Getting to your three numbers
Calories come first, because macros are just those calories divided up. A rough maintenance estimate is body weight in pounds times 14 to 16, toward the high end if you're on your feet all day. For fat loss, the 2013 AHA/ACC/TOS obesity guideline uses a 500 to 750 calorie daily deficit, and the lifestyle arm of the semaglutide STEP 1 trial ran roughly 500 calories a day below estimated needs alongside the drug. If a doctor or a registered dietitian gave you numbers, use theirs over anything on this page.
Then split the total. Protein at 1.2 to 1.6 g per kilogram of body weight, which is about 0.55 to 0.7 g per pound, fat at 0.3 to 0.4 g per pound, and carbs absorb the rest. The table takes the top of the protein range:
| Body weight | Calories | Protein | Fat | Carbs |
|---|---|---|---|---|
| 140 lb | 1,900 | 100 g | 50 g | 260 g |
| 180 lb | 2,200 | 130 g | 65 g | 275 g |
| 220 lb | 2,600 | 160 g | 75 g | 320 g |
Treat these as illustrations. Round to the nearest 5 g, because no app is precise enough to justify 131. If you're pregnant, managing diabetes or kidney disease, or recovering from an eating disorder, get your targets from a clinician before you set anything.
Weighing versus serving sizes, and when each is fine
A $12 kitchen scale set to grams solves most tracking error. Put the bowl on, press tare, add food, read the number. The foods worth weighing are the ones where a small visual difference is a large calorie difference.
Two tablespoons of peanut butter is 32 g on the label. Scooped by eye it comes out closer to 40 g, which is another 48 calories and 4 g of fat, every morning, invisibly. Olive oil is worse, since a generous pour into a pan is easily 20 g instead of 10.
Meat has its own trap. USDA figures put 100 g of raw boneless skinless chicken breast at about 120 calories and 22.5 g protein, while the same breast roasted comes in around 165 calories and 31 g protein per 100 g, because it lost roughly a quarter of its weight in water. Pick raw or cooked, then match the database entry to what you picked.
Serving sizes are fine for the rest: one medium apple, one large egg, a cup of milk, a single-serve yogurt. That's why the food database in LockedIn keeps real serving sizes on 267,238 of its 321,737 foods instead of making you convert a medium apple into grams.
What a barcode scan gets you, and what it doesn't
Scanning tells the app which food you ate. It doesn't tell it how much. A scanned protein bar is one bar and that's easy, but a scanned bag of tortilla chips is 9 servings, and the entry defaults to one. This is where a logged day quietly goes wrong by a few hundred calories.
Labels themselves carry slack. FDA rounding rules let anything under 0.5 g of fat per serving be printed as 0, which is how cooking spray reads as zero calories. Compliance rules also allow the real calorie content to sit up to 20 percent above the declared figure. So a label is a good estimate and nothing tighter than that.
When a store brand isn't in any database, search the generic version instead of skipping the food. "Whole wheat bread, commercially prepared" logged is more useful than a blank line where two slices used to be. Barcode scanning ships in LockedIn using MLKit, and I still weigh the cereal.
Mixed foods, home cooking, and restaurant meals
A chili, a curry, a tray of baked ziti: these break the one-food-one-entry model, and the fix is arithmetic you do once. Weigh every ingredient as it goes into the pot and log each one. Then weigh the finished dish (tare the empty pot first, or weigh the pot before and after). Divide. If the pot came out to 1,800 g and you eat a 300 g bowl, you ate one sixth of everything you logged.
Restaurants are guesswork and you should treat them that way. Find the closest chain entry, assume more oil and butter than you'd use, and round up on fat. A grilled chicken salad at a sit-down place is easily over 700 calories once the dressing is counted.
You log a restaurant meal to keep the week's data continuous, so your averages still mean something on Sunday.
Saved meals are what keep this going past week two
Most people I know rotate through the same fifteen or twenty meals. The same oatmeal, the same lunch salad, the same rice and chicken on training days. Searching for each component every single morning is the friction that ends tracking, and it has nothing to do with motivation.
Build the meal once with the right gram weights (80 g oats, 30 g whey, 15 g peanut butter, 100 g blueberries), save it, and log the whole thing in one tap tomorrow. My own breakfast has been the same four-item entry for months, and logging it takes about two seconds.
Two practical notes. Save the version you eat on a normal day rather than an idealized one, and rebuild the saved meal when your portions shift, since a stale saved meal is a slow, believable error.
Micronutrients: worth a look weekly, not daily
You can hit 130 g protein and 275 g carbs on chicken, rice, and protein powder for a month and end up low on things that matter. Fiber is the usual casualty. The Institute of Medicine set the adequate intake at 14 g per 1,000 calories, so a 2,200 calorie day works out to about 31 g, and a lot of high-protein days land near 12.
The 2019 National Academies review put potassium adequate intake at 2,600 mg a day for adult women and 3,400 mg for adult men, with sodium's chronic disease risk reduction level at 2,300 mg. Most tracked days invert both of those.
LockedIn tracks fiber, potassium, calcium, iron, magnesium, zinc, vitamins A, C, D, E, K, B6, B12, folate, cholesterol, saturated fat, trans fat, sodium, and sugar alongside the macros. Check them on Sunday, notice the pattern, add a vegetable or some beans. Don't try to optimize every one of them daily.
The missed day, and the week that falls apart
You'll miss days. A wedding, a bad Tuesday, a work trip where lunch was whatever was in the conference room. Here's what to do with it.
Log what you remember, mark the day as incomplete in your head, and move on. Don't reconstruct 36 hours from memory, because invented data is worse than a gap. Don't cut the next day's calories to compensate either, since that swing is what turns one missed day into a four-day binge and restart cycle.
Judge yourself on the week. If you hit protein on five of seven days and stayed near your calorie target on four, the system is working. Weight moves on weekly averages anyway, which is why a Tuesday weigh-in that's up 2 lb means very little on its own.
A workable first week looks like this:
- Days 1 to 2: log everything, change nothing, just see your baseline.
- Day 3: set your protein target and hit that number only.
- Days 4 to 7: add the calorie target, save your three most repeated meals.
- Day 8: compare the week's average to your targets and adjust one number.