Day one, day seven, day thirty — three numbers, and what they leave out.
Photo: RDNE Stock project / Pexels
Three numbers that run the industry
Day 1, Day 7, Day 30: these are the three benchmarks that define whether a mobile game lives or dies. They represent the share of players who installed the game and still opened it one, seven, and thirty days later. A studio watches these numbers the way a doctor reads a pulse.
Typical figures are brutal. Day 1 retention for a mobile game that makes it to market often sits in the range of 30–40 percent; Day 7 halves that again; Day 30 can drop to the low single digits. The curve is steep, fast, and almost always concave — the greatest loss happens in the first hours.
Every percentage point gained on Day 1 compounds forward, which is why studios spend disproportionate effort on the opening sequence. The first sixty seconds of a session is where most of that Day 1 drop occurs, not the second or third minute.
What the three numbers hide is the shape of the drop between them. A game can hit a respectable Day 7 figure while losing players violently in the middle of the first week, then stabilising — or it can trickle losses steadily and reach Day 30 with roughly the same proportional base.
A curve that falls inside the first minute was decided in the first minute.
Photo: Image Hunter / Pexels
These produce the same published benchmark and almost nothing alike as a design problem. Aggregated cohort statistics flatten the curve into three points and erase the diagnosis.
They also hide who is leaving. Retention curves are cohort averages — they say nothing about which players churn, and the reasons can be completely different across segments.
Key thresholds
Lifted out of the flow
- Day 1 — share of players who return the day after install; the most-watched single number in mobile publishing
- Day 7 — week-one retention; often roughly half the Day 1 figure
- Day 30 — long-term retention; single-digit percentages are common and considered acceptable
A player who bounces within four minutes is not the same problem as one who played for six days and quietly stopped. Studios with more sophisticated instrumentation disaggregate by acquisition source, device, and progression depth; smaller teams often cannot, and optimise for the number without understanding what drives it.
The metric also has a structural tension with free-to-play as structure: retention is necessary for monetisation but it is not sufficient. A game can hold its cohort loyally and still convert almost no one to payment.
The measurement changes the design, which is both the point and the risk.
Photo: picjumbo.com / Pexels
Designing purely to extend sessions — timers, daily rewards, artificial gates — can prop up a retention curve while hollowing out the actual experience. Ian Bogost ↗ has argued at length that compulsion and engagement are not the same thing, and the retention curve cannot tell the two apart.
What the metric is genuinely good for is early signal. If Day 1 retention collapses below baseline, something in the onboarding is broken — confusing tutorial, sluggish first reward, wrong difficulty gradient. The number points to a wound even when it cannot name it.
The shape between benchmarks — same D7 figure can mask very different weekly drop patterns
That is why studios run soft-launches in smaller markets first: to read the curve before committing to global distribution, and to fix what the numbers indicate without having to say exactly what they mean.