Creativity Inc and Software Dev: Break Large Problems Into Small Ones

As I mentioned in a recent post, I've been reading Ed Catmull and Amy Wallace's new book Creativity, Inc. and continue to find many similarities between how Pixar develops films and how those techniques can be used to develop software in an Agile fashion. This is the first post to explore some of these similarities.

I'm going to start with the following quote, where Mr. Catmull is talking about coloring a bottle:

It was easier to figure out how to color and display each tiny piece

(Excerpt From: Ed Catmull with Amy Wallace. Creativity, Inc., Chapter 1, P. 37, iBooks.)

I like this quote being the source of my first post as this is a fundamental tenet of most Agile development methodologies. In traditional methodologies, the team starts with a requirements specification that can have hundreds or thousands of individual requirements. Trying to analyze a project with a large requirements specification can be overwhelming - it is easy to get lost in the details, and most times the details aren't necessary to come up with accurate estimates or the system design. On the other hand, breaking up a large project into small pieces of functionality (the Agile term for these pieces are stories) makes the project more manageable and easier to understand.

Let me give an example of using stories that ties back to the quote above. I don't know how well this is going to work, but let's just go with it. Assume you work for a company and you're asked to build a customer loyalty system, maybe to be more specific a new frequent flyer program for an airlines company. Just like the bottle in the quote above, thinking about this system as a whole could be overwhelming! Where do you start? What are the most important pieces of functionality? Somehow you need to break up the system in smaller components.

Using a traditional development methodology, you might have a requirements specification with thousands of requirements. This would be akin to shattering the bottle (OK, possibly over the top, but still, work with me) - sure, it might be easy to "color" each tiny little piece but you'll likely never be able to put all the pieces back together. You could also quickly lose sight of the bigger picture.

Instead, an Agile approach would be to define the system in small pieces, each piece representing a specific feature or function the system must perform, such as:

• Users must be able to log in
• Once logged in, users should be able to see their transaction history
• Users should be able to shop for flights using their points/miles
• Users should be able to request credit for missing miles

In this way, it is easier for all interested parties (business owners, project managers, developers, testers) to understand the system. We can also decide on the priority of each story/feature, providing flexibility in building the system. Writing stories is a difficult task but if done well it can be a very powerful and flexible mechanism.

This is just a quick summary of stories and a couple of their benefits. I'll talk a lot more about them in later posts.

Creativity, Inc., Culture and Software Development

Readers of my blog probably know I've been a long-time admirer of Pixar's culture. Yes, my interest in Pixar started with their films, but as early as Monsters, Inc., what really piqued my interest was their culture. Their films were amazing, full of creativity, imagination and heart. How were they able to do this, film after film? And how were they able to grow the company and maintain that fun, passionate environment? I didn't see this happening at other companies, so what was the secret sauce Pixar had that other companies didn't? To learn more about how they fostered creativity, I watched the behind-the-scenes features on their home video DVDs. I listened to talks by Ed Catmull, read interviews with Brad Bird and read books like To Infinity and Beyond! and The Pixar Touch.

So I was excited when I heard Ed Catmull was writing a book! The man who has had the most impact on Pixar's culture, releasing a book on how that culture was built and maintained. The book, Creativity, Inc., is full of inside stories of the studio and how their films are developed. I really believe there will never be another book like this.

So I am trying to not rush through the book.  It helps that I'm not that fast of a reader! But I really want to take my time and appreciate its content as much as possible. As I'm reading it, I continue to be struck by how many of the techniques Pixar uses to develop films have corollaries to my industry of software development, especially Agile development methodologies like XP and Scrum. I realized this well before this book came out and have thought it would be good to write about these corollaries. Now that I'm reading the book, I think I'm ready to write these posts. I expect most of them to be short, using specific quotes from the book to focus on 1 subject or technique. I already have 3 posts in varying degrees of completeness, plus another half dozen or so quotes picked out for future posts, and am only on chapter 5! I should have my first post out within the next day or two.

UPDATE: My first post is now available. You can also use this link to see all posts in this series.

