Somnera
Sleep restriction

Why the sleep diary is the part that actually does the work

It looks like homework. It is the only thing standing between you and guessing — and your memory of last night is wrong in a specific, predictable direction.

6 min read

Last reviewed October 7, 2026

A grid of nights, each drawn as a part-filled bar — the shape a sleep diary makes over a month.

Share

Your memory of last night is wrong, and it is wrong in a particular direction. People with insomnia consistently recall more time awake than a recording shows. That is not carelessness — lying awake is vivid and memorable in a way that sleeping is not.

Which is the whole reason the diary exists. Every decision in CBT-I is made from numbers, and if the numbers come from recollection they are numbers about how the nights felt.

What it is actually measuring

Five things, every morning, from memory, taking under two minutes:

  • What time you got into bed
  • Roughly how long you took to fall asleep
  • Roughly how long you were awake during the night
  • What time you got out of bed
  • How the night felt

Estimates are expected. Nobody is timing anything, and precision is not what makes this work — the trend across a week is.

From those five, two numbers fall out:

Total sleep time — time in bed, minus falling asleep, minus time awake.

Sleep efficiency — total sleep divided by time in bed, as a percentage. Nine hours in bed and six hours asleep is about 67%. Consolidated sleep usually sits above 85%. That single number is what the sleep window is set from and adjusted by.

Consistency beats precision, and it is not close

The most common reason people abandon a diary is the belief that they are doing it wrong, because they do not know exactly when they fell asleep.

Nobody knows exactly when they fell asleep. Falling asleep is precisely the moment you stop being able to observe yourself. Round to the nearest fifteen or thirty minutes and move on.

The arithmetic that matters runs across a week, and a week of rough estimates produces a usable average. A week of four exact entries and three blanks does not, because the average is now built on whichever nights you felt like recording.

There is a sharper reason not to chase precision. Knowing the exact time requires looking at the clock, and looking at the clock at 2am starts the calculation - how long is left, how bad tomorrow will be - which is arousal, and arousal is the thing standing between you and sleep. A diary that makes you watch the clock is worse than a rough one.

Log in the morning, not in the night

Same reason. Filling anything in at 3am means being awake, alert and doing sums at the exact moment you are trying not to be.

Morning has a second advantage. Same-morning recall is imperfect but it is far better than evening recall, and it is done before the day has had a chance to reinterpret the night. By 9pm, a mediocre night that was followed by a difficult day has usually been remembered as a terrible one.

Under a minute, before the day starts, every day. That is the whole habit.

What happens when you miss nights

A gap or two is nothing. Leave the row blank, carry on, and the week still averages out to something the program can work from.

Stopping for a stretch is different, and the failure is specific rather than dramatic. Nothing breaks and nothing is lost - but the plan is only ever as current as the entries behind it, so with none to read there is nothing to adjust against. The window stays where it was set, which means it stops fitting your sleep as your sleep changes. The program does not fail; it goes stale.

If that happens, the fix is a week of fresh entries before changing anything, not starting the whole thing over.

Nothing can be prescribed without it

Sleep restriction means narrowing time in bed to roughly what you are actually sleeping. That sentence contains a number nobody can guess. Set the window from a feeling and you either restrict far too hard — which is unpleasant and unsafe — or barely restrict at all, and nothing changes.

The same for widening it. The window opens up when efficiency holds high across a week. Without a week of entries there is nothing to hold high.

The reason it helps beyond the arithmetic

Something happens after two or three weeks of entries that has nothing to do with calculations.

A bad night feels total while you are inside it. Tomorrow is ruined, this is never getting better, last night was the worst yet. Then you look at the row and it was your third-best week, and last Tuesday — which you have completely forgotten — was fine.

That is not positive thinking. It is the correction of a genuinely inaccurate impression, and it takes some of the force out of the catastrophizing that keeps people awake in the first place. The 3am arithmetic is arousal, and a chart is one of the few things that argues with it credibly.

Why the diary rather than a watch

Wearables estimate sleep well enough for a general picture. Two reasons CBT-I is still built on the diary.

The research is on the diary — the Consensus Sleep Diary is what the trials used, so the thresholds and the protocols are calibrated to it.

And for some people, tracker data becomes its own problem. Waking up, checking a score, and feeling worse because the number disagrees with how you feel is a real pattern with a name in the literature. A diary asks how the night was; it does not grade you at breakfast.

In Somnera

The diary takes under two minutes in the morning. Efficiency, total sleep and the trend across weeks are worked out for you, so you never do the arithmetic. There is a morning reminder if you want one, and it goes quiet on days you have already logged.

Once the program finishes you keep access to the material, and logging continues if you want it — that part is optional and separate, and what the price does and does not cover is written out in full.

If you do one thing

Start the diary before you start anything else. Two weeks of honest entries before a single change is what turns the rest of CBT-I from generic advice into something fitted to your actual nights.

What the diary feeds into is sleep restriction, and the rest of this section covers it. The other habits that decide the outcome are worth reading before week two rather than during it.

For the practical side, how to keep a sleep diary covers what to record each morning.

Somnera is a self-guided education program built on CBT-I, founded and written by Dr. Camilo Ruiz, DO, FACOI, FAASM. It is not a diagnosis and does not replace care from your own clinician. Start with the free assessment.

Citations

  • Carney CE, et al. The Consensus Sleep Diary: standardizing prospective sleep self-monitoring. Sleep. 2012;35(2):287–302. doi:10.5665/sleep.1642
  • Spielman AJ, Saskin P, Thorpy MJ. Treatment of chronic insomnia by restriction of time in bed. Sleep. 1987;10(1):45–56.
  • Edinger JD, et al. Behavioral and psychological treatments for chronic insomnia disorder in adults: an American Academy of Sleep Medicine clinical practice guideline. J Clin Sleep Med. 2021;17(2):255–262. doi:10.5664/jcsm.8986

Frequently asked questions

When you got into bed, roughly how long you took to fall asleep, how long you were awake in the night, when you got up, and how it felt. Estimates are fine — precision is not the point, the trend is.

Total time asleep divided by total time in bed. Nine hours in bed and six asleep is about 67%. Consolidated sleep usually sits above 85%.

For total sleep, often. But CBT-I is built on the diary, which captures your experience of the night — and there is a real risk in fixating on device data that does not match how you feel.

One to two weeks before anything is set, and then throughout, because the sleep window is adjusted from what the last week showed.

Leave it blank and carry on. A gap costs almost nothing. Abandoning the diary because of a gap costs you the program.

Nobody does, and you are not expected to. Round to the nearest fifteen or thirty minutes. Consistency across the week is what the calculation needs; precision on any single night adds nothing and checking the clock to get it makes the night worse.

In the morning, always. Logging at 3am means being awake, alert and doing arithmetic at the exact moment you are trying not to be - and clock-watching is one of the things that turns a brief waking into an hour.

Those are the most important entries you will make. Bad nights carry the information the window is set from, and they are the ones people skip, which quietly biases the record toward looking better than the week was.

More in Sleep restriction

Related posts