Case Study
TalentUp
Salary Platform
A centralized tool for smarter compensation analysis and workforce insights.
TL;DR
TalentUp had valuable salary data but poor usability — 80% of users dropped off after a single search. As sole Product Designer, I led a complete redesign: overhauled the information architecture, built a 200+ component design system, and restructured the product around salary benchmarking as the core value driver.
01 — Problem
A powerful tool that people stopped using
TalentUp had valuable functionality, but poor usability was holding it back. Users — primarily HR professionals — found the interface confusing and the language too technical. Many used their credit allowance for one-off salary searches but rarely engaged beyond that.
Key features like employee management or the dashboard were largely ignored, either due to lack of clarity or perceived irrelevance. The platform lacked consistency in design, and feedback showed that users were unsure how to extract insights or take meaningful action.
80%
of users dropped off after a single search with no further interaction
22%
reported that key terms and features were confusing or not self-explanatory
84%
of active users had never added employees — a core part of the product's value
02 — Research
Understanding what HR teams actually need
We conducted a competitive benchmark analyzing Ravio, Figures, Pave, Compensation IQ, and Gradar — identifying common patterns, value propositions, and usability trends across the landscape.
We also evaluated TalentUp's existing buyer personas, from solo HR professionals to large compensation teams, mapping key user needs and behaviors to uncover the root causes of low engagement.
Lack of HR system integration
Users expect to connect BambooHR, Personio, or SAP to auto-import employee data. Without it, manual uploads become a barrier for larger teams.
Underused core features
Key actions like comparing, filtering, and downloading data were buried across sections — causing fragmented workflows and low engagement.
Lack of trust and transparency
Competitors emphasize confidentiality and compliance. TalentUp lacked messaging around data safety, source transparency, or methodology.
Misaligned value proposition
"Manage Employees" created friction — users perceived it as requiring sensitive input before getting value. The real strength (benchmarking) was barely surfaced.
03 — Process
Rethinking the entire information architecture
Based on the research, we redefined the platform's IA and core functionality. Strategic renaming played a key role — changing "Manage Employees" to "Salary Benchmarking" alone reduced perceived risk and improved clarity significantly.
We began with hand-sketched low-fidelity wireframes to quickly iterate on structure and layout. After defining flows, we moved to digital wireframes and built a design system from scratch using Preline (Tailwind CSS) as a base — the first the platform had ever had.
📊 Salary Benchmarking
Editable tables · Smart lists · Downloadable insights · Data visualizations (equity pay, gender ratio, retirement risk)
🔎 Search Salaries
Advanced filtering · Market Trends · Similar Roles · Bonus and Stock breakdowns
🏢 Company Finder
Salary Tables · Hiring Trends · Company Overviews by Department
⚖️ Compare Salaries
Compare up to 10 positions · Location-specific insights (Cost of Living, Inflation %, Unemployment rate)
04 — Testing
7 HR professionals, 2 flows, 30 minutes each
We conducted remote usability tests to validate new functionalities and ensure clarity across two distinct flows: one where the company is found automatically in the database, and one where the user must manually add employees.
We interviewed seven professionals across HR roles — Compensation & Benefits specialists, People Ops generalists from scaling startups, and external consultants. Sessions lasted 20–30 minutes, covering registration, adding employees, exploring insights, unlocking market benchmarks, and comparing internal data with market trends.
The interviews revealed key areas for improvement: insight visualizations needed clearer structure, users wanted to filter percentile comparisons by department, and some labels created confusion. We iterated on flow structure, rewrote key UI copy, and made targeted design changes in Figma based on these findings.
Remote usability testing with 7 HR professionals across Compensation, People Ops, and consulting roles
05 — Outcome
A more strategic, user-focused product experience
After completing the final design iterations based on testing insights, the project entered the development stage. We worked closely with the engineering team through recurring reviews — tracking implementation, resolving design inconsistencies, and testing flows as they're built.
Enhanced feature comprehension
Clearer structure around Salary Benchmarking improved perceived value of the feature and reduced abandonment after first use.
Increased user trust
Stronger messaging around data privacy, source transparency, and methodology increased user confidence in the platform.
Scalable design system
200+ components built on a consistent, documented library — improving cohesion and accelerating developer handoff.
Streamlined workflows
Tailored to both occasional users and compensation specialists, reducing friction across the key flows.
Dashboard — salary benchmarking overview and team workforce insights
Insights — equity pay, gender ratio and retirement risk data visualizations
Compare Salaries — up to 10 positions with location-specific cost of living and inflation data
06 — Reflections
What this project taught me
This was a full 0→1 redesign of a live product — from strategy and research to high-fidelity UI and developer collaboration. Here's what I took away.
User research is most valuable when it leads to real decisions
Testing with actual HR professionals helped us spot specific usability issues and restructure the product in ways that made a measurable difference.
Copy isn't cosmetic — it's functional
Renaming "Manage Employees" to "Salary Benchmarking" had a direct impact on comprehension and trust. Words are design decisions too.
Prototypes don't need to be perfect — they need to be real
Auto-generated test data let us run focused sessions without overwhelming participants, keeping feedback on usability rather than content.
Design systems enable alignment, not just consistency
A documented component library improved collaboration with developers and ensured visual and functional coherence across the product.
Product quality improves when design stays in development
Weekly build reviews let us catch critical issues early — before they became costly to fix or shipped incorrectly.