9-615-023 R E V : M A Y 1 8 , 2 0 1 5
Professor Amy C. Edmondson, Senior Fellow David L. Ager, independent researcher Emily Harburg, and Research Associate Natalie Bartlett prepared this case. It was reviewed and approved before publication by a company designate. Funding for the development of this case was provided by Harvard Business School and not by the company. HBS cases are developed solely as the basis for class discussion. Cases are not intended to serve as endorsements, sources of primary data, or illustrations of effective or ineffective management. Copyright © 2014, 2015 President and Fellows of Harvard College. To order copies or request permission to reproduce materials, call 1-800-545- 7685, write Harvard Business School Publishing, Boston, MA 02163, or go to www.hbsp.harvard.edu. This publication may not be digitized, photocopied, or otherwise reproduced, posted, or transmitted, without the permission of Harvard Business School.
A M Y C . E D M O N D S O N
D A V I D L . A G E R
E M I L Y H A R B U R G
N A T A L I E B A R T L E T T
Teaming at Disney Animation
Disney Animation produces entertaining films enjoyed worldwide and also happens to be a first-class Studio for technological advancement. Producing animation is a fantastically complicated blend of art and science and it requires many coordinated experiments to create. We view all of the 900 Artists, Engineers, and Scientists in our studio as filmmakers and each has a creative part to play in the process.
— Andy Hendrickson, Chief Technology Officer, Disney Animation
Jonathan Geibel, Director of Systems at Walt Disney Animation Studios (hereafter referred to as Disney Animation), walked through the workspace occupied by the group he had been tasked to lead. Geibel knew he was part of a creative and magical environment. The Disney studio had created more than 53 feature animated films in over three-quarters of a century—beginning with Snow White and the Seven Dwarfs in 1937 through to Frozen, released in November 2013 and awarded the Oscar for Best Animated Feature in March 2014, the first Academy Award in that category for Walt Disney Animation Studios. In late March 2014, Frozen became the highest-grossing animated feature, worldwide, of all time. There was a period in the history of the 90-year-old studio, not so many years ago (and prior to John Lasseter and Ed Catmull’s leadership), when Disney Animation had become more structured and hierarchical, and it wasn’t always easy to work across departments to innovate. Yet the work, which involved both high-tech computer animation and creative storytelling, was more cross-disciplinary and dynamic than ever. Geibel wondered what he and Ron Johnson, whom he hired and teamed up with to re-envision the Systems group within Disney Animation, could do to improve the flow and the efficiency of the organization’s increasingly technical and creative work. Geibel and Johnson had already made dramatic changes in the work structure and in the physical space to promote the effective teamwork that was so essential to producing compelling, engaging, animated films. Now it was time to figure out how well the changes were working and what further changes, if any, were necessary.
This document is authorized for use only by DACHUAN LIN (493954953@QQ.COM). Copying or posting is an infringement of copyright. Please contact customerservice@harvardbusiness.org or 800-988-0886 for additional copies.
615-023 Teaming at Disney Animation
2
Background
The magic of Disney Animation had enchanted audiences for decades. Roy and Walt Disney founded the company in 1923 in Burbank, California, creating Walt Disney Animation Studios (known then as “The Disney Brothers Cartoon Studio”). Initially, the company produced animated short films and, in 1937, released the first full-length animated Hollywood feature film, Snow White and the Seven Dwarfs. With Walt Disney’s legendary talents as a storyteller and innovator, Disney Animation continued to revolutionize feature film, from Fantasia, Pinocchio, and Dumbo to Cinderella, Sleeping Beauty, and The Jungle Book. In the late 1980s, a new generation of filmmakers, who had been inspired by the studios’ earlier films, created such classics as The Little Mermaid, Aladdin, Beauty & the Beast, and The Lion King in a period often considered a renaissance for the studio. (Exhibit 1 shows Walt Disney participating in an early storyboarding meeting.)
