360 ‒ How to change your habits: why they form and how to build or break them
- Guest
- Charles Duhigg
- Read time
- ~4 min
- Uploaded
- 11 Aug 2025
- Duration
- 2:27:34
- Views
- 87.5K
KEYPOINTS
- 01
Every habit is a loop of cue, routine and reward, and habits make up roughly 40–45% of daily behaviour.
Duhigg, author of The Power of Habit, cites Wendy Wood's research at USC for the 40–45% figure. The basal ganglia strengthens the links between cue, behaviour and reward until the behaviour runs automatically. He became interested after an Army major in Iraq told him 'the military is a giant habit change machine,' and after realising that despite professional success he couldn't make himself run or lose weight.
- 02
Positive reinforcement is roughly 20 times more effective than punishment.
In Duhigg's example, you'd need about 20 cane strikes to match the motivation of one post-run smoothie. Negative rewards work best as tension imposed before the behaviour and relieved by doing it, like the interrogator's 'I'll let you out of the box if you talk.' He suggests the brain prices punishments accurately but overvalues rewards, especially unexpected ones—the slot-machine effect—so pair any punishment with a reward.
- 03
Rewards must be immediate, then shift from extrinsic to intrinsic.
Newcomers to exercise don't anticipate the runner's high, so it can't motivate them yet. Start with a smoothie, a long shower or praise. Over time the felt reward—accomplishment, endorphins, relief—takes over. Many people sabotage new habits by accidentally punishing themselves, such as rushing a shower and running late after a morning run.
- 04
Cues fall into five types; stack as many as possible for a new habit.
The five: time of day, place, emotional state, other people, and a preceding ritual. For a new running habit: alarm at 7, shoes and clothes by the bed, a friend waiting at the corner, and a reward decided in advance. Attia finds other people—accountability partners or trainers—among the most powerful cues.
- 05
You don't break a bad habit; you change it.
Neural pathways for habits rarely thin, which is why a lapsed meditator restarts quickly and an ex-smoker still craves a cigarette after breakfast. Identify the cue and the reward, then insert a new routine that delivers a similar reward. Duhigg says physical nicotine dependence fades within about 100 hours; the habit loop is what persists.
- 06
Willpower is like a muscle: trainable, but it fatigues within a day.
Errors cluster in surgeons' second or third operations or very long cases; affairs among lawyers and doctors cluster after 9–9:30 pm. Plan for depleted moments. Duhigg says willpower has low heritability and is learned through parenting, sport and practice.
- 07
Relapse is data; treat yourself like a scientist.
Prochaska's work shows committed smokers quit about seven times before it sticks. Early attempts rely on willpower; later ones succeed with 'implementation intentions,' such as keeping candy ready for the mother-in-law visit that triggers a cigarette. Attia coaches patients never to judge a missed day—just don't let one become two—because self-blame floods the brain with negative reward.
- 08
Mental habits matter as much as behavioural ones, and contemplative routines enable deep thinking.
The mind produces something like 47 thoughts a minute. Serial innovators build rituals—a weekly reflection or a walk before brainstorming—that trigger depth on demand. 'Making the right choice is so much more powerful than executing really well on the wrong choice.' Cal Newport's work explains why walks feel unproductive compared with clearing email, and why they aren't.
- 09
Decision fatigue is real; design which decisions you make, and when.
Obama wore identical suits to save decisions for important ones. Decide in a 'cold' state—choose lunch after breakfast, not when starving. Attia admits he struggles to pick a restaurant late in the day; Duhigg says the fix is to plan for frailties, not deny them.
- 10
Motivation is the prerequisite no coaching can supply, and it changes across life.
Duhigg argues information and reinforcement—whether from a book like Outlive or an AI coach—only help someone already motivated. Tony Robbins' gift is generating motivation, not sophisticated advice. Attia's reasons for training at 52 differ completely from those at 20, so keep experimenting with what motivates you.
EXAMPLES
- 01
Soap in South Asia (Val Curtis's research)
Families saved soap for guests or prayer. Health messaging failed. What worked: a less perfumed scent that felt medicinal, and ads framing handwashing as being a good parent—an identity reward—built on disgust research.
- 02
Marines and IED dust
Training used charges that made dust clouds without noise or fire, because dust is a more reliable cue than sound. The reward was squad-mates saying 'good job,' and camaraderie becomes why soldiers stay.
- 03
Savings experiment (Wharton)
In cash economies in South America, a bank clerk praising savers each week raised savings by about 40% compared with the same box and information alone.
- 04
Qantas Flight 32
After an engine blew apart and 12 of 14 key systems failed, Captain Richard de Crespigny reframed the A380 as a Cessna and landed all 469 people safely; simulators usually crash. Before every flight he quizzed co-pilots on emergencies and argued with their answers.
- 05
Toothpaste mint and vitamin gummies
Mint's tingle made brushing feel like it was working; sugar coatings boosted vitamin use. Both manufacture immediate rewards.
- 06
Katie Milkman's 'temptation bundling'
She listened to Harry Potter audiobooks only on the bike. Attia does the same with favourite podcasts during zone 2.
OTHER RELEVANT
- 01
Start additive habits absurdly small
Attia's prescription for a cardio-hater: three 15-minute rides a week, tracking only time; showing up is the win. For subtractive habits, control your environment—an hour of grocery-store willpower buys 167 easy hours.
- 02
Make a game board for long-horizon goals
Duhigg and Attia both use spreadsheets to gamify saving. Duhigg updates his weekly, tracking 1–3-year earnings next to net worth so market dips don't kill motivation.
- 03
Parent for agency
Praise effort ('you worked hard') rather than traits (Carol Dweck). Narrate your own cues and rewards out loud, and admit failures as data points.
- 04
Find a reward you actually want
Kale smoothies rarely work as rewards; experiment. For a pill with no felt benefit, attach a meaning—'I'm staying around for my kids.' Social praise turns intrinsic fastest; AI encouragement may help some people.