Classic vs Unbounded Retention
You'll learn: How to distinguish two fundamental ways of measuring whether users return—on an exact day versus on or after a target day.
The Core Distinction
When building retention curves, you need to define what "returning" means. There are two standard approaches:
Classic Retention (also called "Return on Day N") measures the percentage of users who come back on exactly Day N after their initial event. If you're measuring Day 7 classic retention, you only count users who returned precisely 7 days later—not on Day 6 or Day 8.
Unbounded Retention (also called "Return on or After Day N") measures the percentage of users who come back any time on or after Day N. For Day 7 unbounded retention, you count anyone who returned on Day 7, 8, 9, or any subsequent day.
Why the Difference Matters
Think of classic retention like asking "Who came to my party on Tuesday?" versus unbounded asking "Who has attended at least one party since Tuesday?"
Classic retention is stricter and shows periodic engagement patterns. It's ideal for products with scheduled behaviors (e.g., weekly grocery shopping apps) or when you need to detect specific day-of-week effects.
Unbounded retention is cumulative and always equal to or higher than classic. It answers "Has this user ever come back since Day N?" and is better for measuring long-term product stickiness and overall user base health.
In cohort tables, classic retention can fluctuate (spike on weekends, dip on weekdays), while unbounded retention either stays flat or increases monotonically.
Key Takeaway: Classic retention measures return on exactly Day N (captures periodic patterns); unbounded measures return on or after Day N (captures cumulative engagement). Choose based on whether you care about timing specificity or overall retention health.