What a rating tells you. And what it leaves open.

Nine criteria, explicit weights

Methodology 1.0 separates purity, measured content, identity evidence, endotoxin measurement, documented product breadth, verification keys, batch attribution, public reports and freshness. The graphic shows the actual maximum points. Together they total 100.

The criteria have different weights. One missing documentation item is not equivalent to an entirely missing measurement category. Open a vendor’s calculation and read its components beside the underlying reports. The weights are a published editorial choice, not a law of nature.

Further reading: Methodology 1.0

Which reports count?

The active calculation considers eligible reports from the accepted laboratories Janoshik and Analiza. It selects the latest valid report for each product. Older reports remain documented but do not add another counted product. Reports older than 365 days leave the active calculation.

The number of uploaded files is therefore not automatically the number of counted products. Multiple reports for one sample or several sizes of a product do not necessarily add the same breadth as an additional documented product. Read the exclusion reasons rather than comparing the largest document counts.

Further reading: Methodology 1.0

No points does not always mean a poor result

An undocumented identity finding earns no positive identity-evidence points. It does not mean that the laboratory identified the wrong substance. An explicitly negative identity result is a different event, treated separately by the methodology.

Other missing measurements follow the same distinction: unknown information remains identifiable as unknown. The score rewards recorded evidence that can be used under its rules. It should not be interpreted as if every absent measurement had returned a poor laboratory result.

Why a score can fall without a new incident

Freshness is evaluated across all counted reports. A report up to 90 days old contributes 10 freshness points, up to 180 days contributes 6 and up to 365 days contributes 3. The vendor’s freshness component is the mean of those contributions.

For two counted products with reports aged 30 and 200 days, freshness is (10 + 3) ÷ 2 = 6.5 out of 10. One new report does not refresh the older one. Crossing an age threshold can therefore change the next calculation without a new adverse finding.

Further reading: Methodology 1.0

What a rank cannot establish

A high score is not medical approval, a guarantee for every batch or advice about personal use. Price, delivery experience and customer reviews are also different from analytical evidence. Consider them separately.

For closely ranked vendors, pay particular attention to breadth, missing fields and freshness. A decimal place can suggest more certainty than the data supports. The most useful next step is often to open the report behind a meaningful component rather than looking at the ranking again.