Clementine

The 3am Report · 2026

It’s not a sleep problem

What half a million nights tell us about the time women don’t get.

  • By Kim Palmer, founder of Clementine
  • Published
  • About 12 minutes to read
  • Free to read, free to quote

The finding

Why I’m sharing this#

I’m sharing this because it underpins our why: why we exist, and why we try to make things that fit around how women actually live. It shapes how we talk to women, too. It’s not just a report. It’s our calling card, and our north star.

It matters for another reason, too. We’ve spent years helping women, and we’ve been quite low-key about it. This was a chance to see exactly how women used our tools, and how they didn’t. And it busts a few myths along the way.

For six years, we ran an app of hypnotherapy audio. It wasn’t sold as a sleep app. It was there to help you sleep, stress less and build your confidence. Sessions for the morning, for work, for a wobbly afternoon, for bedtime. You could use it any time of day.

Women used it for sleep. Overwhelmingly.

About seven in ten of the plays we can name were sleep sessions. More than half of all listening happened between 9pm and 5am. We offered help for the whole day, and women brought it to bed.

That was the first clue. Every time a woman pressed play, the app noted the time. Over six years, that’s 541,097 plays by 25,950 people.

Nine in ten of those people are women. Most are between 25 and 54.

This year I sat down with that data properly for the first time. Not the dashboard. The actual nights.

I expected it to tell me about sleep. It didn’t, really. It told me something I think is bigger.

I don’t think most women have a sleep problem. I think they have a no-time-to-think problem.

The day never gives them a gap to take in what’s happened. So it all arrives when their head hits the pillow. The conversation that went wrong. The thing they should have said no to. Everything that’s happening tomorrow.

And then we tell them a good sleeper drops off in minutes. So on top of everything else, they feel they’re failing at sleep too.

How can you sleep when you haven’t had a minute to process your day?

That’s the story I see in these numbers. I’ll show you the numbers, and you can see if you agree.

Part one

No time in the day

Finding one

Early evening is the time she doesn’t get#

When do women get time to themselves? Not in the early evening: it’s the quietest time of the whole day.

Remember, this app had help for stress and confidence that women could use any time. So you might guess the quietest time would be 3am. It isn’t. It’s early evening. The single quietest hour is 7pm.

For women in the UK, the four hours from 5pm to 9pm are under 6% of all their listening. 3am is about three times busier than 6pm.

The busiest hour of all is 11pm. That one hour is 15% of everything UK women played.

For UK women, the single hour of 11pm is 15% of all their listening, while the four hours from 5pm to 9pm together are under 6%.0%5%10%15%11pm11pm5–9pm5pm–9pmClementine, 2020–2026 · UK women, own local time
UK women, in their own local time. The single hour of 11pm is 15% of all their listening; the four hours from 5pm to 9pm together are under 6%. Clementine, 2020–2026.

Think about what 5pm to 9pm looks like for most women. Tea. Bath. Homework. The last emails. Tidying up. Getting everyone else to bed. There’s no gap for her.

Then 11pm. Everyone else is done. It’s the first hour of the day that’s hers. And that’s when she reaches for help.

She doesn’t get an evening. She gets an 11pm.

Kim Palmer

Finding two

Mornings are for armour#

What do women listen to in the morning? Something that tells them they can handle the day.

The mornings tell the same story from the other end.

Our app has short affirmations with names like “I got this sh*t” and “Today is mine to own”. They’re not for sleep. Thousands of women used them. Where I know where a woman lives and can use her own local time, about half of the days she used them, she did it between 5am and 9am. Most of the rest was during the working day. Hardly any was in the evening or at night.

Before the day starts, thousands of women play a voice telling them they can handle it.

So she gets herself ready for the day. She gets everyone else through it. What she doesn’t get is any time to take it in.

Part two

So the day arrives at bedtime

Finding three

Night is for sorting out the day#

Why do I go over what I should have said at night? Because night is when women sort out the day.

We have sessions about boundaries: learning to say no. I’d have guessed women would use those in the day, before a hard conversation. They don’t.

Where I know a woman’s local time, about 80% of the days she used them, it was at night. Mostly between 9pm and 1am.

