Flagship · Attribution Bench

Hybrid App Measurement Strategy

Nineteen sitting days. A writable sandbox. A measurement contract you defend in week eight. English instruction, Asia/Bangkok hours, with optional Thai office hours for PDPA questions.

Team working with multiple screens in a dark studio

Who this is for

Product analysts, mobile engineers, and growth partners who already ship a hybrid app — React Native or Flutter shell, WebView checkout, mini-app, or an older Cordova remnant — and who are tired of reconciling two event streams. You should be comfortable reading JSON. You do not need to have written an SDK.

Modules

1. Ownership tables

Map every board-facing metric to a runtime. Decide who may fire purchase, who may fire screen_view, and what becomes diagnostic-only.

2. Taxonomy versioning

Prefixes, deprecation windows, and the rule that “temporary” events are still events. You edit a living dictionary, not a spreadsheet screenshot.

3. Consent before init

PDPA timing for Thailand, ATT prompts, and how a WebView that loads later inherits (or violates) the native consent artefact.

4. The JS bridge, narrowly

Pass consent state and a signed runtime flag. Stop treating the bridge as a user-id pipe. Lab includes a deliberately leaky token.

5. Attribution windows

Write the lookback in the contract. Handle missing IDFA/GAID. Refuse to score campaigns when the WebView drops the referrer.

6. Funnels across runtimes

Onboarding in native, payment in HTML. Join keys that survive a process death. Funnel QA that does not “fix” drop-off by blending events.

7. Finance extract

What leaves the warehouse. Duplicate purchases, ghost sessions, and the sentence you will say when a number gets smaller.

8. Defence

Present the contract to faculty playing a sceptical finance partner. Anything you cannot defend is removed before you leave.

Learning outcomes

Portrait of instructor Kanya Wiriyachai

Instructor

Kanya Wiriyachai

Former measurement lead on a Bangkok super-app wallet. Kanya spent 2016–2019 reconciling native Firebase with a WebView pixel that finance refused to retire. She designed the sandbox used in this flagship and still reviews every measurement-contract defence.

Informational pricing

A cohort seat (Attribution Bench) is ฿28,700, excluding VAT where applicable. Signal Sketch recordings of this syllabus are ฿9,400 with one office hour. Private teams use Measurement Residence at ฿86,000. Logicqueuebase does not collect payment on this page. See studio retainers and the refund policy.

FAQ

Do I need a native engineer on the call?

One person who can change the shell helps, especially for the JS-bridge lab. Analysts can complete the ownership table and consent work alone; the sandbox still lets you observe native logs.

Is this useful if we only wrap a website?

Yes. Thin wrappers are the most common enrolment. You will spend extra time on page_view versus screen_view and less time on native gesture events.

What will this course not cover?

We do not teach gaming anti-cheat analytics, ad-network bidding, or creative testing. If your primary job is UA buying, the flagship will feel incomplete — start with a conversation, not a seat.

Language and timezone?

Sitting days are in English. Office hours follow Asia/Bangkok. Thai-language PDPA office hours are available on request.

From people who sat this flagship

Week four’s leaky token was uncomfortably close to our production bridge. We rotated the session scheme before the next store release.

Verified learner · 2025-B

The JS-bridge lab assumed more native access than our vendor SDK allowed, so I had to rewrite two exercises against a wrapper we did not control. Taxonomy week still changed how we name checkout events in the WebView.

Mira S., product analytics, insurance — Bangkok

Defence day was more useful than the lectures. Our CFO proxy asked why install-to-purchase fell. I had the sentence ready: WebView ghosts left the extract. That is now in the board appendix.

Client in consumer banking

Ask about the next cohort