When Disney acquired Pixar in 2006, CEO Bob Iger asked Pixar President Ed Catmull and Pixar Chief Creative Officer John Lasseter to also take on leadership of Disney Animation. In 2007, as part of their changes to their studio, Andy Hendrickson was hired as chief technology officer (CTO). Under the guidance of Catmull and Lasseter and Disney Animation General Manager Andrew Millstein, Hendrickson was tasked to transform the science of animation at Disney. The first step was to find the right people to innovate a new culture.
In 2010, Hendrickson recruited Jonathan Geibel to join the team as the Director of Systems at Disney Animation. Geibel had worked at Disney for 10 years, and had extensive experience on technical teams as a Senior Systems Engineer and a Production Engineering Lead. Geibel expressed interest in testing out new ways of working in his group, inspired by the vision set by the new leaders. As he recalled, “[Catmull and Lasseter] came in and set up a new culture where leadership was trusted; people felt trusted and empowered. They also promoted ‘failing quickly’—the idea that there are both successes and failures when striving to innovate and the best thing is to try and have your mistakes and learn from them early in the process. This new leadership sparked changes like our new teaming structure.” Geibel set out to restructure his team by flattening the hierarchy. He envisioned a dynamic organization with the agility of a startup company, breaking down traditional silos, empowering engineers, and supporting experimentation.
Later that year, Geibel promoted Ron Johnson, who had worked at Disney Animation since 2007 as a Senior Systems Engineer, to Manager of Systems. Geibel and Johnson had spoken frequently about management, and they shared similar beliefs about the operation and leadership of technical teams. As Geibel explained, “We had both gone between being engineers and being managers of engineers several times throughout our careers, and that back and forth allowed for a lot of learning and perspective.”
Walt Disney Animation Studios
Animation—the technique of creating the illusion of movement from static figures—had a long history. One of the first entirely animated films, Humorous Phases of Funny Faces, appeared in 1906,
directed by James Blackton, who was considered the father of American animation.1 Since those early days, animation technique had been dramatically transformed, with Disney Animation being one of the leading innovators of this transformative process.
An animated feature at Disney began with stories that would resonate with audiences. Visual breakdowns of story sequences were created. These “storyboards” illustrated the visual message, emotion, and approximate timing of a scene in a form that was roughly described as a “comic book”
This document is authorized for use only by DACHUAN LIN (493954953@QQ.COM). Copying or posting is an infringement of copyright. Please contact customerservice@harvardbusiness.org or 800-988-0886 for additional copies.
Teaming at Disney Animation 615-023
3
depiction of the film. The sequences were pitched, debated, and refined until the story was clear and the next phase of image production began.
In hand-drawn animation, frames were drawn to depict the characters, props, and backgrounds of a scene. Artists called animators created the look of substance and form, the appearance of cinematic lighting, and the perspective of camera. Animators created slight variations of those drawings, which were different from still-life paintings, to represent visual snapshots of a scene over time. Using the multiplane camera, which was a Disney innovation, the drawings were then recorded in successive frames of film. After developing, the film was projected onto a screen at the rate of 24 frames per second, giving the illusion of movement and bringing the characters to life.
In the hand-drawn style of animation, there were many stages of the process as the drawings were refined. Rough animation used pencils and tracing paper to block out the character movement and timing. Animators often flipped between multiple drawings to gauge motion. In the early years of hand-drawn animation, “in-betweening” and cleanup animation came next, where the rough pencil sketches of major character poses were “filled in” with additional drawings to smooth transitions between poses, and the rough pencil sketches were finalized with clean ink lines. The final ink lines were painted onto transparent plastic “cels” that allowed for the subsequent step of painting in color. Pinocchio, an example of this process, was created by a team that consisted of over 750 individuals, each with a unique contribution.2
In the late 1980s, Disney Animation began to utilize computing to assist in the creation of animation using the Computer Animation Production System (CAPS). CAPS allowed the hand- drawn animation to be created as individual “pieces,” which were managed by the computer as
unique layers.3 Those pieces would be assembled, or composited, into a single image using the repeatability and precision of a computer. CAPS brought Disney to the forefront of hand-drawn technique, advancing from the multiplane camera with a digital computer equivalent, for unprecedented levels of complexity depicted in the films. The Rescuers Down Under, released in 1990, was the first animated feature to be made with CAPS.
