The setting

Cadence is a habit tracking app with about 120,000 monthly active users. You are the product analyst embedded with the growth team, reporting into the head of product, Miriam.

Every mission on this path happens at the same company, so context carries over the way it does in a real job: the data you cleaned in mission two is the data the finance lead questions in mission four.

The missions

1. metric treestarter

Build Cadence's metric tree around the real north star

Every leadership deck opens with total downloads, and downloads keep rising while the business feels flat. Miriam thinks the team is steering by a vanity number and wants a metric tree that connects a real north star to the inputs the team can actually move.

You deliver: A metric tree with a defined north star and its input metrics.

Scored on: North star choice, Correct decomposition, Demotes the vanity metric, Actionable inputs.

Working from: business_context.md, metric_definitions.md.

2. funnel analysisstarter

Find where Cadence loses new users in onboarding

Activation is soft and nobody can say which onboarding step leaks the worst. The growth team keeps guessing. Miriam wants the leak found and quantified before the next sprint plan.

You deliver: A funnel analysis identifying the worst drop-off and a first fix.

Scored on: Correct conversions, Finds the worst drop, Platform split, Actionable fix.

Working from: onboarding_funnel.csv, onboarding_flow.md.

3. retentioncore

Read Cadence's cohort retention and explain the February lift

The monthly review is coming and the retention chart jumped for February cohorts. Miriam wants to know if the lift is real and what caused it before she puts it in front of the board.

You deliver: A retention analysis explaining the curve shape and the February change.

Scored on: Curve described, Quantifies the lift, Links to release, Honest caveats.

Working from: cohort_retention.csv, release_notes.md.

4. experiment readoutcore

Read out the streak-animation experiment before the ship decision

The growth team wants to ship a new streak animation based on a one-page summary that says plus 26 percent engagement. Before the ship meeting, Miriam asks you to check the readout, because shipping on a bad number wastes the next sprint.

You deliver: An experiment readout with a defensible ship decision.

Scored on: Reads the full window, Catches the novelty effect, Checks retention, Sound decision.

Working from: experiment_brief.md, experiment_weekly.csv.

5. decision memostretch

Write the reminders investment decision memo for Cadence

Smart reminders showed a retention lift, but engineering is stretched. Miriam has to decide whether to spend a quarter on a reminders v2 or move on, and she wants a decision memo that pulls the threads together rather than a pile of charts.

You deliver: A one-page decision memo with a clear recommendation and its tradeoffs.

Scored on: Recommendation up front, Evidence grounded, Weighs opportunity cost, Measures and checkpoint.

Working from: cohort_retention.csv, reminders_costs.md, exec_question.md.

How the scoring works

Each deliverable is graded against the rubric written for that mission. Separately, every mission on every path is graded on how you used AI, against the same four criteria:

  • Understood the task. The learner framed the goal for the assistant clearly instead of pasting the brief and hoping.
  • Grounded in the material. The learner directed the assistant into the provided files and based the work on them, not on invented facts.
  • Verified the output. The learner checked claims, numbers, or coverage against the source material before submitting.
  • Iterated with judgment. The learner refined weak parts of the draft with specific follow-ups rather than accepting the first answer.

Both scores, with the work behind them, go on your proof profile. That is what makes a claim like "I can use AI for product analytics" something an employer can check.