snapkin
The habit · Decision 11 of 16
Why we read your week back on Monday
Every Monday the app reads your last week back. It shows which days were close to your goal, which were higher, and what the week added up to against maintenance. This page says why a week and not a day, why Monday, and why the report leaves out your weight.
Last updated .
- What the app does
- Reads the week as one total, shows a higher day inside that week, compares your logging with last week, and leaves weight to the monthly check-in.
- How sure we are Fair , Real, but small, indirect, or thinly measured.
- The trials on spreading a deficit over the week are solid. Nobody has tested a weekly report screen like this one. The evidence on logging is mostly observational, and the four-day minimum and the ±10% band are our own choices.
The evidence for it
The same weekly deficit lost the same weight, however it was spread. Weekends are logged least. And feedback can make things worse, which is why the report is careful how it talks.
Why a week, and not a day
Weight follows the energy you eat and burn over time, not on one day. The cleanest test of that is trials that give two groups the same weekly deficit and spread it differently. Harvie et al. (2011) put 107 women on a 25% cut for six months. One group ate very little on two days a week and normally on the other five. The other ate a little less every day. They lost 6.4 kg and 5.6 kg, and the difference was not significant (p = 0.4). Trepanowski et al. (2017) went further: one group ate 125% of their needs every other day and 25% on the days between. After six months both groups had lost 6.8% of their weight.
Reviews that pool these trials agree. Cioffi et al. (2018) found a difference of 0.61 kg across 11 trials, with a confidence interval that runs from −1.70 to 0.47 kg. Headland et al. (2016) found 0.08 kg in trials of six months or more. So the report reads your week as one total, and shows a higher day as part of it.
Why it expects a higher weekend
People eat more at the weekend. In a national US survey, Haines et al. (2003) found 82 kcal more on each day from Friday to Sunday. An (2016) compared 11,646 adults with themselves and found 181 kcal more on Saturday. In a small diet trial, Racette et al. (2008) measured 236 kcal more on Saturday, and the people dieting stopped losing weight at the weekend.
The weekend is also when logging stops. Krukowski et al. (2013) found that people logged 28% of weekdays but only 17% of weekend days. So the report never treats a day you did not log as a day you ate nothing. It leaves that day out, and it says how many days the result is based on.
Why it counts the days you logged
How often you log matters more than how carefully. In a trial with 220 women, Peterson et al. (2014) found that more food records helped only in weeks with more than three days logged. In Burke et al. (2012), people who logged 60% of the time or more lost more than people who logged under 30%. Ingels et al. (2017) found a real loss only in people who tracked more than two days in three. In 1,685 people, Hollis et al. (2008) found 0.52 kg more weight lost for each extra food record a week.
That is why the logging slide counts days, compares them with last week, and praises the logging itself, including on a day that went over. A week needs four full days before the report reads it back. That number is ours: it is the first count above the "more than three days" Peterson found, not a threshold any study tested.
Why it compares you with last week, and never says perfect
Seeing your progress helps. A review of 138 trials by Harkin et al. (2016) found an effect of d = 0.40 on reaching a goal, and a bigger one when progress was written down. In Tate et al. (2001), people who got weekly feedback on their food diary lost 4.1 kg against 1.6 kg in six months.
But feedback can backfire. Kluger & DeNisi (1996) looked at 607 results and found that over 38% made performance worse. Feedback that threatened how people saw themselves did almost nothing (d = 0.08, against 0.47). Feedback that showed change since last time worked best (d = 0.55, against 0.28). And the belief that you have overeaten can be enough to make some dieters eat more: Polivy et al. (2010) saw it with the same slice of pizza, described as bigger or smaller. A large study of 54,517 people by Westenhoefer et al. (1999) linked flexible control to a lower BMI, and rigid all-or-nothing control to a higher one.
So the report compares you with your own last week, calls a higher day "higher", and ends on next week. It never says "perfect", never calls a day ruined, and a bigger deficit never earns better words. A day under your plan's minimum is never praised: the report first asks whether a meal is missing. The rules come from the eating disorders page.
Why it says "about", and never turns kcal into kilos
A photo log is close on average and rough for one person. In Martin et al. (2012), a photo method was 152 kcal a day low on average against doubly labelled water, but the spread was 694 kcal a day either side. So the report rounds the week to the nearest 100, says "about", and names the number of days behind it. It does not turn the week into kilos of fat. The 7,700 kcal per kilo rule overstates loss, because the body adapts: Hall et al. (2011) work out that a lasting change of 100 kJ a day ends at about 1 kg, half of it after about a year.
Why your weight is not in it
A week is too short for the scale. Turicchi et al. (2020) followed 1,421 people on smart scales and found weight swings about 0.35% inside a normal week, up at the weekend and down on weekdays. That is about 0.3 kg for somebody at 80 kg, as big as a good week's loss. Orsama et al. (2014) found the weekly peak on Sunday and Monday, so a Monday reading would often show a good week next to a bad-looking number. And over two weeks, Bhutani et al. (2017) found 84% of the change in weight was not fat. Your weight trend stays in the monthly check-in, which fits a line through a month of weigh-ins. See why the weight chart draws a line.
Why Monday
The start of a week is when people turn to their goals. Dai et al. (2014) found searches for "diet" 14.4% higher, gym visits 33.4% higher and new goal commitments 62.9% higher at the start of a week. But calling a message a "fresh start" did not help in a trial of 9,143 people (Njuguna et al. 2024). So the report comes on Monday, after the weigh-in, and does not use the slogan.
The evidence against
No trial has tested a weekly report screen like this one. The pieces each have evidence; the whole has none yet. Most of the logging evidence is observational: people who log more may simply be the people who would have lost more anyway. In Gorin et al. (2004), people who ate the same way all week were 1.5 times more likely to keep their weight off than people who were strict only on weekdays. So "the weekend does not matter" goes too far: the week's total is what counts, and a habit of loose weekends still predicts regain. One small trial, Byrne et al. (2018), found two-week diet breaks more efficient, and a larger one, Peos et al. (2021), did not. Uneven plans were also harder to stay on in Trepanowski et al. (2017): 38% dropped out against 29%. The four-day minimum and the ±10% band around your goal are our own choices, not tested thresholds.