UX Team
Skills Assessment
When two UX teams merged, leadership was staffing projects based on guesswork. I built a SQL-powered optimization engine inside Google Sheets that mapped 120 competencies across 21 designers — enabling data-driven team composition for the first time.
The Problem
When two UX groups merged into one Enterprise UX organization, leadership ran a basic skills survey. It was surface-level and provided almost no actionable insight — so team assignments kept happening based on intuition and familiarity rather than actual capability.
Leadership couldn't identify optimal team compositions, understand individual growth areas, make strategic hiring decisions, or leverage hidden strengths sitting right in front of them.
An Unexpected Foundation
My fluency with Google Sheets' QUERY function came from an unusual place: I'd built complex spreadsheets to analyze my own Overwatch gameplay — tracking damage output, kill ratios, and survival metrics to find improvement areas.
The skills transfer was direct: multi-dimensional game stats (damage/kills/survival) mapped cleanly to multi-dimensional UX competencies (expertise/interest/role fit). Performance optimization for personal gameplay became team composition optimization for an organization.
System Architecture
Assessment Framework
- 120 UX competencies — technical, strategic, and domain expertise
- Expertise scale — No Experience · Working Knowledge · Mastered
- Interest scale — Interested · Not Interested
- Each dimension tracked separately to surface "mastered but burned out" vs "interested but developing"
Six Analytical Views
Individual View
Complete skill profile per team member — their full expertise and interest map across all 120 competencies.
Skill Analysis
How all teammates ranked on any specific skill — instantly surfacing who to pull for a given need.
Role Definition
Weighted skill matching against role archetypes — who's actually a researcher vs. who thinks they are.
Service Capability
Team capacity across four core service offerings — where are we strong, where are we thin?
Team Builder
Slot in any combination of teammates, instantly see aggregated strengths, gaps, and role distribution.
Project-Specific
Custom skill matching beyond traditional role definitions — built when leadership requested it mid-project.
The Raw Material
Everything started here — 120 skills, 21 participants, hundreds of rows. Useful for storage, useless for decisions. The QUERY engine turned this into something leadership could actually act on.
Raw survey output: participant IDs, skill labels, expertise levels, and interest ratings — all queryable but requiring a system to make sense of at scale.
Automated QUERY output per teammate: service offering interest levels, role fit scores, and full skill breakdown sorted by mastery — leadership could instantly profile any team member.
The Mobile Vision
The spreadsheet solved the immediate problem. The Figma explorations showed what this system could become — a purpose-built tool for team leads to build and compare project teams on the fly.
Individual profile: service capacity scores, role fit rankings, full skill list with expertise level and interest flags.
Any skill, instantly ranked — who's mastered it, who's interested, who has it but doesn't want to use it. Surfaces the nuance the raw data hid.
Build two teams, choose a skill to sort by — instantly see where each team is strong or exposed.
Side-by-side team skill scores with visual weight bars — data-driven staffing decisions in seconds, not days.
The Team Builder
Leadership could slot in any number of teammates and instantly see aggregated skill strengths, capability gaps, role distribution, and service delivery capacity. A weighted ranking system automatically scored fit — high matches received positive weights, gaps were flagged as risks.
The system operated autonomously — leadership navigated it independently without needing to understand a single QUERY formula. That accessibility was a core design requirement, not an afterthought.
What I Grappled With
The Dunning-Kruger Problem
Self-assessment data has a known distortion: team members with limited expertise often overrate themselves, while highly skilled individuals — aware of how deep their knowledge goes — rate themselves more conservatively. True experts appeared less capable than novices on paper.
I didn't implement peer validation in this version, but recognizing the pattern meant I could brief leadership on how to weight and contextualize results — adding managerial judgment as a calibration layer.
Insights Uncovered
Hidden Technical Depth
Several team members had stronger development backgrounds than leadership realized — opening new possibilities for technical UX work no one had assigned them to.
Training Investment Clarity
Assessment gaps directly informed which NN/g courses to prioritize for the team — turning abstract skill gaps into a concrete training roadmap.
What I'd Do Differently
- Manager + employee dual assessment — both complete independently, then compare the gap as a career development conversation starter
- Automated data refresh — integrate with HR systems rather than relying on bi-annual manual surveys
- Peer validation layer — calibration exercises to normalize how team members interpret skill level definitions
The unrealized potential: using this for career progression conversations where you and your manager each assess you separately, then meet to discuss the differences. Never built it. Next time.