Macro tracking has a retention problem. Most people who start weighing food and logging meals stop within a few weeks — not because it doesn't work, but because the standard approach demands a level of precision that nobody sustains alongside a job, a family and food cooked by someone else.
The fix isn't more discipline. It's accepting that approximate tracking done for a year beats precise tracking done for three weeks, and building a system around that.
What macros are, briefly
Three macronutrients supply energy: protein, carbohydrate and fat. Roughly, protein and carbohydrate provide about 4 calories per gram, fat about 9. Alcohol supplies about 7 and is the reason people's logs mysteriously don't add up.
Tracking macros rather than calories alone matters because the composition changes the outcome. Two 1,800-calorie days — one with 120g of protein and one with 40g — produce different results in body composition, satiety and how well you recover from training, despite the identical calorie total.
Start with protein and ignore the rest
This is the single most useful simplification available, and it's where most people should begin.
Protein is the macro that (a) most people under-consume, (b) has the strongest evidence base for a specific target, and (c) drives the outcomes people usually care about. Carbohydrate and fat ratios matter far less than the internet suggests, provided total energy is roughly right.
How much protein
- General health: around 0.8–1.0g per kg of bodyweight.
- Building muscle or losing fat while training: roughly 1.6–2.2g per kg. Meta-analyses of resistance training studies converge on about 1.6g/kg as the point where additional protein stops producing meaningful extra benefit.
- Older adults: toward the higher end, since the muscle-building response to protein blunts with age.
For a 70kg person training regularly, that's roughly 110–150g a day. Most people eating a typical diet are well below this and are surprised when they first measure it.
Why this is harder on an Indian diet
A standard vegetarian Indian day — poha or upma, dal-chawal, roti-sabzi — often lands between 40 and 60g of protein. Not because the food is poor, but because the staples are carbohydrate-dominant and the protein sources are dilute. Dal delivers protein at a much lower density than most people assume.
The practical fixes are unglamorous and effective:
- Add paneer, curd or Greek yoghurt to meals you already eat rather than restructuring your diet.
- Increase dal quantity rather than variety — double the portion, not the number of dishes.
- Soya chunks are the cheapest high-density vegetarian protein widely available in India by a considerable margin.
- Combine cereals and pulses across the day — the classic dal-chawal pairing improves the amino acid profile, and it need not be in the same meal.
- Eggs, if you eat them, solve much of the problem cheaply.
The tracking method that actually survives
Here's a three-tier approach. Start at tier one. Most people never need tier three.
Tier 1: Protein only, estimated (weeks 1–4)
Log protein. Nothing else. Use rough portion estimates — one katori of dal, one hundred grams of paneer, two eggs. Do not weigh anything. The goal is awareness of a single number, and it takes about ninety seconds a day.
Most people discover they're 40–50g short of target. Fixing that alone produces more change than any amount of ratio optimisation.
Tier 2: Protein and calories, estimated (weeks 5–12)
Add a rough calorie estimate. Still no scale. Accept ±15% error — it's fine. You're tracking a trend, and a consistent 15% overestimate still tells you accurately whether you ate more this week than last.
Tier 3: Full macros, weighed (only if needed)
Reserve this for a specific goal with a deadline — a physique target, a weight class, a plateau you've genuinely failed to break with tiers one and two. It is a temporary tool, not a lifestyle.
Where tracking usually goes wrong
- Logging after the fact. Log before or during eating. Retrospective logging is systematically under-reported — this is one of the most consistent findings in nutrition research, and it applies to everyone, not just people being careless.
- Ignoring cooking oil. The most under-logged item in Indian cooking by a wide margin. A tablespoon of ghee or oil is roughly 120 calories, and a home-cooked sabzi may contain two or three.
- Perfect weekdays, unlogged weekends. Two unlogged days is 28% of your week. It's usually the entire explanation for a stalled result.
- Chasing exact daily numbers. Hit a weekly average. Bodies don't reset at midnight.
- Quitting after a bad day. The log's value is the trend line. One outlier changes nothing; three weeks of missing data changes everything.
Where AI logging genuinely reduces friction
The barrier has always been input effort. A few things meaningfully lower it:
- Photo-based estimation — pointing a camera at a thali is dramatically faster than searching a database for six items, and for tier-one tracking the accuracy is sufficient.
- Barcode scanning for anything packaged, which is exact rather than estimated.
- Meal memory — most people eat a rotation of fifteen to twenty meals. A system that learns yours turns logging into a single tap.
- Recipe-level entries for home cooking, so a household dish is logged once and reused rather than reconstructed ingredient by ingredient every time.
A caution worth stating: nutrition figures generated by a language model can be confidently wrong. Databases built on measured food composition data — for Indian foods, the IFCT tables are the reference standard — are what you want underneath any estimate.
What to actually watch
Daily numbers are noise. Review weekly, and look at four things: average daily protein against your target, average daily calories, how many days you logged at all, and bodyweight trend across two weeks rather than day to day. Water weight will swing two kilos in either direction and mean nothing.
Tracking also only tells half the story — what you do with the intake depends on how you're training, and protein in particular does far less for you without a resistance stimulus.
The bottom line
Track protein first, estimate rather than weigh, review weekly not daily, and treat a missed day as a missed day rather than a failed project. The system that lasts a year is the one that wins, and it will always be the less precise one.
BodyFastLane's nutrition tracking is built around Indian foods and household portions, with photo logging and barcode scanning to keep daily effort low — free on iOS. For medical nutrition therapy or a diagnosed condition, work with a registered dietitian.