How Eric Schmidt managed chaos to make Google innovative

Executive overview

Most managers try to impose order. At Google, the opposite worked. Innovation comes from letting ideas emerge through free-flowing conversation and experimentation across the organisation — then making fast decisions on what to scale.

Eric Schmidt joined Google in 2001 after a bruising turnaround at Novell. He found a company modelled on a Stanford grad school: crowded offices, free food, and engineers who argued with the CEO. His job was not to control that energy but to channel it.

The core insight: your role as a leader is not to generate the best ideas but to build the conditions where they can emerge — and then decide quickly.

Eric Schmidt's entry at Novell and Google

• Novell's financials on day one were worse than advertised; the company entered a full crisis within a week
• Learning to fly planes taught Schmidt to make rapid decisions: decide, decide, decide — then live with the consequences
• At Google, Larry Page and Sergey Brin grilled him for 90 minutes and attacked his work at Novell; he left knowing it was a special company
• Willingness to challenge leaders is rare; most teams tell leaders what they want to hear

Why chaotic organisations outperform

• Ideas rarely spring fully formed; they improve through networks of people giving feedback and criticism
• The heroic lone-inventor story of innovation is false: the best idea may come from person 89, and you cannot predict who
• Sharing problems openly across the company — including to non-engineers — is what enables solutions to emerge
• Mark Pincus (Zynga): a good entrepreneur's instincts are right 95% of the time; their specific ideas only 25% of the time
• Separating a losing idea from a winning instinct is essential — kill ideas quickly, preserve instincts

Google's graduate school culture

• Offices modelled on Stanford grad school: four people per room, free food, casual environment
• The culture of perks and quirks was deliberate — founders were recreating the feeling that discovery is imminent
• Staff meetings ran long and unstructured; business procedures were added gradually

The auction pricing experiment

• Schmidt was convinced a second-price auction model for ads would bankrupt Google; he froze all spending to his personal approval on Friday mornings
• The team pushed it through anyway; after a week of daily crisis meetings, revenue was materially higher than before
• The result revealed that the search ads market was far larger than anyone had estimated
• Lesson: if you're stunned by success, you were managing well — you let the team prove it

Decision-making discipline

• Google adopted a weekly rhythm: staff meeting Monday, business meeting Wednesday, product meeting Friday
• Everyone knew which meeting decisions were made at; any issue could get a hearing within a week
• The YouTube acquisition was decided in approximately 10 days — clarity of ownership and readiness made it possible
• Quick decisions beat slow ones almost every time, even when the quick decision turns out to be wrong

20% time

• Engineers could devote 20% of their week to any project; product leads had to persuade engineers to join their teams — no assigned headcount
• Gmail, Google Maps, Google News, and AdSense all originated from 20% time
• The hidden function: 20% time gave employees a legitimate check on overreaching managers — "I'll give you 100% of my 80%"
• Failure during 20% time was understood as part of compensation; persistence means changing tactics, not grinding the same hill

Scaling at speed

• Once the ad auction model proved out, Schmidt told Omid Kordestani to fly to Europe and not return until offices were open
• Kordestani set up London, Paris, and Hamburg operations in one week by interviewing candidates in hotel lobbies
• Europe became 50–60% of profits in a $100 billion company; waiting a year would have cost enormously
• Knowing when to scale — and moving immediately — is among the most valuable decisions a leader can make

Hiring for chaos

• Google reviewed all hires centrally to avoid "glue people": well-meaning staff who sit between functions without adding direct value
• Persistence is the single best predictor of success in a knowledge economy; curiosity is the second
• Rocket scientists were favoured because they are inherently interesting; Olympians and elite athletes for the discipline demonstrated in reaching that level
• You cannot copy Google's perks and expect the same results; the system only works if you first hire people who thrive in ambiguity

Unexpected discoveries

• Google Earth and Maps gave researchers the first global view of cattle herds; they discovered cows consistently align north-south due to magnetic resonance in their brains
• No manager could have anticipated or scheduled that discovery
• Humans are far more varied in behaviour than leaders expect; staying open to surprise is part of the job