Skip to main content
Gervin Kahunde
Research

Tuition Calculator Post-Launch Research

After a redesigned tuition calculator launched, many visitors left before seeing a cost — and the analytics alone couldn't explain why.

  • Research and analytics
  • Strategy

AI-assisted step Human decision Observed friction

Summary

Challenge
Explain where and why prospective students abandon a redesigned tuition calculator before they see a cost.
Role
I review and log session recordings, look for patterns, and draft the trends and recommendations, with guidance from the research lead.
Team
Research lead, the calculator's product owner, and the analytics team
Duration
Fall 2026, in progress
Methods
Session-recording review, behavior logging, analytics and survey synthesis, recommendations
Outcome
Findings and recommendations are being finalized for the product team.

Context and problem

Cost is one of the biggest factors in whether someone applies. The redesigned calculator added clearer full-time and part-time choices, a time estimate, and a savings field — but after launch, analytics showed many visitors leaving before reaching a result.

Analytics could say where people left. The research needed to show why.

My role and collaborators

What I own
Reviewing session recordings on mobile and desktop, logging what users actually do, identifying trends, and drafting the product recommendations.
What the team owns
The research lead set up the study and guides it; the analytics team supplied the quantitative report; the calculator's product owner decides what changes ship.

Method

  • Watched recordings from both mobile and desktop users, skipping sessions too short to be meaningful.
  • Logged observable actions only — fields completed, steps skipped, where the session ended — never guesses about what someone was thinking.
  • Used the analytics report to target specific groups: people who abandoned mid-form, people who reached a result, and people who went on to request information or apply.
  • Cross-checked patterns against on-page survey responses and earlier research.
  • Helped frame the funnel the analysis follows — calculator visit, request for information, application start, enrollment — including how long after a visit a conversion should count.
  • Paused session review to confirm privacy and consent handling for recordings before continuing.

What we're seeing

  1. Step 01 Friction

    Before the form

    Many visitors leave before starting

  2. Step 02 Friction

    Program choice

    Some can't find or choose a degree type

  3. Step 03 Friction

    Residency

    Uncertainty about which option applies

  4. Step 04 Friction

    Transfer credits

    Entered, then skipped — expecting the cost to change

  5. Step 05

    Cost results

    Out-of-state costs cause confusion

  6. Step 06 Friction

    Next step

    The call to action sits only at the bottom

Where friction repeats, described qualitatively.

Survey responses add context: many people find tuition higher than they expected, and some visitors — such as non-degree students — aren't served by the calculator at all.

Directions for recommendations

  • Show a cost sooner, so people get value before a multi-step form asks more of them.
  • Explain how transfer credits affect cost — or don't — at the point people enter them.
  • Make program selection more forgiving and clarify residency in plain language.
  • Move the next-step action higher, so people who are ready can act.
  • Point visitors the calculator doesn't serve to the right place.

What I learned

Logging only what people do keeps research honest. Some of the most important patterns — like entering transfer credits and immediately skipping — only make sense when you watch them happen.

What's next

  • Finish the session review and present findings and recommendations to the product team.
  • Re-check behavior after the next round of changes ships.

Continue with work that uses a similar approach.

In use

Admissions Application Monitoring

A high-traffic admissions application produces plenty of data. A weekly usability report turns it into decisions, so problems surface from evidence instead of support calls.

My role
Ran the team's weekly usability-monitoring report for a couple of months, plus QA and dashboard work
Evidence
Kept the weekly report on schedule, combining funnel, error, help and session-recording evidence, and added clear week-over-week changes. It now runs with a teammate.
  • Product Operations
  • Research and analytics
  • Quality assurance