Baselines before targets
You build a comparison set from apps that share session shape, not just category labels from an app store. Vanity peers get written out of the set on purpose.
App analytics, taught as a craft
Figures that still hold after Songkran traffic dies down.
Mindtracebase is a small teaching studio for product and growth leads who need Mobile Engagement Benchmarking they can defend in a steering meeting — not a screenshot of a rising DAU line.
See the current studio calendar
2.7×
Week-4 retention spread we document across comparable Thai consumer apps
From the journal
DAU is a crowd count. Engagement is whether anyone came back with intent.
We keep a running note on how festival calendars, prepaid data packs, and late-evening commute windows distort Southeast Asian retention curves. The latest piece walks through a Bangkok grocery app that looked “healthy” until the cohort was cut by acquisition channel.
Read the field note- 61
- Published benchmark tables, each with a written limitation note
- 14
- Bangkok product teams who sat a full studio seat last year
- 22
- Live seats per Engagement Benchmark Studio cohort
- 2018
- Year the desk opened on Vibhavadi-Rangsit
Flagship
Engagement Benchmark Studio
Eight weeks of live critique plus take-home worksheets. You leave with a category baseline, a cohort definition that does not leak, and a one-page narrative your CPO can actually read. We do not pretend the same curve fits a wallet app and a casual game.
Next Bangkok evening cohort opens with Module 03 focused on Thailand and wider SEA seasonality.
Open the studio syllabusHow the desk teaches
We grade the argument, not the chart skin.
Every exercise asks the same question: would this number still be honest if the acquisition mix shifted next Tuesday?
Cohorts you can explain
We drill first-open windows, timezone cuts, and “new user” leakage until the definition survives a sceptical finance partner.
Market weather included
Thai public holidays, campus exam weeks, and prepaid top-up cycles sit inside the teaching examples — not as a footnote after a US-centric template.
A known edge
We say out loud what the studio will not cover: paid user-acquisition attribution, web analytics, and store-listing A/B. Those desks exist elsewhere.
Other programmes
Shorter rooms for a single signal.
Retention Signal Mapping
Isolate which return events actually predict week-4 stay, then drop the ones that only correlate with push volume.
Push Timing & Silent Hours
Build a send calendar that respects Bangkok commute windows and opt-in decay, without treating every tap as consent forever.
Session Quality Lab
Replace “average session length” with a quality score your design lead will recognise from actual task completion.
Voices from seats
What people actually used after class.
“The Engagement Benchmark Studio forced us to throw out a ‘peer set’ of six global super-apps. Once we compared against two Thai lifestyle apps with similar session length, the week-2 drop stopped looking mysterious.”
★★★★☆
Useful, occasionally slow. Module 03 on festival distortion was the piece I still open. I wanted more time on in-app message frequency; we only sketched it.
Invitation
Bring a messy dashboard. Leave with a baseline you can stand beside.
Seats are small on purpose. If the calendar is full we will say so, rather than stretch a critique room past 22 people.