Retention Rate = (Users At End - New Users) / Users At Start x 100
D1 Retention = % of users active on day 1 (from install day 0)
D7 Retention = % of users active on day 7
D30 Retention = % of users active on day 30
Retention Curve = D1, D7, D30 plotted over time; typically steep drop then plateau
Retention measures the percentage of users who stay or return to active use. Cohort retention (D1, D7, D30) tracks a group of users from their install date; it is the best measure of product stickiness.
A worked example
Cohort of 10,000 users installed on day 0:
- D1 (day 1): 4,000 active = 40% retention
- D7 (day 7): 2,500 active = 25% retention
- D30 (day 30): 1,000 active = 10% retention
This curve is typical for consumer apps: steep drop from D1 to D7 (most casual users leave), then flattens. The plateau (D7 to D30) shows your committed users.
Simple mode example: 10,000 users at start, 8,500 at end, 2,000 new during period:
- Retained = 8,500 - 2,000 = 6,500
- Retention = (6,500 / 10,000) x 100 = 65%
Retention benchmarks by category
Typical D30 retention by app category:
- Social apps: 15% to 25%. Users churn fast unless they build a network.
- Productivity apps: 25% to 40%. Often task-specific; users leave after the task is done.
- Dating apps: 20% to 30%. Users leave after matching or dating. High churn is expected.
- Fitness apps: 20% to 35%. Habit-forming but expensive to sustain (gym membership model).
- Games: 10% to 20%. Highly competitive, most users churn fast. Pay attention to D1 (onboarding) and D7 (habit).
- Utility apps: 40% to 60%. Weather, calculator, reference. Sticky if they solve a real problem.
D1 retention is usually 25% to 50% (one-third to half of users return on day 1). D7 is typically 50% to 70% of D1. If your D7 is much lower, first-week experience needs work.
How to improve retention
Measure the curve first, then fix by segment:
- D1 retention below 25%: Onboarding is broken. Make the first session faster and show core value immediately.
- D7 retention sharp drop from D1: Users get it but do not come back. Add reason to return: notifications, social, habit loops, streaks.
- D30 retention below 5%: You are in discovery mode, not yet sticky. Focus on finding one user segment that stays (core users).
- D30 retention 5% to 15%: Early product-market fit. Push the curve up by finding more of your core users and improving their experience.
- D30 retention above 15%: You have retention. Scale acquisition and measure cohort trends month-to-month to confirm the curve is improving.
Questions people ask
What is D1, D7, D30 retention?
Day 1, Day 7 and Day 30 retention measure what percentage of users from a cohort (group installed on the same day) are active on those days. Day 0 is install day. D1 is day after install.
Should I count a user as active if they only open the app?
No, not for product retention. Count them active if they complete a core action (send a message, log a workout, submit a form). Opening without action is meaningless.
How often should I measure retention cohorts?
Weekly at minimum, ideally daily. Cohort tables with weeks on rows (cohorts) and columns (D1, D7, D30) show trends. Improving trends beat absolute numbers; declining trends warn of problems.
What if my D30 retention is negative?
Cannot be negative. If users at end minus new users is below zero, churn exceeded starting users. Measure again; if real, acquisition is outpacing retention and growth is unsustainable.
Is retention rate the same as MAU churn?
No. Retention is cohort-based (percentage of an install cohort). MAU churn is percentage of a month's active users who did not stay. Retention is more precise for product decisions.
Sources
Last reviewed 11 September 2026. Formulas and benchmarks are published on the page so you can check them. This tool gives estimates, not quotes.
