The Sleep Percentage Framework
Stop counting hours in bed and start maximising the percentage of that time spent in REM and slow wave sleep.
- Difficulty
- Easy
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 92%
Ahmed argues sleep is discussed in the wrong framework entirely. Seven hours in bed is not seven hours of sleep: it decomposes into awake, light, REM, and slow wave, and only the last two matter physiologically. REM is cognitive repair and is when deep dreams occur, so anyone who cannot remember dreaming is likely short on it. Slow wave is when the body produces 95 percent of its human growth hormone, which is when you actually get stronger, not in the gym where you are breaking muscle down. He says two people with identical seven-hour blocks can land anywhere from 30 minutes to five and a half hours of REM plus slow wave, and that gap shows up in the rest of their lives. His target is over 50 percent of time in bed spent in REM and slow wave; he personally runs about seven to seven and a half hours in bed while extracting three to four and a half hours of REM and slow wave, matching people who spend eight or nine hours in bed.
Origin
It came out of Ahmed's undergraduate physiology research at Harvard, where as a competitive squash player he read roughly 500 medical papers trying to understand what training was actually doing to his body.
Core principles
- 01Time in bed is an input; REM and slow wave are the output.
- 02Awake and light sleep deliver essentially no physiological benefit.
- 03You do not get stronger in the gym, you get stronger in slow wave sleep.
- 04A third of your life is a black box that dramatically affects the other two-thirds.
- 05Consistency of timing beats almost every other lever available to you.
How to run it
- 1
Reframe the number you are tracking
Stop reporting hours in bed. Ahmed's example is the person who went to bed at 11, woke at six, and calls it seven hours of sleep; in reality that was seven hours in bed split across awake, light, REM, and slow wave stages.
Pro tip Ask yourself whether you remember your dreams. If you never do, you are almost certainly short on REM.
- 2
Measure the stages and compute your percentage
Use a continuous monitor to get actual REM and slow wave minutes, then divide by total time in bed. This converts a vague feeling about sleep quality into a single number you can move.
Watch out Two people with the same time in bed can be at 30 minutes or five and a half hours of restorative sleep, so an unmeasured estimate is close to useless.
- 3
Fix timing consistency first
Go to bed and wake up at very similar times each day. Ahmed singles this out as the lever that will unusually elevate sleep quality before any gadget or supplement is considered.
Pro tip Consistency is free and compounds, so bank it before spending money on the environment.
- 4
Engineer the sleeping environment
Make the room very cold, very dark, very quiet, and high in air quality. Ahmed notes that athletes who are serious about this measure all of those variables and dial them in, and that most athletes are still not sophisticated enough about it.
Watch out Ahmed says athletes remain under-optimised here relative to how much it affects performance, so do not assume elite habits are already good ones.
- 5
Neutralise the pre-bed screen problem
Laptops and phones emit light that tells your eyes and brain to stay awake. Ahmed wears red-tinted blue-light blocking glasses before bed and describes them as a get-out-of-jail-free card that lets him work right up to bedtime without paying for it in sleep quality.
Pro tip Treat this as a hack for the reality that you will answer email late, not as permission to skip the timing work.
- 6
Test supplements carefully and refuse sedatives
Magnesium and melatonin work better or worse depending on the person, so test individually. Ahmed is explicit that harder sleep drugs such as Ambien shorten sleep latency but wreck quality, taking someone who would normally be at 30 percent down to about 20 percent.
Watch out Falling asleep faster is not the goal; a drug that knocks you out while suppressing REM and slow wave makes the number you care about worse.
In the wild
Ahmed says he is still optimising his life to be a CEO rather than purely to be a healthy human, so instead of the eight to nine hours in bed he would otherwise choose, he spends about seven to seven and a half. By applying consistent timing, environment control, and blue-light blocking, he extracts three to four and a half hours of REM and slow wave from that shorter block.
→ He gets the same restorative sleep as someone spending eight or nine hours in bed, buying back well over an hour of waking time per day without paying a recovery penalty.
Ahmed describes professional athletes WHOOP worked with who played video games from 11pm to 2am before a game watched by a million people, without connecting the two. The data showed a red recovery the next day despite no alcohol involved, purely from staying up late staring at a screen.
→ Being able to show the athlete the counterfactual, what recovery would have looked like had they gone to bed at the time they started gaming, made an invisible performance cost concrete.
Common mistakes
Reporting hours in bed as hours of sleep
This is the core error the framework exists to fix. Ahmed's seven-hour example hides a spread from 30 minutes to five and a half hours of restorative sleep, and the person at the bottom of that range has problems throughout their life while the person at the top is high functioning.
Using sleep drugs to fix a quality problem
Drugs like Ambien knock you out so sleep latency looks great, but the quality of the sleep is never that high. Ahmed's estimate is that someone normally at 30 percent REM and slow wave drops to about 20 percent, meaning the metric that matters gets worse while the metric you notice gets better.
Assuming naps carry the same quality as night sleep
Ahmed notes the percentage of nap time spent in restorative stages is often lower, so a casual nap on a couch under-delivers. If you nap, raise the intensity of the setup with a proper room and a sleep mask rather than assuming the time counts.
Copying another person's protocol wholesale
Ahmed is clear that genetics affect how well people sleep and that supplements work better or worse for different individuals. Adopting someone else's stack without measuring your own percentage means you have no evidence it is doing anything for you.
Is it for you?
Best for
Founders and executives who cannot extend their time in bed but need far more recovery out of the hours they do have.
Not ideal for
People with an undiagnosed clinical sleep disorder, who need a physician rather than a self-optimisation loop.
From the transcript
“the thing to understand about sleep is it's all about the percentage of time that you spend in bed getting high quality sleep”
“awake and light are pretty much irrelevant you're not getting any physiological benefit from being in awake or light stage sleep REM and slow wave…”
“slow wave sleep is when your body produces 95 of its human growth hormone... you're not getting stronger in the gym right or working out…”
“if you can get over 50 of the time you're spending in bed to be REM and slow wave sleep like you're doing pretty well”
From the episode
WHOOP Founder Will Ahmed: How I Built A $3.6 BILLION Company
WHOOP Founder Will Ahmed