We built a feature that tells users to stop journaling. It became our highest-rated feature.

One of the hardest conversations we had at Murror was about a pattern we kept seeing in the data.

Some users were journaling every single day. Multiple entries. Long, thoughtful, beautifully written reflections about the same thing over and over — a breakup, a falling out with a friend, a decision they couldn't make.

On paper, these were our best users. High engagement, long sessions, consistent retention. Every growth metric said: these people love the product.

But when we actually read what they were writing (with permission, during research interviews), we realized something uncomfortable. They weren't processing. They were looping. The journal had become a place to rehearse pain without ever moving through it.

So we built something that felt deeply counterintuitive for a journaling app: a gentle nudge that appears after the AI notices repeated emotional patterns. It says something like: "You've been reflecting on this a lot. Sometimes the next step isn't another entry — it's a conversation with someone you trust."

Our growth team was nervous. You're literally telling users to leave? In a product that lives and dies by daily active users?

But here's what happened: users who received the nudge rated the app significantly higher in trust surveys. They didn't churn — they came back, but with lighter entries. Some told us it was the moment they realized an AI product actually cared about them, not just their engagement.

We learned that the most powerful thing an AI companion can do isn't always respond. Sometimes it's point you back toward the humans in your life.

Has anyone else built a feature that intentionally reduces usage? How did your team react?