A member gets a push notification. She orders the usual drink and watches her Star count move. Repeat that for a few months and it stops being a decision. It is a habit. Charles Duhigg called the pattern behind it the habit loop. Loyalty mechanics map onto it, so a loyalty team needs to know how it forms, how long it takes and how to measure it.
What is a habit loop?
A habit loop is the three-part pattern behind an automatic behavior: a cue, a routine and a reward. Duhigg made the term famous in his 2012 book The Power of Habit. He told NPR the cue triggers the behavior, the routine is the behavior itself, and the reward tells the brain whether to keep the habit (NPR, 2012). James Clear, author of Atomic Habits, lists five kinds of cue: time, location, a preceding event, an emotional state and other people (James Clear).
His standard example is toothpaste. Duhigg says the ad executive Claude Hopkins sold Pepsodent by pointing people to the film they could feel on their teeth. The reward was the tingling, clean feeling after brushing (NPR). Over time, Duhigg says, people begin craving the reward when they see the cue, and the craving makes the habit run on its own.
The three-step form is Duhigg’s packaging. Researchers describe the same thing in terms of context. A 2016 review in the Annual Review of Psychology calls habits the efficient, default mode of response (Wood and Rünger, 2016). Wendy Wood and David Neal wrote in 2007 that habits form as people learn links between a response and the features of the setting where it happened. Once the habit exists, the setting triggers the response without a goal in between (Wood and Neal, 2007).
How common are habits, and where do they live in the brain?
Habits fill a large share of daily life. Wood, Jeffrey Quinn and Deborah Kashy ran two diary studies in which people reported their activity every hour. Between a third and a half of the behaviors listed were habits, meaning done just about every day, usually in the same place. The share was 35% in the first study and 43% in the second (Wood, Quinn and Kashy, 2002).
During habits, people tended to think about something unrelated to what they were doing. They also reported less stress than during behavior done less often or in shifting settings (same paper). The habit does the work, and attention goes elsewhere.
MIT neuroscientists tested how deep the pattern runs. Rats learned to run a maze for a reward and kept running it after the rewards stopped. They kept running it even after their chocolate milk was laced with a substance that made them mildly ill, though they stopped drinking it (MIT News, 2012). The pathways that encode habits appear to sit in deep structures called the basal ganglia. A prefrontal region decides moment by moment which habit runs.
Switching that region off dropped the habit almost at once. The rats then formed a new habit. Switching the region off again sent them straight back to the original one. One researcher said the old habit was never really forgotten (same article).
How long does it take to form a habit?
It takes longer than most people expect, and the range is wide. Phillippa Lally and her team at University College London asked 96 volunteers to pick one eating, drinking or activity behavior. They did it daily in the same context, such as after breakfast, for 12 weeks. Each day they rated how automatic it felt. For the 39 people whose data fit the model well, the median time to reach 95% of their personal plateau was 66 days. The range ran from 18 to 254 days (Lally et al., 2010).
One missed day did not derail anyone. A single skipped chance did not change how fast a habit formed. Most volunteers were postgraduate students who picked their own tasks. The numbers fit that group and will differ for others.
How do you break a habit loop?
Change the cue, because the cue is where the habit lives. Wood and Neal point to a study of students who moved to a new university. Students with strong habits whose settings stayed the same kept them. Students with strong habits whose settings changed dropped them, and their actions followed their new goals instead (Wood and Neal).
Goals alone are the weaker tool. Wood and Neal conclude that goals gain little traction on the slowly built links between a setting and a response. They propose changing exposure to the cue instead, such as keeping a tempting snack out of sight. The MIT result points the same way. A broken habit is replaced by a new one and not erased, so the old pattern returns when the cue does.
How do loyalty programs use the habit loop?
Every working program maps its mechanics onto the three parts. The table shows one mapping and the metric a team can watch for each part.
| Part | Program mechanic | What to measure |
|---|---|---|
| Cue | A push notification, a bonus-points day, a streak reminder, a time of day | Share of sends followed by a visit within 24 hours |
| Routine | The purchase, the lesson, the check-in, the scan | Visits per member per week, and weeks in a row with a visit |
| Reward | Points, Stars, a streak count, a tier step, a surprise | Redemption rate and repeat visits after the reward lands |
Starbucks’s published terms contain all three parts. They say special offers go out through the app, including by push notification, and can be customized to a member’s purchase behavior. Some carry the chance to earn Bonus Stars. A Green Level member’s Double Star Day is personalized and selected by Starbucks (Starbucks Rewards Terms). The notification is the cue. The order is the routine. The Stars are the reward.
Duolingo is the clearest consumer case. Its annual report says each lesson awards points and extends a streak. It also sends push notifications with lesson reminders (Form 10-K, 2025). In its second-quarter 2026 letter, Duolingo put its retention rate for current users at an all-time high of 84%. It calls that rate one of its most important leading indicators (Q2 2026 letter). Our post on what Duolingo gets right about loyalty goes further.
How can a loyalty team measure a habit?
Count rhythm as well as purchases. A member who visits most weeks, in the same pattern, is closer to a habit than one who buys twenty times in a single week. The team sets a rule, such as a visit in at least 6 of the last 8 weeks, and tracks how many members meet it.
For illustration, a program has 120,000 active members, and 30,000 meet the 6-of-8 rule. That is a 25% habit base. The team tests a reminder sent at each member’s usual purchase hour. It sends it to 5,000 members and holds back 5,000 at random. If 41% of the notified group visit that week against 38% of the holdout, the cue added 3 points. The holdout is what shows the cue caused the visit.
Two cautions follow from the research. Expect a new habit to take about two months to settle, so judge a new mechanic over a full quarter. And treat any change to the app, the checkout or the earn flow as a change of context. Strong habits in the transfer study broke when the setting changed. Our guide to gamification software covers streaks and challenges in more depth.
How GRAVTY handles the habit loop
Each part of the loop is a rule a program team has to set and change. The GRAVTY technology page says business teams set loyalty logic without code, including rewards, recognition and redemption for any segment. It also says every behavior, influence and transaction becomes first-party intelligence, with targeting and game mechanics built on it. A team can change a cue or a reward and watch the habit base move.
Frequently asked questions
What are the three parts of a habit loop?
The three parts are a cue, a routine and a reward. Charles Duhigg described them in The Power of Habit. The cue triggers the behavior, the routine is the behavior, and the reward tells the brain whether to keep it (NPR, 2012). Some writers add a fourth step, craving.
Who created the habit loop?
Charles Duhigg popularized the habit loop in his 2012 book The Power of Habit. The ideas under it come from earlier habit research, such as the context-cue model of Wendy Wood and David Neal (Wood and Neal, 2007).
How long does it take to form a habit?
It took a median of 66 days in a University College London study of everyday habits, with a range of 18 to 254 days (Lally et al., 2010). The 39 volunteers in that analysis each chose a daily behavior and rated how automatic it felt over 12 weeks. A single missed day made little difference.
Can you break a habit loop?
Yes, and changing the cue works best. Strong habits held when students moved to a new university with the same settings and broke when the settings changed (Wood and Neal, 2007). Habits do not vanish, though. In the MIT rat study, a suppressed habit returned once the new one was switched off (MIT News, 2012).
How do loyalty programs use habit loops?
They pair a cue, a routine and a reward on a steady rhythm. Starbucks sends personalized offers through its app, including push notifications, and awards Bonus Stars (Starbucks Rewards Terms). Duolingo sends lesson reminders and extends a streak with each completed lesson (Duolingo Form 10-K). Teams measure the loop with visit rhythm and holdout tests.
