DOWN DOG — Woohee: UX designer

ABOUT

Down Dog

Problem

While in-person yoga instructors often suggest alternative postures for practitioners with physical limitations, including temporary injuries and disabilities, most at-home yoga apps lack personalized modification options.

Outcome

I redesigned the app with an automated pose-filtering system that enables users to customize their yoga flows based on their physical constraints.

Timeline

01.2025-05.2025

Role

UX Researcher
UX Designer

Team
Solo Project

Final Design

Assessment Onboarding

Start with a guided assessment to shape your flow. Just share your goals, current condition, and any limitations—your practice adjusts from there.

Injury-aware Filter

Add an injury filter to leave out discomfort-inducing poses—bringing clarity and ease to every session.

In-depth Pose library

Follow up the pose library with a supplemental screen—offering clear context, visual cues, and easy-to-understand guidance for every posture you explore.

Here’s how the work came together!

Market Research

Yoga has gained significant popularity in recent years.

With an increasing emphasis on mental well-being and physical fitness, people from all walks of life recognize the benefits of yoga.

🧘🏻‍♀️

1 in 6

of U.S. adults practiced yoga in the past 12 months.

🛁

9 in 10

are taking more action than they were a year ago to improve their well-being in any dimension.

Empathy Interview

Lack of in-app guidance led to unsafe adaptations

I conducted a semi-structured interview with 7 yogis with injuries. Due to a lack of modification, unlike in-person studios, I discovered that yogis often modify poses intuitively or skip, which can often lead to unsafe or inefficient practice.

Competitive Evaluation

What’s missing across platforms?

I compared Down Dog, Pocket Yoga, and Yoga Studio against the core user needs to understand how users with injuries build customized practice and discover friction points in the process.

This analysis revealed shared gaps across all three platforms—the absence of injury-based modification.

Problem Identification

Digital yoga experiences fail to support yogis’ real-life conditions

From in-depth interviews, competitive evaluation, and gap analysis, I was able to find common findings across yoga practitioners–centered around lack of personalization, unclear pose guidance, and inaccessible interfaces.

01

Lack of pose information

Limited information about posture creates frustration among users.

02

No modification for physical condition

No guide or modification to adjust postures based on physical condition.

03

Cognitive burden from manual selection

Unfriendly search bar and endless scroll increase cognitive burden.

Ideation

Improvement opportunities uncovered

Based on key challenges identified from the research phase, I ideated opportunities for each to improve the user experience of the Down Dog app.

Concept Exploration

Divergent concepts were explored with Crazy Eights

To address the challenge, I brainstormed distinct concepts that can satisfy users’ needs. I focused on an injury-based filter after understanding the technical limitations and challenges based on users’ mental models.

Concept Refinement

From sketch to systematic solution

I refined the concept around an injury filter system that allows users to input sensitive areas. This input dynamically filters out unsafe poses or suggests modifications.

I also visualized the improved backend logic, ensuring the experience is both adaptable and scalable - balancing personalization with simplicity.

Usability Testing

Rapid prototyping revealed usability challenges

To validate the effectiveness of the redesigned experience, I conducted a series of usability tests, including A/B testing and multivariate testing.

❌ Light information was not preferred by users as it doesn’t add much value to the pose library.

✅ Medium-length text was preferred by most participants as it was in-depth enough to assist visual demonstration.

❌ Too much detailed information was not needed for users with visual demonstration.

❌ Participants expected to see any kind of visual demonstration to better understand.

❌ Participants wanted to know how to get to the posture with better context and motion.

✅ Proceeded with the design with a short video clip providing better context and understanding.

❌ Participants barely used a search bar as they were expected to already know about the name of the pose.

✅ Added filters that are easily categorized for beginners and intermediates without knowing the exact term.

Solution

Yoga flow made just for you and your body

An injury-based filter is implemented to solve the main challenge of yoga practitioners with various physical conditions. An in-depth pose library and assessment onboarding were added to support the challenge.

This solution not only aligns with the existing back-end system but also emphasizes the importance of taking care of users’ bodies before they realize it.

Validation

Did the new experience actually feel safer and more personal to users?

To validate the effectiveness of the redesigned experience by measuring cognitive load, I conducted eye tracking heatmaps, NASA-TLX, and preference testing with a follow-up interview.

Outcome & Impact

Where It Landed

I successfully presented the project to key leadership and spectators during the Thesis Show. Some of the feedback I got from mentors and faculty includes:

“It has the potential to expand into other areas similar to yoga.”
“Looks great! I really hope the injury filter gets added to the app.”
“Well-rounded presentation, and has a clear problem and a data-driven solution for the situation.”

Reflections

What did I learn from this project?

Validate assumptions early and often.

User testing isn’t just a final step—it’s an ongoing process that grounded our decisions in reality. By validating ideas early, I uncovered usability issues we hadn’t anticipated and ensured that our designs aligned with actual user needs, not just internal opinions.

Work in layers, not lines.

Working on the Down Dog project taught me that research, design, and testing rarely happen in neat phases. Often, I had to juggle refining wireframes, revisiting research insights, and gathering feedback simultaneously—progress wasn’t linear, but layered.