Working out where she should have said no happens in bed, late at night.

The sleep sessions women play most tell the same story. They’re not about relaxing the body. They’re about calming the mind, trusting yourself, feeling more confident. (To be fair, that’s partly because those are the sessions we made most of. But they’re the ones women chose, again and again.)

What she playedWhen, in her own local time
Affirmations for the dayAbout half 5am–9am, a third during the working day
Boundaries and saying noAbout 80% at night

She learns to say no at 11pm, after the conversation.

Kim Palmer

Finding four

Tomorrow keeps her awake#

Why can’t I sleep the night before something important? Because tomorrow keeps women awake as much as today.

It’s not just today she’s processing. It’s tomorrow.

The hard nights aren’t random. They’re the nights before something. These figures are for women in the UK only, because these are UK dates.

The nightCompared with a normal night
The night before A-level results34% busier (every year I looked)
The first Monday back in January31% busier (5 years out of 6)
The Sunday before itUsually busier (4 years out of 6)
Christmas Day48% quieter (every year I looked)
Easter Sunday23% quieter (every year I looked)
For UK women, the night before A-level results is 34% busier than a normal night and the first Monday back in January 31% busier; Easter Sunday is 23% quieter and Christmas Day 48% quieter.an ordinary night lands in hereThe night before A-level results+34%The first Monday back in January+31%Easter Sunday−23%Christmas Day−48%← quieter than usualbusier than usual →Clementine, 2020–2026 · UK women
UK women. Each night against the same night of the week over the four weeks before. Any single night can land a quarter above or below by chance (the shaded band), which is why only nights that moved the same way year after year are shown. Clementine, 2020–2026.

When there’s nothing to get up for, fewer women reach for help. When the diary comes back, more do.

The night before A-level results is busier every single year, even though teenagers don’t use our app. I think that’s mums, lying awake for their kids.

The quietest nights of the year are the ones with nothing to get up for.

Kim Palmer

Finding five

The worst night of the week is Monday#

What’s the worst night of the week for sleep? Monday, not Sunday.

Everyone talks about Sunday-night dread. In our data, Monday is the night women reach for help most. Saturday is the quietest.

For women in the UK, Monday night is 57% busier than Saturday night. It holds for women abroad too, in their own local time.

For UK women, Monday night is 57% busier than Saturday night, the quietest night of the week.050Sat150MonMondaySatSaturdayClementine, 2020–2026 · UK women
UK women. Monday night against Saturday night, the quietest of the week. A night runs 5pm to 5am, so a 3am play on Tuesday counts as Monday night. Clementine, 2020–2026.

By Monday night, there’s a whole day to process and a whole week ahead. Research on work stress finds the body’s stress response is higher on weekdays, and higher on Monday than Sunday for people under pressure at work.

And Saturday? It looks like the one night she has nothing to get through.

Saturday is the only night with nothing waiting in the morning.

Kim Palmer

Part three

And then we tell her to switch off

Finding six

Nobody gave her the time, then we timed her#

How long should it take to fall asleep? Longer than you’ve been told.

So she lies down with a whole day still to take in. And what does she hear? That a good sleeper drops off in minutes. That she should be asleep by now. Track it. Score it. Do better.

That’s not even true. A panel of sleep experts in the US (the National Sleep Foundation) says taking up to 15 minutes to fall asleep is a sign of good sleep, and up to 30 minutes is fine at any age. It’s only when it regularly takes 45 minutes to an hour that they call it a sign of poor sleep.

The worry itself makes it worse. Sleep doctors have a word for people who get more anxious because of their sleep tracker: orthosomnia. Some of them spend longer in bed chasing a better score, and sleep worse.

Here’s the thing I find most telling. The advice sleep clinics actually give for a racing mind isn’t “try harder to sleep”. It’s to make time to process the day before bed. The NHS suggests writing a list of what you need to do tomorrow before you go to sleep. One sleep clinic technique asks you to set aside about fifteen minutes early in the evening to go through what’s on your mind. In one lab study, people who wrote a to-do list at bedtime fell asleep faster than people who didn’t.

