A Trackable Parent Dashboard

Before the dashboard, every session ended in a raw transcript: pages of back-and-forth dialogue with no structure, no learning framework, and no signal of what was mastered or still shaky. To answer “is this working?”, a parent had to read the whole thing — so most just guessed.

To find where trust breaks down, we triangulated across theory, business, and real users — a literature review of learning-science frameworks, internal expert interviews at Varsity Tutors, a survey of 104 parents, and 1:1 in-depth interviews — to understand what parents care about, what confuses them, and where they feel uncertain.

What I owned

• Synthesized survey + interview data into the framework below, with the team, via affinity diagramming
• Owned the data strategy — how a raw session transcript maps to skills, mastery, and learning-science principles
• Designed the MVP: homepage, session reports, and an AI-driven skill-mastery breakdown
• Ran 1:1 think-aloud testing and made the hard calls that turned a confusing v1 into an actionable v2

Jobs to be done

What parents hire the product to do

Synthesis pointed to three jobs behind the renewal decision — and the product was failing all three.

Journey map

Where the journey breaks down

We mapped the full journey from search to renewal and found four pain points — and the biggest gap lands exactly where it hurts most: during and after sessions, when parents most want to understand learning but the system gives the least clarity.

Getting from problem to product meant four decisions where the obvious answer wasn’t the right one — spanning the co-creation workshop through the final design.

We tested the first MVP with four parents of students in grades 5–8, think-aloud. Three issues came up every time: parents didn't know where to start; the information felt scattered, abstract, or too technical; and the biggest one — “I understand it, but I still don't know what to do next.” The problem wasn't the data. It was how the data was structured and how action was guided. That reframed the MVP around three jobs:

Explorations

Five to six structures, compared

Before landing here, we explored five to six information structures — session-centered, skill-centered, and emotion-signal-first — and compared them on clarity, focus, and how well each guided action. The winner mirrored how parents actually make decisions.

← Swipe to browse →

The homepage's job is to guide, not to show everything at once. A personalized opening message, clear entry points to the session view and overall progress, and quick access to history and schedule. Earlier versions packed in more data — and just raised cognitive load.

The goal: let a parent understand how a session went in about 30 seconds. Three high-level metrics summarize performance, with hover explanations that clarify the data logic without interrupting the read. Below, skill highlights — and a cross-session view and at-home cards that came straight from testing.

If Session Overview answers 'what happened this session,' Skill Breakdown answers 'where does my child stand across the subject?' An AI agent breaks down each practice, scores it on accuracy and repetition, and sorts skills into levels parents can read.

Testing validated the shift from more data to the right data — structured around the questions parents ask, and guided toward action. The MVP (homepage, session reports, and skill breakdown) is now in Varsity Tutors' development pipeline. Given more time, the next steps are: a human-in-the-loop to validate and improve AI accuracy; expanding beyond Algebra to more subjects; and larger-scale testing with real Varsity users to measure long-term value.