D1 / D7 / D30 retention
Early retention predicts everything downstream.
- Formula
- % of install-day cohort active on day N
- Unit
- %
- Models
- Subscription, Gaming, Media
| Social / D1 | 50%–70% | a16z; GameAnalytics 2025 |
| Social / D7 | 35%–50% | a16z; GameAnalytics 2025 |
| Social / D30 | 20%–30% | a16z; GameAnalytics 2025 |
| Gaming | 22% | a16z; GameAnalytics 2025 |
| Gaming | 3%–4% | a16z; GameAnalytics 2025 |
| Gaming | <3% | a16z; GameAnalytics 2025 |
What it is
D1 / D7 / D30 retention measures the percentage of the install-day cohort that returns to the app on day 1, day 7, and day 30 respectively. It is the standard early-lifecycle engagement signal for consumer apps, games, and subscription products.
How to calculate it
For each day-N milestone, divide the count of cohort users who were active on exactly day N by the total users who first appeared on day 0 (install day). Each milestone is calculated independently on the same starting cohort.
Why it matters
Retention at D1, D7, and D30 is a leading indicator of long-term LTV and payback: users who survive to D30 convert to paid tiers, engage with monetisation mechanics, and generate word-of-mouth at materially higher rates than those who churn in the first week. For gaming and consumer subscription products, these milestones also predict whether paid acquisition will ever be profitable.
Benchmarks & pitfalls
According to a16z and GameAnalytics 2025 data, social/consumer apps benchmark at D1 50–70%, D7 35–50%, D30 20–30% across an OK-to-Great range. Games are considerably lower: median D1 ~22%, D7 ~3–4%, and fewer than 75% of users remain by D28 (below 3%). The widely-cited '40/20/10' rule (D1/D7/D30) represents top-tier performance, not median — do not benchmark a typical game against it. Be precise about whether you track 'active on day N' (exact-day) vs. rolling windows; the two definitions can differ significantly.