Reviews
Alumni are asked to name the module they used and to mention friction if they felt it. The notes below are edited only for length and for product names the companies asked us to withhold.
★★★★☆
The week-one spreadsheet work felt slow compared with jumping into Mixpanel, and I still think that onboarding could be tighter. Once we reached confidence intervals on drop-off, the Studio finally matched what our squad needed — especially the PM readout template from module five.
Nalinee Preecha, product analyst · BangkokActivation Path Diagnostics did one useful, slightly uncomfortable thing: it showed that our “first save” event was firing on tap, not on success. We had been celebrating a funnel that engineering already knew was wrong. Two afternoons. Enough.
Marek · Chiang Mai4.0 · alumni board
Taxonomy week was the part I expected to skip. I could not. Our Thai and English event names had drifted far enough that “bookmark” and “save_item” were splitting one action. Cohort Retention Mapping later made more sense because the feature funnel finally had a single start event.
Anonymous client in travel super-appsI wanted vendor certification. This is not that. If you already know your tool and cannot explain a feature to a designer, go. If you need a badge for LinkedIn, look elsewhere.
Somsak Wattana, growth PM · Feature Funnel Reporting StudioOffice hours saved the Research Partnership. Our schema was messier than the intake form admitted. Araya refused to build a vanity funnel over broken events and made us rewrite the stage contract first. That delay annoyed the CPO for a week and then paid for itself.
Ploy, analytics lead · regional fintechA consumer lending squad arrived with a single “activation funnel” that mixed KYC completion, first transfer, and first bill-pay. Leadership saw a steep drop between step two and step three and had already commissioned a redesign of the transfer screen.
Inside the Feature Funnel Reporting Studio they wrote three feature contracts. KYC success was defined as an approved identity check, not a submitted form. First transfer required a settled transaction, not a confirmation tap. Bill-pay was treated as a separate feature with its own start population: users who had already funded a wallet.
After rebuild, KYC drop-off sat inside a range that no longer justified a redesign. Bill-pay still leaked, mostly between “biller selected” and “payment authorised,” which pointed at a list-sorting problem rather than a trust problem. The squad cancelled the transfer redesign and moved a designer onto biller search. They still use the two-page readout format for monthly reviews.
A grocery app’s “favourite store” feature looked healthy in a dashboard and unused in support tickets. Activation Path Diagnostics asked them to define start as “user opened store sheet with intent to favourite,” which they could not observe. They only had a tap on a heart icon.
The clinic outcome was not a prettier funnel. It was an instrumentation ticket: a visible start event and a success event that waited for the server ack. Two sprints later the funnel existed. Adoption was lower than the old heart-tap chart had implied. The PM stopped promising “favourite-led retention” in the quarterly plan.
Neither company paid for publication. Both reviewed these summaries before we posted them.