My Role
UI/UX Designer (part-time)
Timeline & contribution
April (solo): public user profile and task posting flow
June - July (collaborative): worked with the second designer on other screens and flows. Not detailed in this case study
August (solo): accessibility audit
Team
2 Co-founders
Second designer
Collaboration
async feedback and regular team calls
Note: This case study focuses on my solo contributions.
BUSINESS CONTEXT & GOALS
Aidly is a C2C platform connecting people who need help with everyday tasks to those willing to provide services for a fee.
The platform was launched in January 2025 with no design system, Figma files, or accessibility standards. When I joined in April 2025 as the first designer, the product was functional but lacked visual consistency and accessibility considerations.
The product needed to:
Increase user acquisition (both task posters and service providers)
Reduce friction in core user flows
Build trust between strangers engaging in service transactions
PROBLEMs
User problem
10-field modal for posting tasks created friction and confusion
Inconsistent UI design confuses users and makes the platform look untrustworthy
Color contrast issues, missing keyboard navigation, and unclear error states excluded users with disabilities
Business problem
Platform needed to convert visitors into active task posters and service providers
No design foundation: Without standards or component libraries, every new feature introduced new inconsistencies
IMPACT
Prevented user exclusion and expanded addressable market
Documented 41 accessibility issues; 20+ were fixed by the developer.
71% of users with disabilities abandon inaccessible sites (industry benchmark).
Created design foundation
Built first Figma component library, color system with contrast documentation, and accessibility tracker to create a scalable foundation for future work.
Made public user profile clearer
Reduced profile complexity using minimal color, strong hierarchy, and role-based tabs.
Improved scannability and helped users assess trust faster on a platform where strangers hire each other.
Redesigned a 10-field modal into a 4-step flow, reducing cognitive load through progressive disclosure and a clear progress indicator.
APPROACH & RATIONALE
I had to be strategic about where design effort would have the most impact. Before I joined, the startup had no designer, no researcher or design system. It was just me working part-time while the developer was trying to ship fast.
Without budget for research or time for extensive user testing, I relied on what I had: competitive analysis, UX best practices, and accessibility standards.
I focused on 3 high-impact areas:
1
Public user profile
Trust matters on a platform where strangers hire each other. Users needed to quickly assess if someone was reliable.
execution
Before starting, I conducted a brief audit of the existing homepage to understand what patterns existed.
I found multiple corner radius values, insufficient color contrast, and no clear distinction between primary and secondary text, all of that creating a broken visual hierarchy.
Next, I set the base and created a minimal style guide for typography and WCAG-compliant color variables.
Why? To work faster, eliminate redundant decisions and create consistency across design.
Users can be Taskers (complete tasks for others), Posters (post tasks requesting help), or both.
Every user has a public profile that other users can view to see what tasks they’ve posted, what tasks they are currently working on, and the reviews they’ve received.
Before
Visual noise and broken hierarchy (competing colors, misaligned elements)
No CTA for active posts
Badges were styled like buttons, causing confusion about what was interactive
After
Reduced cognitive load with minimal color palette and clear visual hierarchy
Introduced trust signals (user bio, reviews, verification badge) to build trust between users
Added CTA to active posts so users could apply directly from someone's profile instead of returning to homepage and searching for the post
Clarified roles with tab-based interface so users instantly see if someone is a Tasker and/or Poster
Badge styling distinct from buttons
Before
10 fields modal with a mixed layout creates confusion and cognitive overload
No clear error guidance
Tooltip hides essential information, increasing user errors
After
Full-page layout to eliminate scroll conflicts, especially on mobile
Reduced cognitive load by breaking the form into 4 logical steps with a visible progress indicator
Eliminated disappearing placeholders in favor of persistent helper text to minimize memory strain and prevent input errors
Reduced frustration by making errors visible
Supported users with color vision impairments by designing descriptive error messages placed directly above relevant fields, using icons and clear language
Note: This design was later simplified to single page after iterations with team. We removed non-essential fields to reduce time and abandonment.
As the platform stabilized, I wanted to ensure it was usable by everyone. Because of limited time, I selected 17 WCAG criteria with the highest potential impact to test on the homepage.
Issues found were documented in a Notion table along with the proposed solution.
Examples of issues found:
Skip heading hierarchy
Redundant alt text
Insufficient color contrast
Unlabeled form fields
Interactive elements couldn't be reached via keyboard
Focus indicators at 1.5:1 contrast instead of required 3:1
Layout broke and text overlapped at 200% zoom
Missing skip link
Screenshot of Notion
Learnings
A few improvements don't make a product accessible.
Should have established metrics from day one (form completion rate, task posting frequency). Validate assumptions and reduce guesswork about which problems mattered most with user interviews.
Building reusable components made everything faster later, when I needed to iterate and keep the design consistent.
I thought my 4-step form was pretty good. Then we made it one page, and it was actually better. Learned not to fall in love with my own solutions.
Things changed constantly. Developer couldn't implement everything perfectly because there just wasn't time. Not every detail is worth holding up a launch for. Get the structure right, worry about polish later.
Using NVDA confirmed what I already knew but needed to experience: the same product feels completely different through assistive technology.
It reinforced why I do this work. Not for compliance checkboxes, but because tiny actions open the digital world to more people. That's the impact I want to keep making.