In other words: the experts know the mind needs time before sleep will come. Most women just don’t get that time.

Nobody gave her the time. Then we timed her.

Kim Palmer

Part four

The other one

Finding seven

3am finds women at every age#

Is waking at 3am a menopause thing? In our data, it happens at every age.

Not everything in the data is about bedtime. There’s a second pattern, and it’s different.

Waking at 3am and not getting back to sleep is often talked about as a menopause thing. I expected our data to show that. It didn’t.

I matched about a third of our listeners to an estimated age. Then I looked at how much of each woman’s listening happened after midnight.

Roughly a fifth of a woman’s listening happens after midnight at every estimated age: 22% for 18 to 24, 18% for 25 to 34, 17% for 35 to 44, 18% for 45 to 54, 20% for 55 to 64 and 24% for 65 and over.0%10%20%30%About a fifth22%18–2418%25–3417%35–4418%45–5420%55–6424%65+Clementine, 2020–2026 · estimated ages · 65+ is a small group
Share of each woman’s listening after midnight, by estimated age (Mailchimp’s estimate, matched for about a third of listeners). The 65+ group is small: treat it with care. Worked out before the time-zone correction, which moved the overall after-midnight share by only about a point. Clementine, 2020–2026.

It’s roughly a fifth at every age. It rises a bit from 55, which fits what’s known about sleep getting lighter as we get older. (I worked these out before I corrected for time zones. When I did correct, the overall after-midnight share only moved by about a point, so I don’t think it changes the picture. But treat the exact figures as approximate.)

The 3am listening barely changes across the week, either. For UK women, bedtime listening is about twice as busy on the busiest night as the quietest. Between 1am and 5am, the difference is only 13%. Whatever wakes her at 3am doesn’t take weekends off. So I don’t think this one is about her day. It seems to be something else.

Where perimenopause fits. It matters, just not as the whole story. Perimenopause can start in the late 30s or early 40s. Mine did, in my early 40s. That’s years before most women expect it. Almost half our listeners are at an age where it could be part of the picture, and lots won’t have made the link. Waking in the night is one of its most common signs.

And the busiest decade. Women aged 35 to 54 do the most listening at bedtime and the least after midnight. Work, kids, often parents too. Their hard time is the end of the day. Which takes us back to part two.

It isn’t her age. It’s every age.

Kim Palmer

Part five

It comes in patches

Finding eight

Nobody sleeps badly every night#

Do people with bad sleep have it every night? Almost none of them do.

This is the one that surprised me most.

Of the 26,000 people who used our app, 64 used it most nights. Sixty-four.

I looked at the 9,288 people who played something at least five times:

Of the 9,288 people who played something at least five times, 45% came and went, 44% had one burst then stopped, 11% used it steadily now and then, and under 1% used it most nights for two months or more.45%Came and went44%One burst, then stopped11%Steady, now and thenUsed it most nights: 64 people (0.7%)Clementine, 2020–2026 · 9,288 people, 5+ plays
How the 9,288 people who played something at least five times used the app. Under 1% used it most nights for two months or more. Clementine, 2020–2026.

The typical woman did three quarters of her listening inside a single fortnight.

One woman’s year, anonymised: quiet for three and a half months. Then fifteen plays in six weeks in February and March. Then quiet again. You can see exactly when life got heavy. And when it eased.

We don’t sleep badly every night. We sleep badly in patches, when there’s more to process than usual. The week before the big thing. The January wall. The six weeks when life fell apart a bit.

She didn’t lapse. The patch ended.

Kim Palmer

Finding nine

The run-up to Christmas#

Is sleep worse in the dark months? It’s the run-up to Christmas, not the dark.

I expected the dark months to be the story here. It’s more interesting than that.

For UK women, the busiest months aren’t the darkest ones. They’re November and December. Each woman listens about 20% more than in an average month. July is the quietest. January to March are about average.

(The app as a whole was busiest in January to March. But that’s because more women joined or came back in the new year, not because each woman listened more.)

So is it the dark, or is it the time of year? Our women abroad help answer that.

