Google gets cited as a culture example because of the free food. Patagonia gets cited because of the environmental mission. Netflix gets cited because of "freedom and responsibility." Spotify gets cited because of squads and tribes.
All of those are downstream of the actual thing. The thing is the set of default behaviors the team runs on when nobody is watching. Free food is a benefit. The mission statement is a poster. The values list, the one with five adjectives and an oxford-comma layout, is marketing. None of those three are culture.
Culture is what happens in the first thirty seconds after someone screws up. It is who speaks first in a design review when the senior engineer has been quiet for ten minutes. It is what the team does when the deadline slips. It is whether the new hire asks the dumb question in the meeting or waits to DM one person afterward. Those defaults are the culture. Everything else is the wallpaper on top.
This matters because company culture is the number one factor people cite when they leave a job. A 2025 MIT Sloan analysis found culture to be 10.4 times more powerful than compensation in predicting voluntary turnover (MIT Sloan Management Review). And companies that rank in the top quartile of employee experience, the consumer-facing output of culture, have net profit margins 8.4 percentage points higher than the bottom 75 percent (MIT CISR, 2017). Culture has financial weight. The question is how to actually produce it.
The Four Examples Worth Studying
There are four companies whose culture has been documented in enough detail to actually learn from. Each took a different approach. Each produced different outcomes. All four made their default behaviors explicit and wrote them down in a way the rest of the organization could practice against.
Netflix: The Memo and the Keeper Test
Netflix's 2009 culture deck, written by Reed Hastings and Patty McCord, is the most-read slide deck in Silicon Valley history. Sheryl Sandberg called it the most important document to come out of the Valley. It has been revised four times since the original release (Quartz).
The core behavior Netflix specified was the "keeper test." Managers were told to ask themselves a question about every direct report: if this person came in tomorrow and resigned, would I fight hard to keep them, or would I accept the resignation with some relief? If the answer was the second one, Netflix expected the manager to let the person go. Today. With a generous severance package. The second stated behavior was "freedom and responsibility," meaning fewer rules, higher expectations, and more discretion pushed to individual contributors.
What it produced. A high-velocity organization where performance standards were extreme, decisions moved fast, and top talent had broad autonomy. Netflix's per-employee revenue became one of the highest in the industry.
What it cost. Former Netflix employees have described the experience as "terrifying, with anxiety 24/7" (The Week). Constant evaluation produced fear of failure, which stifled the creativity and risk-taking the company said it wanted. Institutional memory was hit by regularly letting long-tenured people go. Recruitment got harder as the company's reputation for harshness spread. The keeper test, applied poorly, became cover for ordinary workplace politics.
The Netflix approach worked for Netflix's specific context: high-growth streaming with a need for fast, high-stakes creative decisions. It did not transfer cleanly to other contexts. Companies that tried to copy it without the extreme severance packages and the hiring rigor got the downsides without the mechanism.
Amazon: The Six-Pager and the Leadership Principles
Jeff Bezos made the six-page narrative memo Amazon's official meeting format on June 9, 2004. He sent a now-famous email announcing that no Amazon team member would be allowed to bring PowerPoint or bullet lists into a meeting. Every idea had to emerge from a densely written, narrative memo (CNBC).
The meeting format became the first 30 minutes of silent reading, followed by discussion. Bezos's rationale: "The reason writing a good four-page memo is harder than writing a 20-page PowerPoint is because the narrative structure of a good memo forces better thought and better understanding of what's more important than what." Bezos said a great memo should take "a week or more" to write and should be shared, revised, set aside, and edited with a fresh mind.
Alongside the memo culture, Amazon published its 16 Leadership Principles: Customer Obsession, Ownership, Invent and Simplify, Are Right A Lot, Learn and Be Curious, and so on. The principles were referenced in hiring, promotion, performance reviews, and everyday operational decisions.
What it produced. A scaling engine that could handle enormous operational complexity without losing decision quality. Meetings were slower on the front end and faster on the back end. Memos captured enough context that decisions held up when later challenged. The leadership principles gave everyone a shared vocabulary for difficult trade-offs.