In 2014, when Disney Animation created a film using computer animation, production started with artists creating models built using computer-Aided Design (CAD) systems that were cousins to similar systems used in architecture and engineering. Those models were then applied to a mathematical model skeleton. Animators breathed performance into character by animating the model to move the animated figure in a scene.
Computer simulation had become omnipresent in animated features, creating such character items as hair and cloth. For example, one of the most challenging aspects encountered in the making of the film Tangled was ensuring that Rapunzel’s hair appeared natural and real. This required much back-and-forth iteration between artists and technical engineers in order to develop tools that could accurately represent the flow and movement of hair.
The current frontier used simulation to create cinematic lighting. The physics of photon transport and responses to light energy by materials were captured in a mathematical model that Disney used to represent the world of an animated feature. Artists used digital virtual equivalents of real stage lights and theatrical techniques that mimicked the real world to create cinematic lighting. Additionally, cameras were represented in digital scenes with realistically simulated parameters. Every nuance was computer simulated, including lens distortion, chromatic aberration, and depth of field. All of this simulation required an extraordinary amount of computing capability.
This document is authorized for use only by DACHUAN LIN (493954953@QQ.COM). Copying or posting is an infringement of copyright. Please contact customerservice@harvardbusiness.org or 800-988-0886 for additional copies.
615-023 Teaming at Disney Animation
4
By 2013, Disney Animation contained a 30,000-core computer cluster for simulation and lighting, 6 petabytes of data storage, a 1.5 megawatt data center, and over 1,800 Macintosh and Linux workstations. Driving all of this capability were 900 highly skilled artists, engineers, and scientists telling stories through words, music, and images with the most technologically advanced tools in the world.
Disney Animation Systems Group Structure (prior to 2010)
After Disney’s acquisition of Pixar in 2006, Catmull and Lasseter, along with Millstein, assumed responsibility to lead Disney Animation. By 2014, Disney Animation comprised several departments. The Technology department was headed by Hendrickson and included both Software and Systems departments.
Prior to 2010, the 50-person Systems department was organized in traditional units, each consisting of 15–20 technical staff. (Exhibit 2 illustrates a traditional team structure in the Systems group.) Managers coordinated projects from their upstairs offices. Meetings were held one-on-one with group members, and every few weeks the entire group met in a formal conference room. Final technical decisions were often made by managers and implemented by group members. Individuals were hired for a specific role or promoted from within. Top-down reorganizations were occasionally made to accommodate changing technologies. At the time of these shifts, individuals were often reassigned into new groups, reporting to a new manager.
Once individuals were assigned to groups, task boundaries were clearly drawn. Geibel reflected:
In the past, we rarely saw resources (budget or people) flow between silos, even when it was obvious that moving resources to other areas was in the best interest of the studio. The studio environment is highly dynamic and changes rapidly, and our organizational structure was getting in the way of our ability to react to the changes. After observing the situation for a few years, it became obvious why this was happening. Managers were being judged by how effective their silos were performing. If a manager gave away resources to another silo, they would have fewer resources to accomplish their goals. This caused a mind-set for managers to hoard and protect these resources throughout the year. The competition for resources became extremely apparent during budget season when managers would fight with one another to obtain the most resources they could for the goals they set out for the upcoming years.
It wasn’t just managers who felt the boundaries between groups. Jim Bette, a systems engineer, explained, “It was very siloed, as in ‘this is our space, that is yours.’” Manager Matt Schnittker added, “It felt like giving was a bad thing. You wanted to protect your domain and hold onto your resources.”
