Applicant Tracking System.
Focused an eight-year-old ATS on one recruiter segment and validated the redesign before production.
I led research and product design, choosing one recruiter segment over a compromised redesign for four personas.
Key results
- +21% — Perceived utility gain (UMUX-lite, pre-production benchmark, n=6) (Measured)
- +13% — Perceived usability gain (UMUX-lite, pre-production benchmark, n=6) (Measured)
- 100% — All 6 participants completed all 10 benchmark tasks without assistance (Measured)
What I measured
- Perceived utility (UMUX-lite) — +21% (Baseline vs. redesigned prototype, 6-user study)
- Perceived usability (UMUX-lite) — +13% (Baseline vs. redesigned prototype, 6-user study)
- Candidate review speed — ~30% faster (Prototype benchmark study, n=6 — observed task timing across 10 tasks)
- SUS baseline — Established (Run alongside UMUX-lite to validate it as future benchmark)
- Task completion — 100% (All 6 participants completed all 10 tasks without assistance)
Competitor products failed recruiters the same way
Recruiters had moved real work into spreadsheets, weakening product adoption, reporting, and compliance.
Eight-year-old product
Technical debt and entrenched workflows.
Four personas
The client could fund one.
Shadow tools
Recruiting work happened outside the system.
No telemetry
Research had to drive prioritization.
I have 800 profiles for this role and a day to get through them. I end up eyeballing it.
Our real tracking happens in Excel. The ATS is mostly for compliance.
In-house recruiter, mid-size company
Competitor ATS user (interview)
Choosing one of the four personas
Research found four personas. I chose in-house recruiters at mid-size companies—the core market—and documented the rest.
“Refresh the ATS so it’s competitive again.”
“Use the redesign to test a new recruiter workflow. Serve one persona well instead of compromising across four.”
Competitor research confirmed the problem was category-wide.
Serving all four personas would dilute the workflow.
Optimize for the core segment; preserve the other journeys for expansion.
Stakeholders observed the evidence.
In-house recruiters
Co-designed and tested the core workflow.
Competitor ATS users
Confirmed category-wide pain.
Tech lead
Vetted phased delivery and feasibility.
Client stakeholders
Observed research before approving scope.
A job-creation wizard that starts from a template
I replaced scattered job creation with templates, defaults, contradiction checks, and multi-board posting.
Group workflows by recruiter decision, not the data model, and place frequent actions beside the information that triggers them.
800+ candidates without bulk actions
Manual entry and no automation
Shadow tracking in Excel
No hiring reports
What I measured — and what I couldn't
Directly observed metrics from a 6-user usability study combining SUS and UMUX-lite.
Not claimed
No production or financial instrumentation
+21% perceived utility in the pre-production benchmark
+13% perceived usability in the same study
= Evidence for the redesign direction, not production adoption or revenue
The study compared the existing product with the redesign across 10 tasks and 6 participants
Whether recruiters moved work out of Excel was not measured
Retention, sales, and financial outcomes remained confidential or out of scope
Excel displacement
Whether recruiters abandoned shadow spreadsheets was not measured.
Production performance
The evidence remains a pre-production study.
Other personas
Suitability outside the chosen segment remains unmeasured.
What worked · what I'd do differently
What worked
Running SUS and UMUX-lite together to bootstrap future measurement.
Running both instruments established UMUX-lite as the practical ongoing benchmark.
What I'd do differently
Negotiate post-launch adoption metrics — especially Excel-displacement.
Negotiate post-launch telemetry for the outcome that matters: work moving out of Excel.
FAQ
Why does +21% perceived utility matter for an ATS?
ATS adoption is notoriously low; recruiters revert to spreadsheets, losing visibility and compliance. A +21% utility gain means recruiters believe the tool genuinely helps — the lever the redesign is designed to pull: keep the work in the system, keep reporting accurate, keep compliance intact. That stays an intended mechanism rather than a result: this was a pre-production benchmark, so production adoption, reporting accuracy and compliance outcomes were not measured. For the client, that is a premium-priced product differentiated by research-backed design.
You focused on a single persona — was that the right call?
Given the budget, yes. Serving all four personas equally would have meant a mediocre experience for everyone. Optimizing for the core market (in-house recruiters at mid-size companies) solved their problems instead of half-solving everyone's. The other three personas are documented as validated future-expansion artifacts.