Creativity, Inc.

January 2, 2016

My notes from reading Creativity, Inc., a book by Ed Catmull (president of Pixar) in which he describes his career and its main focus: to develop a portable set of blueprints for creating a thriving creative company.


The author (Ed Catmull, computer scientist and president of Pixar) believes that figuring out how to build a culture that fosters creativity is the biggest mission of his career. That the cultural blueprints he seeks to define will, if achieved, outlive and expand beyond Pixar.

Getting Started

Unhindered communication, regardless of job title, is paramount. Don't create physical or social cues that undermine this (like having meetings at a long, skinny table with the leaders in the middle, and everyone else either being close the center or far away).

Bad management decisions can lead to multiple symptoms - things we notice and say "hey, that's fucked up". When trying to correct the decision by reversing the symptoms, you have to make sure to get all of them.

He mentions a particular college CS department as an early model for the kind of culture he wanted to create: attract a bunch of smart people, give them computers, and allow them to work on what interests them. He was surrounded by people who inspired him. They didn't know what was impossible.

When faced with a challenge, get smarter.

Hire the best people, even (especially) if they're better than you.

At NYIT, he created an environment similar to academia: hire smart people, flat org, just let them work on what's interesting. It was fun, but he ended up with a bunch of individuals working on individual projects. When he started at Lucasfilm, he wanted to figure out how to create a team with a common goal.

Lucasfilm was between Silicon Valley and Hollywood, but part of neither. It was its own protected environment where experimentation was valued, but the for-profit pressure was also in the air. People knew they were solving problems for a reason.

Visual polish doesn't matter nearly as much as getting the story right

There is nothing quite like ignorance combined with a driving need to succeed to force rapid learning.

Steve Jobs would sometimes be a jerk just to see how people responded - to see if they disagreed and were willing to stand up for themselves. It was like a form of sonar: send out a sharp impulse and learn about your environment by what bounces back.

Don't be seduced by simple answers ("start with a high price"), they can prevent you from asking the more fundamental ones ("how do I meet customer expectations"). This is especially dangerous when you feel in over your head and seek the advice of someone wiser.

Give ownership and responsibility of the quality of the product to everyone, especially those most closely involved with making it. Anyone on the floor can pull the cord that stops the assembly line. It's practical, but more importantly: people feel pride when they fix what is broken.

Make this part of your culture: "You don't have to ask permission to take responsibility."

When Toy Story launched and was a huge hit, Ed felt lost. He had achieved his dream of 20 years, and now he was... what? A manager? Why? He wasn't driven by the idea of "just" keeping a company running.

They moved on to Bug's Life, and Ed was shocked to find out that production managers had hated working on Toy Story. They felt disrespected compared to the artists and technicians. Ed had an open door policy, and couldn't believe that not a single production manager had brought this up.

Two things. First, they had hired experienced production managers who were used to working as freelancers; they didn't voice their objections until they were asked to stay and knew that their jobs were safe. Second, even though they hated it, they felt like they were making history and loved working for John Lasseter. The lesson: when times are generally good, people won't complain about what's bugging them because they don't want to be seen as complainers. This can fester until it destroys morale and the company with it.

Realizing that he had entirely missed the fact that 1/3 of his staff hated their jobs, Ed saw that there were real challenges in running the kind of company he wanted to run. Figuring these out became his new inspiration. How to establish a culture that outlived him, which enabled creative people to do work that made money and contributed to the world.

Two mantras came from Toy Story: "story is king" and "trust the process". The author then warns that maxims or rules are often useless platitudes that impede thoughtfulness.

Disney asked Pixar to make Toy Story 2 direct-to-video. They agreed, then realized it was a mistake. No-one wanted to make B quality work. They decided not to compromise on quality and to make it a theatrical release instead.

Bug's Life launched, and John turned his attention to the progress on TS2. It was a disaster. They had tried having John create the vision, then step away, allowing a less experienced team to handle the directing and execution. In the end, they had to replace the project leadership and redo everything. Why couldn't the junior team make art, even though it was John's story and they were building on everything from TS1?

The Braintrust came in, tweaked the story to have more drama, and then the whole company worked their asses off for six months to ship. One third of the team got RSI during this period, and one guy almost accidentally killed his baby by leaving him in the car because he was so fried. In the end, they met their deadline and some critics felt that TS2 was even better than the original.

If you give a good idea to a mediocre team, they will screw it up. If you give a mediocre idea to a brilliant team, they will either fix it or throw it away and come up with something better. The takeaway here is worth repeating: Getting the team right is the necessary precursor to getting the ideas right. Getting the right people and the right chemistry is more important than getting the right idea.

