01
The install as an event
First open versus download versus registration. What finance usually means when they say “install,” and what the MMP actually stored.
Flagship program
Install Attribution Intensive
Twelve weeks of campaign-to-install attribution for people who already live inside app analytics. Evening sittings in Bangkok, remote dial-in allowed. Informational fee: THB 42,000 per Measurement Desk seat.
Modules
Sequence, not a buffet
02
Click IDs and residue
Deterministic handshake versus probabilistic leftover. When a device graph is a clue and when it is theatre.
03
MMP configuration without the sales deck
Lookbacks, partner mappings, deduplication, and reinstall rules. You change one knob and redraw ownership.
04
SKAN windows
Postbacks, coarse conversion values, and writing an iOS claim that does not invent users.
05
View-through in regional inventory
Video networks common to Thailand and nearby markets. When an impression may share credit and when it must not.
06
Fraud fingerprints
Click flooding, farms, and hijack patterns. How they distort campaign-to-install tables before anyone notices CPI.
07
Incrementality as a corrective
Holdouts and geo splits that can overrule an MMP. Also: experiments too thin to cite.
08
The install graph for finance
Capstone. One annotated window, assumptions named, unknown installs left unknown.
Outcomes
What you should be able to do
Rebuild a window
Take a redacted export and walk an install back to click, impression, or an explicit unknown — without using last-click as a silent default.
Interrogate a setting
Explain how a lookback or deduplication rule moved ownership, in language a controller can mark.
Write a SKAN claim
Separate what a postback supports from what it merely fails to contradict.
Refuse a screenshot
Spot when an agency or vendor summary cannot be reconstructed from the underlying rows.
Instructor
Nalinee Prasert
Nalinee spent seven years on MMP solution desks covering Thailand and Vietnam before opening Vault Nodecore. She teaches configuration as a set of claims, not as product training. Students meet her on Tuesday evenings in Bangkok; remote desks join the same room.
She does not consult on media buying. If your question is “which bid to raise,” she will send you back to your agency.
Informational pricing
Measurement Desk seat: THB 42,000. Field Ledger notes only: THB 9,800. Attribution Residency (intensive plus private graph clinic): THB 78,500. There is no checkout on this site. Confirm a seat through the desk. See fees and refunds.
FAQ
Before you enquire
Do you teach paid media buying?
No. We do not teach creative testing, bid strategy, or how to lower CPI. If cheaper traffic is the goal, this intensive will disappoint. We only cover how installs are attributed after a click or impression exists.
Must I sit in Bangkok?
No. Two evening cohorts a year are hosted in Phra Nakhon with a remote dial-in. Recordings go to Measurement Desk seats. Field Ledger is notes-only.
Which MMP do you require?
None. Bring whatever export you already have, redacted. The intensive is vendor-neutral. We will not ask you to switch tools as a condition of passing.
What if my catalog is almost all Android?
That is common in this region. The SKAN weeks will be thinner for you; the click-ID, fraud, and incrementality weeks will not. Say so on your enquiry so we can set expectations.
From this intensive
How seats described it
Week three on lookbacks changed how I write the Monday deck. I still dislike presenting unknown installs, but at least they are labelled now instead of being dumped into brand.
8/10 · internal LMS
Nalinee is dry. The capstone rubric is stricter than our agency’s. I needed two resubmits on the finance note.
Short note: the fraud block paid for itself when we found click flooding on a network we had trusted for a year.