The setting

Cadence is a habit and workout tracking app on a monthly and annual subscription. You are the growth marketer, owning the path from install to paid retention, working alongside a small product team.

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. funnel auditstarter

Audit Cadence's install-to-subscribe funnel for March

Cadence spends more on installs every month, but paid subscribers are not growing to match. Before anyone argues about ad spend, you need to know where in the funnel users actually fall out. The team keeps guessing it is the paywall, but nobody has looked at the whole path.

You deliver: A funnel audit with step-by-step conversion rates, the worst leak identified by the numbers, and a verdict on whether the paywall is the real issue.

Scored on: Step-by-step math, Names the worst leak, Benchmark comparison, Paywall verdict.

Working from: funnel_steps.csv, funnel_notes.md.

2. activation ideastarter

Propose an activation fix for the Cadence onboarding drop-off

The funnel audit pointed at onboarding as the biggest leak. Now the team needs to know which onboarding screen loses people and what to do about it. Guessing wastes a sprint, so you are going to pin it to the screen-level data and what users are saying.

You deliver: An activation brief naming the drop-off screen, the supporting feedback, one specific fix, and its predicted effect on completion.

Scored on: Names the screen, Links the feedback, Specific fix, Predicted effect.

Working from: onboarding_events.csv, user_feedback.md.

3. experiment designcore

Design the A/B test for Cadence's new paywall timing

The team wants to move the paywall from first app open to just after the user logs their first workout, betting that people pay more readily once they feel value. Before shipping it to everyone, you need a clean experiment so the result is trustworthy and does not quietly wreck retention.

You deliver: An A/B test design specifying variants, primary metric, guardrail, sample size or runtime, and the analysis rules.

Scored on: Clean variants, Primary metric, Guardrail metric, Sizing and rules.

Working from: paywall_baseline.csv, hypothesis_notes.md.

4. retention loopcore

Design a retention loop to fix Cadence's week-two churn

New users are dropping off hard in the second week. The subscription math only works if people stick around, so closing the week-two gap matters more than any new acquisition channel right now. You have cohort retention and the reasons people give when they leave.

You deliver: A retention loop design targeting the week-two cliff, with a named trigger, action, variable reward, and reinvestment step.

Scored on: Finds the cliff, Top churn reasons, Complete loop, Targets the window.

Working from: retention_cohorts.csv, churn_reasons.md.

5. growth memostretch

Write the quarterly growth memo for Cadence leadership

Leadership sees installs and revenue going up and assumes growth is healthy. You can see that the growth is bought, not earned: acquisition costs are climbing while activation and conversion stall. The quarterly memo is your chance to reset the strategy toward efficiency before the burn gets worse.

You deliver: A quarterly growth memo with the honest read on efficiency and exactly three ranked, evidence-backed bets for next quarter.

Scored on: Real trend, not top-line, Quantified efficiency decline, Three ranked bets, Ties to prior findings.

Working from: growth_metrics.csv, growth_context.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 growth marketing" something an employer can check.