When the Label Lies: How Analytical Testing Laboratories Verify Potency Claims on Finished Herbal Supplements

Here’s a number that’s harder to shake the more time you spend in this industry: roughly 1 in 4 finished herbal supplements, when tested by an independent analytical testing laboratory, don’t match what’s printed on the label. That’s not a fringe finding from a single advocacy group. ConsumerLab’s 2024 annual review, which covers thousands of supplement products across categories, found that approximately 23% of products evaluated failed on potency, identity, or contamination criteria. For botanical products specifically, the failure rate runs higher — driven by the complexity of plant matrices and a false confidence that comes from trusting a supplier’s certificate of analysis as a substitute for real finished-product testing.

If you manufacture or distribute a herbal supplement and your label says “standardized to 5% withanolides” or “500 mg full-spectrum turmeric extract (95% curcuminoids),” that’s not a marketing claim. Under 21 CFR Part 111 and the FD&C Act’s misbranding provisions, it’s a legal representation about your product’s composition. And if an independent lab — or an FDA investigator — pulls your finished product and gets a different number, the problem starts with 483 observations and can end with a mandatory recall.

What “Standardized” Actually Means — and Where It Falls Apart

The word “standardized” on a botanical label is supposed to signal consistency: a guaranteed minimum percentage of a specific marker compound, achieved through controlled extraction and confirmed by testing. In practice, the traceability of that claim varies enormously across the supply chain, and the gap between “supplier tested it” and “it’s in the capsule” is wider than most brands realize.

Take ashwagandha root extract. A label claiming 2.5% withanolides means the product should contain at least 2.5 g of withanolides per 100 g of extract. The industry-standard method for verifying that is HPLC — high-performance liquid chromatography — run against a certified withanolide reference standard. A qualified analytical testing laboratory runs the finished dosage form, integrates peak areas for the relevant withanolide compounds, and compares against a calibrated standard curve. The result is a quantitative, method-validated potency number with a defined uncertainty range.

What you often receive from a raw material supplier’s COA is something considerably less rigorous. Some use UV-Vis spectrophotometry, which is faster and cheaper but far less specific — it measures absorbance at a wavelength that overlaps with dozens of other compounds in the botanical matrix. Others do use HPLC but without method validation documents you can audit, and against reference standards you can’t verify. The COA number looks authoritative. Without knowing which method was used, what reference standard was calibrated against, and what the method’s linearity range and detection limits were, you’re making a business decision on an opaque data point.

We see this gap show up in finished-product verification work with real consequences. A brand sources an ashwagandha extract certified by their supplier at 5.0% withanolides. They formulate at 300 mg per capsule, targeting 15 mg withanolides per serving. When finished capsules come through an analytical testing laboratory for third-party verification, the result comes back at 8.4 mg per capsule — 44% below the label claim. The raw material COA wasn’t falsified. The supplier ran a real test. But the method wasn’t validated against a finished capsule matrix, some withanolide degradation occurred during encapsulation, and nobody verified the finished product before it shipped. The label claim was never anchored to actual finished-product data. That’s where the liability lives.

What an Analytical Testing Laboratory Actually Needs to Get the Method Right

Potency verification isn’t one test — it’s a method selection problem. The right approach depends on the marker compound, the matrix, and the precision needed to defend the claim. Getting this wrong at the method level means the number you report is meaningless regardless of how carefully you ran the instrument.

For curcuminoids in turmeric products, HPLC with UV detection at 425 nm is well-established and USP-referenced. The three major curcuminoids — curcumin, demethoxycurcumin, and bisdemethoxycurcumin — are quantified individually and summed as total curcuminoids. A properly run assay on a 500 mg capsule claiming 95% curcuminoid extract should return a total curcuminoid content within ±10% of label claim to pass under standard potency specifications.

St. John’s Wort is more complicated. Hypericin, the marker associated with the herb’s pharmacological activity, is photosensitive and degrades on exposure to light and elevated temperature. A raw material testing at 0.3% hypericin at receipt can test significantly lower 12 months later under suboptimal storage conditions. This is why a one-time raw material COA tells you nothing meaningful about what’s in the tablet sitting on a retail shelf a year after manufacture. Finished-product potency testing at or near the labeled expiration date is the only data that actually matters for label defense.

