She Scores Strangers' Sleep For A Living. Then She Scored Her Own: Nine Wake-Ups A Night.
A 16-year sleep technologist ran the most honest experiment of her career, on herself. The chart on her fridge shows what eight weeks did.

For 16 years I have sat in a dim control room watching strangers sleep, scoring their nights page by page. Spindles, stages, arousals. I can look at a squiggle and tell you whether you dreamed.
Last winter I finally pointed the equipment at myself. Nine wake-ups. The woman who teaches sleep, surfacing nine times between midnight and six.
What did not do it
Understand: my sleep hygiene was already perfect, it is literally my profession. Cool room, dark room, no screens, fixed bedtime, the whole catechism. Hygiene is the fence around the garden. My fence was immaculate and my garden was dying anyway.
The one-subject study
So I ran it the way I would want it run. Baseline week, honestly scored: 8.7 wake-ups a night on average. Then one change and only one change: 400mg of magnesium bisglycinate, two gummies, 9:45pm. I chose bisglycinate because the absorption picture is clear, and because the oxide form is, professionally speaking, a joke with a label.
Why magnesium at all? Because it sits at the bottom of the wind-down chemistry, because I drink four coffees a shift and coffee spends it, because I am 54 and the years spend it, and because shallow, fragmented, surface-y sleep is exactly what the shortage looks like on a chart. I had been reading the signature for years on other people.
The numbers
Week two: six wake-ups. Week four: three. Week six: two. Week eight: one brief surfacing at 4:10, back down inside a minute. I printed the chart and put it on the fridge, which in my field is the equivalent of framing a marathon medal.
One subject, no control group, I said it before anyone in the comments does. But sixteen years of reading nights teaches you the difference between wishful thinking and a real change in architecture. This held. It is still holding.
What I take is Kaive, for a reason that matters at 10pm: it is a sugar-free raspberry gummy I look forward to, and the intervention you do not skip is the only one that works. No melatonin in it, so the mornings are clean. Third-party tested, which I verified, because of course I did.
Your own baseline week
You do not need my equipment. A notepad and honesty will do: count the wake-ups for a week, then change one thing and count again. Thirty nights with a full money-back guarantee makes it a free study with one subject who matters. Put your chart on the fridge. Tell them a technologist sent you.