Francis Phan

Product Designer
























Work
About
Email
Linkedin

Showtime CMS Asset Manager 

Rebuilding the B2B Operations Platform Behind 350+ Live Entertainment Venues

B2B SaaS @ AXS (Carbonhouse Team)
Timeline: 5 Months (Q3 2025 - Q4 2025)
Platforms: Desktop
Role: End-to-End Product Designer
Team: 4 Engineers, 1 PM, and Stakeholders
Tools: Figma, Figma Make, Notion, Microsoft Clarity, Microsoft Forms, Claude






Problem: Venue staff managing high volumes of event media relied on a hidden tag system and limited file details, so they kept duplicate files and clicked through image after image to find the right one. Newer users struggled most, and modern competitors put retention at risk.

Opportunity: Make media management feel as familiar as the tools staff already use.

Solution: I rebuilt the Asset Manager’s information architecture and tagging model from the ground up, adding folders, filters, a metadata-first list view, and multi-select to support high-volume workflows.

Impact: 27% fewer asset-related support tickets and 41% fewer repeat corrections.



A Foundational Tool in the Live Entertainment Industry

Showtime is an events-driven CMS that powers content operations for more than 350 venues worldwide. First conceptualized in 2005, its structure and functionality stayed largely unchanged even as live entertainment workflows grew more complex. That long history made it reliable, but also hard to adapt.

Modernization began in early 2025 with a full dashboard redesign and a new native design system, giving the team a consistent foundation for building new workflows. The Asset Manager was next.




An Industry Standard With Clear Opportunities

Showtime is used daily by non-designers at venues around the world, but behind this high engagement, friction was slowing users down.

Even with a refreshed look, the platform struggled to support users’ tedious, high-volume tasks.
I led mixed-methods research across 34 venues in four countries, pairing 30 surveys and interviews with heat-mapping and session recordings. Some staff managed up to 100 assets a month, but inconsistent tags and limited file details forced them to keep duplicates and click through image after image to find the right one.

93% experienced friction managing media. 80% create their own media in tools like Adobe and macOS, and expected Showtime to work the same way.

These findings became the foundation of the Showtime roadmap.



User Problem


More experienced users rated the platform higher, while newer and younger users rated it lower, highlighting a reliance on learned behavior rather than intuitive design. This indicates that core workflows don’t align with modern tools users interact with elsewhere, increasing ramp-up time and creating unnecessary friction. 

Business Problem


With newer CMS platforms offering workflows better aligned to venue operations, Showtime faces growing competitive pressure and retention risk. To maintain adoption and meet business goals, the platform must deliver clear value by reducing workflow friction and improving user efficiency.



Vision

An Asset Manager built for non-designers: familiar on day one, fast at 100+ assets a month, and safe enough that cleanup never breaks a live page.





New Asset Manager Experience

The Asset Manager modernizes asset workflows by aligning with today’s file-management standards, empowering users to organize, discover, and deploy assets with confidence.




From Tags to True Organization

Previously, asset organization relied on a hidden, faux folder system built on tags, which often felt opaque and difficult to scale as asset volume grew. To better align with users’ mental models, I led the shift toward fully customizable folders.

The new structure allows users to create and nest folders, apply metadata, and layer additional tags as a secondary organizing system. This approach mirrors familiar workflows in file-management tools while maintaining the flexibility needed for event-driven workflows. This makes it easier for users to find, manage, and reuse assets at scale.
before

after



Metadata-First Design for Faster Asset Discovery

64% of users rely on metadata to find assets, so I made a detailed list view the default, while keeping a grid view for more visual work. I refined which details appeared over rounds of usability testing, and gave alt text its own field instead of defaulting to the file name, so teams could set naming conventions and keep assets organized.

before

after



Streamlining Repetitive Tasks with Multi-Select

Cleaning up assets one at a time was tedious, a common frustration among clients. I introduced multi-select so staff could delete, organize, or download assets in bulk, cutting out repetitive work and freeing them up to focus on their events.



Protecting Live Pages in the Deletion Flow

Faster bulk deletion raised a harder question: what happens when someone deletes an asset that’s still published?

Event pages referenced images directly by URL, so deleting a published asset broke the page. Speed was everyone’s first instinct, so I looked back at site audits we’d run on long-running venue sites. They were already full of broken pages, and faster deletion would only add to them.

I worked through two directions with engineering:

Warn, then allow. A warning flags published assets, but staff can override it. Fast, but easy to click past in a bulk action, leaving broken pages on the venue’s public site.

Soft delete with a safeguard. Every deletion goes to a trash first, a pattern staff already knew from their desktop tools, with 60 days to recover anything removed by mistake. Published assets can’t be permanently deleted until they’re unpublished.

I chose the soft delete, prioritizing reliability over speed. To offset the extra step, the error lists each affected asset and where it’s published, so staff know exactly where to go to fix it.

Client feedback and input from the director of technology confirmed the decisions, and users came to us with fewer questions.



explored

shipped




Supporting High-Volume Workflows with Advanced Filters

While introducing folders and clearer metadata improved overall organization, it became clear that users still needed more precision when managing large volumes of assets. To address this, I designed refined filters and sorting, allowing users to quickly sort, locate, and interact with assets within the Asset Manager.


before

after - filtering options
after - filtered results



Simplifying Collaboration with Partners

Collaborating with industry partners is now easier and more efficient. Asset specs, previously hidden within the asset manager, are now accessible in the sidenav, where users can generate customized spec lists to share with the teams producing their assets. This improvement directly supports the 80% of users who receive a share of their assets from promoters, reducing friction and speeding up asset handoffs.


before
after


AI-Assisted Design, Built on a Stronger System


AI-assisted design became a core part of how I tackled the CMS’s complexity. Rather than building every modal state by hand, I used Claude to write targeted prompts for Figma Make, scaffolded designs quickly, then refined them in Figma: hierarchy, states, and component consistency. The refined designs went back into Figma Make as interactive prototypes, giving stakeholders something real to react to.

But AI is only as good as the system it builds from. As new features pushed the CMS into new territory, inconsistencies in component behavior and color usage were compounding. I updated existing components, built new ones like the image editor, and wrote the missing documentation, so every AI-generated draft started closer to the finish line.

The result was faster sign-off, tighter iterations, and a smoother handoff to development, an investment that paid off on every product and feature after.
Figma Make’s Initial Output
Shipped



Results & Reflection  


27% fewer asset-related support tickets. Replacing hidden tags with real folders and a rebuilt IA meant staff stopped needing help to find and organize their media.

41% fewer repeat corrections. A metadata-first list view, shaped by multiple rounds of usability testing, helped staff pick the right image the first time.

What I’d do differently. I’d measure the business side from the start. Asking venues why they use Showtime, whether they’d consider switching, and whether updates bring in new customers would have let me tie the redesign directly to retention, the risk that started this project.

What carried forward. The design system documentation I built here became the foundation for Showtime Audits, turning a system for one product into one shared across the product family.

What’s next. With assets finally organized, the next step was putting them to work. I’d started designing an AI-assisted page builder that combines the data we aggregate with what venues tell us, guided by UX best practices. Staff assemble pages from ready-made building blocks, so they can reshape their sites much faster.



Want To Learn More? Reach Out For More Information.







All Projects | Showtime CMS | Exponential Pathways | Insomniac