Changes in Structure and Workspace
Geibel and Johnson began the transformation of the Systems group by seeking ideas from various sources to guide their approach. For example, they were influenced by the work dynamic typical in startup companies and management concepts such as the Matrix Model. Wary of large, sweeping reorganizations that changed too many variables at the same time, they created their own approach to transform the group. Geibel outlined the change process that Johnson and he finally adopted:
This document is authorized for use only by DACHUAN LIN (493954953@QQ.COM). Copying or posting is an infringement of copyright. Please contact customerservice@harvardbusiness.org or 800-988-0886 for additional copies.
Teaming at Disney Animation 615-023
5
Our approach to change is based on principles of the scientific method. Organizational changes are done methodically, with the intent to evolve over time. We study and discuss current problems that exist in the organization. We make one change at a time so that we know what worked, what did not work, with learning being at the center. We observe how it affects the system, we gauge its impact, we learn, we generate another theory, and debate the results. If something works, we do more of it. If something does not work, we learn from it and try something different. Everyone is involved in the process, not just management. Our technical staff members are contributing just as many ideas as the management team. We try to ensure critical thinking is happening at all levels. If you think a process could be made better, we encourage people to speak up, offer new ideas, run experiments, and share the results. The same art of experimentation and problem solving that happens on technical problems is now being applied to evolving the way we organize ourselves and build teams.
One of the first theories they tested was the use of small autonomous teams to break the formation of previously established groups, with each team focused on a particular specialty. The teams would be responsible for specific tasks or problems that arose in the studio, for example, networking and technical support. There were many questions related to size, structure, and composition. Geibel, together with Johnson, engaged in a series of experiments. For example, Geibel explained:
We set up a weekly hour-long leads meeting, which we intentionally made optional so each week we’d have a different number of people in the room. Some of the most interesting moments happened during meetings where people would stagger in at different times. We noticed in the jump from six to seven participants the quality of the discussions dropped dramatically and often people would go silent. We saw that happen over and over and came up with the idea not to have more than six people in the room when we were having a working meeting.
Geibel concluded that their organization should consist of small, autonomous teams of two to six people, and began by selectively assigning employees to various “specialty” teams.
Shortly thereafter, Geibel and Johnson observed that employees were confused about who was coordinating the work and how to allocate their time. This led them to define clear roles for team members. Each team would consist of a lead and several primary members, and occasionally would also comprise a few secondary members. (Exhibit 3 presents a sample team and role assignment grid.)
Team Leads
The teams would be headed by a team lead, someone who had been identified as a technical subject-matter expert and who had demonstrated a strong vision for the area. Geibel created a two- page lead doctrine, shown in Exhibit 4, outlining expectations for those in a lead role. He explained:
We’re not going to be rewarding them for sitting in an office and dictating orders. The lead is not administrative in nature. They’re doing the work alongside of everyone else and leading by example. Hopefully 80% of their job is doing the work and 20% is establishing a vision and evangelizing their ideas to key stakeholders across the studio. The goals of the leads were to be transparent with their ideas and to be open to feedback from their peers and management. It is not mandatory to execute on any or every piece of feedback, regardless of the source; it’s up to the lead and the team to decide which ideas are best to move forward on.
This document is authorized for use only by DACHUAN LIN (493954953@QQ.COM). Copying or posting is an infringement of copyright. Please contact customerservice@harvardbusiness.org or 800-988-0886 for additional copies.
615-023 Teaming at Disney Animation
6
Bert Faerstain, a member of the Media Engineering Team and team lead of the Displays Working Group, explained:
I head up the displays working group for the studio. I have substantial background and interest in this area and many of my coworkers already knew that this was my hobby for years. I research this world in my spare time outside of work, so it’s my passion and they knew this. My team and I are tasked with continually advancing our display standards for the studio, while weighing options to balance our requirements, the direction technology is headed, and overall value.
Catmull had inspired the idea of having team leads through his “director-driven studio” model, where an individual was in charge of the story, pitching to peers for voluntary feedback that their teams would sort through to look for new ideas to help propel their films forward. Storytelling was in the hands of the filmmakers themselves.
