Reed Hastings: How Netflix Built a Culture That Fires Good People

KEYPOINTS

• 01

  Treat loyalty as a stabiliser, not the goal.

  Netflix's early struggle was balancing loyalty against performance. Hastings settled on the pro-sports-team bargain: you play for your position each season, but the team can be close and supportive. Loyalty means one bad day doesn't get you fired and a bad day from your boss doesn't make you quit. Ultimately the company valued growth and achievement over lifetime loyalty. He sees friendships and marriage as different: there, loyalty comes before performance.

• 02

  The keeper test flips the default from 'bad enough to fire' to 'worth fighting to keep'.

  Ask: if this person told you they were quitting, how hard would you work to change their mind? If you wouldn't, give them a generous package and recruit someone you would fight for. The industrial norm is that 'you nearly have to commit a crime to get fired'; the sports-team norm is 'I think I can upgrade.' The nuanced version is an honest conversation first, so nobody is surprised.

• 03

  Generous severance is really a tool for managers.

  The money goes to the departing employee, but managers are usually chosen for liking people and find it hard to let them go; a big package makes it easier for everyone. Departing employees sign a release, so Netflix has seen hardly any lawsuits. Hastings sees performance improvement plans as a false system of legal protection, since so few people on one turn around. Long-tenured people in a rough patch get context and are often moved to a new role.

• 04

  Manufacturing drives out variation; innovation needs more of it.

  For about 200 years manufacturing drove GDP growth, and its logic of consistency and error elimination shaped everything, including factory-like classrooms. Process is the language of reducing variation, which is right for safety-critical work like airlines. In inventive businesses, where the main threat is someone inventing nearby, errors are part of the process and too much process kills innovation.

• 05

  Freedom and responsibility beats policy plus enforcement.

  Most companies set per-city expense limits — $250 here, more in Tokyo, less in Omaha — and need a bureaucracy to enforce them. Netflix replaced all of it with 'travel as you would on your own nickel.' Occasional abuse costs far less than enforcement, and it treats people like adults. You still set context with principles and examples.

• 06

  Hiring is imperfect, so take chances and judge by performance.

  Five or ten hours of interviews rarely reveal much. Hastings stays open-minded, expects a three-month honeymoon before cracks show, and coaches first. His highest-value step is a meal: how a candidate treats service staff shows their character. He reads persistence from track record — job moves every 18 months versus stints of 6, 3 and 7 years.

• 07

  Make reference checks real by finding people loyal to you, not the candidate.

  Supplied references are anodyne. He combs LinkedIn for people at the candidate's previous employer, starts a conversation, and only then reveals he's calling as a reference. His recurring question: would you hire them again?

• 08

  Two religions: customer satisfaction and operating income; everything else is a tactic.

  Customer satisfaction drives the top line, operating income the bottom. Echoing Bezos — firm on principles, flexible on tactics — he notes it's easy to grow one at the other's expense. The art of business is improving both faster than competitors.

• 09

  Big creative bets are judgement calls, and information grows by stage.

  Ted Sarandos won House of Cards, a roughly $100 million bet, by bidding for two seasons, which HBO's precedent wouldn't match. With a partial script and some casting, no data can tell you how good it will be. Each stage — script, cast, dailies, edit, test screenings — improves the odds. Netflix expected Squid Game to be big in Korea; its global success surprised them after 20 years.