Where she livesHer busiest monthsHer quietest months
UKNovember and DecemberJuly
US and CanadaNovemberJanuary, August
Ireland and EuropeDecemberAround average all year
Australia and New ZealandJanuary (their summer)March to April, July

In Australia and New Zealand, winter is June to August. If darkness were the main thing, we’d see their listening rise then. It doesn’t. (There are only about 170 of them, so I wouldn’t lean on it hard.)

What every group shares is the end of the year. The run-up to Christmas. The list, the school events, the work deadlines, the family.

That fits everything else in this report. It doesn’t look like the dark that keeps women up. It looks like how much is on.

It isn’t the dark. It’s how much is on.

Kim Palmer

Finding ten

Things that turned out not to be true#

Does the full moon affect sleep? Not in our data, and nor do payday, storms or heatwaves.

I checked a load of things people swear keep them awake. Most of them don’t show up at all.

What people sayWhat the data shows
The full moon70 full moons. No real difference.
Valentine’s nightSlightly quieter than normal, not busier.
Payday68 month-ends. No difference.
StormsSix named storms. No pattern.
Budget day, election nightNothing.
Eurovision, the football, WimbledonNothing you’d notice.
HeatwavesNo difference.
Blue MondayNothing.

What keeps women awake isn’t the moon, the weather or the news. It’s everything they haven’t had time to think about.

It isn’t the moon. It’s everything she hasn’t had time to think about.

Kim Palmer

Questions

The short answers#

Why do I wake up at 3am?

Waking in the night isn’t a sign something is wrong with you: it happens to women of every age. In our data, women did roughly a fifth of their listening after midnight at every age from 18 to over 65, and 3am listening barely changed between weeknights and weekends. So it isn’t only about the day she’s had. Perimenopause, which can start in the late 30s or early 40s, makes night waking more common. See finding seven.

Why can’t I sleep when my mind is racing?

Because the day hasn’t been processed yet. The best-known explanation for insomnia is a busy mind going over things, keeping the body switched on. In our data, sessions about saying no were played at night about 80% of the days they were used. Sleep clinics suggest making time for your thoughts before bed, like writing tomorrow’s to-do list. See finding three and finding six.

Why is it worse the night before something important?

Because tomorrow keeps you awake as much as today. For UK women, the night before A-level results was 34% busier than a normal night, every year, and the first Monday back in January 31% busier. Christmas Day, with nothing to get up for, was 48% quieter. See finding four.

Is waking at 3am a sign of perimenopause?

It can be, but it isn’t the whole story. Perimenopause can begin in the late 30s or early 40s, years before most women expect it, and waking in the night is one of its most common signs. But in our data, women aged 18 to 24 did just as much of their listening after midnight (22%) as women in their fifties (18–20%). See finding seven.

Method

How I counted

I want to be straight about what this is.

Where it’s from.
Every play in the Clementine app from September 2020 to September 2026. I only counted plays of a minute or more, so taps and quick previews don’t count. That left 541,097 plays by 25,950 people.
Times.
Times are each woman’s own local time where I know where she lives (see the note on time zones below). A “night” runs from 5pm to 5am, so a 3am play on Tuesday counts as Monday night.
Unusual nights.
I compared each night with the same night of the week over the four weeks before. Any single night can be a quarter higher or lower just by chance, so I only report things that happened year after year.
Who they are.
The app never asked for age or gender. Mailchimp estimates both for the people on our mailing list. By that estimate, nine in ten are women and most are 25 to 54. For the age chapter, I matched about a third of listeners to an estimated age by their email address. These are estimates, not facts people gave us.
Counting women, not just plays.
Where a few very keen listeners could skew things, I counted women and nights instead of plays.
What’s missing.
Three weeks in July 2024 when the app wasn’t recording properly are left out. So are a few nights that look like marketing emails, not real life. A third of plays since 2023 are for sessions that have since been retired and have no names.
Privacy.
All of this was done on my own computer. Nobody is named or identifiable. The email files used for the age match are being deleted.

And once more: this is data. It shows when women reached for help. It can’t tell us why. Where I say what I think is going on, that’s me reading the pattern, backed where I can by other people’s research.

