of bureaucracy to reclaim startup speed and flexibility. But here’s where it gets controversial — can massive corporations truly think like startups again, or does their success make unlearning impossible?
Learning to Unlearn
According to Nadella, thriving in the AI era demands exactly that — the courage to unlearn what once worked. “The most important skill for long-term relevance,” he said, “is to be a ‘learn-it-all,’ not a ‘know-it-all.’” It’s a mindset shift that sounds simple but is painfully hard in practice. Leaders have to let go of the formulas that built their companies’ success to make room for new ones.
He also highlighted that empathy and emotional intelligence are now critical leadership tools. In an age dominated by algorithmic decision-making, understanding people — not just data — may be what separates the thriving companies from the obsolete ones.
And Nadella isn’t just preaching theory; he’s reorganizing from the top down. A leaked Microsoft org chart recently showed he now has 16 direct reports — carefully chosen executives expected to move fast, tear down silos, and embed AI into every layer of the company’s operations.
The Four Fixes Every Company Needs
So what’s going wrong with corporate AI projects? Nadella says most fail because companies treat artificial intelligence like a routine IT upgrade. “If you approach AI that way,” he warned, “you’re going to fail by definition.” Instead, he outlines four foundational steps every organization must tackle:
- Rethink workflows entirely from the ground up — not just add AI on top.
- Adopt modern AI tools, not legacy ones repackaged for today.
- Train employees continuously to work effectively with these systems.
- Ensure company data is open, connected, and not trapped in outdated infrastructure.
Only businesses that rebuild these core systems, he argues, will see meaningful results from AI. And only leaders willing to unlearn their old habits will have any hope of leading that transformation.
So here’s the debate worth having: Can a company as vast as Microsoft ever truly think and move like a startup again — or is that an impossible dream for any organization that’s already “made it”? Share your thoughts — is Nadella’s strategy visionary, or is it wishful thinking?
