The partner-priority lecture saved us from double-counting LINE ads against our MMP. I still argue with creative about view-through, but at least the numbers now share a window.
Pimchanok T. · UA lead, Bangkok e-commerce
Bangkok · measurement school
We teach growth teams how a tap becomes a number they can defend.
Model Routehub exists for people who sit between paid user acquisition, product analytics, and finance. The work is Mobile Attribution Analytics: identity graphs, MMP configuration, privacy-safe postbacks, and the quiet art of reconciling three dashboards that refuse to match.
Open the flagship cohort outline
2,180practitioners through our labs since 2019
37MMP and ad-network setups reviewed in class
71%of alumni still use the reconciliation workbook six months later
Flagship
Attribution Measurement Lab
An 11-week working studio, not a slide dump. Each week you bring a real campaign from your stack — or a sanitised sample we provide — and rebuild the credit path from click to recognised revenue. SKAN, Android Advertising ID realities in Southeast Asia, fraud signals, and incrementality sit beside the unglamorous work of event taxonomies.
Seats are priced as Signal Desk, Route Studio, or Full Corridor. None of those include a checkout here; write to the desk if a cohort date fits.
How the room works
We mark the argument, not the screenshot.
A pretty MMP export is not a passing grade. Instructors ask where the conversion value was mapped, which partner was last-click versus view-through, and whether finance would accept the same window. Students leave with annotated workbooks and a vocabulary that survives a sceptical CFO.
Cohorts stay small on purpose. Bangkok evenings for the in-person room; a single remote seat per table so the discussion does not splinter into chat noise.
What you leave with
Skills that show up on Monday morning
- A defensible event map Name events once, version them, and stop the product team from silently renaming “purchase_complete” mid-quarter.
- MMP hygiene without folklore Install referrers, reattribution windows, and partner priority lists written down, not remembered by whoever last spoke to the account manager.
- Privacy-safe reading SKAN postbacks and coarse conversion values interpreted as ranges, not fake precision. Android realities for Thailand traffic included.
- Incrementality as a design problem Geo holdouts and PSA-style tests scoped to the budget you actually have, including the honest case where the test is too small to run.
From the room
Notes from people who sat the lab
Wan’s SKAN labs assumed more iOS volume than our Android-heavy catalogue had, so I skimmed those weeks. The reconciliation spreadsheet is what I open every close.
Client in ride-hailing ops
★★★★☆
Dry in the best way. No motivational posters. I wanted more time on device-farm fraud; the rest of the MMP module was exactly the mess we live in.
Arun · Chiang Mai · 4/5 on the alumni board
Field notes
Recent writing from the desk
What probabilistic matching still does after IDFA
It did not vanish. It got noisier, more jurisdictional, and easier to over-claim.
Reading an MMP discrepancy report without panic
Three columns that almost never agree, and the one question finance actually asks.
Why Thailand UA often needs a different window
LINE, delayed checkout, and prepaid SIMs do not behave like a US last-click textbook.
Ask whether the next Bangkok intake has a seat.
Tell us your MMP, your store mix, and what “attributed revenue” means in your finance file. We will say if the lab is the right room.