Content quality is significantly impacted due to a lack of features, a rigid workflow, and a tool that is progressing at a snail’s pace
Design strategy
Build a no-code modular tool that can scale as new feature requests arise
Goal
Enable enterprise customers to quickly and easily create quality VR training content for their employees
Context
Strivr is a startup that Creates VR training for Fortune 1000 companies. I was the senior director of design, reporting to the CPO. I managed a design team of 2 on this project, as well as a CG manager, who had 4 ICs on their team
Creator tool
This tool is used by Strivr's internal content creation team, who build immersive training content that is pushed to VR headsets for enterprise learners to experience. It is also used by customers to create their own content that is pushed to their headsets.
1
Discovery: user research
To understand how this tool impacts the business and adds value to users, I talked to our internal users, customers, stakeholders, and team leads to get the information I need. This was critical since the redesign had to be thought of through this lens.
Click on images to enlarge
Tool evaluation
I conducted user interviews, contextual inquiry sessions, and a heuristic analysis to gauge the tool
User Empathy
I had monthly shadow sessions with users as they create content, as well as individual one on ones with the users to understand more about how they are using the tool and the challenges they are facing
Performance Check
Met with content team leads regularly to discuss content quality, completion times and other operational challenges when creating content
Takeaway
Lack of features coupled with a hard to use interface result in below-par training content
Takeaway
Frustrated team that can’t showcase their ability by creating the content they envision
Takeaway
A rigid workflow prevents users from working in parallel, leading to an inefficient pipeline and increased production timeline
Problem Statement
Content quality is significantly impacted due to a lack of features, a rigid workflow, and a tool that is progressing at a snail’s pace
2
Identifying the key pain points of the tool
I grouped my findings in two categories: Operational and design.
Operational
Pain point
Business impact
Rudimentary built-in UI components
Slow development cycle, very low velocity
Unfamiliar framework (WPF)
Engineers not experienced in using framework
Windows only
Narrows the user base significantly
Spaghetti code that was hard to build on
A lot of bugs and feature building limitation
Design
Pain point
Business impact
Rigid workflow
Longer production time, increasing sales cycle
Limited features
Less competitive in the VR training space
Inconsistent interaction/visual design
User frustration and higher error-rate
Video-based only
Very costly (requires actors, location, crew etc)
No in-headset preview of content
Slow / painful iteration cycle
3
Solution: Proposal to the leadership team
By talking to different stakeholders and users, I came to the realization that it is faster and more cost effective to completely reimagine the tool. My proposal was to build it using web technology, and completely redesign the interaction model. Below is how this solution solves the challenges identified above
Operational
Pain point
Solved with
Rudimentary built-in UI components
Available UI libraries ready to go
Unfamiliar framework (WPF)
Engineers experienced in web technology
Windows only
Cross platform
Spaghetti code that was hard to build on
Opportunity to rebuild the right way
Design
Pain point
Solved with
Rigid workflow
Building collaboration tools to enable an efficient pipeline
Limited features
Designing modular features that scale easily
Inconsistent interaction/visual design
Building and utilizing a design system / UI kit
Video-based only
Reducing cost/time by building virtual experiences
No in-headset preview of content
Enables preview mode easily using web APIs
Additional Design Challenge
One of the major hurdles we had from a design perspective is that our users didn’t know all the additional features they needed.
Depending on market trends, VR training could go in a number of directions, so we had to anticipate and prepare for that when designing this new tool
4
Green light: Design leadership for building faster, better
I ran a number of design sprints with the design team, content creation team, product and engineering leads, and various stakeholders to come up with the final design. The design requirements I defined to the team were as follows:
Scale
The tool had to accommodate for unknowns, like the "unknown features" outlined above as it scales
Speed
The design had to enable parallel development cycles (We already had a tool, lacking as it was, so investment into a new one had to be justified cost-wise)
I proposed a node-based approach where the functionality of a feature is completely encapsulated into that feature’s node. This meant that when a new feature is defined, it would all be coded into a node that seamlessly integrates into the tool without adding any UI work. Operationally, each engineer could tackle a node independently, increasing velocity significantly.
This was then vetted and iterated on by prototyping and testing sessions, including building a whole existing experience using this approach in Figma.
5
Leading the implementation approach
To increase velocity, I defined clear milestones for the engineering leads. These were meant to reduce repetitive coding of common behaviors and UI, and to make sure that multiple engineers can work on different tasks independently. The goal was to be more effective and increase velocity significantly. This was also communicated with my designers.
I set up multiple sessions with engineering leads and the product manager to go over my proposed plan, and to align on resources. The proposal was this:
1. Extract common denominators in the UI and build them once to achieve a unified, consistent interaction model.
2. With the modular node system, one engineer owns one feature. This parallel workflow would enable us to build multiple features concurrently
3. UI and functions are now separated operationally, and velocity increases significantly
CG environments and characters
While doing all this, I was also directing the CG team on building the 3D assets that were to be used in this tool, which we now called Creator 2.0. The key there was to build one generic character that can be modified visually so that all the same animations worked without extra work, and for the environments, having different areas for interaction so the same 3D model can be used in multiple modules. But more on that in a separate case study.
Click on image to enlarge
Click on images to enlarge
Usability testing
Aaaah... reaping the rewards of happy, engaged users. Once the Creator tool MVP was complete, we ran usability testing by building a complete training experience and gathered feedback on how to streamline the process even more. Those tasks were added to the backlog for future development.
This design, among other advantages, enabled multiple users to work in parallel on the same document, reducing the dependence on previous steps to be completed before starting other steps. It also enabled collaboration and exploring new ideas while still hitting deadlines.
Results & positive reviews
Wasn't sure you would read this far, I guess I should keep going...
Key results
1. I pushed for a web-based authoring tool to enable multi-platform support and enable a faster development cycle 2. Users loved the ability to mix and match nodes to push their narratives and achieve more engaging experiences "It’s incredibly cool how the playback nodes and playback phases work so that we can intelligently structure playback sequences for branching experiences! "Spatial layout makes branching a breeze and I know exactly where all my nodes are flowing 3. Due to a more intuitive interface, customers were able to build experiences without our help. Customers support calls dropped by 70% ""Creator is the best tool out there for VR content creation. it's a fantastic tool!!"" 4. Production times dropped by more than 50%, due to a visual interface, ability for customers to quickly test out ideas, and collaborate efficiently 5. Considerable drop in production cost, due to reducing reliance on other software and faster development time
