Updated for 2026
How we test and rank food tracking apps
Every score on this site traces back to a documented test. We publish our rubric, our weights and our formulas so that anyone — including the developers we review — can check our work. We take no affiliate money.
What is the 100-point rubric?
We score each app from 0–100 on eleven segments, then combine them with fixed weights that total 100%. The result is a single overall score, but the segment scores are where the real information lives — two apps can tie overall while being right for completely different people.
| Segment | Weight | What it measures |
|---|---|---|
| Accuracy | 16% | How close logged calories and macros land to a weighed reference. |
| Ease of Use | 14% | How fast and frictionless it is to log a typical day of eating. |
| International Food Database | 12% | Coverage and data quality for foods outside the US, including regional and restaurant items. |
| Data Display & Visualization | 9% | How clearly the app turns logged data into trends, dashboards and reports. |
| Nutrition Coaching | 9% | The quality of guidance that nudges behaviour, not just records it. |
| Diet Analysis | 8% | How deeply the app interprets nutrient quality, gaps and patterns over time. |
| Meal Feedback | 7% | The usefulness of the per-meal response you get right after logging. |
| Design | 7% | Visual craft, information architecture and how pleasant the app feels in daily use. |
| Meal Planning | 7% | Tools for planning meals ahead, building recipes and generating shopping lists. |
| Value | 7% | What you get for free versus paid, and whether the subscription earns its price. |
| Workout Planning | 4% | How well the app logs activity, plans training and reconciles it with energy balance. |
| Total | 100% |
How do we measure accuracy?
Accuracy is the heaviest segment because every other feature depends on it. We build 40 reference meals, weigh every component on a calibrated kitchen scale, and compute ground-truth energy and macros from USDA FoodData Central. We then log each meal in the app and measure the mean absolute percentage error (MAPE) between the app's totals and ground truth.
The accuracy formula. We convert error into points with accuracy_points = clamp(100 − MAPE × 4, 0, 100). An app that is on average 5% off scores 80; one that is 15% off scores 40. We test photo logging, barcode and manual search separately, then weight them by how people actually log.
How do we score logging ease and speed?
We run a 20-task battery — barcode scans, text searches, photo capture, recipe builds, restaurant meals and re-logging recent foods — and stopwatch every interaction on both iOS and Android. Friction matters as much as raw speed, because the best tracker is the one you keep using. We note paywalls, ad interruptions and dead-ends that break the flow.
How do we test AI photo recognition?
For apps with photo logging, we use a 30-plate battery spanning different lighting, angles, portion sizes and mixed dishes. We grade both identification (did it recognise the foods?) and portion estimation (did it get the amount right?). Portion estimation is where most AI trackers struggle, so we weight it heavily.
How do we score the food database?
We search 120 foods spanning North American, European, Latin American, South Asian, East Asian, Middle Eastern and African cuisines. We grade hit rate, the quality of the returned entries, and crucially the ratio of verified entries to unvetted crowd-sourced guesses. A huge database full of duplicates scores worse than a smaller, curated one.
How do we score coaching, feedback and diet analysis?
We feed each app two identical weeks of realistic logs and grade what it does with them: does it set adaptive targets, surface micronutrient gaps, and give per-meal feedback that is specific and actionable? Our registered dietitians judge this the way they would judge a junior colleague — is the advice specific, safe and useful?
How do we score value and pricing?
We map each free tier against the paywall, compare annual prices across the category, and weigh features-per-dollar against how aggressively an app pushes ads, upsells and dark patterns. A more expensive app can still win on value if it earns the price.
How do we handle independence and conflicts of interest?
We buy every app with our own money, including the paid tiers. We take no affiliate commissions and accept no sponsored placements. Developers can submit corrections of fact, but they cannot review scores before publication or pay to change them. Read our full no-affiliate disclosure and editorial policy.
How do we use AI in our own work?
We use AI tools to help organise notes and draft summaries, but every score is set by a human reviewer, every accuracy figure is measured by hand, and every published claim is checked by an editor. Details are in how we use AI.
Quality control and corrections
No score is published until a second reviewer has checked the underlying tests. When we get something wrong, we fix it visibly and log it on our corrections page. Apps change; so do our scores.
Who runs the testing?
- Dr. Elena Marsh, PhD, RD — Editor-in-Chief & Lead Nutrition Analyst.
- Marcus Bell, MSc — Head of Testing & Data.
- Priya Anand, MS, RDN — Senior Reviewer, Coaching & Meal Planning.