3 May 2026
Soft-launch markets that disagree on Day-3
Two early markets can both be ‘right’ and still send opposite signals. Choose which question you are answering.
Soft launch is supposed to reduce surprise. Instead, many teams face two charts: Market A with longer sessions, Market B with steadier Day-3 return. UA wants the longer sessions. Production wants the return curve. Both open the same export and leave with different stories.
Pick the decision, then the metric
Before comparing markets, write the decision in one sentence: Are we validating session depth for mid-core progression, or validating early return for a hyper-casual loop? Session trend work without that sentence becomes a beauty contest.
For progression-heavy titles, we weight session length bands and first meaningful peak. For short-loop titles, we weight Day-1 and Day-3 return and session frequency.
Holiday and payday noise
In Malaysia and neighbouring markets, public holidays and payday weeks bend both session length and return. We ask studios to flag those dates in the intake form. Ignoring them is how a “winning” market quietly fails two weeks after scale.
A Soft-Launch Cohort Review ends with a go / hold / retest line—not a blended average that hides the disagreement.