Flagship programme
Predictive CLV for Commerce Teams
Eight live sessions in English. You finish with a two-page valuation note on a population you can name, a window you can defend, and a limitation you are willing to say out loud.
Modules
- Population, not vibes. Who counts as a customer, which orders enter, how gift cards and employee purchases leave.
- Cohort grammar. Acquisition month versus first category, first discount depth, and first fulfilment promise.
- RFM as a leftover, not a religion. When recency still ranks and when discounting has already poisoned it.
- Buy-till-you-die assumptions. What the classical models expect, and where Thai holiday sparsity breaks them.
- Survival without a PhD. Reading a curve, naming censoring, refusing to extrapolate past the window you have.
- Margin overlay. Contribution, returns, COD failure, payment fees — on the same sheet as repeat probability.
- Channel humility. Paid, organic, and marketplace fragments; what you may not claim.
- The sentence finance will sign. A memo with ranking, exclusions, and the assumption that flips the top two segments.
What you should be able to do
- Write a population rule a warehouse analyst and a CFO both recognise.
- Choose a cohort split that matches a decision, not a chart style.
- State whether your extract can support a predictive claim or only a historical ranking.
- Put contribution next to repeat so a “high CLV” segment cannot hide unprofitable replenishment.
- Present a two-page note without a single unexplained acronym.
Instructor
Arisa Thongchai
Programme lead. Former CRM analytics lead at a Bangkok multi-brand retailer; now faculty at Metric Cloudhub. She still argues, politely, with anyone who brings a single blended lifetime number to Lat Phrao.
Visiting sessions on survival notation are occasionally taught with Luca Moretti, independent statistician.
Informational fees
Published so finance can budget. Payment is not taken on this website. Invoices follow a confirmed seat via the desk.
- Cohort Seat — ฿42,000: eight live sessions, peer critique, one faculty review of your memo, Reading Room access for twelve months.
- Practice Bench — ฿67,500: Cohort Seat plus four private clinics on your extract.
- Bangkok Lifetime Value Studio weekends are optional; Cohort Seat students are not charged extra tuition for studio attendance.
Questions we actually get
Do I need SQL or Python?
No. Spreadsheets are enough if you can join orders to a customer key. If your identifiers are broken, the programme will expose that early — which is useful and not glamorous.
Is this suitable for a brand with only nine months of data?
You can still learn the grammar. You cannot honestly publish a long-horizon predictive CLV. We will make you write that limitation into the memo. That is a real constraint, not a marketing footnote: sparse windows produce rankings, not forecasts you should take to a lender.
What language is the classroom?
English, with Thai retail examples. Faculty can clarify terms in Thai in office hours, but materials are not bilingual.
Can three people share one seat?
No. A seat is one person, one memo. A colleague may sit in on Practice Bench clinics if named on the invoice.
From people who sat this block
Module four on buy-till-you-die assumptions was the first time I understood why our beauty replenishment looked fine until month four. I still dislike how much manual cleaning the extract needed.
★★★★☆Arisa will stop a pretty chart mid-sentence. The memo format is now our internal template. Live slot timing is the weak point if you close a store.
Not a software tour. The channel humility session saved us from claiming marketplace “new buyers” we cannot actually see across storefronts.