How to reduce early churn and activate the second session

So, your app has got new users. They breeze through onboarding, love their first session, and then… churn. Like, ghosted-you-after-one-date gone. That’s Death Valley, the spot where subscription apps crash, burn and churn users.

I talked to 50+ app product leaders in early 2025. Four out of five confirmed: “2 sessions completed within the first week” is their critical activation metric. Hit that, and users retain longer, trials convert better, and subscription apps make more realized LTV. 

Yet most apps see 77% of their users churning by day 3.

Day 3!

All that ad money, all that effort that went into creative strategy, organic and ASO to acquire users, iterating onboarding to increase trial starts or conversions: gone for most users you acquired.

You might think: “But I can hit $3M ARR just by focusing on user acquisition, onboarding and the first session!”

Sure, maybe.

But that plan is going to hit a wall, be sure about that. CAC goes through the roof because your user acquisition machine will get less effective over time. 

The bridge from the first to second session doesn’t start after users leave your app. It starts within that crucial first session.

Brilliant (Duolingo for computer science & math) mastered the bridge. Here are five easy tricks from them to reduce churn and keep your users coming back.

Make the first session quick, valuable, and fun to stay ahead of churn

Getting more users back into the second session starts with the first session.

First sessions aren’t what they used to be. Way back, you could make people spend 10-15 minutes in the first session and still see high completion rates.

Needless to say that those good days are over in the fast-dopamine-hunting world we live in (thanks, TikTok).

The magic is balancing this triangle:

• Short enough to feel quickly to accomplish (Brilliant’s took me 2min 14sec; done before my coffee is cold)
• Valuable enough to provide real learning (proof users get what they signed up for, for Brilliant → become a problem solver)
• Fun enough to enjoy the process (never forget: as a health/education app, your app feels like work to users; so make that work fun; Brilliant is a math app but doesn’t feel like homework)

It might take a couple iterations to get it to work. But it’s definitely worth it.

Metrics to watch:

• 1st lesson start rate (if it’s below 60%, simplify your first home landing)
• 1st lesson completion rate (aim for >60%)

Obsessively celebrate them

This is the most crucial part. So let’s pay extra attention to it.

We’ve all read Hooked by Nir Eyal, and know about the habit loop. If you don’t, here’s a recap of the framework that you should never forget if you’re serious about building sticky apps. The framework describes how habits are formed and maintained.

It includes: 

• Trigger: starts the habit, can be external (push, notification) or internal (boredom, anxiety). Example: a push by Duolingo’s Duo makes you anxious to lose your streak
• Action: behavior you perform in response to the trigger (open Duolingo)
• Variable Reward: keeps users hooked (=wanting more). Streaks, animation, haptic feedback, makes your brain swim in dopamine and crave more. Better if not predictable, that’s why Duolingo adds so much variety to their celebrations.
• Investment: make users invest time, effort or data to increase the odds of the next trigger (complete the next lesson).

What we want is to make users feel accomplished, and make them want to feel this more often. Good dopamine, reward state of the habit loop.

Brilliant throws a party every time you finish. It’s a 3-part celebration:

• Nice work: we all love a good tap on the shoulder, and a buzz in your hand with haptic feedback
• Recap learning: makes progress visible; reminds you what you’ve just learned
• Streak counter: shows you the streak count, nudges you to the next mini-milestone (day 3, to boost activation), and taps into identity (become a problem-solver)

Such a strong flow, Brilliant’s product growth team kills it here.

Because identity-based goals (become a problem solver) trigger behavior change stronger than outcome-based goals (reach 3-day streak).

It’s a 20-second experience for Brilliant. You don’t have to go as sophisticated as they went. Free Lottie files (lottiefiles.com) does the job already. Use Rive (rive.app) if you want to step up.

“Is consciously triggering a user’s dopamine evil?”, you might ask yourself.

It’s not. Your users want to spend more time in your health/education app instead of wasting hours on TikTok. It’s your duty to help them stay on track to something that’s worthwhile for them. If you’re in gaming, social networks, or entertainment, this doesn’t apply to you (you’re most likely already aware of these dynamics).

Remind them why they’re here

At this point, your users are pumped and feel hyped. They’ve just learned something new (education app), feel calmer (meditation/prayer app), or feel exhausted but in a good way (fitness app).

It’s a state we love. 

Think back to the last time you were in that state. Completed a 5k run, and a good feeling rushes to your body. Imagine your running buddy now asks you: “Want to run in two days again?”.

