30 nights of sleep hypnosis: here's what changed

Night one, zero expectations

We'll be upfront. We went in skeptical. One of us on the team had tried sleep apps before and said they all felt like spa music with a British accent. Their sleep onset latency was averaging 45 minutes, with two wakeups a night, and they'd been telling themselves this was just what being 38 felt like.

First session was 18 minutes. Phone face down, eyes closed, no judging. About halfway through our tester stopped hearing the words and woke up at 6:45 AM in the same position. Their watch said sleep came in 11 minutes. We assumed it was a fluke.

What we tracked

We kept a spreadsheet for 30 nights. Four columns: sleep onset latency, number of wakeups, total sleep time, subjective quality (1 to 5). We also noted which session was played, caffeine intake, and whether there had been alcohol.

• Onset latency: dropped from 45 min average to 14. That's the biggest shift.
• Wakeups: went from 2.1 per night to 0.8.
• Total sleep: up about 40 minutes most nights.
• Subjective quality: from 2.6 to 3.9. The one we trust least, but it matches the objective numbers.

What didn't work

A few things didn't move. 4 AM gym days still wrecked sleep quality, no matter what session ran. Alcohol still knocked an hour off total sleep time. And the nights a session was played during stress (deadline, sick kid) worked worse than the normal nights. The tool doesn't substitute for the basics, it sits on top of them.

Would we keep doing it? Our tester already is. Day 47 and still going as we write this.