I removed streaks from my cognitive training app. Here's what happened to DAU.
Every habit-formation app has a streak counter. Duolingo made it religion. Snapchat weaponized it. Every meditation app copied it. We put one in WoW Brain too — a clean number in the top right, days-in-a-row, reset to zero if you skipped.
Eight months in we deleted it. This post is why, what happened, and what we shipped instead.
The interviews that changed my mind
For four weeks I ran fifteen-minute calls with users who had come back to WoW Brain after a break — people who broke their streak and returned anyway. I wanted to understand the recovery.
Almost every conversation had the same shape. The user missed a day. They opened the app the next morning, saw the streak had reset to zero, and felt a small hit of shame. Some pushed through and started over. But about half of them stopped opening the app for another week or two, and told me later they were "avoiding the app because I felt bad about breaking my streak."
The streak wasn't holding them accountable. It was creating avoidance behavior. The exact opposite of what we wanted.
The counter-intuitive finding
Here's the thing that surprised me. The users who kept coming back after a break weren't the ones with high previous streaks. They were the ones who had a specific reason to open the app that day.
"I want to see what my Cognitive Radar looks like after the vacation."
"I want to write in the evening ritual about the argument I had yesterday."
"I want to do the memory trainer because I forgot two names at a meeting today."
The streak was doing nothing for these users. Their motivation was intrinsic — tied to a specific need on a specific day. The streak counter was a summary of past behavior; their return was driven by present relevance.
What we shipped instead
We removed the streak counter and replaced it with three things.
One: a "welcome back" that references what the user missed. If you didn't open the app for three days, tomorrow's greeting says "You skipped Wednesday and Thursday. Yesterday your evening ritual would have said X" — where X is a real question tied to what the user wrote in their last session. It's not shame; it's continuity.
Two: a Cognitive Radar delta. The Cognitive Radar shows eight cognitive dimensions on a spider chart. After a gap, we show the user how the chart shifted from their last session. Not gamification — just a mirror of what actually happened to their cognitive form during the break.
Three: a "why open the app today" line. Every morning notification includes one specific reason to open, generated by the AI from yesterday's evening ritual. "Your evening note yesterday said you wanted to remember to try the Word Sense trainer today. That's ready."
These three replacements do the job the streak was supposed to do — return the user tomorrow — but without shame as the mechanism.
The numbers
Two months after removing the streak counter:
- DAU/MAU ratio went from 0.34 to 0.41 — meaning users who used the app in a given month came back more days on average.
- 7-day retention for users returning from a break went from 22% to 47%. This is the big number. Users who lost their streak used to fade out at 22%; now they come back at more than double the rate.
- Weekly active users — the metric we were most afraid to lose — went up 18%. Not down. Up.
The streak counter was suppressing our metrics, not driving them.
What this means if you're building anything similar
I'm not going to argue every habit app should kill its streaks. Duolingo makes billions on their streak mechanic, and their team knows more about this than I do. What worked for us was:
- Users who came back after a break were driven by present relevance, not past performance. We built around that.
- Shame-based mechanics work for extrinsic tasks (learning French for a test); they backfire for intrinsic ones (learning your own attention patterns).
- Continuity mechanics — "here's what you missed, here's what's ready for you today" — beat gamification for a cognitive-training audience.
If you're running a habit product, take a week and interview users who returned from a break. What you'll hear might change your roadmap.
If you want to see the mechanics in action, wow-brain.com has a 7-day free trial, no card required. The welcome-back logic runs from day one — even if you skip day two, day three will show the delta.
Curious to hear how other teams have handled this. Streaks work for some products and destroy others; I'd love to know which side of the line your product is on and why.