01 · Choose a surface that can fail
We reject “the whole homepage.” You bound a feature with an entry point, a successful first action, and a repeatable job. If the surface is too vague, the rest of the course cannot start.
Flagship programme
Adoption Signals
Twelve weeks on measuring feature adoption inside a live app. You pick one surface, instrument it honestly, and defend a number in front of peers who will not let you hide behind activation rate.
What you will be able to do
- Separate exposure, trial, and habitual use for a single feature without inventing new vanity charts.
- Write an event map that still works after a navigation redesign.
- Set an adoption floor before launch and treat a miss as information.
- Build a cohort recipe that accounts for late language packs and delayed payment rails in Thailand.
- Deliver a one-page readout a product council can fund, pause, or retire.
Informational fees
Field Ledger access to recordings and templates: THB 18,400. Cohort Atelier with live critique: THB 41,900. Adoption Stewardship (Atelier plus two instructor reviews of your live readout): THB 63,500. These figures are published for planning only; there is no checkout on this site.
See fee bands for what each includes. Refunds follow the refund page.
Modules
02 · Exposure is not a pageview
Deep links, empty states, feature flags, and Thai/English copy swaps all change who was genuinely shown the feature. You will log those conditions instead of pretending traffic equals opportunity.
03 · First successful use
Trial is a completed job, not a tap. The module covers partial completions, retries after failed payments, and why “opened the screen” is almost never trial.
04 · Habit windows
You pick a return rule that matches the job: four distinct days in 28 for daily tools, two billing cycles for monthly ones. We critique windows that are too kind.
05 · Adoption floor and confounders
Before you look at the number, you write the floor and the known confounders: seasonality, a competing promo, a store outage. Looking first is how vanity metrics get born.
06 · Council readout
One page. Floor, evidence, what you still cannot see, and a decision request. Peers try to punch holes in it. That is the exam.
Instructor
Nalinee Prasert
Nalinee led measurement for a Bangkok super-app’s payments surface before joining Devtracebase. She still reviews live readouts for Stewardship students and refuses charts that cannot name an exposure denominator.
Guest critics from Thai banks and marketplace teams join two evenings per intake.
Student notes on this syllabus
“Module 02 made me throw out three ‘impression’ events that were firing on a cached Thai homepage nobody had scrolled.”
Ploy, product manager in Phuket, wrote that the habit window critique was “slower than I wanted, and I am glad it was.” An anonymous client in ride-hailing said the council page format is now their internal template, with one edit: they added a cost line we do not teach.
Questions we actually get
Do I need SQL?
You need to query something: warehouse SQL, a product analytics UI, or an event debugger. We do not teach SQL from zero. If you cannot yet pull a count by day, take a query course first.
Can I use a toy app?
No. The flagship requires a live feature you are allowed to inspect. Sandbox apps hide the messy flags and copy swaps that make adoption measurement hard.
What this course will not do
It will not teach paid acquisition attribution, store listing experiments, or how to raise install volume. If your problem is “not enough new users,” this room will feel like the wrong address, because it is.
Is the Bangkok evening seat mandatory?
No. Remote students join the same critique calendar. Studio visits at Asoke are optional office hours, not a separate curriculum.