This Day in Pixar History: Pixar's 1997 Annual Report

© Pixar

I've written a number of articles on Pixar's financial reports when it was a public company. Today, let's turn time back about 26 years to look at their 1997 annual report, which came out in early 1998. As with many of Pixar's early annual reports, it started with an informative and entertaining letter from CEO Steve Jobs. They also often included a nice little gift - the 1997 annual report came with a VHS copy of the Academy Award winning short film Geri's Game. So I will try and minimize the financial information (much of it has already been covered in the individual quarterly posts I've done) and focus on the other content.

In their 1996 annual report, Pixar warned of a significant decrease in revenue for the upcoming year due to a decrease in revenue from Toy Story. This was logical; Toy Story was released more than 2 years earlier in late 1995, and Pixar's revenue in 1996 were over $35 million, primarily from the film. But in usual Pixar fashion, they were conservative in their guidance for 1997 and the company's revenue were almost as much as in 1996, $34.7 million. In fact, film revenue was higher in 1997 than 1996: $26.9 million versus $18.8 million, or an increase of 43%!! Not too bad given the expectations of a significant decline! The increase in film revenue was due to the Feature Film agreement between Pixar and Disney: as Toy Story revenue began to accrue, Disney was allowed to capture the majority of it to offset their marketing and distribution costs. As Disney's outstanding costs declined, Pixar received a larger percentage of the revenue. By the middle of 1997 Disney had recovered all their costs, allowing Pixar to capture a proportionally higher percentage of the Toy Story home video and merchandise sales.

© Pixar

You might then ask why were overall revenues down in 1997? This was due to 2 areas. First was in patent revenues - in 1996 Pixar received patent revenue of over $9 million from Silicon Graphics (SGI), which dropped to only $1.7 million in 1997. The second area of decreased revenue was in animation services, such as television commercials. Pixar decided to get out of doing animation services for external customers in 1996 to focus on its feature films which caused this revenue drop.

Pixar's gross margins continued to increase, which is amazing since they already were quite high. Overall gross margins increased from 86.6% in 1996 to 92.7% in 1997. Much of the increase came from Pixar getting out of animation services, which had the lowest gross margins of all their segments. Patent licensing revenue had no associated costs and the software segment (which derived revenues from sales of their RenderMan application) had very low costs (1.8% in 1997 versus 3.4% in 1996). Cost of film revenue also dropped to 5.5% versus 8.2%, mostly due to Disney recovering all their costs in mid-1997 which allowed Pixar to receive a proportionally higher amount of the revenue.

Overall, 1997 turned out to be a better year financially than 1996, except for the bottom line. Pixar ended up paying quite a higher amount of taxes ($9.9 million) in 1997 than 1996 ($2.0 million), due to the utilization of net operating loss carryforwards during 1996. In the end, Pixar reported net income of $22.2 million ($0.46/share) in 1997 versus $25.3 million ($0.54/share) in 1996. Still, I'd consider those pretty good results given the guidance Pixar gave at the beginning of the year!

Pixar's cash position also improved in 1997, growing from $161 million in 1996 to $176 million, even with the much larger outflow of cash Pixar experienced. Pixar spent $10 million on computers and other property to run the studio and $7.7 million on the new Emeryville studio. In addition, with the new Co-Production agreement Disney and Pixar signed in early 1997, Pixar was responsible for half of all film development costs, which totaled a little over $27 million. These costs were more than offset by the higher revenues and the $15 million Disney invested in Pixar on the signing of the Co-Production agreement.

© Pixar

OK, enough of the financial information. As I mentioned at the beginning of the post, Steve Jobs started the annual report with the shareholder letter, which he wrote after watching the 1997 Academy Awards. Pixar won their third Oscar that year, this time a Best Animated Short Film award for Geri's Game. Jobs congratulated director Jan Pinkava,  producer Karen Dufilho and the entire Geri's Game team. Besides the Oscar, Tom Duff, Eben Ostby and Bill Reeves each won an Academy Scientific and Technical Achievement award for their work on Pixar's Marionette 3-D Animation System. In addition, Tom Porter won a Scientific Academy Award for his work on digital painting. The addition of these awards brought Pixar's total count of Academy Science awards to 18.