Not just smart people, but people who work together well to form a smart team. Ideas necessarily come from people, so get good people and you will have good ideas. There are no single ideas. Ideas take shape and get painted with details through a long process involving a whole team. eg: the iPhone; one "thing", but composed of millions of hardware, software, and design decisions.

Find, develop, and support good people, and they in turn will find, develop, and own good ideas.

After the TS2 death march, Ed resolved to focus on long-term employee health. Putting people first. Encouraging balance and a healthy life outside of work.

"Story is king" and "trust the process", simple and often repeated, didn't seem to have much value. It hadn't helped the TS2 team write a good story. "The process" had literally almost killed a baby. In fact, having the mantras gave a false assurance that things would be okay. "Trusting the process" made them passive and sloppy, if anything.

People glom onto words and stories that are often just stand-ins for real action and meaning.

Protecting The New

The Braintrust is a rotating group of storytellers at Pixar that meets to give iterative feedback on projects. Braintrust meetings are bluntly candid in a way that probably wouldn't scale to the whole company. Braintrust members focus all their energy on the product, none on politics or personal agendas.

Pixar movies all start out sucking. The only way they end up great is through an iterative process of candid feedback and improvements.

The Braintrust works for two reasons. First: it's made up exclusively of storytellers. People who have been through the process themselves. Second: the Braintrust has no actual authority. Directors are free to decide for themselves how to deal with the feedback they receive. Removing power dynamics makes it easier to focus on the idea, not the person.

Ideas only become great when they are challenged.

Good feedback is constructive, and should inspire the recipient. It may have suggestions, but cannot be proscriptive. The Braintrust would often identify a problem with a script and offer some solutions, but then the director - the person most intimate with the story - would take that feedback and come up with an entirely different, artistically brilliant way to solve the problem.

Instead of saying, "The writing in this scene isn't good enough," you say, "Don't you want people to walk out of the theater and be quoting those lines?" It's more of a challenge. "Isn't this what you want? I want that too!"

When building your own Braintrust, there are only two qualifications: find people who make you think smarter, and who come up with a lot of solutions quickly.

Failure is a manifestation of learning and exploration. If you aren't experiencing failure, then you are making a far worse mistake: You are being driven by the desire to avoid it.

You want to create a culture where no-one is afraid of making mistakes. Otherwise everyone will just do safe, mediocre work.

When experimentation is seen as necessary and productive, not as a frustrating waste of time, people will enjoy their work - even when it is confounding them.

However, some times there are too many failures and a director needs to be replaced or a project shut down. The best gauge of this is the crew: do they still have faith in the director? Are they confused? Also, it's a red flag if the director does not address issues brought up in Braintrust meetings. The Braintrust is a stand-in audience; if they say something doesn't work, it needs to be fixed somehow.

When there is failure, you need to take time to reflect on it and figure out what you can learn from it.

Trust is the antidote to fear, and creates an environment where the kind of ambitious work that results in mistakes can happen. Build that trust by being as transparent as possible, and making your employees feel like co-owners of the process.

Disney had a resurgence, and then went back into a slump. When they had some hit films, they hired more people, and then there came a pressure to keep them busy with projects, which led to mediocre work. They always had to "feed the beast." How could Pixar avoid this trap as they continued to grow?

New Pixar films start out as "ugly babies", not miniature versions of the finished product. Original ideas need protection during this phase - it's too easy to kill them before they can mature because they look ugly now. This is especially dangerous when companies focus on trying to make their process better (more efficient. feed the beast.) and forget that what they're actually trying to make is better product.

In an unhealthy culture, each group believes that if their objectives trump the goals of the other groups, the company will be better off. In a healthy culture, all constituencies recognize the importance of balancing competing desires - they want to be heard, but they don't have to win.

The Beast and the Ugly Baby are both important. In many companies the Beast takes over, because he's aligned with making the money flow, whereas the value of the Baby is almost always unclear at first. If either constituency "wins", it's bad for the company. There needs to be a dynamic balance between the forces.

Every creative organization ... is an ecosystem. "You need all the seasons," he says. "You need storms. It's like an ecology. To view lack of conflict as optimum is like saying a sunny day is optimum."

An example of "protect the new": Ed wanted to start an intern program, but none of the departments thought it was worth the expense. He used corporate funds to hire some interns that would be "free" for any department that wanted them. Eventually, people saw the benefit of having interns, and departments started volunteering the funds to hire more.