You’re likely to say yes. In fact, you’ll probably never be more likely to say yes.

Let’s use the high user psych to help them feel this again, and start building a habit.

Brilliant strengthens the habit loop (investment) by making users commit to a streak-goal. Users have to do that, there’s no way to skip this screen. Smart.

They tap into the goal gradient effect here: we’re more motivated to complete a task as we get closer to a goal. The 3-, 7-, and 14-day goals are broken up into incremental milestones (“Great”, “Amazing”, “Phenomenal”).

Once users select a goal, they have to commit to it (CTA). When we commit to something, we feel a psychological need to follow through. It’s self-imposed accountability, and can be a powerful nudge to make users want to come back.

Capture commitment

Building habits is tough.

Your job is to make establishing a habit around your core app activity easy-peasy.

As we’re in the high user psych state, this is the perfect time to help users feel the good dopamine from your app again.

That’s why Brilliant prompts the notification permission on the next screen.

“Well done! Want to do this more often? “Let us help you.”

Again, well played, Brilliant PMs!

I’m sure they see high notification adoption rates here because people are buzzing.

There’s also a new way to nudge users back into your app: lock distracting apps like TikTok and Instagram until your users activate the second session in your app. “Complete your next Spanish lesson in Duolingo to unlock TikTok.”

Essentially, this is your native screen time feature for retention. When users open social media during a specific time (e.g. morning), they see a full-screen nudge that reminds them to do a session in your app. It’s like your app telling users, “Hey, finish your lesson in my app first”. You can build it yourself or launch it with Wellspent’s SDK in a couple days. Full discloser: my team built the SDK.

Preview the second session

I’m repeating myself. But because this is so essential to building habit-forming products, I want to make sure it sticks:

It’s your job to make it as easy as possible to activate the second session. Your responsibility.

If you only take away one thing, I hope it’s this.

Brilliant shows what the exact next lesson is about and allows users to start it with one tap.

What happens from a psychological point:

• Zeigarnik Effect: We are more likely to remember and feel compelled to complete unfinished things. By showing the user what’s coming next, Brilliant creates an open loop: they’ve started their learning journey, and the app very subtly reminds them that there’s more to explore. This creates a psychological itch to return and close the loop by starting the next session.
• 1-Tap-Access: You probably can’t hear it anymore but I’ll double down: reduce friction. Friction kills your growth. Lower the activation energy to start the next loop by making it dead-simple to activate the second session. Brilliant does it well: one CTA, name of the lesson highlighted. As a user, I know exactly what to do. No brainer.

What you want to prevent is decision-fatigue: “What am I supposed to do now?”. It’s the number one killer for habit formation. 

The next step is crystal clear. Perfect starting point for the next habit loop.

You’re on the right track when you get >50% of the first session completions into the second session.

Conclusion

Duolingo’s CPO Cem Kansu says you’re in the habit-building business, with three products: onboarding, core activity, and retention mechanics. 

There’s a ton of validated retention mechanics out there, so you don’t have to reinvent the wheel. These five tips are relatively low effort, potentially high impact things you can launch within days or weeks. I genuinely hope this is useful for you to drive activation.

Most apps die in death valley. Your app doesn’t have to.

To dive deeper into how your retention strategy impacts churn and subscription growth, check out this guide on cohort analysis for subscription apps.

Big props to Brilliant for lighting the way.

Key takeaways

1. Make the first session short, valuable and fun: short enough to feel quickly to accomplish, valuable enough to provide real progress, fun enough to enjoy the process.
2. Obsessively celebrate them: Most crucial part: dopamine high lifts user psych. Make them want to feel this more often. Go all in on animations, haptic feedback.
3. Remind them why they’re here: Leverage high user psych to commit them to the next milestone. Identity-based goals (become a problem solver) > outcome-based goals (complete three lessons).
4. Capture commitment: Leverage high user psych to get them into push notifications or your screen time feature to remind them of the core activity within your app.
5. Preview the second session: Show the exact next lesson with one-tap access. make it crystal clear what comes next to prevent decision fatigue.

About the author

Selcuk Aciner is the co-founder of Wellspent. Their Screen Time SDK helps subscription apps lift retention by allowing them to lock distracting apps until their users complete a session in their app. Wellspent is backed by the founders and builders behind Blinkist, mysugr, Tomorrow’s Education, and Refurbed.

LinkedIn

Website