A few other pieces of information Jobs shared:

• Hiring 97 employees during 1997 for a total of 391.
• Expecting to break ground on the new Emeryville studio that summer with a move-in date of early 2000.
• Investing over $8 million annually on research.
• Growing the size of their RenderFarm to 1000 Sun processors and having storage capacity of over 5 terabytes.
• Highlighting that Toy Story 2 had been upgraded to a full theatrical release, and that their still secret 4th film (Monsters, Inc.) was in development and was hoped to go into production by the end of the year.

Jobs was very clear on his goal for Pixar - to make it the second greatest feature animation studio in the world, only behind Disney Animation. As part of reaching this goal, Jobs stated they were trying to release one animated film per year for the next 3 years (A Bug's Life in 1998, Toy Story 2 in 1999 and Monsters, Inc. in 2000). But Monsters, Inc. would end up not being released until 2001 and Pixar did not accomplish the goal of 3 films in 3 years until 10 years later.

© Pixar

Jobs went into great detail on the making of A Bug's Life.  He highlighted how there were over a dozen major characters, and that each one was more complex than any characters in Toy Story. He talked about how a new subdivision surface technology developed for Geri's Game was used to bring more subtlety and lifelike expressions to the characters. He also explained how they were using simulation software for moving crowds of hundreds of ants or creating lifelike movement in blades of grass (Simulation would reach a new level of complexity and use in Monsters, Inc.). Finally, he discussed how the lighting team was challenged to create more sophisticated lighting to support the outdoors environment the film takes place in. He said the results were "breathtaking". Jobs was also excited about the wide-screen nature of the film, stating that it would look "epic".  A Bug's Life was the first film entirely transferred to film via lasers, and Pixar had to develop their own laser film recorder to perform the transfer. And with all the complexity and larger cast of characters, Jobs stated they were using 10 times more processing power to create the film as they did on Toy Story just three years earlier.

© Pixar

Jobs also explained how it was decided to upgrade Toy Story 2 to a full theatrical release. They originally felt that, with most of their key people from the original Toy Story working on A Bug's Life, they would not be able to find and recruit enough talent to meet the higher standard demanded of a theatrical release. But since the success of Toy Story, Jobs stated Pixar had "become one of the hottest places to work in our industry," and had pulled together a team capable of delivering the necessary quality, at that time being led by Ash Brannon and Colin Brady. The decision to expand Toy Story 2 to a full theatrical release occurred after a November, 1997 meeting in which teams from Disney and Pixar watched the completed story reels and felt the story was strong enough to receive a full theatrical treatment.

Readers of this blog are probably familiar with the story that, less than a year after Jobs wrote this letter, Pixar would realize the story wasn't as strong as originally thought, and in late 1998 production was stopped and the story underwent a major overhaul, with John Lasseter, Lee Unkrich and others coming on board to make sure the film was delivered on time.

© Pixar

To finish this post, many of you have probably seen the image above of Ed Catmull, Steve Jobs and John Lasseter. Interestingly, the image's origination was in this annual report but in a slightly different fashion. You can see the original image below, which is of Pixar's executive team at the end of 1997, including CFO Lawrence Levy and Vice President of Production Sarah McArthur.

© Pixar

© Pixar

This Day In Pixar History: Tia Kratter Quits (and Quits Again)!

It's April Fools Day, so what better way to celebrate it than with a story of April Fools pranks! Tia Kratter is a long time Pixar employee, starting her career as a digital painter on Toy Story in 1993.  Since then she has been the shading art director on a number of films including A Bug's Life, Monsters, Inc., Cars and most recently Brave.

Growing up in a family of practical jokers, she decided to play an April Fools joke on her co-workers. So on April 1, 1998 she sent a company-wide email saying she was quitting. She fooled quite a few people (she even got a "Goodbye Tia" cake!), and has quit every year since (at least through 2009). She even fooled CEO Steve Jobs on 3 separate occasions! You can tell Tia is a true Pixarian, as her reasons for quitting are always original, such as becoming the art director for Tron 2, going to work in a wax museum and (my favorite) air brushing monster trucks!

Not to be outdone, her co-workers did get her back one year. I won't say how - if you want to find out go to the Disney/Pixar official YouTube channel and watch the "Tia Quits" Studio Story, which is narrated by Tia Kratter and Pixar producer Jonas Rivera.

I would love to hear if Tia has kept this tradition going and if so, what new ways she has come up with for quitting!