MochiMost calorie counting advice falls apart because it's all-or-nothing. Either you're weighing every almond, or you've quit by Friday. Here's the boring middle path that actually works.
Step 1: figure out your maintenance calories
Maintenance is the number of calories you'd eat to stay exactly where you are. There are formulas (Mifflin-St Jeor is the most accurate one), but the quick version is:
- Sedentary woman: bodyweight in lbs × 12-13
- Sedentary man: bodyweight in lbs × 14-15
- Active (3-5 workouts/week): add ~200-300 calories
A 160 lb woman who walks the dog and hits the gym twice a week is around 2,000 maintenance. Most calorie tracker apps will calculate this for you in onboarding, so you don't have to do mental math.
Step 2: set a deficit you can actually live with
To lose ~1 lb a week, eat ~500 calories below maintenance. To lose ~0.5 lb a week (more sustainable, less hangry), aim for ~250 below. Going below 1,200 calories for women or 1,500 for men is almost always a bad idea, you'll lose muscle, energy, and the will to keep going.
Slow is fast. The person who loses 0.5 lb a week for a year ends up 26 lbs lighter. The person who crash diets for three weeks and quits ends up exactly where they started. We've all been the second person at least once. Be the first one this time.
Step 3: decide what you're actually going to log
Biggest mistake beginners make: trying to log every leaf of spinach. The 80/20:
- Always log: meals, snacks ≥ 100 calories, drinks with calories (latte, juice, beer, smoothie)
- Don't bother: black coffee, plain tea, water, raw veg, gum, condiments under a tablespoon
This catches 95% of your actual intake without the friction that makes people quit by week two.
Step 4: pick a tool that doesn't punish you
The single biggest predictor of whether you'll keep counting calories is how long it takes to log a meal. If it takes more than ten seconds, you'll quit. This is why photo-based AI calorie trackers (like, well, Calchi) have changed the game, you snap a photo of your lunch and it's logged before your fork hits the plate.
Snap a photo of your meal. We'll do the math.
Step 5: trust the trend, not the day
Bodyweight bounces 2-5 lbs day to day from water, sodium, sleep, hormones, and bowel movements. The actual signal is the 7-day moving average. If your weekly average is trending down, the deficit is working, even if today's number is up. Don't make decisions based on a single weigh-in. Ever.
Step 6: plan for the days you won't log
You're going to have weddings, vacations, and "the kind of Tuesday where you just don't feel like it." That's fine. The point isn't to log forever. It's to log long enough to internalize what 500 / 700 / 1,000-calorie meals actually look like. After 8-12 weeks of consistent tracking, most people can eyeball their intake within 200 calories and only log occasionally for accountability.
Common stuff that quietly wrecks people
- Liquid calories: a daily 250-cal latte is 91,000 calories a year, or about 26 lbs of body fat
- Cooking oil: a tablespoon of olive oil is 120 calories. Most "splashes" are 2-3 tablespoons.
- Eating back exercise calories: fitness trackers overestimate burn by 20-50%. If you must eat them back, eat back half.
- Weekday saint, weekend warrior: five days at -700 followed by two days at +1,000 is just maintenance with extra steps. Spread it out.
The bottom line
Calorie counting works. It's not magic. It's just paying attention to something we'd otherwise ignore. The trick is to set up a system so low-friction you can't talk yourself out of it tomorrow. Pick a tool that lets you log a meal in seconds, set a deficit you can actually live with, and trust the trend over time.