Technicolor

Technicolor

CineGlass | enterprise end-to-end asset driven workflow management system for content creators and distributors in the film production and entertainment industry.

Business Problem

Having visibility across the entire feature film production workflow, from script through distribution, had never been attempted but was badly needed to allow for effective forecasting, resource allocation, and competitively scheduling releases. The purpose of the CineGlass system was to provide that transparency, to minimize waste and redundancy, maximize efficiencies, reduce operational costs, and improve workflow planning.

Constraints

CineGlass operated as a small 5-8 person start-up within a 40,000+ persons global corporation. During production, Technicolor’s entire film services division shut down, signaling tremendous changes for the industry and sparking sweeping uncertainty. Our product would touch almost every division throughout Technicolor. Success required ardent executive support and company-wide commitment to integrate into decades old embedded, and mostly analog, practices.

User Experience
Strategic Services

UX Strategy
Design Leadership
Systems Analysis
Ethnographic Research
State Flow
Usability Testing

Discovery

Working closely with the internal team of film production veterans, I gathered a detailed understanding of the film production workflow, the project’s underpinning technology and intended users.

Details

Local Production Staff Interviews

To learn firsthand about Technicolor’s departmental organization and general production workflow, I interviewed local post production staff. I became familiar with existing tools; but more importnatly I learned about the ingrained protocols and both the desire for, but resistance to change.

Provisional Personas

Provisional personas, based on subject matter experts, anticipated stakeholders’ behaviors, needs and goals. These were later validated and revised according to findings from ethnographic fieldwork.

White Board Sketches

Design collaboration white boarding with the core team in order to iterate quickly on potential interface requirements and further define technical data structures.

Wireframe/Sketch of the Pipeline

Preliminary wires explore possible pipeline visualizations according to developing technology. In turn, the resulting design explorations informs the further development of the technology.

Notifications System

By rapidly documenting high level discussions from the team’s explorations of detailed tracking of the assets, their states and smart systems warnings; the team could understand the required architecture to support the system even before coding.

Rapid Collaborative Ideation

Early sprints were intimate with the core team: myself, the system engineer (my supervisor) and the lead developer. We had daily, sometimes hourly conversations on the 2 large white board walls that completed my windowed corner office--otherwise known as the war room. The room was dynamic, alive and always open when someone had a new idea to test out or get feedback on. It was about rapid iteration of ideas, sketches (lots of them), wireframes, code and collaboration.

Strategy

UX Strategy starts with establishing alignment around the project goals and measurable success criteria. I took on the task of leading the develop of the primary strategy with the executive team and production teams.

Details

Team Mission and Values

A series of questions were provided that required deep thinking into what we were doing, why, for whom, questions about the industry, our competition, etc. We worked with sticky notes in the war room as a team to find answers and assimilate them into a set of core values, and a mission and vision statement.

Business Case Hypothesis

Lean UX starts with an idea, an assumption, a business hypothesis and then methodically iterates to either validate the idea or ‘pivot’ based on the collected user data.

Initial Interface Design

Using the business knowledge described in this section, and the assumptions we felt comfortable making about the user’s needs and goals, and the foundational understanding of the underlying technology, it was time to take the first stab(s) at the interface

Design Leadership

By leading collaborative design exercises that included the executive team, not only did we achieve a shared vision, but I was able to elevate the influence of user experience design. The eXperience CineGlass newsletter that I started also helped to educate and inform about the value of design thinking and user experience.

For the production team, inclusion in the early stages of design thinking with the development of a set of unified core values and, mission and vision statements gave teammates a deeper sense of connection and ownership of the design. All members of the team, regardless of roles, were included.

Research

With personas identified, the hypothesis assembled, and the underlying architecture built, it was time to “get out of the building” and into the field to assess the user’s actual needs in relation to the product hypothesis.

Details

Data Architecture Validation

The architecture needed validating too. With real data, it became obvious that our initial thoughts about the pipeline as the primary view required rethinking.

User’s Mental Model

Our assumption that the workflow revolved around receiving and sending assets was spot on, but we learned a lot about how the users actually wanted to see the data. What we learned is that the user’s mental model expected to see something more like the arrival and departure boards at an airport.

Color Study Across Exisiting Products

It was interesting to document the color uses across the systems used in the various departments. The color was all over the place, lacked standards and there was no palette that would provide uniform transition for all departments. It also emphasized the fragmentation of the existing systems and the need need for unification.

Christian’s Dashboard

We invited Christian, the producer, onto our team as an integral contrib- utor. With a good understanding of what we were intending, Christian gave us his idea about what he thought the dashboard would look like. Caution when allowing the client to design!

Contextual Studies

Using ethnographic research, user interviews and task analysis I was able to form a deep understanding of the trailer localization production and distribution process along with the needs, behavior and goals of the people making it happen. This intimate knowledge allowed me to establish empathy with our users and ultimately led to my insights into the eventual dashboard solution.

Design

The pipeline view turned out to be too complicated to function as the main dashboard as we had anticipated. Although the pipeline remained a sound concept for the data architecture, we also learned that our user’s mental model was more akin to the arrival and departures boards at the airport. We also discovered that our personas could be simplified to focus on only three uses cases. We had what we needed to tackle the user interface and iterate using real data and ongoing feedback from our user.

