Retentiecurve

Retentiecurve, or retention curve, is a graphical tool used in product analytics to visualize how user engagement evolves over time after initial activation. It represents the retention rate on the vertical axis and time since activation on the horizontal axis, typically for a defined cohort of users.

Calculation and data: For a cohort of users acquired on day 0, the daily retention at day t is the proportion of those users who are still active on day t. Retention(t) = active_in_cohort_on_day_t / cohort_size. The curve can be drawn for daily, weekly, or monthly intervals and for multiple cohorts to compare changes.

Interpretation and patterns: The curve typically starts near 1.0 at day 0 and declines over time as users churn. Early drop-offs are common and reflect onboarding friction; slower decay later suggests ongoing engagement by a subset of users. Shapes may be exponential, multi-phase, or approximate a power-law tail depending on the product and user behavior.

Uses and applications: Retention curves are used to assess product changes, onboarding effectiveness, and feature impact; compare segments (e.g., paid vs. free, new vs. existing users); estimate long-term value and churn risk; inform experiments and roadmap decisions.

Relationship to other methods: It complements cohort analysis and churn rate calculations. For statistically rigorous handling of censored data and variable follow-up, survival analysis methods (e.g., Kaplan-Meier) can be applied.

Limitations: Retention curves depend on how activity is defined and data collection. They can be sensitive to window choice, reactivation events, and inactive users who still appear in the dataset. Censoring and selection bias should be considered when comparing curves.