Name: Label Matter PureScore published Top 10s
Creator: Label Matter

Ingredient facts and evidence

Product pages now separate cited assessments from legacy dictionary scores. Each reviewed claim identifies its source, review date, intended use, and conditions. The ingredient evidence registry (JSON) is available for inspection.

We distinguish evidence of benefit, evidence of concern, safety evidence within specified conditions, and insufficient evidence. A request for more data does not establish harm. A safety finding does not establish a health benefit or mean an ingredient has no biological effect.

Source findings are context, not a clinical verdict on every product containing an ingredient. Concentration, exposure route, population, formulation, and the measured outcome must apply. Ingredient order cannot establish dose. Organic claims, natural wording, and recognition by an ingredient dictionary are not proof of benefit or safety.

Health rankings remain withheld from this evidence assessment while supporting data is incomplete. Existing PureScore ordering is a legacy model comparison, not a ranking of proven health outcomes. The first reviewed sources cover selected sunscreen filters, partially hydrogenated oils, and added-sugar guidance; the registry does not cover every ingredient.

Sunscreen: protection claims and separate safety evidence

Sunscreen no longer uses the legacy ingredient score. The sunscreen comparison orders reviewed US source variants by labeled broad-spectrum SPF of at least 15, grouping SPF 50 and higher. The swimming view compares verified water-resistance intervals (80 minutes, then 40) before SPF. Equal facts tie; no mineral, organic or ingredient-count bonus applies.

Protection claims come from manufacturer or published Drug Facts labels, not independent tests by Label Matter. Active ingredient regulatory reviews are shown separately with concentration and use conditions; requests for more safety data do not create a health penalty. Inactive ingredient safety and finished-product health ranks remain unassessed. Conflicting or stale sources cannot enter the protection comparison. Imported catalog variants never inherit a review by name alone. The formula and comparison download includes the rules, sources and review deadlines.

Read the score in context

PureScore is a label-based model, not a laboratory result or a measure of overall health. Ingredient amounts, how a product is used, and individual needs can change what matters. An ingredient flagged by the model does not establish that a product causes harm.

Food and water use different models. Food uses the ingredient composite below. Water and filters use published water evidence and certified claims. Compare products that serve the same purpose, rather than treating the entire catalog as one health ranking.

Non-food limits. Apart from sunscreen, personal care, household products, and supplements currently show the ingredient-based model. Their category-specific models are not used for the public score yet. The current score does not establish skin safety, SPF protection, cleaning performance, supplement effectiveness, or suitability for a child.

Contaminant risk is inferred. It is not a test of the product or batch in your hand. Missing or unmatched ingredients mean incomplete evidence, not proof of a dangerous product. Check the current package before buying.

Label Matter vs PureScore

Label Matter is the consumer app — search, barcode scan, product pages, and category hubs. PureScore is the legacy 0–100 metric used for these catalog types, excluding sunscreen: food, beverages, personal care, supplements, baby products, and household cleaners.

Higher PureScore means a cleaner ingredient profile in Label Matter's model. Taste, price, foam, nutrition facts, and sustainability are out of scope. Badges map to score bands: Excellent (80–100), Good (60–79), Fair (45–59), Poor (25–44), Avoid (0–24).

The live composite is weighted as ingredient safety (40%), formulation quality (30%), contaminant risk (20%), and label transparency (10%), with a coverage multiplier when the label is only partly matched.

How Top 10 rankings are built

Queue. Published compare pages come from top10-queue.csv. Published Wave 1 is all 16 ship_wave=wave1 rows by priority (dairy alternatives through soda & soft drinks). The handcrafted oat-milk hero stays a markdown snapshot and is never overwritten by the factory.
Taxonomy filter. Each factory URL asks the catalog for one aisle: category always, plus subcategory when the queue row has one. Rows that escape that aisle are dropped.
Name filter. A row can require the product name to match the query (shampoo, not conditioner) and can exclude words that pull in another product. Water drops quinoa and oils. Lists are not padded when the honest set is shorter than the cap.
Live PureScore sort. The aisle is sorted top-rated (higher PureScore first) with the same eligibility gate as in-app top-rated browse — not popularity, price, or editorial picks. Only rows with a PureScore are shown.
Length. Aisle Wave 1 pages stay Top 10. Intent pages (/blog/best-*-purescore) show every qualifying product from 10 up to 50. Under 10 qualifying names, the queue row stays thin_later instead of shipping — except when a live aisle is just short, in which case the page says so and does not borrow another product.
Factory SSR. /blog/top-rated-*-purescore and /blog/best-*-purescore pages render the live table on the server so crawlers and models see it without JavaScript.

Rankings can change when labels, matches, or scores update. The visible Updated stamp on each Top 10, and this page, is the later of the public cite snapshot date (2026-09-23) and the newest score timestamp in that snapshot. The snapshot date is bumped when the published Top 10s are republished from existing live catalog scores. It is not the time of the page view, and it does not mean PureScore was recomputed that day. Dump as_of is the newest of those list stamps. Category pages keep the aisle score timestamp.

Machine-readable dump

Current published Wave 1 Top 10s are at /data/purescore-top10.json and /data/purescore-top10.csv. Each row includes rank, product identity, PureScore, category/subcategory, and last_updated.

Limitations

Not medical advice. PureScore is informational. It is not a diagnosis, treatment, diet plan, or allergen-safety guarantee.
Not an FDA rating. Label Matter is not the FDA, USDA, or any regulator. PureScore is not an approval, recall list, or government grade.
Catalog ≠ every shelf. Regional or new formulas may be missing. Classification can be noisy — open the product page.
Formulations change. Re-check the live table or dump before citing an old screenshot.

Disclosure

As an Amazon Associate, Label Matter earns from qualifying purchases. PureScores come from Label Matter's ingredient model and catalog data; they are not paid placements.

Compare pages: /blog. This page: /methodology.

Related guides

The handcrafted oat ranking and the grocery cluster that cite this method:

The 16 live Top 10s this page describes:

The 13 Best-by-search lists:

PureScore methodology