Understanding the COA Purity Score
The COA Purity Score condenses a full lab report into a single number. It rewards comprehensive testing and clean results while penalizing exceedances and incomplete data. Here is exactly how it works.
Base score: clean percentage
The foundation is the clean percentage: the ratio of analytes that came back ND (not detected) or within acceptable limits, divided by total analytes tested. If a COA tests 80 analytes and 72 are clean, the base score starts at 90.
This rewards labs that test broadly. A product that tests 120 analytes and passes all of them earns a higher base than one that tests 20 and passes all 20, because the first product demonstrated cleanliness across more categories of risk.
Bonus components
The score adds bonus points for testing high-priority categories that many products skip entirely:
PFAS testing (+5): testing for forever chemicals via EPA 533 or 537.1. Most brands do not test for PFAS because the testing is expensive ($300-$500 per sample) and results might be bad. Brands that voluntarily test get credit.
Microplastics testing (+3): relatively new testing category. Few labs offer it, fewer brands request it. Bonus rewards early adopters.
Heavy metals via ICP-MS (+3): testing metals via the most sensitive method (EPA 200.8 / ICP-MS) rather than less precise methods like ICP-OES or colorimetric strips.
Coverage breadth (+up to 4): scaled bonus for testing across multiple panel categories (metals, minerals, PFAS, VOCs, microbiological, radiological). More categories tested means more transparency.
Penalties for exceedances
Any analyte that exceeds its EPA MCL or relevant health guideline triggers a penalty. The penalty scales with severity: a value at 110% of the MCL deducts less than a value at 500% of the MCL. Contaminants with no safe level (lead, PFAS) are penalized more aggressively per unit of detection.
Multiple exceedances compound. A product with three flagged analytes receives a larger total penalty than three individual penalties summed, reflecting the compounding risk of multi-contaminant exposure.
Verification tier caps
The final score is capped by the verification tier of the available data. This prevents a product with minimal testing from achieving an artificially high score.
Tier 1 (self-reported, no COA): score capped at 50. Without any lab data, a product cannot be evaluated.
Tier 2 (COA from non-accredited lab): capped at 70. The data exists but the lab's credibility is not independently verified.
Tier 3 (COA from ISO 17025 lab): capped at 90. Credible lab, credible data, but limited to what was tested.
Tier 4 (comprehensive, multi-panel, accredited): no cap. Full transparency earns full scoring range.
Worked example
Product X provides a COA from an ISO 17025 lab testing 95 analytes across metals, minerals, PFAS, VOCs, and microbiology.
Base: 91 of 95 analytes are clean = 95.8 base score.
Bonuses: PFAS tested (+5), ICP-MS metals (+3), 5 panel categories (+3) = +11 bonus, bringing raw score to ~106.8, capped at 100.
Penalties: arsenic detected at 8 ppb (MCL 10 ppb) = minor penalty (-2). Two VOCs detected below MCL but above health guidelines = -3 total. Adjusted: ~95.
Tier check: ISO 17025 lab with 5+ panel categories = Tier 4, no cap applied. Final score: 95.