The Pixar Braintrust had one rule that made everything else work: the directors sitting around the table could tear a film apart, scene by scene, but the filmmaker on the receiving end was under no obligation to change a single frame. Ed Catmull, who described the system at length in his 2014 book Creativity, Inc., was emphatic on this point. The group offered diagnosis. It never prescribed. The director kept full authority over the picture.
That asymmetry — total candor paired with zero authority — was the entire mechanism.
Why the powerlessness was the point
Catmull’s argument was simple and slightly counterintuitive. The moment a feedback session carries the threat of a mandated change, the people giving feedback start performing. They soften. They calculate. They reach for the version of the note that won’t humiliate the director, won’t sound aggressive in front of the CEO, won’t get them blamed if the suggested fix flops. Inside one meeting, the candor evaporates and what remains is a polite committee.
Strip the authority out and the calculation goes with it. A peer telling a director that the second act doesn’t earn its emotional climax is not issuing an order. The director can nod, take notes, and walk back to the cutting room and ignore every word. So the feedback-giver has no reason to hedge. And the receiver, knowing nothing is being imposed, has no reason to defend.
This is the same principle that workplace researchers later formalized as psychological safety. Harvard Business School professor Amy Edmondson coined the term “team psychological safety” to describe the shared belief that it is acceptable to take risks, voice concerns, and admit mistakes without fear of negative consequences.
How the room actually worked
A Braintrust session happened every few months on every film in development. The assembled group was a rotating set of Pixar’s most experienced storytellers. They would watch the latest reel of the film in trouble. Then they would talk for two or three hours.
The notes were direct. Catmull has described meetings where directors heard that their protagonist was unlikable, that the emotional arc collapsed in the middle, that an entire sequence didn’t belong in the movie. There were no soft preambles. The room operated on the assumption that everyone present had earned the right to be there by having made films themselves, and that the work — not the maker — was the subject.
The director sat through it. Took notes. Asked clarifying questions. And then — this is the part most companies skip — went away and decided alone what to do. Some notes got integrated. Some got rejected outright. Nobody in the room had standing to follow up and demand compliance. The studio’s executive layer, Catmull included, deliberately stayed out of the prescriptive role.
The films that came out of this process — Toy Story 2, Finding Nemo, Ratatouille, Up, Inside Out — were not the films that walked into those rooms. Most were in serious trouble at some point in production.
What most companies get wrong about “candid feedback”
Walk into almost any corporate environment and you will hear leaders insist they want honest input. Then watch what happens when someone gives it. The feedback is met with a clarifying question that is really a counterargument. Or it is accepted in the room and quietly remembered at performance review time. Or — most commonly — the senior person in the meeting agrees with the criticism and then issues a directive based on it, which means the next person who speaks up is now responsible for an action item they didn’t ask for.
Every one of those reactions teaches the room the same lesson. Speaking up costs something. As Forbes contributor Rodger Dean Duncan put it, psychological safety is fundamentally about being able to take a risk without penalty or punishment — and “penalty” includes the soft punishments of being volunteered, blamed, or quietly downgraded.
When individuals believe their answers will be evaluated, they shift toward what they think the evaluator wants to hear. Put that same person in a meeting where the boss controls their bonus, and the distortion compounds.
Catmull’s insight was that authority itself is the contaminant. You cannot have honest feedback and binding feedback in the same conversation. Pick one.
The two-job problem
Most managers running creative or strategic review meetings are quietly trying to do two jobs at once. They want to hear the truth about the work. They also want to be the person who decides what happens next. Those goals fight each other.
If the room knows the manager will decide, the room starts speaking to the manager rather than to the work. People stop saying what they actually see and start saying what they think will land. The smartest person at the table becomes the person best at reading the manager’s face.
Catmull’s design splits the jobs. The Braintrust diagnoses. The director decides. Neither role is allowed to creep into the other. A peer cannot mandate. A director cannot punish a peer for the diagnosis. The structural separation is what allows both roles to be performed well.
This is also why simply telling a team “please be candid” almost never works. The instruction does not change the underlying incentive. As leadership experts have noted, psychological safety is a structural condition, not a personality trait of the boss. You can be the warmest, most empathetic leader in the building and still preside over a meeting where nobody tells you the truth, because the formal authority you carry into the room is doing the talking.
How to borrow the mechanism
Pixar’s setup is hard to copy whole, but the underlying move is portable. A few practical translations show up in companies that have actually tried it.
Separate the review meeting from the decision meeting. The first one is for diagnosis only — no commitments, no action items, no “so what are you going to do about it.” The owner of the work goes away and comes back later with their own plan.
Strip seniority signals from the room. If the CEO is present and visibly weighing in, the room is no longer giving peer feedback — it is auditioning. Catmull himself usually sat in but kept his opinions calibrated, knowing his title bent the conversation.
Protect the receiver’s authority publicly. The director’s right to reject every note has to be obvious and defended, or the next director will start pre-complying.
And accept the trade. Some good notes will get ignored. Some bad films will get made anyway. The system tolerates that loss because the alternative — a room full of people performing agreement — produces worse films more reliably. Building psychological safety from day one points to the same compounding effect: teams where speaking up is genuinely costless make better decisions over time, not because any single meeting is better, but because the information loss across hundreds of meetings is smaller.
The Braintrust outlasted Catmull’s tenure as Pixar president. The films kept coming. The mechanism kept working. And the rule that made it work — that the people with the sharpest things to say had no power to enforce a single one of them — turns out to be the rule most companies are least willing to adopt.
