Johanna Huarachi's Portfolio*

TLDR

Overview

I redesigned the bill-splitting experience for Billclub.io, a startup that helps groups divide costs by scanning receipts. My contributions include running usability tests with 5 users, identifying that 80% dropped off during bill assignment, redesigning the entire flow to be collaborative instead of solo, and validating the solution with 6 participants. The result: task completion increased by 60%.

ROLE

Design Systems Leads

UX/UI Designer

ROLE

Design Systems Leads

UX/UI Designer

ROLE

Design Systems Leads

UX/UI Designer

TIMELINE

12/1/23 - 6/15/24

TIMELINE

12/1/23 - 6/15/24

TIMELINE

12/1/23 - 6/15/24

TEAM

Johanna Huarachi

Kien Tran

Henry Zhao

TEAM

Johanna Huarachi

Kien Tran

Henry Zhao

TEAM

Johanna Huarachi

Kien Tran

Henry Zhao

SKILLS

User Research

Usability Testing

Design Systems

SKILLS

User Research

Usability Testing

Design Systems

SKILLS

User Research

Usability Testing

Design Systems

PROBELM

Users were dropping off mid-flow. The host had to manually assign every item to every person while friends waited. Most gave up.

How can we redesign bill splitting so the entire group shares the work instead of one person doing everything?

Users were dropping off mid-flow. The host had to manually assign every item to every person while friends waited. Most gave up.

How can we redesign bill splitting so the entire group shares the work instead of one person doing everything?

Users were dropping off mid-flow. The host had to manually assign every item to every person while friends waited. Most gave up.

How can we redesign bill splitting so the entire group shares the work instead of one person doing everything?

RESEARCH

Key insights about the flight search experience

4 out of 5 users abandoned at the assignment screen.
I ran usability tests with 5 users. This pattern showed up: users scanned their receipt, items populated, and then assignment. Billclub.io asked them to go through each item and assign it to each person. For a group dinner, this meant dozens of decisions. Only 1 user finished.
The solo host burden was exhausting.
One person had to remember what everyone ordered, make decisions for the whole group, and tap through every single item. It turned a social moment into an admin task.
Users wanted speed and trust.
Talking to young adults revealed what mattered: split bills as fast as possible and trust the math. These principles became the foundation for everything that followed.
4 out of 5 users abandoned at the assignment screen.
I ran usability tests with 5 users. This pattern showed up: users scanned their receipt, items populated, and then assignment. Billclub.io asked them to go through each item and assign it to each person. For a group dinner, this meant dozens of decisions. Only 1 user finished.
The solo host burden was exhausting.
One person had to remember what everyone ordered, make decisions for the whole group, and tap through every single item. It turned a social moment into an admin task.
Users wanted speed and trust.
Talking to young adults revealed what mattered: split bills as fast as possible and trust the math. These principles became the foundation for everything that followed.
4 out of 5 users abandoned at the assignment screen.
I ran usability tests with 5 users. This pattern showed up: users scanned their receipt, items populated, and then assignment. Billclub.io asked them to go through each item and assign it to each person. For a group dinner, this meant dozens of decisions. Only 1 user finished.
The solo host burden was exhausting.
One person had to remember what everyone ordered, make decisions for the whole group, and tap through every single item. It turned a social moment into an admin task.
Users wanted speed and trust.
Talking to young adults revealed what mattered: split bills as fast as possible and trust the math. These principles became the foundation for everything that followed.
4 out of 5 users abandoned at the assignment screen.
I ran usability tests with 5 users. This pattern showed up: users scanned their receipt, items populated, and then assignment. Billclub.io asked them to go through each item and assign it to each person. For a group dinner, this meant dozens of decisions. Only 1 user finished.
The solo host burden was exhausting.
One person had to remember what everyone ordered, make decisions for the whole group, and tap through every single item. It turned a social moment into an admin task.
Users wanted speed and trust.
Talking to young adults revealed what mattered: split bills as fast as possible and trust the math. These principles became the foundation for everything that followed.

