# Log without unlocking: why we put BurnWeek on the iPhone lock screen

September 14, 2026 · From the Lab · 8 min read · https://burnweek.fit/blog/log-from-lock-screen/

> App loggers managed 92 days, paper loggers 29. The difference was not intelligence but distance from the hand, so the log moved to the lock screen.

**Key takeaways**

- Since July 2026 an iOS 18 "Log food" control on the Lock Screen, Control Center or Action Button opens BurnWeek already listening; the meal is logged with a range and spoken back.
- In a 128-person randomized pilot, app users logged 92 days over six months versus 35 on a website and 29 on paper, and lost 4.6 kg versus 2.9 and 1.3 kg (Carter 2013).
- In a post hoc analysis of 96 dieters, app users reported 1,437 kcal/day at six months versus 2,049 on paper, but logging frequency and BMI did not differ (Turner-McGrievy 2013).
- A 2018 review of 20 momentary diet-assessment protocols found real-time mobile logging can cut burden and recall bias; only 6 were validated against a traditional method (Schembre 2018).
- Capture gives up after 10 seconds of silence and refuses transcripts naming no food, so a pocket misfire cannot log a phantom meal.

## The cost of a log is measured in seconds, and adherence pays it

Since July 2026, an iPhone running iOS 18 can log a meal in BurnWeek from the Lock Screen: a "Log food" control sits where the flashlight or camera shortcut normally lives, or in Control Center, or on the Action Button. Press it, say what you ate, and put the phone away. The app opens already listening, stops on its own when you stop talking, logs the estimate without asking you to confirm anything, and reads the result back as a spoken calorie range with protein, even if the screen has gone dark in your pocket. A notification card holds the entry with Edit and Undo.

We built it because the research on food logging keeps returning the same finding: the act of logging predicts weight change, and the thing that kills logging is not difficulty but the small, repeated cost of doing it. Across the 22 studies in a 2011 systematic review, people who monitored their diet lost more weight, and the reviewers rated that evidence weak, because adherence was almost always self-reported ([Burke et al., 2011](https://pubmed.ncbi.nlm.nih.gov/21185970/)). The [beginner's guide to counting calories](/blog/how-to-count-calories) covers what that review does and does not license. This post is about the narrower engineering claim: cut the seconds, and more logs get made.

## Apps beat paper because they are closer to hand, not because they are smarter

The cleanest test of that claim is a 2013 pilot trial from Leeds. 128 overweight adults were randomised to six months of self-monitoring by a smartphone app, a commercial website, or a paper diary. The interventions were deliberately similar in content; what differed was the device. At six months, 93 percent of the app group was still in the trial against 55 percent of the website group and 53 percent of the diary group, and the app group had logged on a mean of 92 days compared with 35 for the website and 29 for paper. Weight change followed the same order: 4.6 kg lost in the app group, 2.9 kg with the diary, 1.3 kg with the website ([Carter et al., 2013](https://pubmed.ncbi.nlm.nih.gov/23587561/)). Two things to hold in mind: the authors built the app they tested, which they disclose, and self-monitoring declined over time in all three groups. The app slowed the decay; it did not stop it.

A second study from the same year points the same way with a weaker design. In a post hoc analysis of a 96-person podcast-delivered weight loss trial, participants who chose an app to track diet reported a lower energy intake at six months (1,437 kcal/day) than those who chose a paper journal (2,049 kcal/day), though how often they logged did not differ significantly by method and neither did BMI ([Turner-McGrievy et al., 2013](https://pubmed.ncbi.nlm.nih.gov/23429637/)). People picked their own method in that analysis, so it cannot separate the tool from the kind of person who picks it.

| Study | Design | Days of dietary recording (6 months) | Weight change |
| --- | --- | --- | --- |
| Carter 2013, app arm | Randomised, n=128 | 92 (SD 67) | −4.6 kg |
| Carter 2013, paper diary arm | Randomised | 29 (SD 39) | −2.9 kg |
| Carter 2013, website arm | Randomised | 35 (SD 44) | −1.3 kg |
| Turner-McGrievy 2013 | Post hoc by chosen method, n=96 | No significant difference by method | No BMI difference by method |

Neither trial tested lock-screen logging; the Leeds trial's app still had to be unlocked and opened. But the mechanism both point to is the same one a lock-screen control attacks. A phone is already in your hand at a meal; a paper diary is in a drawer. The control removes the remaining steps between the two.

## What the control does, step by step

The trigger differs by phone and the rest is the same. On any iOS 18 iPhone, the "Log food" control can replace a Lock Screen shortcut or sit in Control Center. On an iPhone 15 Pro or later it can be the Action Button. Older phones can use Back Tap, after a one-time step of adding the shortcut to your own library, and "Log food with BurnWeek" works as a Siri phrase on every model. Siri only acts as the trigger; BurnWeek handles the transcription itself, so the food you name is understood the same way it would be in the app.

From the press onward:

1. The microphone is live before the app has finished appearing, so the first word is not lost.
2. Provisional captions show as you speak. When you stop, the capture ends after about a second and a half of silence, with a hard cap of 45 seconds.
3. The transcript goes to the estimate. Nothing asks you to confirm; the meal is logged with the likely value and its range.
4. The result is spoken back: the range, the protein, in the language you spoke or the one you chose in settings. This works with the screen off and the silent switch on.
5. A notification card carries the entry. Edit opens the meal; Undo deletes it in the background without opening the app.