Fear of change is real, but more accurately: people fear experimenting with unproven ideas.

In many ways, it reminded me of Musical Chairs: We cling as long as possible to the perceived "safe" place that we already know, refusing to loosen our grip until we feel sure another safe place awaits.

The challenge: how do you know when you're avoiding change for the wrong reasons? Some ideas get enshrined to the point where it's taboo to even talk about changing them ("but it's the Pixar way to do X like this!"). When is that good (protecting your values), and when does it hold you back?

Rather than fear randomness, I believe we can make choices to see it for what it is and to let it work for us. The unpredictable is the ground on which creativity occurs. ...change is our friend because only from struggle does clarity emerge.

Up went through four major revisions. Change was an inevitable part of the process, and lead to a great film. The key was to focus on the emotional core of the story they wanted to tell; that never changed - the implementation details did.

Removing fear of change gets people to think more freely. Thinking freely leads to good ideas that people are excited about following, even though it means making changes. Remove fear with phrases like "I'm not actually suggesting this, but what if we...." Also, leave time for play and experiments.

If you don't try to uncover what is unseen and understand its nature, you will be ill prepared to lead.

Ed believes that smart companies do dumb things when their leaders aren't actively looking for blind spots. Corollary: there are problems, right now, in your company that are invisible to you, and may one day bring you down.

When someone is promoted, they lose touch: people are less candid around them. If you associate corporate rank with personal value, people will manage upward and mistreat those "below" them.

When viewed from a single vantage point, a full picture of the dynamics of any group is elusive.

If you become successful, it becomes even harder to know what's going on at your company: your success convinces you that you're doing the right things. When you're convinced you're right, you don't consider alternative viewpoints.

The point is to recognize that you (as an individual) will always have blind spots. Trying to remove them all is futile. Instead, make sure everyone in the company feels comfortable speaking up, and recognize that different viewpoints are additive, not competitive.

Building And Sustaining

Adding people to a group makes it less flexible, and the additional viewpoints can get in the way of collaboration. Pixar has a few strategies for dealing with this problem.

Dailies, or solving problems together - "Dailies" are morning meetings to share and analyze the team's ongoing work. It makes the work better, but it also gets people in the habit of being open with one another. Everyone shares incomplete work, making themselves vulnerable, and no-one is mean with their feedback.

Research trips - When people at Pixar feel that they're being derivative - copying other people's art - they go out to real locations and study them (MIT for Monsters, Inc; Parisian kitchens for Ratatouille; etc). It's what keeps them "creating rather than copying". All those details and new ideas are brought back into the film.

Craft is what we are expected to know; art is the unexpected use of our craft.

The power of limits - It's important to make implicit limits visible and quantified so that everyone knows what they're working with. "Want this feature? Ok, what do you want to take out to make room?" Have the people doing the work manage the constraints, not an outside committee or manager.

Integrating technology and art - Three leaders of Pixar: Ed the computer scientist, John Lasseter the artist, Steve Jobs the businessman. Pixar's tech isn't just modeling and rendering, they also build tools that make it easier to seek and give feedback on projects and concepts, speeding up the improvement loop.

Short experiments - Pixar makes short films that run before their feature films. They don't have any commercial value, but they send a message within the company that artistry is important. They also provide a cheap place to make those "necessary mistakes" discussed earlier.

"Better to have train wrecks with miniature trains than with real ones."

Learning to see - Biases get in the way. Even when drawing: your brain has a distorted idea of what things are "supposed" to look like, and that makes it harder to draw what you actually see.

Postmortems - Five reasons: consolidate what's been learned, teach others who weren't there, don't let resentments suffer, use the schedule to enforce reflection, and to pay it forward by preparing for the next project with the new wisdom from the last.

A large portion of what we manage can't be measured, and not realizing this has unintended consequences. The problem comes when people think that data paints a full picture, leading them to ignore what they can't see. Here's my approach: Measure what you can, evaluate what you measure, and appreciate that you cannot measure the vast majority of what you do.

Continuing to learn - Pixar offers free classes in all kinds of things to its employees - mostly artistic, like sculpture and ballet. Having people from different parts of the company learn together, where they're relaxed and there's no hierarchy, is immensely valuable. It also taught people humility and respect for other people's jobs.

In my experience, creative people discover and realize their visions over time and through dedicated, protracted struggle.

Share your solutions as well as your problems. A team will follow a leader that makes mistakes, as long as they own those mistakes, share their reasoning, and decisively steer a new course.

