Editorial standards

Every Labelgrade page is held to the same standards. This is the operating manual.

Sourcing

• The primary source for nutrition data is USDA FoodData Central. Every product page lists the USDA FDC ID and a direct link to the source entry.
• Secondary sources (when USDA does not cover a product) include Open Food Facts and the current manufacturer label image. Use of secondary sources is disclosed on the page.
• Retail pricing comes from major US retailers' public-facing product pages, sampled at the date stated on the page. Pricing is not used in the Labelgrade score until it is verified across at least two retailers.

Accuracy

• Every numeric claim in a published post must correspond to a value in the underlying data source.
• We use the exact USDA per-serving values. When the USDA-reported serving size differs from the typical retail cup or package size, we note both in the body of the post.
• Estimates, ranges, or "approximately" qualifiers are used only where the underlying data warrants them, and never to fudge a number.
• We do not extrapolate values that aren't reported. If USDA's entry does not list (for example) fiber, we write "not reported" — never zero by assumption.

Voice and tone

• Direct. Every post leads with the answer to the implied question in the headline.
• Factual. No hype adjectives ("amazing," "the best," "incredible") in our voice. We may quote the manufacturer's own marketing language with attribution when relevant.
• No medical claims. We do not say a food "supports immunity," "builds muscle," "fights inflammation," or otherwise makes structure/function or health claims unless the FDA has approved the specific phrasing for the specific nutrient content of that product. When in doubt, we state nutritional facts and let readers draw conclusions.
• FDA-compliant nutrient content claims only. We use phrases like "good source of protein" only when the FDA's threshold (≥10% Daily Value per reference serving) is met. We use "high in" only at ≥20% DV.

Conflicts of interest

• Labelgrade participates in affiliate programs with retailers (Amazon Associates and others). Pages with affiliate links display an FTC-compliant disclosure block above the fold.
• Affiliate relationships do not affect Labelgrade scores. The score is computed by the formula at /methodology applied to public nutrition data, before any commercial consideration.
• We do not accept compensation from brands in exchange for coverage, higher scores, favorable framing, removal of unfavorable content, or any other editorial concession.
• We do not run sponsored content, sponsored "best of" lists, or pay-for-placement comparisons. If this changes, this page will be the first place it's disclosed.

Corrections

• If we get something wrong, we fix it within 24 hours of confirmation and log the correction on the public corrections page.
• Corrections to numeric claims are never silent. The corrections page describes what changed, why, and when.
• Minor copy edits (typos, formatting, link fixes) are not logged — only changes to factual claims or interpretations.
• If a reader requests anonymity in the credit for a correction, we honor it.

Updates and refreshes

• Pages in the top 10% of monthly traffic are re-verified monthly. All other pages are re-verified at least quarterly.
• Every re-verification updates the "Last verified" timestamp on the page even when no values change. This protects the integrity of the date, which AI search engines use as a freshness signal.
• When a manufacturer reformulates a product, we update the post within 7 days of detection and add a note to the corrections page.

Reader requests

Readers can request products, corrections, additional comparisons, or clarification on any page. We typically respond within 48 hours. Submit via the contact page.

What we will not do

• Publish health, fitness, or medical advice tailored to individual readers.
• Recommend supplements or products based on commercial relationships.
• Use AI-generated images of real branded products.
• Make claims about ingredients we cannot verify from a public source.
• Run pop-ups, interstitials, or other intrusive ad formats that interfere with reading.
• Use clickbait headlines that misrepresent what's on the page.

Last updated: 2026-05-27. We update these standards as the operation matures. Material changes will be noted on the corrections page.