Case study · State of Illinois · 2018

Illinois College2Career

A state website that helps high school students compare Illinois colleges by fit, cost and what graduates go on to earn. We set out to help students pick a career. The research said they needed something else first.

Work from 2018. Diagrams and the screen image redrawn in 2026. See the originals
My roleUX architect: research, wireframes, prototype, visual design and development oversight
TeamA developer, a data scientist, a project manager from the Governor's Office, the IT team manager and me
ClientIllinois Department of Employment Security, with the state's higher education boards
ShippedA live responsive website, six months from kickoff
The problem

Students were leaving Illinois for college and not coming back.

The state wanted to show what its own colleges offer: programs, price, and the jobs and wages that follow. The starting idea was that if students could see Illinois job data tied to areas of study, choosing a career path in the state would get easier.

Process

Six months, kickoff to release.

Discover30 interviewsPersonasJourney mapsDefineGoal alignmentwith the customerDesignSite map and flowWireframesMockupsTestMarvel prototype23 usability sessionsAffinity mappingLaunch and learnBeta with schoolsWeekly analyticsRefinementKICKOFFRELEASE · 6 MONTHS
Round one · discovery

First we watched how students actually look for a college.

30juniors and seniors
3 high schools · observed with minimal prompting
  • Most used Google, not a college search tool.
  • Many leaned heavily on parents and friends.
  • They started researching too late.
  • Undecided students looked at cost before careers.
  • They cared more about an interesting path than about pay.
Three people to design for
The undecided studentWants college, has no major in mind.
The decided studentKnows the major, needs the right school.
The parentWants a good match the family can afford.
Journey maps of the existing process added three more: involve students early, too much data confuses, and with many metrics a best match isn't obvious.
The turning point

The research changed what the product was for.

Students weren't ready to choose a career. They were trying to find a college that fit and that they could pay for. I took that back to the customer, and they agreed to reorder the product around it.

The brief we started with1. Choose a career path2. See which majors lead there3. Find a collegeThe brief after research1. Find a best-match college2. See what it costs3. Explore careers and earnings
Round two · usability testing

Then we put a prototype in front of students.

23juniors and seniors
2 high schools · moderated, open and directed tasks · clickable Marvel prototype
  • Most found it easy to navigate and use.
  • Affinity mapping of their own words surfaced the rest.
What testing foundThe results screen felt overwhelmingly busy.
What changedCut the data shown at once and redesigned the results layout.
What testing foundThe colored arc gauge was hard to read at a glance.
What changedReworked the gauges with a colorblind-friendly palette.
What testing foundStudents tried to tap the tutorial. It was a static picture.
What changedRevised prompts and affordances so instructions don't look like controls.
What testing found"All" did not mean "undecided" to them, and the tabs weren't understood.
What changedRenamed and restructured the choices around how students described themselves.
After launch

A soft launch, then weekly numbers.

BetaStudents and administrators from several high schools and colleges tried it. We followed up by email and in person, and fixed issues in setup, data graphics and table customization.
AnalyticsWeekly Tableau reports showed where visitors spent time, what they skipped and which metrics they cared about.
RefinementFixed the most serious problems first, then tuned the gauge colors and the start screen for search. Stakeholders approved the next release.
Looking back from 2026

What I'd keep, and what I'd do differently.

Keep
  • Research before the brief hardens. Thirty conversations moved the whole product.
  • Taking an uncomfortable finding to the customer early, with evidence.
  • Testing with the real audience, in their schools.
Do differently
  • Agree on success measures before launch, so there are outcome numbers to report.
  • Test the data-heavy results screen earlier, at lower fidelity.
Archive · the original 2018 project files

The real artifacts, as made at the time.

The diagrams above were redrawn in 2026. These are the originals: personas, journey map, wireframes, mockups, test materials and the screens that shipped. Select one to see it larger.

{{cur.t}} {{cur.n}} of {{count}} Open full size
{{cur.t}}
More from before space Home