Case Study
Scheduler: Improving Content Variety for Atmosphere TV
Timeline
January 2024, Research
February 2024, Design
March 2024, Beta Testing
April 2024, Shipped
My Role
Lead designer
Lead Researcher
Atmosphere TV delivers streaming content to thousands of venues across the country—but despite a robust content library, many business owners were seeing the same few clips on repeat. This case study details how I led a UX initiative to uncover the root of this content fatigue, and designed a scalable scheduling solution that increased content discovery, improved engagement, and balanced the needs of both small businesses and enterprise users.
Through user interviews, prototyping, and real-world testing, we reimagined how content is selected, scheduled, and controlled—transforming the way businesses use Atmosphere to entertain their customers.
Note: Some details have been modified or generalized to respect company confidentiality.

Final Shipped Design of Scheduler
The Challenge
"We're getting the content on repeat, and it's not keeping customers engaged."
This comment from a restaurant owner during a customer interview captured the core challenge facing Atmosphere TV. Business owners were growing frustrated with the limited content variety on their screens, and engagement was suffering. Driving owners to turn off Atmosphere TV in favor of other content providers.
As the UX lead on this project, I needed to understand: Why were users experiencing content fatigue despite having access to a robust content library? What was preventing them from discovering new channels? And most importantly, how could we design a solution that worked for our diverse user base?
Discovery Phase
Uncovering the Real Problems
I designed a research study to dig deeper into these content diversity issues, focusing on four key questions:
- What pain points existed in content discovery?
- What does the current mental model around our content look like?
- How (and would) a scheduling feature add value?
- Where in the existing user experience would this fit?
Prototyping Process
With these questions in mind, I created a low fidelity prototype to test with users and begin uncovering what would be most helpful.
The prototype introduced a popup onboarding flow with two paths: one allowed users to build their own schedule by entering content goals and selecting preferred content types—Atmosphere would then generate a tailored schedule. The other offered pre-curated schedules based on the user's venue type.
Each of the steps in this flow were designed to help us uncover the answers to our key questions, and seeing which path users took helped us to understand how custom the scheduling solution would need to be.

The resulting low fidelity prototype that users could interact with
Over the course of two weeks, I conducted eight 60-minute interviews with a diverse set of Atmosphere TV users, split evenly between enterprise accounts and small-to-medium businesses (SMBs). The venues ranged from casual dining restaurants to university lounges, car service shops, and upscale bars.

Dovetail overview of interview notes and images of me in action (clearly I talk with my hands)
Key Insights
The interviews revealed a key insight: most users rated their content awareness as just 2 out of 5. Despite having access to a wide range of channels, many users simply didn't know what was available or how to find content that matched their needs. Furthermore, users were discovering one form of content and stopping there, leading to longer term content fatigue.
"There's probably great content out there, but finding it feels like a burden. I don't have time to dig through everything, so I stick with what I know."
Other users were frustrated with the lack of control over content selection, some feeling like they had no control over what local managers were choosing.
"It's a constant battle between corporate and local managers over content control. We need standardization, but local flexibility too."
When discussing the scheduling feature, users were excited about the potential to customize content for different times of day, but had concerns about the complexity of setting up and managing schedules.
"I'd love to have more control over the content, but I don't have time to constantly monitor the content and make changes."
I organized the 264 insights from these interviews into 14 themes

