← Testing & Certification
THE STANDARD — DOCUMENTS

What a certificate of analysis proves, and what ours do not

A certificate of analysis proves one thing: a named laboratory measured named properties on one named sample, on a stated date. It does not prove the sample was your product. TARE holds three, dated December 2025 to March 2026 — one names its own fill lot, two name the ingredient supplier’s raw material — and each prints the axes left unrun.

A certificate is a composition and contamination document: it says what is in a material and how much, and is silent on every question about outcome. What separates a real one from a specification sheet is set out in the definitional article. This page is the test itself, and TARE’s own score on it.

Six questions to ask of any certificate of analysis

Ask them in this order, of any brand, including this one. Five of the six are answerable from the face of the document in a minute.

  1. Whose sample was tested? The finished product you would receive, or a raw material the ingredient supplier bought before anything was filled.
  2. Does it name a lot, and is it the lot in your container? A certificate with no lot code names nothing. One naming a raw-material lot names something real that is still not the thing on your shelf.
  3. Who is the customer of record? A certificate is addressed to whoever commissioned the work. If that is not the brand selling to you, the brand did not commission it.
  4. Which of the four axes were run — identity, potency, heavy metals, microbiology — and does any result column read Not Tested? A printed table with an empty result column is not a result.
  5. Is each figure a result or a specification? A specification is a target the maker set; a result is what an instrument returned. Only one is evidence, and most documents in this category carry only the first.
  6. What is the date, and does it describe the product on sale today? A certificate can be genuine, complete, and about a formula that has changed since.

A seventh is asked of the answers rather than the paper: does the brand publish the questions it fails? The six above can be copied in an afternoon. The table below cannot, because filling it in honestly is the part nobody wants to do.

How TARE’s three certificates answer those six questions

Every cell is transcribed from the document itself. Where an axis was not run, the cell says so: the absence is the finding.

TARE’s three certificates, scored against the six questions · all three read 27 August 2026; the magnesium document re-read 4 and 5 September 2026
 Under Load — collagenBase Work — creatineNight Work — magnesium
Whose sampleTARE’s own fill lot 121125450 — the lot the tub was filled fromthe supplier’s raw material, lot CM20251225A — a sealed poly bag from Accobio USA Inc.the supplier’s raw material, lot MGL260212 — a sealed poly bag of powder
Customer of recordnot a field on this document — the laboratory belongs to the company that makes the powderVox Nutrition — not TARE, not the manufacturerVox Nutrition — not TARE, not the manufacturer
LaboratoryViaCore Nutrition, El Reno OK — the laboratory of the company that makes the powderEthos Analytics, Phoenix AZ — independent of the makerEthos Analytics, Phoenix AZ — independent of the maker
Work dated2025-12-222026-02-182026-03-09
IdentityNOT RUN — no infrared scan, no chromatographic fingerprint. The ingredient line reads “collagen peptides”, which is a statement rather than a testMEASURED — scored from the assay row below; TARE argues against itself on that under the table †NOT RUN — no identity section exists. Nothing on the document distinguishes glycinate from oxide, citrate or a buffered blend
PotencyMEASURED — protein 9.8 g against a 9 g specification, method ISO 14891:2002, in the document’s 10 g servingMEASURED — creatine monohydrate 99.87% against a specification of more than 99.5%, HPLC, limit of quantitation 100 ppmNOT RUN — no potency section, no assay, no elemental magnesium result; Serving Size printed N/A
Heavy metalsMEASURED — lead 0.07 mcg against a 0.5 mcg Proposition 65 limit by ICP-MS; mercury, arsenic and cadmium below the detectable limitMEASURED — ICP-MS on a numeric USP <233> page, every line a passNOT RUN — the Result column reads “Not Tested” on arsenic, cadmium, lead and mercury, with a dash under Pass/Fail
MicrobiologyMEASURED — five rows, every one a pass, by AOAC and VIDAS methodsMEASURED — Petrifilm/qPCR, every line a passMEASURED — Petrifilm/qPCR, seven organisms, tested 2026-03-09, pass
Lot Mark3 / 4, on TARE’s own fill lot4 / 4, on the supplier’s raw material, not on a TARE tub1 / 4, microbiology only, on the supplier’s raw material

The objection TARE prints against itself. Identity and potency for Base Work are both scored from one HPLC row. On a one-ingredient powder an assay reporting 99.87% creatine monohydrate says what the substance is as well as how much of it there is, and TARE counts it as both — while printing that someone could reasonably call it 3 / 4.

A Lot Mark counts what a document measured; it does not weigh who measured it or what was in the bag. Base Work’s 4 / 4 is the best count in the range and sits on raw material tested for another company. Under Load’s 3 / 4 is the weaker count and is the only one of the three run on a lot that went into a TARE container. Neither number means anything read apart from the row saying whose sample it was, which is why the provenance sits in the same table rather than a paragraph away.

What NOT RUN means, and where ours read it

NOT RUN means no row of any kind exists on the document, and the absence is the finding. It is not a failure and not a pass. It is also not the same as a figure with no test behind it — a composition line or a limit — which is a weaker state again.

Four cells above read NOT RUN: Under Load’s identity, and Night Work’s identity, potency and heavy metals. The magnesium document prints a heavy-metal table with arsenic, cadmium, lead and mercury on it and the words Not Tested in every result cell; the LOQ, LOD and Limit columns beside them are the method’s floors and the acceptance limit, and none is a result. That is why Night Work’s mark is 1 of 4, and why no measured magnesium figure appears anywhere on this site.

What TARE does not have

Where the documents themselves are published

Each is published whole and unaltered, under written permission given by the manufacturer on 17 August 2026. This page is the comparison; Testing & Certification is the archive.

Related: how a specification differs from a result, and the sixty-second label procedure. Both are written to be used on any brand, which is the only way either is worth anything.