← Back to Portfolio

Data Systems · UX Research · System Design

Turning a Team Audit Into a Strategic Advantage

A SQL-powered skills system that replaced gut-feel staffing with real data — and surfaced capabilities leadership didn't know they had.

I built this on my own time because nobody else was going to. A SQL-powered team optimization system across 120 skills for 21 UX professionals — self-initiated, self-taught, and used by leadership for real staffing decisions.

Project Overview

Two UX teams merged. Nobody actually knew what the combined team could do. Leadership was assigning projects on instinct and hoping for the best.

I built a SQL-powered assessment system in Google Sheets that mapped 120+ skills across 21 people — and handed leadership a tool they could actually use to staff projects.

Project Details

Role
Data Analyst, UX Researcher, System Designer
Duration
2–3 months from basic survey to full optimization system
Team Size
21 UX professionals across merged teams
Tools
Google Sheets, Advanced QUERY functions, Weighted analysis formulas
Skills Assessed
120 UX competencies across multiple dimensions
Team Talent app — Details tab showing individual skill profile, Skill tab showing teammates ranked by a single skill, and Team Overview tab showing aggregate team capability

The shipped Team Talent app — Details (individual profile), Skill view (who's strongest at what), and Team Overview (aggregate capability across 120 skills)

A Merged Team With No Real Map of Itself

The merged UX organization faced a critical knowledge gap. With team assignments based on intuition rather than data, leadership couldn't effectively:

  • Identify optimal team compositions for projects
  • Understand individual capabilities and growth areas
  • Make strategic hiring or training decisions
  • Leverage hidden strengths within the team

The Core Problem: We needed to transform a superficial skills inventory into actionable team optimization insights that could drive real business decisions.

Raw Google Sheets data showing rows 2396–2437 with participant IDs, skill names, proficiency ratings, interest levels, and multiple comparison columns

The raw data underneath — 2,500+ rows of participant/skill combinations, each rated across expertise, interest, and multiple role dimensions. Useful only once systematically queried.

An Unexpected Foundation

From Overwatch to UX Team Optimization

My expertise with Google Sheets' QUERY function came from an unexpected source: analyzing my own Overwatch gameplay statistics. I had built complex spreadsheets to track damage output, kill ratios, and survival metrics to identify improvement areas in my gaming performance.

Key Skills Transfer:

  • Multi-dimensional analysis: Gaming stats (damage/kills/survival) → UX skills (expertise/interest/role fit)
  • Performance optimization: Personal gameplay improvement → Team composition optimization
  • Pattern recognition: Identifying gameplay weaknesses → Discovering organizational skill gaps

The System I Built to Fix That

Comprehensive Skills Database

I created a centralized data system capturing comprehensive skills assessment across 120 different UX competencies. Each team member rated their expertise level and interest level for every skill, creating a rich multi-dimensional dataset.

Assessment Framework

Expertise Scale: No Experience | Working Knowledge | Mastered

Interest Scale: Interested | Not Interested

Scope: 120 UX skills covering technical, strategic, and domain expertise

Google Sheets individual participant view — Participant 06 showing service offerings interest level, role fit averages, and full skill breakdown across Interested/Not Interested/Regardless columns

Individual participant view — service interest, role fit averages, and full 120-skill breakdown across proficiency and interest dimensions

Google Sheets dropdown showing the full list of 120 UX skills available for selection in the analysis views

The 120-skill taxonomy — every competency available as a query parameter across all analytical views

SQL-Powered Analysis Views

Using Google Sheets' QUERY function extensively, I built multiple analytical views that could slice and dice the data:

Individual View

Complete skill profiles for each team member

Skill Analysis

How all teammates ranked across specific skills

Role Definition

Weighted skill matching against role requirements

Service Capability

Team capacity for four core service offerings

Team Comparison

Custom team builder with strength analysis

Project-Specific

Flexible skill matching beyond traditional roles

Google Sheets skill analysis view — Brand/Logo/Identity Design selected, showing all 21 participants sorted into Interested/Not Interested columns with proficiency sub-columns and team averages

Skill analysis view — select any skill to see all 21 teammates sorted by proficiency and interest, with team averages at the top

Google Sheets team builder — dropdown skill priority selectors for High and Low priority, with teammates automatically ranked by weighted match score

Team builder — set high and low priority skills, and teammates are automatically ranked by weighted fit score for that project configuration

Custom Team Builder: Leadership could slot in any number of teammates and instantly see aggregated skill strengths, capability gaps, role distribution, and service delivery capacity across all 120+ competencies.

