The setup
The four rules are simple: no liquid calories, no snacking, all meals within a 10-hour window, and daily exercise. Each day, you log whether you followed each rule. Yes or no. No partial credit.
After 30 days of logging, you have 120 individual data points — four rules times 30 days. Patterns emerge that are impossible to see in the moment but obvious in aggregate.
The hardest rule is not what you expect
Most people, before they start, predict that exercise will be the hardest rule to follow. It requires the most effort, the most time, and the most deliberate action.
The data says otherwise.
Across our beta users, No Snacking has the lowest follow rate of the four rules — followed on roughly 71 percent of logged days compared to over 85 percent for exercise.
The reason, on reflection, makes sense. Exercise is a discrete event. You either went for a walk or you did not. The decision happens once and then it is done.
Snacking is different. It is not one decision but dozens — every time you pass the kitchen, open the fridge, or sit near food at your desk. Each of those moments is a separate decision point. More decision points means more opportunities to slip. This is one of the more counterintuitive things why diets fail reveals: it is rarely the hard things that break us.
"Exercise has higher follow rates than No Snacking — not because it's easier, but because it's one decision per day instead of dozens."
Saturday is the hardest day
The day of week data is consistent across users. Weekday performance is significantly stronger than weekend performance — typically 15 to 20 percentage points higher.
Saturday is the hardest day of the week. Sunday is the second hardest.
The mechanisms are straightforward. Weekdays have structure — mealtimes are anchored to work schedules, snacking opportunities are limited by meetings and tasks, exercise fits into commutes or lunch breaks. Weekends have none of this. The structure disappears and with it the automatic compliance that structure creates.
The practical implication: if you are going to plan ahead, plan for Saturday. It is where most streaks end.
The habit curve is not linear
Most people expect habit formation to feel progressively easier — a smooth curve from effortful to automatic. The data shows something different.
The first week is hard. The second week is slightly easier. Then, typically around days 10 to 14, there is a dip — a period where the novelty has worn off but the habit has not yet formed. Follow rates drop slightly. This is the danger zone.
Users who get through days 10 to 14 show significantly improved consistency from day 15 onwards. The habit is beginning to form. The decisions require less conscious effort. The rules start to feel like the default rather than the exception.
This matches the research. Phillippa Lally at University College London found that habit formation takes on average 66 days — but the curve is not linear. There is typically a plateau in the middle before automaticity kicks in. Knowing this in advance makes the plateau survivable.
"Around days 10 to 14, the novelty has worn off but the habit hasn't formed yet. This is the danger zone — and knowing it exists makes it survivable."
Streaks are more motivating than results
One of the most consistent findings from user data is that the streak mechanic drives behaviour more reliably than weight change does.
Weight loss is slow and variable. A good week on the rules might not show on the scale at all, due to normal weight fluctuation. The feedback loop is too slow and too noisy to be reliably motivating.
The streak is immediate and precise. Yesterday you had a 12-day streak. Today you can have a 13-day streak or a 0-day streak. That binary choice, available every single day, creates a form of motivation that compounds over time. The longer the streak, the stronger the pull to maintain it.
Users with streaks above 7 days show dramatically higher 30-day retention than users who never built a streak past 3 days. The streak is not just a feature. It is the primary retention mechanism.
Missing a day is not the problem. Not coming back is.
The most important data point from 30 days of tracking is not the follow rate. It is what happens after a missed day.
Users who return to the app the day after a streak break show retention rates almost identical to users who never broke their streak. Users who do not return within 48 hours of a streak break show dramatically higher long-term churn.
The missed day is not the problem. The what-the-hell response — the decision not to come back — is the problem. This is the same pattern described in why diets fail: one bad day becomes the end not because of the day itself, but because of what follows it.
This is why the streak freeze mechanic exists. Protecting a streak from a single missed day removes the what-the-hell trigger entirely. The streak continues. The motivation continues with it.
What 30 days actually produces
At 30 days of consistent tracking, most users report three things:
First, the rules have become largely automatic. They are no longer a conscious daily effort — they are the default behaviour that requires active effort to deviate from.
Second, the relationship with food has changed qualitatively. The constant background noise of food decisions — should I have this, can I afford these calories, is this on the plan — has largely disappeared. Three meals. No snacking. No liquid calories. The decisions are made.
Third, the results are visible. Not dramatic — this is not a crash diet — but consistent and continuing. The trajectory is clear.
Thirty days is not the destination. It is the point at which the habit is sufficiently formed to become self-sustaining.