Welcome to ChurnJitsu, your weekly briefing on the art of retention.
Let’s dive in.
🥊 TL;DR
The average app loses about three in four new users within three days. Quettra’s 2015 study of 125 million Android phones put average day 3 retention at 23.4%.
The drop is still there today. Adjust’s benchmarks show North American apps keeping 23% of users on day 1 and 5% by day 30.
The fix that moves retention: get every new user to one real result in their first session, and give them a reason to open the app tomorrow.
🎯 Master Class
The first session decides whether you keep a user.
The numbers show how fast new users leave. In a 2015 analysis by Andrew Chen and Quettra covering more than 125 million Android phones, the average app kept 23.4% of its users active on day 3. The top 10 apps kept 71.5%.
A gap that size comes from how the first session is built. Prettier tooltips can’t close it.
The drop looks the same a decade later. Adjust’s retention benchmarks put North American apps at 23% of users on day 1 and 5% by day 30.
So treat onboarding as an activation contract (a promise that the user gets something real before you ask them for anything). They gave you an install, and the first session is where you pay it back. Three checks tell you whether you’re keeping that promise.
1. Promise: show the payoff on the first screen
Put the one job people installed your app for on the first screen. A calorie tracker gets them logging food. A game gets them playing. An AI photo app changes their photo before it asks them to rate the app, invite friends, turn on notifications or join a Discord.
Every request you make before they see value gives them a reason to close the app. Clear those out and more users reach the moment that brings them back.
2. Proof: find the step where you lose people
Get each user to one real outcome as fast as your product allows. Every app is different, so there’s no universal target for that speed. Measure your own.
Track four steps: install → first open → first meaningful action → first completed outcome. Then split your Day 1 and Day 7 retention by whether users reached that outcome.
You’ll see which step costs you the most users, and that settles more than another week of debating button colors.
3. Return: give them a reason to come back tomorrow
Before the first session ends, hand the user something that pulls them back: a streak they’d hate to break, new content tomorrow, a result that improves with more data, a reply from a friend, a daily challenge or a price alert.
Ask for notification permission after they’ve created something worth being notified about. Ask on the first screen and most users have no reason to say yes.
What it’s worth
Firebase’s Doodle case study shows the payoff. Doodle first made its app more stable, then A/B tested (showed different versions to different users and compared the results) its onboarding and poll creation flows.
One flow produced 42% more polls, Doodle’s core engagement metric. Firebase also reports 6% better 7-day retention and 15% more daily active users. The gains came from improving the one action people open Doodle to do.
Old flow: Install → tutorial → account creation → permissions → empty dashboard → “We miss you” push two days later.
New flow: Install → useful action → visible result → saved progress → permission request in context → reason to return.
The new flow puts a result in front of the user before the first ask. Paid installs only pay off when those users come back, so build Day 2 before you buy Day 0.
Your move this week: pull Day 7 retention for last month’s new users and split it by whether they completed their first outcome. The gap between those two numbers is what fixing your first session is worth.
Inman