The Details That Made It Actually Useful

Self-Assessment Psychology

The Dunning-Kruger Effect in Practice

Team members with limited expertise often overrated their abilities, while highly skilled individuals who understood the depth of their knowledge gaps rated themselves more conservatively. This created data distortions where true experts appeared less capable than novices on paper.

Calibration Considerations

While I didn't implement peer validation in this version, recognizing these psychological patterns informed how leadership interpreted the results and highlighted the need for managerial context when making staffing decisions.

From Spreadsheet to App

As the system matured, I designed a companion mobile app — Team Talent — that made the data accessible to leadership without requiring any Sheets literacy. Three views covered the core use cases.

Team Talent Details tab — Reinhardt Wilhelm profile showing service scores and role scores, with sortable skill list showing proficiency level and interest per skill

Details tab — individual profile with service and role scores, searchable and sortable skill list with proficiency level and interest indicator

Team Talent Overview tab — Typography skill selected, showing avg skill level 192 and avg interest level 190, with all teammates ranked by proficiency

Overview tab — select any skill to see the full team ranked by proficiency, with interest indicators showing who actually wants to use it

Comparison View — Design Iteration

The comparison view went through three distinct layouts before landing on the final approach. The challenge was representing two teams' skill levels against each other in a way that was scannable at a glance.

Three design iterations A, B, C of the team comparison view — A uses mirrored bar charts from center, B uses single-direction bars, C uses a more compact version of B

Three comparison view iterations — A (mirrored bars from center), B (single direction with clearer labeling), C (refined spacing). B and C tested better for quick scanning.

Team Talent final comparison tab — Team 1 vs Team 2 with horizontal bar charts showing relative skill levels across multiple competencies

Final comparison view — team members stacked per side, skills listed with bar charts showing relative proficiency. Add Teammate+ supports multi-person team comparisons.

Comparison view with sort filter open — sort by Team 1 or Team 2, Skill Rating, Ascending or Descending

Sort and filter interaction — leadership could re-sort by either team's strength, making it easy to identify where one team significantly outperformed the other

Query Interface Design

Rather than forcing non-technical stakeholders to write QUERY formulas, I created pre-built analytical views that automatically surfaced insights. Leadership could simply navigate between views to find answers like:

  • "Which team members have mastered accessibility but aren't interested in research?"
  • "What's our team's overall capacity for usability testing projects?"
  • "How do these three potential team configurations compare for a complex research project?"

Results & Impact

21
Team Members
UX professionals assessed across merged teams
120
Skills Analyzed
Comprehensive competencies across all UX disciplines
6
Analytical Views
Different ways to slice and analyze the team data
2x
Assessment Cycles
Bi-annual updates with archived historical data

Hidden Technical Capabilities

The assessment revealed that several team members had stronger development backgrounds than leadership realized, opening up new possibilities for technical UX work and cross-functional collaboration.

Strategic Decision Support

Assessment results informed Nielsen Norman Group course selections for the team, demonstrating how individual skill gaps could drive organizational learning investments.

What Version 2 Would Look Like

  • Automated Data Refresh: Implement more sustainable data maintenance processes, possibly integrating with HR systems rather than relying on manual bi-annual surveys.
  • Dual Assessment Framework: Build a system where both employees and managers complete assessments, then use the comparison view to facilitate career development conversations around discrepancies.
  • Enhanced Validation: Consider peer validation for critical skills or implement calibration exercises to help team members understand skill level definitions more consistently.

Unrealized Potential: We could have used this for career progression conversations where both you and your manager assessed you, then used the comparison view to come together and discuss any differences — but I never built that feature.

Skills Demonstrated

Advanced Spreadsheet Engineering
Data Architecture
Google Sheets QUERY Functions
Stakeholder Research
System Design
Strategic HR Analytics
Self-Directed Learning
Weighted Algorithm Design
Multi-dimensional Analysis
Organizational Problem-Solving
Change Management
Data-Driven Decision Making

View More Case Studies