SOLUTION

Stop making one person do all the work. Let people claim their own items. Turn bill splitting into something the group does together.

Host scans the receipt
The host scans. AI extracts all items and costs automatically.
Host verifies and adds everyone
The host scans. AI extracts all items and costs automatically.
Everyone claims their own items (The Fix)
The host scans. AI extracts all items and costs automatically.
Final amounts calculate automatically
Everyone sees who's paid and who hasn't. The math is seamless..
Host scans the receipt
The host scans. AI extracts all items and costs automatically.
Host verifies and adds everyone
The host scans. AI extracts all items and costs automatically.
Everyone claims their own items (The Fix)
The host scans. AI extracts all items and costs automatically.
Final amounts calculate automatically
Everyone sees who's paid and who hasn't. The math is seamless..
Host scans the receipt
The host scans. AI extracts all items and costs automatically.
Host verifies and adds everyone
The host scans. AI extracts all items and costs automatically.
Everyone claims their own items (The Fix)
The host scans. AI extracts all items and costs automatically.
Final amounts calculate automatically
Everyone sees who's paid and who hasn't. The math is seamless..
Host scans the receipt
The host scans. AI extracts all items and costs automatically.
Host verifies and adds everyone
The host scans. AI extracts all items and costs automatically.
Everyone claims their own items (The Fix)
The host scans. AI extracts all items and costs automatically.
Final amounts calculate automatically
Everyone sees who's paid and who hasn't. The math is seamless..

VALIDATION

We tested high-fidelity prototypes with 6 participants, comparing legacy vs. redesigned collaborative flow.

The numbers shifted dramatically.

Legacy flow: 1 out of 5 completed.
Redesigned flow: 5 out of 6 completed.

That's a 60% increase in task completion.

What users said:

"This is easier when everyone can just grab what they ordered"

When the work was shared across the group, users stayed engaged. Collaborative assignment eliminated the bottleneck.

LEARNINGS

Key Takeaways

Trust your research

When 4 out of 5 users showed the same behavior, I had enough data to make a decision. Working as the only designer on a startup team taught me to move fast and trust what users were telling me.

Advocate for fundamental change

The hardest part was convincing the team to rebuild the core flow instead of polishing what existed. In a startup, time is everything. I had to show clear evidence that incremental fixes wouldn't solve the real problem.

OTHER WORKS

Overview

Users were dropping off mid-flow. The host had to manually assign every item to every person while friends waited. Most gave up.

Users were dropping off mid-flow. The host had to manually assign every item to every person while friends waited. Most gave up.

Users were dropping off mid-flow. The host had to manually assign every item to every person while friends waited. Most gave up.

Key insights about the flight search experience

4 out of 5 users abandoned at the assignment screen.

The solo host burden was exhausting.

Users wanted speed and trust.

4 out of 5 users abandoned at the assignment screen.

The solo host burden was exhausting.

Users wanted speed and trust.

4 out of 5 users abandoned at the assignment screen.

The solo host burden was exhausting.

Users wanted speed and trust.

4 out of 5 users abandoned at the assignment screen.

The solo host burden was exhausting.

Users wanted speed and trust.

Stop making one person do all the work. Let people claim their own items. Turn bill splitting into something the group does together.

Host scans the receipt

Host verifies and adds everyone

Everyone claims their own items (The Fix)

Final amounts calculate automatically

Host scans the receipt

Host verifies and adds everyone

Everyone claims their own items (The Fix)

Final amounts calculate automatically

Host scans the receipt

Host verifies and adds everyone

Everyone claims their own items (The Fix)

Final amounts calculate automatically

Host scans the receipt

Host verifies and adds everyone

Everyone claims their own items (The Fix)

Final amounts calculate automatically

We tested high-fidelity prototypes with 6 participants, comparing legacy vs. redesigned collaborative flow.

The numbers shifted dramatically.

What users said:

Key Takeaways

Trust your research

Advocate for fundamental change

← PROS Fare Finder

DNC →