What it cost. Writing great memos is exhausting. The memo format incentivized a narrow, polished style of thinking that sometimes suppressed lateral creativity. The leadership principles, applied literally, became weapons in political fights. Glassdoor reviews and the 2015 New York Times exposé on Amazon's working conditions surfaced a pattern of employees pushed past sustainable limits in the name of Customer Obsession and Ownership.
Bridgewater: Radical Transparency
Ray Dalio introduced radical transparency to Bridgewater Associates in 1993, after three of his top confidants sent him a memo saying his bluntness was hurting the firm. Dalio decided to systematize the bluntness instead of softening it. He wrote down his lessons as a Philosophy Statement that eventually became 300+ Principles governing how Bridgewater employees were expected to operate (Verified Investing).
The behaviors Bridgewater specified: every meeting was recorded. Disagreements were conducted openly, on the merits, without deferring to hierarchy. Employees ranked each other in real-time public scorecards called "dots" based on how well they embodied each principle. The goal was "thoughtful disagreement," meaning surfacing conflict fully and resolving it on evidence, not politics.
What it produced. A culture where bad ideas got killed fast and where junior people could, in theory, challenge senior people without career consequences. Bridgewater became the largest hedge fund in the world, with Dalio publicly attributing the success to the Principles.
What it cost. High turnover in the first 18 months as new hires experienced what Dalio called "amygdala hijack," their brains interpreting public criticism as a threat. A 2016 sexual harassment claim described the culture as "a cauldron of fear and intimidation," with constant video surveillance and orientation videos showing senior executives reducing a female manager to tears (Westport Journal). Rob Copeland's book on Bridgewater argued that the more time employees spent on the Principles and their associated dots and trials, the worse the fund's investment performance got.
Radical transparency worked at Bridgewater for people who could handle the intensity. For everyone else, it produced the behaviors it was supposed to prevent.
GitLab: Handbook-First Remote Work
GitLab runs the world's largest all-remote company with 2,375 employees in more than 70 countries as of 2025 (up from 1,350 in 2021). The operating mechanism is the GitLab Handbook, which spans 2,700+ pages and serves as the single source of truth for how the company operates (GitLab Handbook).
The specified default behaviors: handbook-first (if a policy, process, or decision is not in the handbook, it is provisional), public by default (internal information is accessible unless explicitly private), asynchronous communication over meetings (organizers must justify any meeting), and a bias toward writing things down rather than explaining them verbally.
What it produced. A fully remote company that scaled from early-stage startup to publicly traded at massive distribution without the coordination collapse most remote companies hit around 100 employees. New hires could onboard themselves. Cross-functional collaboration worked because everyone had the same written source of truth.
What it cost. Writing for the handbook is slow. Decisions that could be made in a five-minute conversation take longer because the documentation burden is nonzero. Employees who prefer live dialogue found the written culture isolating. And the handbook, despite being called a handbook, reached encyclopedia scale, which made it harder to navigate for anyone who had not been inside the company for months.
What These Four Examples Have in Common
Four very different cultures, and on the surface nothing in common. One is a hedge fund. One is an entertainment company. One is a global retail platform. One is an open-source developer tools company. Their specific practices look nothing alike.
Look closer and the shared mechanism appears.
All four specified their default behaviors explicitly. None of them said "be excellent" and hoped for the best. They wrote down the exact behaviors they wanted and gave people a language to reference those behaviors in daily work. Netflix had "freedom and responsibility" and the keeper test. Amazon had 16 Leadership Principles and the memo format. Bridgewater had 300 Principles and the dotting system. GitLab had 2,700 pages of handbook.
All four practiced those behaviors continuously. Specification without practice is just a poster. Every one of these companies had a daily mechanism that pulled the written behaviors into real decisions: the keeper test conversation, the six-pager meeting, the recorded Bridgewater session, the handbook-first policy. The practice was the culture.
All four paid a cost for the intensity. There is no free version. You pick your default behaviors knowing some people will love them and some people will bounce hard. The people who stay are the ones for whom the defaults match their own.
