Journal · March 2026
Where Thai subscription apps lose payers between trial day 3 and day 7
Most weekly packets in Bangkok subscription teams still show a single “trial conversion” tile. It averages day 0 with day 13. That average is polite and almost useless. When we sit with Funnel Bench cohorts and ask them to split the trial clock, the bruise almost always appears between day 3 and day 7.
Day 0–2 is noisy: people tap “start trial” to dismiss a modal, billing partners queue confirmations, and restore-purchase retries create ghost starts. Day 8 onward is a different story — reminder emails, price anchoring, the approaching charge. The middle is where product believes the person is “in trial” while finance has not yet seen a partner-accepted row.
Partner latency belongs on the same chart
Thai billing routes often confirm a trial hours after the in-app tap. Some partners batch. If your analytics tool stamps trial_start at tap time and your revenue export stamps it at confirmation, the day-3 to day-7 window will contain people who, legally and financially, were never in trial. Email calendars that fire “you’re halfway through” on day 4 then look aggressive, even cruel.
In Trial Window Diagnostics we ask learners to plot two lines: UI acknowledgement and partner acceptance. The gap is not a rounding error. It is the difference between a nurture sequence and an accidental lie.
What to stop averaging
Stop reporting a single trial-to-paid percentage in the Monday packet. Show day-3 still-active among partner-accepted trials, and day-7 separately. If those two numbers diverge, your emails, your paywall reminders, or your partner are doing work you have not named. Naming it is the whole point of subscription funnel tracking.
We do not claim a “correct” conversion rate for Thai streaming or wellness apps. We claim that an averaged tile hides the only interval where a small instrumentation change still matters.