America's Most Optimistic Serving Size Award 2026

What we examined

The proposed award asks which eligible US product has the largest gap between its label serving and the median portion reported to PerBiscuit. It is not an award for calories, sugar or overall nutritional quality.

What the label shows

The label serving is the quantity used for the product's nutrition calculation. It is product data, not a recommendation about how much an individual should consume.

What it will be compared with

The comparison quantity will be the median of valid, product-specific US portion reports. An assumed amount, a full pack or a per-100g comparison value cannot stand in for those reports.

Data classification: label servings are product or third-party database data. Portion reports are user-reported PerBiscuit data. Ratios, intervals and ranks are PerBiscuit calculations.

Eligibility rules

• The exact product variant and pack must be available in the US market.
• The label serving, units, source and evidence-check date must be recorded.
• The product must have at least 30 valid reports from at least 20 distinct sessions across at least 30 days.
• Reports with incompatible units, unresolved product matches or failed submission checks are excluded.
• At least ten products must independently qualify before a ranked shortlist is published.

How the gap is calculated

Reported-portion ratio = median reported portion ÷ label serving

A ratio of 2.0 means the sample median is twice the label serving. It does not mean every contributor selected that amount.

Confidence rules

Each eligible product will receive a 95% bootstrap confidence interval for its median reported-portion ratio. The sample size, collection dates, median and interval will appear beside the result.

A sole winner will be named only if the leading product's lower interval bound is above the runner-up's upper bound. Otherwise, the result will be described as tied or inconclusive.

Results

Interpretation

If a winner is eventually declared, it will reflect the largest supported gap in the available PerBiscuit sample. It will not establish manufacturer intent, label illegality or an effect on an individual's health.

Limitations

• PerBiscuit contributors are a self-selected sample.
• Reports may change as the contributor base and product range grow.
• Country, variant and package mismatches can create misleading comparisons.
• The proposed confidence method has not yet received independent statistical review.

Why the information may be useful

A carefully qualified result may identify labels that readers often need to rescale. The publication threshold is designed to stop an entertaining idea from outrunning its evidence.

Method and sources

See the complete provisional award specification, the source categories and the review process.

Last reviewed and corrections

The proposed rules were last reviewed on July 14, 2026. The status will remain provisional until the threshold and review requirements are met. Please submit concerns through the corrections process.