The Part Most Companies Skip
Here is the pattern. Companies read one of the four examples above, pull out a surface practice (the values list, the memo format, the handbook, the transparency), and implement it without the rest of the system. They get the symptom, not the mechanism. Three months later the initiative dies and the culture is exactly where it started.
The mechanism is specificity plus practice. The companies that actually changed their culture did two things. First, they defined a small, concrete set of default behaviors, specific enough that you could run a scenario past them and know what the right move was. Second, they ran the team through those scenarios repeatedly until the defaults were automatic.
Amy Edmondson's research at Harvard on psychological safety is the academic version of the same finding. Posters and workshops do not produce safety. Repeated practice of specific behaviors does: asking dumb questions, admitting mistakes in public, disagreeing without punishment. Teams that practice those behaviors get them. Teams that only talk about them do not (HBR).
Practice is the missing word in most company culture content. Culture is a drill, and teams that only talk about their values never run the drill.
Why Offsites Do Not Work
The dominant American approach to company culture is the offsite. Two days, a nice location, a ropes course or an escape room or an improv workshop, a catered dinner, and a flight home. Every HR leader has done one. Most have done several.
The research on offsites is brutal. A 2024 meta-analysis on team-building interventions found that single-session team-building events produce effects that decay within 2-4 weeks and leave no measurable impact on actual team behavior or performance outcomes (Klein et al., 2009 meta-analysis). The reason is the mechanism. Offsites give you one practice session. Culture is produced by thousands.
This is why remote team culture built on forced fun falls apart so quickly. The forced-fun session is the equivalent of a single simulator flight for a pilot. The pilot who takes one flight and never takes another one does not become a better pilot. The team that does one offsite and returns to its actual defaults does not become a better team.
The Flight Simulator Model for Company Culture
The field that has solved the culture-as-practice problem best is commercial aviation. Pilots do not become a good crew by reading a memo about crew resource management. They become a good crew by logging hundreds of hours in a simulator where specific coordination scenarios are played out, debriefed, and replayed.
The simulator does three things the offsite cannot. It repeats the same scenario enough times for the behavior to become automatic. It observes what the crew actually did without distorting the behavior by announcing "this is the measurement." And it separates practice from live performance so mistakes do not cost lives.
QuestWorks is the flight simulator for team dynamics, applied to knowledge work teams. Teams play voice-controlled scenario challenges on QuestWorks' own cinematic platform. Each scenario is designed around specific team behaviors: conflict resolution, decision-making under pressure, information handoffs, backing up teammates in ambiguous situations. The system generates behavioral data on how the team actually coordinated. QuestDash surfaces the patterns as aggregate trends and strengths-based callouts. HeroGPT, the private AI coach, helps individual players develop the behaviors the simulator surfaced. Participation is voluntary and never tied to performance reviews.
The mechanism is the same one Netflix, Amazon, Bridgewater, and GitLab used. Specify the behaviors. Practice them. Debrief them. Run them again. What QuestWorks does is make the specification and practice accessible to companies that are not willing to spend ten years writing principles documents or build a 2,700-page handbook.
The Working Definition
Great company culture is a set of default behaviors specific enough that the team knows what the right move is in ambiguous situations, and practiced often enough that the behaviors happen automatically under pressure.
Everything else is marketing. The values poster is marketing. The mission statement is marketing. The culture deck is marketing. The free lunch is a benefit. The offsite is an event.
The behaviors are the culture. The practice is the production mechanism. The companies that get this right, Netflix, Amazon, Bridgewater, GitLab, and a handful of others, produced real cultures with real outcomes and real costs. The companies that skip the practice layer produce culture-shaped content that looks fine in marketing and collapses under pressure.
If you want to diagnose your own culture before deciding what to practice, the starting point is asking what your team's actual defaults are when nobody is watching. The gap between what the team says it values and what it actually does under stress is the work.
QuestWorks starts at $20 per user per month with a 14-day free trial. It integrates with Slack for install and invites, then runs on its own platform for the scenario practice. Your team plays. The behavioral defaults surface. You start running drills on the ones that matter. That is what great company culture actually requires, and most companies skip it.
Try it free