"If you're sailing across the ocean and your goal is to avoid weather and waves, then why the hell are you sailing? If your goal is to make it easier and simpler, then don't get in the boat."

Most directors at Pixar have some mental model that they use to think about the process of making a movie: walking through a maze, sailing a ship, climbing a mountain. As a leader, you're taking your team on a journey to an unknown destination.

Testing What We Know

When Disney bought Pixar, Pixar's leaders came up with a list of 59 items that they promised would not change as a result of the merger.

Arriving at Disney, Ed was immediately struck by the fact that their culture emphasized mistake avoidance. Their office was also laid out in a way that elevated and separated senior employees from everyone else.

They kept Pixar and Disney completely separate: different offices in different cities, no shared production work. It avoided a bureaucratic nightmare, but more importantly it forced each studio to stand on its own. If there was a problem at one studio, having the other studio bail them out might mask the problem instead of forcing it to the surface. Finally, it let each studio know that when they finished a film, they had done it themselves.

Disney's feedback system was broken. Directors got three sets of notes, from different departments. The notes were from people who had never made a film before, and they were mandatory. The creatives didn't have a voice.

They started doing Braintrust-style meetings, but at first there was little trust, a wariness around hurting feelings, and as a result the feedback was superficial and useless.

Tangled became the first number one movie for Disney in 16 years. When a movie makes enough money to trigger bonuses, John and Ed personally go around to each employee to hand them their bonus check and say "thank you".

Easy isn't the goal. Quality is the goal.

As Disney started to turn around, Ed realized that he had achieved his goal. The Pixar culture was portable. Applied to a new group of people at a dysfunctional company, it had allowed and inspired them to do great creative work.

I don't believe in simple, prescriptive formulas for success. I wanted this book to acknowledge the complexity that creativity requires.

Now that Pixar was established and successful, new employees were afraid to bring up new ideas. Who were they to challenge how things were done at the mighty Pixar?

Increasingly, we sensed that our people, having enjoyed years of success, were under a great deal of pressure not to fail. Nobody wanted to have worked on the first movie that didn't open at number one. The specter of past excellence was sapping us of some of the energy that we'd once used to pursue excellence.

In addition to employees that were increasingly afraid of taking risks, Pixar was facing incrementally increasing production costs that made it harder to work in small teams on unconventional ideas. They came up with Notes Day: they would shut the studio work down for a day, and have all employees come to the office and spend the time talking about how to fix these problems.

In the lead-up to Notes Day, thousands of ideas were submitted. The leadership narrowed it down to a list of ~120, using the criteria "could a group of 20 discuss this for an hour?".

Pixar was currently spending about 22,000 person-hours on each film. When brainstorming Notes Day ideas, they had asked their employees to imagine what would need to be done to get that number down to 18,500. The most popular discussion topic, which was proposed by multiple employees, was "how could we make a film with 12,000 person-hours?". They were being even more ambitious, and the idea of succeeding under greater constraints was immensely interesting to people.

From the ideas, discussion sessions were set up, and people attended whatever interested them. Each session was run by a trained facilitator, and everyone who attended had to fill out an "exit form" designed to translate the discussion into action items. The exit forms were structured but open-ended, and customized for each session. Each form asked the questions "who's the best audience for this idea?" and "who should pitch this idea?". In addition to capturing good ideas, they wanted to identify passionate champions for those ideas.

John opened Notes Day by addressing the whole company and saying that, of the suggestions sent in, there were two and a half pages' worth about him personally. He told everyone that it was tough to read, but that he valued it greatly. That some of the discussions that day were going to seem personal to some managers, but that they needed to be tough and encourage honesty anyway. With his speech, John led by example and set the tone for the day.

After Notes Day, the people identified by the "who should pitch this idea?" questions on the exit forms started meeting with the exec team. They worked on honing their pitches, and were put in charge of evangelizing that change. This ensured that Notes Day wasn't just a "feel good" exercise, but actually resulted in change.

I believe the biggest payoff of Notes Day was that we made it safer for people to say what they thought. Notes Day made it okay to disagree. That and the feeling our people had that they were part of the solution were its biggest contributions.

What made Notes Day work so well?

  1. A clear and focused goal ("how can we cut production costs by 10%?").
  2. It was championed by the highest-ranking members of the company. This made it clear to everyone that Pixar was serious about it.
  3. It was led from within. Notes Day was all about the people who worked at Pixar, and what they could do to make their company better.

Unleashing creativity requires that we loosen the controls, accept risk, trust our colleagues, work to clear the path for them, and pay attention to anything that creates fear.