Ten Questions I Wrote Before Wearing an Anti-Snoring Device, and the Answers Twelve Weeks Later

On the first night I sat on the edge of the bed with the device still in its box, a bathroom scale under my heels, and the Notes app open on my phone. I typed ten questions. Not a review, not a plan — just the things I genuinely did not know and did not want to guess at, so that in three months I could come back and read my own handwriting instead of reconstructing the story from memory.

Twelve weeks later, here are those ten questions and what I can actually say about each. One answer contradicts what I expected. One of them involves a boat, a generator that shuts off at ten, and seven strangers asleep four feet away. And one of them is an admission that for two weeks I believed numbers that were quietly wrong.

For context: I am thirty-four, I work at a desk, I weighed 96.8 kg on the first morning, and I have had a blocked nose most springs since I was a teenager. My wife has been sleeping with earplugs for two years. That is the whole setup.

1. Why did you start a scale and a device on the same morning?

Because I did not want to hand the device credit for something my waistline did, and I did not want to blame the device for something my waistline caused. Weight and snoring move in the same direction often enough that if you only track one of them, you will end up telling yourself a story you cannot check.

So I made one rule and kept it for eighty-four days: every morning, in the same order, before coffee — step on the scale, export the previous night from the app, write both numbers into one row of a spreadsheet. Same scale, same spot on the floor, same two-minute window. Boring on purpose.

What that rule produced: 96.8 kg down to 88.4 kg, which is 8.4 kilos, mostly from walking 8,000 steps a day and cutting liquid calories. No structured diet, no gym. It also produced a table in which the two variables visibly disagree with each other for three separate weeks, and those disagreements are the only reason this post is worth reading.

2. Where does it sit, and what does the first hour feel like?

Front of the throat, roughly two finger-widths above the collarbone, snug enough that swallowing meets a little resistance but never a pinch. My test, which took me nine nights to invent: turn your head hard left and hard right. If the band rotates or the front pad drifts off-centre, it is too loose. Mine drifted for over a week, and I spent that week concluding the device did not work.

The first hour is the honest part of the experience. The plastic is cold for about four minutes. You become extremely aware of your own swallowing, which is a thing you cannot un-notice once you notice it. I lay there cataloguing it like an idiot until maybe the fifth night, and then one night I simply did not, and that was the end of that.

One detail nobody mentions: put it on while you are still upright, reading or tidying up, ten to twenty minutes before lights out. Wearing it awake while doing something else lets the awkwardness burn off before you lie down. Wearing it in the dark, already horizontal, means the last thing your brain remembers is fumbling with a clasp at the side of your neck.

3. What does it actually do at two in the morning?

It listens. When it hears the pattern it is looking for, it gives a mild vibration against the throat — not a jolt, not an alarm, something closer to a phone buzzing in another room. The idea is that you shift position without fully surfacing. Sometimes I remember a vague buzz and a vague decision to roll onto my side. Mostly I remember nothing at all, and only find out in the morning that it fired six times.

Early on I woke up properly every time it went off — three or four full wake-ups a night, which felt like the device was costing me more sleep than it saved. That faded by the middle of week three. The trade was worth it, but I want to be clear that the first fortnight is a trade and not a gift.

4. How long before anything changed?

Night one: 31 minutes of snoring logged, which was roughly my baseline. Night two: 29. Night three: 24. Night four, after a late dinner and two beers: 33, which taught me more about the device than the good nights did. The first night under fifteen minutes came on night twelve, and it felt like a fluke. It was not a fluke, but it was also not a trend yet — the next night was 27.

The useful rule I settled on: never judge a single night, only a seven-night average, and only compare averages to averages. Snoring responds to what you ate, how late you ate it, whether the window was open, whether you had a cold. A single bad night is weather. Three bad weeks is climate.

5. Do you trust the app’s numbers?

Not completely, and this is the part I did not expect to write. Around week eight the app was reporting 15 minutes a night and my wife was still complaining, so I set an old phone on the nightstand to record audio for two nights and compared.

On the second of those nights the app logged 12 minutes. The recording had audible breathing — not the rattling kind, the quieter open-mouth kind — for something closer to forty. The microphone counts what sounds like snoring above a threshold. It does not count a mouth hanging open and breathing softly through it, and that was most of what was left of my problem.

So from week nine I added a second measure that no algorithm can fake: a dry-mouth score, one to five, rated the moment I woke up, before water. One is “normal mouth”. Five is “woke up twice to drink and my tongue was stuck to my palate”. It is crude and it is subjective and it caught exactly what the app missed — which is that the last eight weeks of improvement were slower and realer than the dashboard suggested.

