Lifecycle charts in Metrics Explorer help you understand how usage changes over time by classifying your unique units (for example user_id) within each time interval based on whether they used an event recently, returned after a gap, continued to use it, or churned.
The first step to creating a lifecycle chart is to decide if you want to use a metric or an event. Lifecycle is designed for a single underlying event and count-style metrics that are composed of a single count type event
Select what you want to count uniquely (for example user_id, stable_id, or another unit). The chart reports how many unique units fall into each category for each time interval.
Pick an interval unit (day / week / month) and a number of intervals per bucket (1–48). Each bar on the chart represents one interval bucket, and the chart shows one data point per bucket across the selected date range (max 1 year).
Each interval bucket classifies unique units into four categories (mutually exclusive):New: Used during this interval, and did not use at any point earlier within the lookback window (up to 1 year before this interval).Returning: Used during this interval, did not use in the immediately previous interval, and did use earlier within the lookback window.Recurring: Used during this interval and the immediately previous interval.Dormant: Used in the immediately previous interval, but not during this interval (often displayed below the axis to make churn visually obvious).
X-axis: Time, grouped into your chosen interval buckets.Y-axis: Count of unique units.Stacked bars: Show how the composition of usage changes over time (new vs returning vs recurring), while the churn component (dormant) highlights drop-off between adjacent intervals.