At 21:30 local time, if any of the day's logs had a wide range relative to its likely value, a digest notification lists them so you can fix the ones that matter. Low-confidence entries do not block the log; they queue for a moment when you have a screen in front of you.

There is also a text path that needs no app at all. A standing, quiet notification accepts a typed reply; type "two eggs and toast" into it and the meal is logged from the notification, with the same card and the same Undo. Camera and Scan chips on the capture overlay hand off to the composer for a photo or a barcode if you want them, and Cancel and Start over do what they say. A capture also gives up after ten seconds without speech and refuses to log a transcript that names no food, so an accidental press in a pocket cannot log a phantom meal.

## A log made at the table is an observation; one rebuilt at night is a recall

The second reason to make logging instant is not adherence but the quality of what gets logged. Dietary assessment research splits methods into records, made at the time of eating, and recalls, reconstructed afterward, and treats the gap between them as a source of error rather than a matter of convenience. A 2018 systematic review of 20 mobile "ecological momentary" diet assessment protocols concluded the approach has the potential to reduce both participant burden and recall bias, while noting that only six of the 20 studies had actually compared it against a traditional method ([Schembre et al., 2018](https://pubmed.ncbi.nlm.nih.gov/30459148/)). That is a promising direction with a thin validation base, and we cite it as such.

What is not thin is the size of the error a recall carries. The classic demonstration is a 1992 study in which obese adults who reported eating under 1,200 calories a day and not losing weight turned out, under 14 days of measurement, to be underreporting intake by an average of 47 percent ([Lichtman et al., 1992](https://pubmed.ncbi.nlm.nih.gov/1454084/)). The [accuracy of calorie counting](/blog/how-accurate-is-calorie-counting) piece works through where that gap comes from. Some of it is portion misjudgement; a lot of it is meals that were never written down. A log you can make in the four seconds between finishing a sandwich and standing up is a log that gets made, and a meal that is logged, even roughly, is a meal that is not forgotten.

That is also why the lock-screen flow never asks a question. An estimate-then-confirm loop would put the screen back in your hand and the app back between you and the next thing. The range is the confirmation. If it is wide, the digest brings it back at night; if it is wrong, Edit is on the card. The [mindful eating with numbers](/blog/mindful-eating-with-numbers) pillar makes the longer case for logging as an act of noticing rather than an act of accounting; the lock-screen control is that case reduced to one press.

## What this does not fix

Speed does not repair a wrong portion. Saying "a bowl of pasta" gets an estimate for a bowl of pasta, which is a wide range for a reason, and the range says so. Weighing fixes that, and the [scale integration](/blog/smart-scale-and-chatgpt-logging) is a different tool with its own post. Speed also does not make the decay in the Leeds trial go away; it moves the curve, and the reason to move it is that each extra logged day is worth something even when the total number of days falls short of perfect. And the control cannot start recording from a truly locked phone on its own: the press wakes the app, Face ID does its work without a gesture, and recording continues after the screen locks again. The design goal was never zero taps. It was one press and no babysitting.

## FAQ

### Does logging food from the lock screen work if I am not on iOS 18?

The Lock Screen and Control Center control needs iOS 18. On earlier versions, or on phones without an Action Button, the same capture starts from a Back Tap shortcut or from the Siri phrase "Log food with BurnWeek". Siri only triggers the capture; BurnWeek handles the transcription itself.

### What happens if the lock-screen capture hears nothing, or hears something that is not food?

It gives up after ten seconds without speech, and if the transcript names no food or drink it shows "Didn't catch any food — try again" and logs nothing. That guard keeps an accidental press in a pocket from logging a phantom meal.

### Does a spoken log without a confirmation step make the estimate less accurate?

The estimate is the same one you would get by typing the sentence; what changes is that the app logs the likely value with its range instead of asking you to approve it. If the range is wide, the 21:30 digest lists that meal so you can correct it, and Edit on the notification card opens it at any time.

### Can I type a meal into the lock-screen notification instead of speaking?

Yes. A standing quiet notification accepts a typed reply, and the meal is logged from the reply without opening the app. The same result card, with Edit and Undo, follows.

## Sources

- [Burke LE, Wang J, Sevick MA. Self-monitoring in weight loss: a systematic review of the literature. J Am Diet Assoc. 2011.](https://pubmed.ncbi.nlm.nih.gov/21185970/)
- [Carter MC, Burley VJ, Nykjaer C, Cade JE. Adherence to a smartphone application for weight loss compared to website and paper diary: pilot randomized controlled trial. J Med Internet Res. 2013.](https://pubmed.ncbi.nlm.nih.gov/23587561/)
- [Turner-McGrievy GM, et al. Comparison of traditional versus mobile app self-monitoring of physical activity and dietary intake among overweight adults participating in an mHealth weight loss program. J Am Med Inform Assoc. 2013.](https://pubmed.ncbi.nlm.nih.gov/23429637/)
- [Schembre SM, et al. Mobile ecological momentary diet assessment methods for behavioral research: systematic review. JMIR Mhealth Uhealth. 2018.](https://pubmed.ncbi.nlm.nih.gov/30459148/)
- [Lichtman SW, et al. Discrepancy between self-reported and actual caloric intake and exercise in obese subjects. N Engl J Med. 1992.](https://pubmed.ncbi.nlm.nih.gov/1454084/)

Source: BurnWeek — "Log without unlocking: why we put BurnWeek on the iPhone lock screen", https://burnweek.fit/blog/log-from-lock-screen/. Licensed CC BY 4.0: free to quote or reuse with a link to this page.