Ed Catmull, Pixar: Keep Your Crises Small

Key Questions:

  • Why do successful companies fail?
  • Is our central problem finding good people or good ideas?

Pixar got a few things right:

  • Artists and Technology people were peers – same compensation, socialized together, worked closely together.
  • People felt comfortable about expressing their problems
  • Brain Trust. Remarkable at telling stories. They had complete trust in each other. Necessarily honest. [Similar to Blizzard.]
  • Review Process. They reviewed the material every day – even if it wasn’t done. [Always be ready to ship. Agile-like concept application as well.] When you get over the embarrassment of showing incomplete work, you get over the embarrassment and become more creative. [Iterative process.]
  • Don’t confuse the organization structure with the communication one. Communication needs to be able to happen between anyone in the company at any time. You still need people controlling activities, but not relationships.
  • Managers hate being surprised. It’s a sign of disrespect. That someone else talked about the problem ahead of time is a good thing. Get over it.

Success hides Problems.

Adding a group of new people. They loved working on Toy Story and at Pixar. As a result they were willing to put up with a lot of things which they didn’t like. When you’re healthy, and have lots of resources, you don’t have address the problems. Often, people let that get in the way of diving deep and solving the problem. Just being aware of it is NOT enough.

Two different Standards of Quality.

This isn’t good for your soul. Don’t confused incomplete with poor quality. You want to see continuous improvement. If you don’t see that, that’s how you know you won’t hit your quality target.

Don’t be afraid to throw away your work if it’s not good enough. Stick to your quality standard, even if you don’t have enough time. Do it anyway. [Ugh. Here comes crunch.] Adopted two ways of how they work as a result. Limit the number of hours people can work. Perks galore. Focus on keeping people physically fit.

Toy Story 2 Story

You need to have people believe that the characters have a real choice to make – or you don’t have a movie. Life goes on. It changes. You can’t hang on to that. These things, which we can all appreciate, are what turns it into a real movie instead of a cheap followup.

Good ideas or good people?

If you have a good idea and you give it to a mediocre group, they’ll screw it up. If you have a mediocre idea, and you give it to a good group, they’ll fix it or they’ll throw it away and do something else.

It’s not just one idea. It’s thousands of idea in any successful product. You have to get most of the right.

The goal of development is not to find good ideas – it’s to put groups of smart people together. As a result, the development group is there to help to support. The measure of success is how well the team gets together over time.

The only failure is that you don’t progress. Not that your initial product was high quality or commercially successful. [This is way the blockbuster-only strategy is bad. Mobile is a godsend for this. 52 games before Rovio got to Angry Birds…]

Good Artists Steal

Copying is a way to learn, but it’s shallow. You should remake bad movies – not good ones. Find ones with a good idea, but poor execution. You could copy the technology, but our competitors couldn’t copy the process Pixar used to write the story.

Things don’t get easier.


First ones were successful. Highly valued. Safe. People got tired. Defensive. Show-off what they did. Not really in-depth analysis. They change how they do them every single film. Currently, they are asked to pick the 5 things they wouldn’t do and the 5 things they would. Get a lot of facts about the process. This leads to a new theory about how to do things. About 1/3 third of the new things is wrong. That’s OK. [Think Fast & Slow concepts abound here.]


  • Constant Review
  • Safe to tell the truth
  • Communication should not mirror the organization hierarchy
  • People, and how they function, is most important
  • Do not let success mask problems.

Pay attention to what affects behavior

Everyone says “The story is the most important thing.” That statement doesn’t affect behavior because there are clearly movies with bad stories.

So it doesn’t matter that it’s true.

Once can articulate an important idea in a concise statement, then one can use the statement without having a fear of changing behavior.

Why do companies fail?

Organizations are inherently unstable, but they fall slowly. Most people don’t notice it. Look for the hard truths with constant assessment.

Kinds of Crisis

  • If you don’t like what you see. Changes are just hard work. What makes it a crisis is having to rearrange people to affect those changes. It’s an emotional thing to do. Action taken, or not, this is a self-imposed crisis.
  • If your audience does’t like what they see.

Keep your crises small

Question: How do you document postmortems? Turn it into something actionable?
Staff members who are notetakers. People in preparing for it get a lot of information. People who are presented to are the leaders of a follow on film, so that’s the knowledge hand-off. Having the discussion also surfaces other data.

Question: Unintelligible.
The process of gathering the evidence is mixed. Sometimes you haven’t kept tracked by all of the detail. People are fascinated [and surprised] by the information because sometimes it doesn’t match their intuitive understanding. Sometimes facts give you pictures you can’t see otherwise. [You have to specifically hire for this.]

Question: Where do the ideas come from?
It’s important to give credit to a team of people. You can’t do that to an idea. Phrase it this way to build comradery. You have to support them. People first. Ideas are the result of the process. It’s thousands of decisions…

Question: Unintelligible.
The notion of self-abasement is lacking. Some people get it. Some people don’t. There are no guarantees. It requires consent diligence and thought process.

Question: Unintelligible.
It’s a bad idea to target films to kids. It’s talking down to them. Kids live in an adult word and they’re used to things which they don’t understand.

Other Pixar specific questions which are answered by reading The Pixar Touch: http://www.amazon.com/Pixar-Touch-David-Price-ebook/dp/B0010SKT0M/ref=sr_1_1?ie=UTF8&qid=1391543842&sr=8-1&keywords=pixar+touch. Waste of his time.

Question: Is Renderware good long term for Pixar?
With any piece of software it requires a huge amount of work to keep up. It’s to Pixar’s advantage that there is an industry wide standard. Their competitive advantage is how the people work together – not the software.