What people said after they took the worksheets home.
These notes name real modules. They are not ranked. One of them is deliberately mixed — glowing copy helps no one choose a seat.
I came for Cohort Signal Lab because our “day-7 paid” number made marketing look like genius and finance look confused. The introductory-offer distortion module is the one I still quote in stand-up. We now split trial convertors from standard convertors before anyone mentions North Star.
8.1 / 10 · internal alumni survey
Store Receipt Literacy sounds dry. It is. That is the point. I finally stopped calling Play’s one-time products “subscriptions with a weird name.”
Client in travel super-apps: Refund Forensics Workshop did not magically lower our refund rate. It did stop us from shipping a third paywall variant the same week a billing outage hit Thailand. That alone paid for the seat.
Attribution Atelier was slower than our agency wanted. The facilitators refused to declare a winner between MMP A and MMP B until we wrote down the actual click-to-receipt windows. Frustrating in week two. Correct in week six.
Case note · utilities
When “annual attach” was mostly a grace-period illusion
A three-person monetization pair from a Thai utilities app joined Ledger Studio after their annual plan attach rate jumped following a paywall rewrite. The Lab assignment asked them to rebuild the cohort with billing retry and grace included.
The attach rate halved. The rewrite had coincidenced with a store-side retry change, not with copy. They kept the new paywall anyway — it read clearer — but stopped celebrating the number. Their metric contract now shows attach net of grace, with a footnote finance initialled.
Case note · education
SKU Experiment Studio and a price ladder that failed on purpose
An independent language-learning publisher used SKU Experiment Studio to test a mid-tier monthly next to an existing annual. Store review delayed the mid-tier by nineteen days in one country and three days in another.
The studio’s isolation checklist had already flagged staggered approval as a kill criterion. They paused the read, rather than “learning” from a contaminated mix. The mild reservation they wrote in the alumni thread: the homework still assumed a single review queue, which Play did not honour that month. We updated the worksheet.
