Reviews

What people said after the homework, not after the welcome slide.

Every note below names a Proxy Routecore module. Length and tone vary on purpose. One reservation is not an accident.

“Week four of Cohort Telemetry Studio — the one where you rebuild D7 without leaning on D1 — is the only analytics class that has ever made our publisher call shorter. We still argued about device-class cuts, but at least we argued with the same definitions.”

Kittipong R. · product lead, mid-core RPG, Bangkok

★★★★☆

Soft-Launch Instrumentation was tight. The Thailand store snapshot template saved us a week. Deducting a star because the sample BigQuery views assumed more SQL than the syllabus admitted up front.

Platform-style note · verified alumni tag

“I came in for the Live-Ops KPI Clinic. We left with a written contract to stop reporting session length as a north star. Uncomfortable. Useful.”

An · Chiang Mai

Client in casual puzzle publishing: the Monetization Funnel Reading track did not magically lift ARPU. It did show that our first-purchase prompt was firing on a tutorial retry event. That was enough.

Case study 01

Idle title, six people, no analyst on staff

A Bangkok idle studio joined the Live-Ops Residency after their D3 looked “fine” in a blended chart and terrible once cut by install week. Mentor hours focused on the Session Ledger homework and a single offer-conversion event that had been duplicated under three names.

They did not hire a data team afterwards. They did freeze a 19-event dictionary and stop shipping unnamed tutorial steps. Retention on the post-rewrite install weeks sat 0.7–1.1 points higher at D7 — modest, and they still say the tutorial copy needs another pass.

Attribution: studio ops lead, first name only, at their request.

Workshop table with notes and laptops

Case study 02

People gathered around a bright meeting table

Publisher-facing readout after a Thailand cell

A remote student on Player Journey Mapping Lab brought a SEA soft-launch that mixed TH, VN, and PH installs in one cohort. The lab forced a store split before any narrative about “SEA performance.”

The Thailand cell showed a healthier tutorial complete rate and a worse first-purchase conversion — the opposite of the blended story they had been sending upstream. They used the Telemetry Desk checklist (see Telemetry Desk) to rewrite the weekly note. No claim that UA spend changed overnight; the win was an honest slide.

See current programs