Listing to Loyalty Studio
Trace a shopper from search through PDP, cart, and the second order — then name the three leaks that actually matter.
Read the syllabus
Bangkok studio · two-sided apps
The quiet metrics that decide who wins a marketplace.
GMV is a loud number. Search-to-PDP drop, seller reply latency, and repeat-order windows are quieter — and they usually explain why a marketplace app stalls. Mirror Cloudcore trains teams to read those signals without drowning in dashboards.
Open this season’s programmes
18
seats in each Listing to Loyalty Studio — small enough that every funnel map gets a critique.
11cohorts since 2019
37marketplace categories studied
1,640alumni across product & ops
71%still using the same event dictionary after 90 days
Programmes on the floor this quarter
Short studios, not slide libraries. Each one is built around Marketplace App Analytics work you can take back to a live catalogue.
Search Ranking Diagnostics
Learn to separate relevance bugs from inventory gaps when in-app search conversion slips.
See all programmes
Seller Take-Rate Lab
Model how commission, ads, and fulfilment fees change seller quality — and what that does to buyer trust.
Enquire about a seat
How the room works
We critique live funnels, not hypothetical start-ups.
Most marketplace teams already have a warehouse. What they lack is a shared language for two-sided behaviour: a seller who never answers chat, a listing that ranks but never converts, a promo that buys GMV and burns repeat rate.
Sessions in Bangkok (and remote seats for the same cohort) mix short lectures with workbook critiques. You bring a sanitised extract or a public marketplace case. The group names the metric, the counter-metric, and the decision it should unlock.
How the studio is runNotes from people who sat the work
These refer to modules that actually exist on this site — not a generic “bootcamp.”
“The search-to-PDP map in Listing to Loyalty forced us to stop celebrating ranking gains that never produced add-to-cart. Painful, and overdue.”
“Seller Take-Rate Lab was slower than I wanted in week one because our fee table was a mess. Once we cleaned it, the buyer-trust discussion finally had numbers attached.”
What teams leave with
Capabilities, not a new tool vendor.
An event dictionary that two sides can share
Buyer search, listing quality, chat latency, and fulfilment states named so product and ops stop arguing about definitions.
A leak ranking, not a 40-slide funnel
You leave with three leaks worth fixing this quarter, each tied to a counter-metric so a “win” cannot hide churn.
Experiment hygiene for two-sided apps
Why a listing A/B test contaminates seller behaviour, and how to fence cohorts before you read a p-value.
Seasonal reading for Thailand calendars
Songkran, mid-year sales, and payday windows treated as structural, not noise to be smoothed away.
From the journal
Field notes on Marketplace App Analytics — written for operators, not for keyword theatre.
Why GMV hides marketplace churn
A rising checkout total can still mean fewer loyal buyers. Here is the split we ask every cohort to plot first.
Read the essay
Reading search-to-PDP drop-off
When in-app search looks healthy and product pages do not, the leak is usually one of four places.
Read the methodBring a live catalogue. Leave with a leak ranking.
If your marketplace app already has traffic and still feels opaque, write to the studio. We will tell you honestly whether a programme fits or whether you only need a diagnostic conversation.
Write to the studio