Reviews
We ask alumni for specifics: which module, which metric, what still feels unfinished. Glowing blurbs without a product attached do not get published here.
Activation Pathways forced us to define first-value as “first correct attempt with a hint unused,” not “signed up.” Our day-1 chart looked worse. Tutor complaints about the third screen dropped the following term.
Kannika Prasert, learning designer Vocational English app
8.4 / 10
Dense homework. Waranya does not flatter the dashboard. Would sit Retention Cartography next, not another growth workshop.
Anonymous client in university bridging courses
Freemium Quiet Hours was shorter than I expected. The useful part was moving the paywall to after Unit 1 instead of mid-lesson. Revenue per learner did not jump; refund requests did fall.
Mint Phuket
I still think the academic brief in week eight is too much writing for a product squad. We nearly skipped it. The board read that document and ignored our slide deck, so I was wrong.
Daniel R., analytics manager Regional K–12 publisher
Cleaned our hint versus skip events. That was the whole win. Not romantic, very usable.
First name + city: Bee, Khon Kaen
We came from a general analytics bootcamp. Datafeaturebase assumes you already ship. If you do not have events yet, wait.
Chaiyaporn Wattana, founder Adult coding prep
Case notes
A consumer-style language app counted completion at 90% video progress. Tutors insisted learners were skipping the in-app check. During module 01 the team split video_end from check_submit. Week-2 “completion” fell from a flattering number into the low fifties. The same cohort’s week-4 return, measured only among check-submitters, rose after they moved a hard item out of lesson 3.
They did not hire a warehouse engineer. They exported CSV from the tool they already paid for. The academic partner vetoed two slides that still used the old definition; those slides never went to the board.
A university bridging app showed a steep February drop every year. Growth advisors had recommended push notifications. Cartography overlaid the Thai exam calendar and family travel weeks. Most “churned” learners returned in the teaching week after exams, often on a sibling’s phone.
The team stopped spending on February win-back campaigns and instead instrumented device switches. The reservation they still hold: the lab did not fix their login, which remains fragile on shared Android phones. They know that is product work, not a metric trick.