Dovetail diagram of key quotes and themes
Understanding the feedback
The research presented an interesting challenge: while 75% of users valued the ability to schedule content by time of day, they had varying responses:
- Enterprise users wanted robust controls and standardization across locations
- SMBs wanted simplicity and quick setup
- Venues with dynamic audiences (like restaurants with different breakfast/lunch/dinner crowds) needed more scheduling options
- Venues with consistent audiences felt they didn't value schedules at all
Above all, it was clear that an automated solution was not clicking with users. So many of them just wanted to build their own. So, I decided to pursue a dual-track solution:
- Atmosphere owned "curated schedules" for key venue classifications.
- This would allow us to provide a dynamic content experience for venues who didn't need or want to invest time in customization
- Full custom scheduling capabilities for users who needed precise control
- Highly customizable, allowing users with particular content needs to create schedules that perfectly match their business needs
Designing a Solution
After several rounds of iterations and sketches, I decided to move schedule creation into a full-page experience rather than a popup or multi-step flow. I wanted users to have the flexibility to choose their starting point. The designs below allow users to create schedules with custom names, venue selection, and channel choices. I also included a thumbnail to visually represent the content within each schedule as users add them.

Early phase sketches of scheduler layout

User flow of the entire scheduling experience and its impact across existing pages and flows

Empty starting view and finished filled out view

Different states of the schedule naming component

Adding venues to the schedule

Different states of the adding venues flow

Adding channels to the schedule
Beta Testing with Real Users
With designs reviewed and finalized, I established a structured beta program to refine the solution in real-world conditions. We recruited 20 active Atmosphere TV users representing different business types and sizes. I designed the beta program in three distinct phases:
Phase 1: Consultation (30-minute sessions)
I met with each participant to understand their specific content and scheduling needs and business goals. These conversations helped identify use cases I hadn't anticipated.
Phase 2: Async Testing
Participants were sent a set of curated schedules and provided feedback through UserZoom, helping identify content misalignments.
Phase 3: Follow-up Session (30-45 minutes)
After participants used a feature-flagged beta version, I conducted follow-up sessions focusing on real-world implementation, fleet management needs, and future feature priorities.

Affinity diagram of quotes from interviews
Unexpected Insights
The beta program revealed several critical needs that hadn't emerged in initial research:
- Tedious scheduling: Some users found filling out the entire 24 hour schedule to be a repetitive and time-consuming task
- Schedule Recovery: Nearly every user had concerns about returning to scheduled content after manual changes.
- Permission Controls: Many enterprise users wanted to completely lock out local managers from changing channels when a schedule was in place.
- Multi-Day Programming: Several users requested having different schedules for different days of the week
Design Iterations from Feedback
With these insights, I made many key improvements to the designs.
Enterprise Management
To better support large organizations, I made changes that supported multi device changes, and permissioning.


Default Channels
To solve the problem of tedious schedule creation, I thought up the idea of a 'default channel', essentially a channel that filled the gaps and worked as a fallback between the other scheduled events.
This helped users fill the entire schedule without needing to drag in the same channel multiple times, and made it much clearer what would be playing at any given time in the schedule.

Default channel drastically reduced the amount of time required to create a schedule.
Content Cues & Explanations
To add clarity to the content channels, we added channel taglines that highlighted a key element of what the channel offered. This was a big hit with our beta users.

Channel taglines in our Wine & Dine curated schedule
Impact and Results
The final solution transformed how Atmosphere TV users interacted with content:
- Content awareness increased from 2.0 to 4.3 (out of 5)
- Schedules quickly gained traction, with a significant portion of our audience switching to scheduled content.
- The feature has shown low usability risk, with minimal customer support requests since launch.
Most importantly, the scheduling feature became a powerful tool for reducing content fatigue. Users who adopted scheduling reported significantly higher satisfaction with content variety. One customer shared:
"This finally gives us the control we need while making sure our customers see fresh content throughout the day. It's the best of both worlds."

Modified final design with a new color scheme & modern navigation
Reflections
This project taught me several valuable lessons about designing for diverse business users:
- One size doesn't fit all: Different business types require varying levels of scheduling complexity
- Trust + control balance: Users appreciate automated recommendations but need modification capabilities
- Enterprise needs differ significantly: Large organizations require robust fleet management and permission controls
- Beta testing provides invaluable refinement: Real-world testing revealed needs that weren't apparent in initial research
Thanks for reading till the end!
If you have any questions, please reach out to me on LinkedIn