// Independent Testing · No Affiliates · No Sponsored Placements Methodology · Editorial
// LAB REPORT — CTL-LAB-Q2-2026-LI

Lose It! Accuracy Lab Report

Published May 8, 2026 · Updated May 23, 2026 · Tester: J. Bremer, RD (Calorie Tracker Lab tester)

TL;DR

  • App: Lose It! (version 15.7.0 (iOS) · 15.6.2 (Android))
  • Test date range: 12 February 2026 – 22 April 2026
  • Pooled MAPE (40 meals): ±7.7%
  • Median logging speed: ~45s per meal (search-and-log, barcode where available)
  • Key finding: Lose It! sits in the middle of the cohort. The single-foods and basic packaged-goods accuracy is materially tighter than MyFitnessPal's; the restaurant-chain and complex-mixed-dish accuracy is materially looser than Cronometer's or MacroFactor's. The product's strength is not per-meal precision — it is the friction-free onboarding that gets a never-tracked user to log day one through day thirty without burning out.
  • Dataset: 2026 Calorie Counter App Accuracy Benchmark v1.2 · raw CSV

Test snapshot

App version15.7.0 (iOS) · 15.6.2 (Android)
Operating systemiOS 18.4, Android 15
Localeen-US
TesterJ. Bremer, RD (Calorie Tracker Lab tester)
Test window12 February 2026 – 22 April 2026
Meals logged (n)40
Reference standardUSDA FoodData Central + published packaged-food labels + chain nutrition

Per-meal results (40 meals)

Meal Cat Ref kcal Lose It! est. Abs err %
Chicken breast, grilled, 4 oz boneless skinless single 187 193 3.2%
Banana, medium, 118g single 105 105 0%
Egg, large, hard-boiled single 78 75 3.8%
Almonds, 1 oz / 28g single 164 160 2.4%
Oatmeal, plain rolled, 1/2 cup dry single 150 148 1.3%
White rice, cooked, 1 cup single 205 216 5.4%
Broccoli, steamed, 1 cup single 55 58 5.5%
Atlantic salmon, baked, 4 oz single 233 238 2.1%
Whole milk, 1 cup / 244g single 149 139 6.7%
Avocado, 1/2 medium single 161 164 1.9%
Chobani Greek Yogurt, Plain Non-Fat, 5.3 oz packaged 80 76 5%
Cheerios, 1 cup / 28g packaged 100 87 13%
KIND Dark Chocolate Nuts & Sea Salt bar, 40g packaged 200 217 8.5%
Quest Protein Bar Cookies & Cream, 60g packaged 190 185 2.6%
Lay's Classic Potato Chips, 1 oz / 28g packaged 160 160 0%
Coca-Cola Classic, 12 fl oz can packaged 140 140 0%
Skippy Creamy Peanut Butter, 2 tbsp packaged 190 204 7.4%
Nature Valley Crunchy Oats & Honey bar (2 bars) packaged 190 185 2.6%
Bumble Bee Solid White Albacore Tuna in Water, 1 can packaged 100 103 3%
Eggo Homestyle Waffles, 2 waffles packaged 180 167 7.2%
McDonald's Big Mac restaurant 590 676 14.6%
Starbucks Grande Latte, whole milk, 16 fl oz restaurant 190 186 2.1%
Chipotle Chicken Burrito Bowl (white rice, black beans, salsa, lettuce) restaurant 660 536 18.8%
Subway Footlong Italian BMT on Italian Herbs & Cheese restaurant 820 580 29.3%
Olive Garden Lasagna Classico lunch portion restaurant 580 559 3.6%
Domino's Hand-Tossed Cheese Pizza, 1 slice (large) restaurant 230 216 6.1%
Sweetgreen Harvest Bowl restaurant 705 791 12.2%
Cheesecake Factory Grilled Chicken Tostada Salad restaurant 1380 1495 8.3%
Panera Mac & Cheese, large bowl restaurant 970 991 2.2%
Five Guys Hamburger with lettuce, tomato, onion restaurant 700 644 8%
Chicken stir-fry over brown rice (1.5 cups, home recipe) mixed 520 505 2.9%
Spaghetti with marinara sauce, 1 cup pasta + 1/2 cup sauce mixed 380 403 6.1%
Mixed garden salad with vinaigrette + grilled chicken mixed 410 424 3.4%
Beef tacos, 2 corn tortillas with ground beef, cheese, lettuce mixed 530 675 27.4%
Homemade pepperoni pizza, 2 slices, thin crust mixed 620 536 13.5%
Stovetop mac and cheese, 1.5 cups mixed 580 676 16.6%
Strawberry banana protein smoothie (1 scoop whey, 1 cup almond milk) mixed 280 320 14.3%
Breakfast burrito, eggs + bacon + cheese + tortilla + salsa mixed 540 480 11.1%
Beef stew, 1.5 cups (chuck, potato, carrot, onion, broth) mixed 460 384 16.5%
Pad Thai with shrimp, restaurant-style 1.5 cup portion mixed 720 645 10.4%