Whenever new technologies would emerge that required the formation of a new team, Geibel or Johnson would approach the broader group with the opportunity to lead the new team. They would then look for the person who had the most compelling vision and whom they felt could effectively drive a team forward, regardless of his or her current title. Titles played very little role in deciding team leads. Geibel and Johnson wanted to generate the best ideas and to execute on them.
The creation of the team lead position offered technical individuals the opportunity to move up within the company. Technical team leads enjoyed the same status and compensation as the traditional “personnel” managers. Engineer Jim Bette described how this change affected his career growth motivation:
Now I feel like you don’t have to be in management to move up—you’re all more on the same playing field; it’s more based on what you bring to the table. You don’t have to be a manager; you can be a technical lead, and technical leads are now considered the top of the organization. This has freed up a lot for the managers—managers can maintain people’s career growth, handle reviews, and make sure people have what they need.
Primary and Secondary Team Members
Employees could also be assigned as primary or secondary members to specific teams. Every employee played a primary role on one team and had the option of playing secondary roles for any number of teams. Employees were expected to focus most of their time and effort on their primary team and help out, when needed, on those teams to which they had been assigned as secondary members. As Johnson explained:
Secondary players helped us to tap into peoples’ other skills and passions across the organization. We wanted to enable everyone to utilize all of their skills, whether it was a specific technical skill or a soft skill. Sometimes we’d add a secondary player to a team simply because they asked good questions and got the team thinking differently. Other times people were brought onto teams because of their attention to detail or project management abilities. Whatever the various teams needed, we could often find someone with a specific skill somewhere in the organization to augment the teams with. People seem to want to have secondary roles and to help other teams, which is great, but it is always expected that your primary team is where your first responsibility lies.
This document is authorized for use only by DACHUAN LIN (493954953@QQ.COM). Copying or posting is an infringement of copyright. Please contact customerservice@harvardbusiness.org or 800-988-0886 for additional copies.
Teaming at Disney Animation 615-023
7
In the previous structure, team members worked only in the areas supervised by their managers. As Geibel explained:
We had network engineers who had a passion for software development. We had Windows systems administrators who also happened to have skills in database administration. We had a systems engineer who also had a passion for mobile software development. In the old organization we would have had to pick one of these skills and now we can leverage all of them. It’s very motivating for our staff to know that they can contribute to and be a part of so many different areas across the organization.
To avoid problems found in matrix organizations, such as employees feeling conflicted about demands among competing managers, the use of primary and secondary designations established expectations that held employees accountable first to their primary team. Team leads understood these expectations, and never battled over the time of an engineer who might be playing a secondary role on the team but was primary on another team. Team members liked the opportunity to manage their own time across multiple teams and pick areas they were passionate about.
Geibel and Johnson spent considerable time on team membership issues, but this did not guarantee performance. Geibel explained, “One year, a new team member joined a very high- performing team, where everyone knew each other very well. By adding this team member, the work dynamics on the team changed drastically. I provided a great deal of feedback to the individual, and made sure we allowed enough time to try and make the personalities gel, but we eventually had to pull them off the team.” Johnson reflected on the decision to pull the person off the team: “You have to be very creative in these moments not to punish that person. It’s all about finding the right fit. With new people coming in, we move them around and find the most effective place for them.”
Other Structural Changes
Geibel and Johnson applied the same teaming structure to their own management team, creating a small six-person management group made up of a diverse set of managers who were passionate about engineering and business operations and who trusted and respected each other. Because the team leads were responsible for creating a vision and running day-to-day operations for their areas, the managers were able to focus on career and team coaching and technology investment strategies.
Managers were expected to lead by example and spend at least 20% of their time being hands-on. Geibel and Johnson found that the best way to earn the respect of employees was by working alongside them. This also helped ensure a connection to what was happening “on the floor.” The management team also acted like a venture capital firm, deciding which technologies merited investment of both dollars and staff time, and setting expectations for each team, so as to align team goals with the broader direction of the studio and the vision of the CTO.