Sources

The research I’ve leaned on

  1. Harvey (2002), Behaviour Research and Therapy — the best-known model of insomnia, where a busy mind keeps the body switched on. Source →
  2. Johnson & Whisman (2013), Personality and Individual Differences — women tend to go over things in their heads more than men (a small but consistent difference). Source →
  3. Schlotz et al. (2004), Psychosomatic Medicine; Devereux et al. (2011), Psychological Reports — the body’s stress response is higher on weekdays, and on Mondays for people under work pressure. Source →
  4. Heacock et al. (2022), Frontiers in Physiology — people sleep longer on most public holidays, Christmas especially. Source →
  5. Office for National Statistics (2023, 2015) — time use and leisure time in the UK. Source →
  6. Ohayon et al. (2017), Sleep Health — the National Sleep Foundation’s view of what normal sleep looks like, including how long it’s fine to take to drop off. Source →
  7. Baron et al. (2017), Journal of Clinical Sleep Medicine — case reports that named “orthosomnia”. Source →
  8. Carney & Waters (2006), Behavioral Sleep Medicine — “constructive worry”, setting time aside early in the evening. Source →
  9. Scullin et al. (2018), Journal of Experimental Psychology: General — writing a to-do list at bedtime helped people fall asleep faster. Source →
  10. NHS Every Mind Matters — make a list of tomorrow’s jobs before bed. Source →
  11. Kravitz et al. (2008), Sleep; Baker et al. (2018), Nature and Science of Sleep — sleep across the menopause transition. Source →
  12. Johnsen et al. (2012), BMC Psychiatry — more sleep problems in winter in northern Norway. Source →

For press

Everything you need to file this today

All of it is free to use with a link back to this page. No embargo, no form, no sign-up.

Press contact: Kim Palmer · hello@clementineapp.co.uk · usually back within the day

  • The full report as a PDF, with page numbers and the contact on the cover on request
  • All the charts as a zip: 1200px PNG and SVG on request
  • Three high-resolution photographs: Kim, the kit, the cards on request

40-word bio

Kim Palmer founded Clementine in 2017 to make hypnotherapy for women something they would actually use. Over six years its app recorded 541,097 plays by 25,950 people, nine in ten of them women. She now makes a screen-free way to end the day.

60-word boilerplate

Clementine is a British company making things that help women sleep. Founded in 2017 by Kim Palmer, it spent six years as a hypnotherapy app for women. It now makes the Unwind & Sleep Kit: a headband you can sleep in, and session cards you tap to play. There is no subscription. You buy it once. clementinerituals.com

Eight stories in here, if you want one

Every one of these is a finding in the report above, with the number attached. Take any of them and run it your way.

And what else we can give you

  • Real women, on the record. We have a community who talk to us honestly about their nights, and plenty are willing to be quoted or photographed. Tell us the angle (a new mother, a woman in perimenopause, a carer, someone on shifts) and we’ll find her. We’ll need a few days’ notice, and we won’t put anyone forward who hasn’t agreed to it.
  • Qualified therapists for comment. The hypnotherapists who write and record our sessions are available to talk to: on what women actually bring to them at night, and on what does and doesn’t help.
  • Kim for interview, including early mornings, and happy to be put up against a sleep scientist.

Free to use, with a link to this page.

The PDF is a courtesy, not the destination: every link we send points here.

Sign-off

What I’m doing with this

Honestly, everything. We’ve rebuilt the whole business around what these numbers showed us.

If women reach for help at the end of a day that never gave them a minute, the help can’t be one more screen to scroll. If bad sleep comes in patches, it shouldn’t be one more subscription to pay for in the months you don’t need it. So what we make now fits around how women actually live.

And we’re sharing it, with you and with the press, because some of this matters well beyond Clementine. Women aren’t failing at sleep. Most of them just haven’t had a minute to think. That’s worth saying out loud.

Clementine makes hypnotherapy sessions for women, and a headband you can sleep in. We’ve been doing it since 2017. What we make →

Want the findings as they come?

One email a month while we write next year’s. Not a sales list.

Email us to join

Your basket

Your basket is empty.

See the Unwind & Sleep Kit