Pooled accuracy breakdown

Slice Pooled MAPE n
Overall (40 meals) ±7.7% 40
Single foods ±3.2% 10
Packaged goods ±4.9% 10
Restaurant chains ±10.5% 10
Mixed home recipes ±12.2% 10

Restaurant-chain meals are the widest category — chain-specific entries lag the actual menu data on several major US chains, particularly when chain portion size changes between the database snapshot and the test date. Dataset-published value: ±7.7%. Arithmetic mean from raw CSV: ±7.7%.

Failure modes

The four meals where Lose It! produced its widest absolute percentage error:

Logging speed sidebar

Median per-meal logging time: ~45 seconds. Search-and-log workflow with optional barcode and Snap-It photo capture. Barcode scans drop to ~9s. The Snap-It workflow runs ~12-18s but produced enough accuracy variance in the test set that the protocol used search-and-log as the default; Snap-It results were recorded as a sensitivity check (see methodology v1.0 §3.5). Cohort comparison: PlateLens ~3s, MacroFactor ~38s, Cronometer ~42s, MyFitnessPal ~47s.

Where Lose It! wins (and where it doesn't)

Lose It! leads the cohort on two measurable axes:

What it does not lead on:

Compared to the other four in this cohort

Lose It! ranks fourth of five on pooled accuracy — behind PlateLens (±0.7%), Cronometer (±2.8%), and MacroFactor (±2.9%); ahead of MyFitnessPal (±9.7%). Per-meal accuracy is not where Lose It! competes. It competes on first-month retention for the beginner tracker and on price. For a reader who has never tracked, who is unsure whether they will keep doing it, and who does not want to spend $60-80/year to find out, Lose It! is the rational entry point. The data we have on retention vs accuracy suggests that for that user, the price-and-friction profile matters more than the 5-percentage-point MAPE gap to Cronometer.

Contextualising the cifra

The ±7.7% pooled MAPE in this Q2 benchmark places Lose It! in the middle of what the DAI 2026 May validation study (624 meals, 244-patient cohort, 86-nutrient panel, 96% adherence at 12-week) called the "consumer search-and-log" band — typically 6-10% pooled MAPE for crowdsourced databases with a verified-barcode layer. The 7.7% here is consistent with that pattern. For an absolute-beginner tracker running an aggressive deficit, ±8% noise on a 1,800-kcal day is roughly ±144 kcal of daily uncertainty, which is small enough that a consistently logged 500-kcal deficit will still produce a measurable weekly weight trend over a 4-week window.

Re-test schedule

Lose It! retests quarterly. Next retest: Q3 2026 (August collection, September publication). Earlier retest triggered if the Snap-It photo recognition model ships a major version update. See RSS and update log.

Limitations

Sources & methodology

Editorial standards. Lab reports apply the published v1.0 accuracy protocol. No sponsored placements; no affiliate revenue.