Details

Final Personas

Our research validated a lot of our assumptions, and provided some insight as well. We learned that although the details of each technician’s job varied, their needs in regards to workflow tracking were the same. We simplified and focused on three main personas: producer, technician and executive.

Interaction Flow Diagram

The interaction flow diagram studies the relationships between all stakeholders and the systems they share. Form a systems point of view, we needed to identify where CineGlass should be implemented and how best to optimize the work flow.

Pipeline Workflow Diagram

The dependecies between each activity needed to be clearly identified. Using color, distinctions were made between the original version (OV), dubbed and subtitled version workflows. Primary activities that display in the dashboard are in red.

Producer’s View

The early concept of the producers dashboard included, from top to bottom, “Christian’s View” (from the work session with our client, Christian), a Filtered Pipeline View and an area for the Arrivals and Departures View.

Technician’s (Vendor’s) View

The Vendor’s View would display the status of the assets required to start and activity (inputs) and the assets it would deliver (outputs) for a given activity. This wire is little more than a rough concept, we knew very little of how this work at this point.

Executive View

PWe knew there needed to be a high level view for the executive making decisions about scheduling in the theatres, resource allocation and budgeting.

Research Analysis

Based on the research findings a set of NEEDS is created by the researchers and adjustments are made to the personas as needed.  Collaboratively, with the entire team, the needs are married with the business GOALS to determine a set of USE CASES for the intended product.   determine a set of USE CASES for the intended product.  

Iteration

We ideated, built and tested rapidly so that our ideas were being tested early as the product was being built. User research and testing was happening in a parallel track to development. The constant feedback loop helped us to better understand the user, while building a better product itself, ensuring a well defined and well tested product/market fit.

Details

System State Diagram

Our goal was always to acheive “at-a-glance” visibility. The incography for the states needed to be simple, bold an informative. State design was also integral to the design of the other dashboards.

Deliverables Dashboard

The producer had provided his idea for what he thought he wanted for the deliverables dashboard. Even after many iterations, in the end this did not prove to be the most effective solution.

Arrivals and Departures

The importance of this view was the primary nugget we returned with from our initial research. The Arrivals and Departures Dashboard drove the iconography of the state more than any.

Production Pipeline

What started as the key visual, ended up being the least important. Although still useful, considerable care went into how it would be filtered to be readable.

In-depth Explorations

In between the major milestones displayed above, there were pages and pages of alternatives explored. Here are a few samplings that explored the various relationships between the dashboards, the pipeline, notifcations, and the state.

Lean UX

Depending on what stage of development the product is in, a DESIGN TEST is devised.  It could be as simple as soliciting feedback on design comps or even sketches, a wireframe click through or prototype, to full fledged usability testing within a product release cycle.  The point is that whatever stage of development the product is in, iteratively test the hypothesis as quickly as possible and don’t get invested in anything, design, code or concept until is has been tested.  Be prepared to adjust the hypothesis, or ‘pivot’, at any point based on user feedback.  In Lean UX, the user is always king.

Deployment

Having been working closely with the client, we had many elements of the product wrapped up pretty tight, except for one thing... The client agreed with the direction we had taken, had even led us, but the main dashboard wasn’t quite aligning with their expectations. After several collaborative sessions together, trying myriad variations, it became clear to me that the linearity of the dashboard didn’t match the client’s mental model. We needed to take a completely fresh approach.

Details

Breakdown, Reorganize, and Regroup

Stepping back with only excel spreadsheets as a tool, I listed out every activity, regrouped, and re-organized based on my understanding of the key needs of the client and the manner in which they were accustomed to thinking about the production.

Group by the 3 Key Activities: Source, Translations, DCP

There were two primary concerns for each title production. 1) Knowning when three key activities were complete: Receviing the source files, Translations, and mastering of the DCP...

Group by the Formats: Broadcast and DCP

...and 2) being able to track these key activities for the Broadcast version and the Digital Cinema Package.

Stacking for At-a-Glance Visibility

By pattern matching, creative re-organization and grouoing, I was able to create a nearly one-to-one relationship between the production of the Broadcast and the DCP. By stacking the two workflows, at-a-glance, each could be monitored simultaneously.

Structured Pipeline

The new groupings also made the pipeline view easier to manage. It also became clear at-a-glance what assets where shared and where the ciritcal dependencies existed.

Retrospective

There is no such thing as a captive audience and you can’t expect the user to be forgiving. When striking the balance between production demands and the needs of the user, always err on the side of the user.

The ultimate fate of a product, project, or creative decision may not be in the control of the designer. My job is to diligently know and empathize with the user, to faithfully integrate the business objectives, and to provide insightful, informed and seasoned solutions. I have also learned to be persuasive by utilizing responsive communication skills and backing up my design choices with sound conclusions. I make my arguments and then everything else I let go. Regardless of the fate of any project, at the end of the day, I know when I have elegantly created a solid design and that needs to be my sole consolation.