Ginkgo biloba follows a similar pattern, with an additional layer of complexity. Its two primary marker classes — flavonol glycosides (specification: 24%) and terpene lactones (specification: 6%) — require separate methodological approaches. Terpene lactones, specifically bilobalide and the ginkgolides A, B, and C, have poor UV absorption and are typically quantified via HPLC with evaporative light scattering detection. Running only the flavonoid fraction and reporting a single composite number would systematically undercount the specification. This is exactly the methodological gap that separates a purpose-built botanical analytical testing laboratory from a general food-chemistry lab that added botanicals to its service menu.

DSHEA, 21 CFR Part 111, and the Compliance Weight Behind Every Label Claim

Under DSHEA — the Dietary Supplement Health and Education Act of 1994 — manufacturers bear full pre-market responsibility for the accuracy and safety of their products. FDA doesn’t review or approve supplement labels before they reach retail. But it enforces aggressively after the fact, and the mechanism is 21 CFR Part 111, the Current Good Manufacturing Practice regulations for dietary supplements.

Section 111.75 is the relevant provision. Before releasing any batch for distribution, manufacturers must verify that identity, purity, strength, and composition meet their established specifications. “Strength” maps directly to potency. The regulation requires that the finished product — not the raw material, not the blend at some intermediate stage, but the actual dosage unit — meet its own label specification. The regulation doesn’t mandate a specific analytical method, but it requires the method be scientifically valid, appropriate to the purpose, and applied to the finished product.

Relying solely on a supplier’s raw material COA as the strength verification record is a documented GMP failure. FDA has cited this exact gap in 483 observations and, in more serious cases, warning letters. Between 2022 and 2025, FDA issued more than 30 warning letters to dietary supplement manufacturers citing specification failures involving identity, purity, or strength — a significant portion of those in herbal supplement categories. A common pattern in the inspection narratives: raw material test records on file, no finished-product verification data in the batch record.

The economics are straightforward. A USP Reference Standard for a botanical marker compound typically runs $300–$800 per unit. A finished-product HPLC potency assay from a qualified analytical testing laboratory adds $150–$400 per sample, depending on compound complexity and matrix preparation. Against FDA’s estimate that the average dietary supplement recall costs roughly $10 million in direct expenses — not counting the brand damage that follows — the per-batch cost of verification is noise.

What a Finished-Product Verification Protocol Actually Looks Like

The brands that handle this consistently well build potency verification into the finished-product specification document during product development, not after the first batch ships. The specification document lists the marker compound or compounds, the quantification method by reference (HPLC method number, pharmacopeial chapter, or validated in-house method), the acceptance criteria — typically label claim ±10% for most botanicals, ±20% for compounds with documented natural variability — and the testing frequency per the batch release protocol.

They send samples of the finished dosage form to an independent analytical testing laboratory. Not the raw material. Not the pre-blend. The actual capsule, tablet, or softgel that will carry the label. They test samples drawn from multiple points in the batch to capture homogeneity variation. And they retain those test records in the batch record, available for regulatory review on request.

For brands shipping samples to our Chicago receiving hub, a finished herbal supplement potency panel on common marker compounds typically returns in 5–7 business days from receipt. That’s fast enough to complete before distribution timeline pressure becomes a factor, and before the batch ships to retail or Amazon fulfillment.

The final piece: real-time stability confirmation. Accelerated stability data at 40°C/75% RH for 6 months — per ICH Q1A(R2) — can support initial shelf-life dating. But whether marker compound potency holds across the labeled shelf life under real-world retail storage conditions requires time-point data at the actual temperatures and humidity levels your product will see. An ashwagandha capsule that passes potency at month 0 may not pass at month 24, depending on packaging material, desiccant choice, and storage environment. The label claim has to hold for the full duration it appears on the label.

None of this requires an elaborate quality system to implement. It requires a specification document, a validated method, a qualified third-party lab, and a batch release decision that waits for the data. That’s a deliberate choice, not a complicated one.

---

Written by Nour Abochama, VP Operations, Qalitex | Quality Consultant, Ayah Labs. Learn more about our team

Ship your sample to our Chicago facility — get a Qalitex CoA in 5–7 days. Contact us

Related from our network

• ISO 17025-Accredited Herbal Supplement Testing — Qalitex Laboratories — Full-panel raw material and finished-product testing with USP-compliant HPLC methods, traceable CoA documentation, and California lab accreditation.