Voices

What people said after the sitting

These notes refer to programmes that exist on this site. We keep the milder complaints in, because a room that only collects praise is not a room you can trust.

The Event Cartography week on property hygiene was the piece I still use. We deleted fourteen souvenir properties from the checkout object. Engineering was annoyed for a sprint and grateful after that.

Week four of Chart Table ran hotter than I wanted. I skipped two optional studios because of a launch, and the taxonomy review came back thinner for it. The naming rules still sit above my desk.

Rated 8 / 10 on a private alumni card

Clear on what App Analytics can and cannot answer. Less useful if you hoped for warehouse modelling; they say so up front, which I wish more studios did.

I came for Experiment Table expecting a statistics primer. I got a method for walking a sceptical director through a non-result without either of us performing confidence.

Retention Drawing made me redraw the first-week curve by acquisition surface. The “loyal” segment was mostly people returning from a single email. Uncomfortable, and correct.

Night Watch on our onboarding map took longer than the brochure implied, because our event names were already a palimpsest. The written readout was the artefact our board actually read.

Case notes

Two products, mapped in public enough to learn from

Product team on a video call reviewing work

A fintech onboarding that looked healthy

A United Kingdom lending app arrived with a three-step funnel that “converted.” Event Cartography split the second step into document-upload versus affordability-check. The drop was not in UX copy; it was in a retry event firing as success. After the sitting, the team kept the new names and retired the old funnel from the Monday pack.

Limitation they accepted: we did not rebuild their warehouse models. The product analytics layer was the only surface in scope.

Learner looking at a screen during a study session

A language app that could not see returning teachers

Retention Drawing for an education product showed a smooth curve until we cohorted by role. Teachers were collapsing into the same user_id path as students after a class-code join. The “win” was an identity stitch, not a curriculum change. The founder still uses the role property as a load-bearing dimension.