Walk a trade-show floor and every botanical-ingredient stand will tell you their material is "standardised." The word does a lot of work — and almost none of it consistently. For a buyer comparing two extracts of the same plant, "standardised to 10% mitragynine" from one supplier and "standardised to 10% mitragynine" from another can mean materially different things in the bag.
This piece explains what standardisation should mean for a botanical extract, what a credible Certificate of Analysis (CoA) looks like, and the three failure modes we see most often.
The plain-English definition
A standardised botanical extract is one in which the concentration of a defined marker compound is held within a stated tolerance, measured by a validated analytical method, on every batch released. All three parts matter:
- Defined marker compound. The marker should be the compound that drives the activity of interest, or — when that compound is contested or multi-factorial — a representative phytochemical fingerprint. "Standardised to total alkaloids" without specifying which alkaloids is not standardisation; it's a guess.
- Validated analytical method. HPLC (high-performance liquid chromatography) is the workhorse for most botanical markers, with HPTLC and LC-MS as supporting techniques. A method is "validated" when its specificity, linearity, precision, accuracy, and limits of detection have been documented against reference standards.
- Every batch released. Per-batch testing, not "representative batch" or "lot-to-lot type" claims. A standardisation programme that doesn't produce a CoA for the batch in front of you is a marketing claim, not a quality control.
What a credible CoA looks like
The single most useful question a buyer can ask is: "Show me the CoA for the actual lot you're proposing to ship." A credible CoA names:
- The lot or batch number that ties the certificate to the specific material
- The marker compound and its assayed concentration with a clear ± tolerance
- The analytical method used (e.g. HPLC-UV with a citation to the validated procedure)
- Identity verification — typically TLC, HPTLC, or DNA barcoding confirming the plant species
- Contaminant screening — heavy metals (lead, arsenic, mercury, cadmium), pesticide residue panels, microbial limits, aflatoxins where relevant
- The assay date and the analyst or laboratory signing off
A CoA missing any of these is not necessarily fraudulent — but it is incomplete, and the gaps are exactly where standardisation claims fail.
Three failure modes we see often
Marker spiking. An extract that doesn't naturally hit the standardisation target is "topped up" with isolated marker compound — sometimes synthetic, sometimes from a cheaper source plant. The marker assay passes; the broader phytochemical profile is wrong. Fingerprint chromatography catches this; a single-marker assay does not.
Batch averaging. A supplier produces 100 kg of material, tests one composite sample, and issues the same CoA against every 1 kg sold. If the source biomass is heterogeneous (most field-cultivated botanicals are), the actual standardisation across the population is much wider than the certificate suggests.
Method drift. The HPLC method on the CoA was validated five years ago against reference standards from a supplier who no longer ships them. Without periodic re-validation, the assay reports a number whose meaning has slowly diverged from what it originally measured.
What we do at Phyto-Medichem
Every lot leaving our facility carries a per-batch CoA generated from our in-house HPLC laboratory and cross-checked against an independent ISO/IEC 17025-accredited reference. The methods are version-controlled against current pharmacopoeial standards and re-validated annually. Identity is confirmed at intake and again at finished-extract release.
Standardisation is, fundamentally, an accountability claim. The supplier is asserting that what's on the label is what's in the bag, traceable to a specific batch, measured by a specific method. If any of those three legs is missing, the word is doing nothing.