Looking for the deep version? The full 8-parameter rubric, source citations, validation against MARS and COSMIN, and per-app benchmark matrix all live on our complete methodology page (v3.1). This page is the short version.
How much testing goes into a Fitness Tracking Guide review?
Days minimum hands-on per app
Reviewers in our 2026 nutrition cohort
Meals logged in the 16-week pilot
Nights vs. a polysomnography reference (sleep)
Blind-test meals for photo-AI accuracy
App reviews published, all independently tested
What editorial standards do we hold every review to?
- Hands-on testing first. Every recommended app has been used in real life by at least one reviewer for a minimum of 45 consecutive days. We do not rank apps we have not used.
- Accuracy verified against reference instruments. Calorie estimates are checked against weighed meals and the USDA FoodData Central database; sleep against a polysomnography-grade reference; running against a GPS reference watch; strength logging measured per-set.
- Adherence measured, not assumed. Our 2026 nutrition cohort followed 127 reviewers across a 16-week pilot, logging more than 21,800 meals. Adherence is reported as share of days reviewers were still logging at week 16.
- Photo-AI accuracy blind-tested. Across 360 reference meals, AI photo loggers were scored on portion and preparation identification under blinded conditions.
- Transparent scoring. Each app is scored out of 10 across accuracy (30%), adherence and friction (30%), value (20%) and ethics and privacy (20%).
- Conflicts disclosed. Any relationship with an app maker, and any affiliate links, are disclosed on the relevant page. Affiliate status never influences a score.
- Vendor figures attributed. Where a vendor reports proprietary stats (e.g. accuracy or processed-log counts), we cite them as the vendor's own reported figures rather than as our independent measurement.
How do we test fitness apps in each category?
- Nutrition & weight loss — 45-day snapshot with 5 reviewers + a 16-week / 127-reviewer adherence cohort.
- Sleep — 128 consecutive nights with 5 reviewers against a polysomnography-grade reference.
- Running apps — 15 weeks of mileage across 8 reviewers (18–62 mi/week) vs. a Garmin Forerunner reference watch.
- Workouts & strength — 15 weeks of programmed lifting across 6 reviewers across multiple protocols.
- Intermittent fasting — 14-week study with 6 reviewers across 16:8, 18:6 and 24h protocols.
- Menstrual cycle — 4 reviewers, ≥4 cycles each; privacy posture weighted at 40% of the category score.
- Meditation — 45+ days of daily use by 4 reviewers per app.
How to contact Fitness Tracking Guide
Tip, correction, or app you want us to review? Email hello@fitness-tracking.com.