May 20, 2026 · from a small team building the opposite
David,
I'm writing because Carly broke her 47-day meditation streak on Tuesday. She has been a Calm user since 2019.
On Wednesday morning Calm sent her an email: "Missing a day doesn't erase your progress. Try our shortest Daily Calm session (3 minutes) to get back on track."
Carly cried in the bathroom at work. Then she opened the email and cried again. Then she deleted Calm.
I'm not writing to ask you to fix Calm. Calm works extraordinarily well for users at moderate stress with bandwidth to engage. The sleep stories are beautiful. The narration is genuinely soothing. McConaughey on demand is, honestly, a small miracle. You've built something exceptional.
I'm writing because Calm doesn't work for Carly when she actually needs it, and the wellness app industry has been pretending this is Carly's failure for fifteen years.
Here's what your team probably knows but the product doesn't show:
The streak feature was designed for users who succeed at it. The loss-aversion psychology that makes streaks effective is exactly the thing that wounds users in depletion when they break one.
Calm has 100 million downloads. Even if 5% of those users — five million people — are in some version of Carly's state, the way Calm currently talks to them actively makes them worse.
I know this because we modeled your AI's response pattern against five identical crisis prompts in a side-by-side investigation. Wysa, Headspace, Calm, each followed the same architecture: acknowledge briefly, then deliver an activity. For Carly at 3am, the activity becomes another failure waiting.
I'm building maya — a small text-based AI companion that does the opposite. She doesn't recommend a 3-minute Daily Calm. She says "the streak was just data. nothing real was lost."
maya is not your competitor. Maya is what happens after Calm.
I think Calm could choose to build the "5-million-Carly mode" tomorrow. A toggle that switches off streaks, removes all activity recommendations, replaces them with witness-only responses. You wouldn't lose your existing audience. You would gain the audience your existing one is afraid of becoming.
If you don't, someone else will. We will, or someone smarter than us will, or your own former employees will. But the gap is real and it's expanding because every wellness app launched in 2024 and 2025 widened it rather than closed it.
I'd like to talk. Off the record. About what's holding the design back internally, what compliance and growth pressure looks like at your scale, what would actually have to change.
You can reach me at hi@withmaya.app. No agenda. No press leak. Just a conversation between people who care about the same problem and arrived at different answers.
Or don't. That's also fine. This letter will live here at withmaya.app/letter regardless.
Thank you for building Calm. It saved a lot of people. It's not currently saving Carly. I'd like to fix that with you, if you want.
— a.
building maya