In 2025, sleep scientist Dr Rebecca Robbins of Brigham and Women's Hospital and colleagues analysed over three million nights of sleep data from 21,000 people worldwide. Their study, published in Scientific Reports, found the snooze button was pressed on nearly 56% of all mornings, and 45% of people snoozed on more than 80% of their mornings, losing around 20 minutes a day [1]. Five full days a year.
The decision to snooze isn't really made by you. It's made by a groggy, half-conscious version of you experiencing sleep inertia: the fog that sleep researchers Patricia Tassi and Alain Muzet described as impairing reaction time, memory and judgement for 30–60 minutes after waking, sometimes longer [2]. Asking that version of you to make a good long-term decision is like asking someone at 2am after three pints. Willpower isn't the tool for this job.
Longer than the snooze interval, which is the whole problem. When Megan Jewett's team at Harvard woke people after a full eight hours in bed and tested them across the morning, alertness and thinking recovered along a curve that was steep for the first half hour and did not level off for two to four hours [5]. The nine minutes a snooze buys you are spent entirely inside the worst part of that curve. You wake up from the second alarm no further along than you were, having restarted the process from scratch. There is a longer piece on sleep inertia if you want the full picture; the short version is that the first minutes after any alarm are the low point of your judgement for the entire day.
A few details from Robbins' study are worth sitting with [1]. Snoozing was heaviest on weekdays and lightest on Saturdays, which suggests it is a response to alarms set earlier than the body wants, not a personality trait. It was heaviest among people who slept longest, which sounds backwards until you remember that the people sleeping nine hours are often the ones whose clocks run late. And the sleep between snoozes was, in Robbins' words, "fragmented, light sleep" that "disrupts some of the most important stages of sleep". Snooze three times and you've woken up four times in one morning.
For balance: a 2024 study led by Stockholm University's Dr Tina Sundelin found habitual snoozers lose only about six minutes of sleep from thirty minutes of snoozing and suffer no measurable harm to mood or cortisol; they even did slightly better on cognitive tests straight after waking than when woken abruptly [3]. Snoozing won't ruin you. It just quietly taxes you, every single day, and it hands the most important decision of the morning to the least capable person in the room.
The oldest advice in the book, putting the alarm across the room, works because it stops negotiating with sleepy-you and simply requires a behaviour: stand up. Every reliable fix for snoozing has the same shape.
Make getting up the path of least resistance, and make staying down annoying.
That's the principle behind mission-based alarms: dismissing it doesn't end the morning. The stop button only quiets it for ninety seconds, and then it's back. A photo of your journal and pen. A few hundred counted steps. A timed breathing session. By the time the mission is done, sleep inertia has lifted and the decision to get up has already been executed by the only version of you that could be trusted with it: the one who set the alarm last night.
People who stop snoozing rarely describe it as discipline. They describe it as identity, what James Clear calls identity-based habits: every action is "a vote for the type of person you wish to become" [4]. Each morning the alarm wins, that identity gets one vote stronger. The trick is to rig the first vote so it cannot be lost from bed.
Not dramatically. A 2024 study found habitual snoozers lost about six minutes of sleep and showed no harm to mood or stress hormones from thirty minutes of snoozing. The cost is quieter: fragmented light sleep, around twenty minutes a day for heavy snoozers, and a decision made in the worst-judgement window of the day that often ends with going back to sleep.
Because sleep inertia impairs judgement and memory for the first minutes after waking, and the decision to lie back down is made in that window. If you can dismiss the alarm from bed, nothing stands between you and the second sleep. Fixes that work make dismissal require getting up.
The worst of it is the first fifteen to thirty minutes; measured alertness and performance keep improving for two to four hours after waking. That is why the snooze decision is never a good one: it is made at the low point.
Yes, better than any amount of resolve, because it stops negotiating with the sleepy version of you and requires a behaviour. Its weakness is that the behaviour ends the moment you have the phone in your hand. Mission alarms extend it: the alarm is not done until something is.