PlateRover guide

Why Do Nutrition Apps Show Different Numbers?

Different numbers do not automatically mean one app is broken. The apps may be using different food records, serving weights, preparation states, calculation methods, or update dates.

The serving may not be the same

One app may call a banana “medium” at a specific gram weight while another uses 100 g. A cup can also vary with how food is chopped, packed, drained, or cooked. Compare grams before comparing nutrient totals.

For packaged foods, check whether each app is showing per serving, per package, or per 100 g.

Preparation changes the food record

Raw and cooked foods should not be interchangeable. Water loss can concentrate nutrients per 100 g, while boiling can add water. Oil, breading, salt, drained liquid, and edible skin also change the correct record.

A generic “chicken breast” result might describe raw meat, roasted meat, meat with skin, or a prepared product. Those are different foods for database purposes.

Databases represent different evidence

USDA FoodData Central contains several data types, including Foundation Foods, SR Legacy, survey foods, and branded foods. Some values come from analysis, some are calculated, and branded data can originate with manufacturers.

Apps may also accept community entries. A record without a clear source or exact description deserves more scrutiny than an official record or current package label.

Labels round and products change

Nutrition labels follow rounding rules, so multiplying a rounded per-serving number does not always reproduce an unrounded database total. Manufacturers may also reformulate a product before every database receives the new label.

When logging a packaged product, use the current label in front of you and verify the number of servings. For generic food, choose a record whose preparation and serving weight match.

Example: two rice entries may disagree because one describes 100 g cooked and another describes one cup. Convert both to the same gram basis before deciding whether the underlying data differ.

A simple record-selection checklist

  • Match the exact food and preparation.
  • Compare the gram weight, not only the household unit.
  • Prefer a current product label for the exact packaged item.
  • Prefer an authoritative generic-food source with an identifier.
  • Keep estimates appropriately rounded.

Sources and further reading