- Formula
- % of cohort active in week N
- Unit
- %
- Models
- Media, Subscription
| W1 | 40%–75% | a16z |
| W4 | 20%–50% | a16z |
What it is
W1 / W4 retention is the percentage of the original cohort that remains active in week 1 and week 4 after first use. It is a weekly-cadence complement to daily retention curves, more appropriate for content and subscription products where daily engagement is not expected.
How to calculate it
For week N, count cohort users who were active at least once during calendar week N and divide by the cohort's starting count. Both milestones use the same original cohort.
Why it matters
For media and subscription products, weekly retention at W1 and W4 maps directly to billing-cycle behaviour. A user retained through W4 has demonstrated enough habit formation to be likely to renew at their first billing date, making W4 one of the most reliable predictors of annual LTV in subscription businesses.
Benchmarks & pitfalls
A16z benchmarks W1 at 40% (OK) / 55% (Good) / 75% (Great) and W4 at 20% (OK) / 30% (Good) / 50% (Great). No asOf year is provided for these specific bands, so treat them as directional reference points rather than precisely dated norms. Ensure your definition of 'active in week N' is consistent — some teams use any session, others require a substantive action (e.g. content play, purchase); mixing definitions across cohorts will distort trends.