6. So what do twelve weeks actually look like?

Weekly averages, not nightly numbers. Weight on Sunday mornings, snoring minutes from the app’s weekly export, dry-mouth from my own seven ratings. Week ten includes four nights aboard a boat, which I will get to.

Thirty-one minutes to sixteen is a 48 percent drop. Dry mouth went from 4.3 to 2.0. Nights under fifteen minutes went from none to five out of seven. The shape of the curve is more interesting than the endpoints: fast at first, flat between weeks seven and nine, wrecked in week ten, then recovered within six days of getting home.

7. Did the weight do the work, or the device?

Honestly, I cannot separate them, and anyone who tells you they can is selling something. What I can do is show where they disagree, which is roughly what a controlled trial would do if it were run by an impatient man with a spreadsheet.

The one thing I will claim: the first four weeks of improvement happened before any meaningful weight came off, so the device was doing something on its own. The rest, I genuinely do not know, and I have stopped needing to know.

8. What went wrong the week you were on a boat?

Six days away, five of them on the water, four nights in a lower-deck bunk eight feet from seven other people I had met at the dock that morning. This is the part of the twelve weeks I was most worried about, because the whole thing works on the assumption that you can lie down in a quiet room at a reasonable hour with a power outlet nearby.

A boat offers none of that. The bunks are short, so you sleep with your head propped and your chin toward your chest, which is the exact angle the device is fighting against. The air is humid enough that the band loosens overnight. The generator goes off at ten, so the power strip in the galley is the only charging option and by six in the morning there are three phones and a camera battery on it.

• Night zero (motel near the dock): unfamiliar bed, four pillows, one of which I folded into a wedge. Fine. 19 minutes.
• Night 2 was the only night in twelve weeks I took it off mid-sleep. Not because of the device. Because the swell picked up at two in the morning and I was green enough to stop caring about my throat, my numbers, or the opinions of strangers.
• Night 3 packing lesson: a folded life jacket under the pillow made a better wedge than the towel I tried on night one, and cost nothing.

Two things surprised me. First, nobody said anything. Seven strangers, four nights, a man strapping a small device to his throat in a shared cabin — and the only comment came on night three, when the guy in the bunk above asked whether it was a sleep tracker and then asked if it worked. Second, the recovery was fast: my first three home nights after the trip averaged 17.2 minutes, better than the week before I left. Four disrupted nights cost me four entries in the table and nothing else.

9. What annoyed you that had nothing to do with snoring?

A short and petty list, because these are the things that decide whether you keep wearing something in month four:

• Charging cadence. Every third night, reliably. Miss it and you find out at 2 a.m. when nothing happens. I now charge it on Sunday and Wednesday mornings, while I am in the shower, which is the only scheduling system I have ever successfully maintained.
• A notification at 01:50. Some firmware update or sync event put a bright screen in a dark room twice in week two. Do not disturb, permanently, from then on.
• The band seam. It frayed at week ten near the clasp, on the side I sleep on. A needle and four minutes fixed it. If you are not going to sew, buy a spare band before you need one.
• No case. It came in a box, not a case, and a box does not fit in a dopp kit. It has travelled in a sock ever since, which works and looks ridiculous.

10. What would you tell someone starting tonight?

• Track one other variable besides snoring. Weight, dry mouth, how many times you woke up — anything you can rate in five seconds. One number alone will lie to you.
• Give it two full weeks before you form an opinion. The first three nights are not data, they are admin.
• Check the fit with a head turn, not with your feelings. Loose feels comfortable and does nothing.
• Put it on before you lie down. This single change is worth more than any setting.
• Expect a bad week around week seven or eight. Mine was a combination of late dinners and a soft-snoring blind spot, not a broken device.
• If you travel, bring a spare band and a wedge substitute. Hotel pillows are too soft to give you the angle you need.
• Ignore single nights. Look at seven-night averages, and compare them to other seven-night averages, or you will spend your life chasing noise.

Two closing notes. This is one person’s twelve-week log, not medical advice, and results vary a great deal between people — my numbers include a weight change I cannot factor out, and a self-rated dry-mouth score that would make a researcher wince. Second: loud snoring, waking up gasping, or a partner who reports pauses in your breathing are reasons to see a clinician rather than to buy a gadget. A device like this is a comfort tool, not a diagnosis.