Journal · 08 Jan 2026
Reading uninstall waves in Southeast Asian markets
After Songkran, after Tet, after payday weekends, product channels fill with uninstall charts. The default story is “seasonal.” Sometimes that is true. Often the wave is a specific acquisition surface that never emitted the events your home feed expects, so the new install looks empty to every downstream segment and to the human who opened it.
Funnel Fracture Clinic grew out of this pattern. A regional retail partner treated a post-festival uninstall spike as weather. The concentration sat in users whose first session skipped category browse because a campaign lander jumped straight to a coupon sheet. Organic home users still browsed. The app was two products on day zero.
Signatures we look for
Dual SIM devices that appear as two advertising ids and one account. LINE-origin landers that suppress the first open event. Cash-on-delivery users who install, browse, and leave without an account because the campaign promised a guest checkout the app no longer supports. None of these are mysterious if you keep lander id on the first-session event.
Storage pressure on lower-end Android also creates genuine uninstall waves that have nothing to do with product love. Those waves correlate with device RAM more than with feature flags. Mixing them into a “churn segment” for win-back push is how you annoy people who deleted you to install a banking app their employer required.
What not to do
Do not answer an uninstall wave with a reinstall promo until you know which first-session events were missing. Promo spend on a broken lander teaches the ads system to buy more of the same emptiness. Fix the event parity between lander and home, then decide if the campaign creative still deserves budget.
A longer outline of the retail case sits on our reviews page. The lab will not take synthetic uninstall data. If you cannot show first-session payloads, wait.