How to Build a Raw Material Specification Sheet That Actually Protects Your Brand

A few years ago, a mid-sized supplement contract manufacturer received a shipment of ashwagandha root powder. The supplier’s COA looked fine — identity confirmed, assay within range, microbiology passing. Thirty days later, a batch of finished capsules was recalled after consumer complaints about a bitter, solvent-like taste. The root cause: the incoming specification sheet had no acceptance criterion for residual solvents, because the supplier had never volunteered that data, and no one had ever asked.

The specification sheet is where most of that risk either gets captured or gets ignored.

Across the supplement and cosmetic raw material space, inadequate specifications are far more common than adulterated materials. The adulteration gets the press. The spec failures silently drive most of the batch rejections, OOS investigations, and customer complaints that eat into margins every quarter.

Building a specification sheet properly — not copied from a COA, not templated from a competitor’s SOP, but built from the actual regulatory and scientific requirements for your specific ingredient — takes a few hours of deliberate work per material. That investment pays back every time a non-conforming lot is caught at incoming inspection instead of in finished product.

Here’s how to do it.

---

Start With Regulatory Minimums — Then Decide Where to Go Further

Depending on your market and product category, different regulatory frameworks establish baseline testing requirements for raw materials. For dietary supplement manufacturers selling in the US, 21 CFR Part 111 is the floor. Under 21 CFR 111.75(a)(1), you are required to conduct identity testing on 100% of incoming dietary ingredient lots — not a statistical sample, not a periodic audit, every single lot. That’s a harder requirement than most manufacturers realize until an FDA investigator asks for records.

For finished dosage forms regulated as drugs in Canada, Health Canada’s Good Manufacturing Practices (Division 2, Part C of the Food and Drug Regulations) requires identity confirmation and a documented specification for every starting material. The EU’s GMP Annex 8 (Sampling of Starting and Packaging Materials) adds sampling plan requirements on top of that.

Compendial standards — USP, Ph.Eur., BP — provide monographs for hundreds of common ingredients. A USP monograph specifies exactly which tests apply, what the acceptance criteria are, and which methods must be used to generate them. If a USP monograph exists for your material, it is the baseline. You can tighten the limits, but you generally shouldn’t relax them without documented scientific justification.

Where regulatory frameworks are silent — and they’re silent on a lot, especially for novel botanicals and proprietary extracts — you set limits based on a combination of historical batch data, published literature, and in-house risk assessment. For a new material with no history, use your analytical testing laboratory to generate baseline data on 5–10 representative lots before setting acceptance criteria. Limits set before any empirical testing are guesses. Limits set from real data are defensible.

---

The Seven Parameters Every Specification Sheet Needs

A raw material specification isn’t just a list of numbers. It’s a document that links each quality attribute to a defined test method and an acceptance criterion. Without all three elements — attribute, method, criterion — you don’t have a specification, you have a wishlist.

These are the parameters that should appear on virtually every raw material specification for supplement or cosmetic ingredients:

1. Identity Specify both the genus and species (for botanicals), part used, extraction ratio (for extracts), and acceptable test methods. HPTLC and FTIR are most common. For high-risk botanicals with documented adulteration history — echinacea, elderberry, valerian — include a secondary method like DNA barcoding or HPLC marker profiling. The method choice matters: some markers are species-specific; others are present across an entire genus.

2. Assay / Potency For standardized extracts, specify the marker compound, the minimum percentage, and the HPLC method including reference standard source. “Ashwagandha extract, ≥5% withanolides” is incomplete. “Ashwagandha root extract, ≥5.0% withanolides by HPLC per validated in-house method REF-012 using a USP reference standard” is defensible.

3. Physical Parameters Moisture/water content (by method — USP <731> Loss on Drying or Karl Fischer per USP <921>), particle size distribution where relevant, bulk density, color, and odor. These matter more than people give them credit for: moisture content above 10% in a botanical powder is a significant microbial growth risk factor before the material even arrives at your facility.

4. Microbiology Reference USP <2021> for Total Aerobic Microbial Count and Total Combined Yeast and Mold Count, and USP <2022> for Microbiological Examination of Nonsterile Products. Specify pathogen absence requirements per USP <62> (Salmonella, STEC, Staphylococcus aureus, and Pseudomonas aeruginosa for topical products). Typical limits for oral botanical powders run ≤100,000 CFU/g TPC and ≤1,000 CFU/g TYMC, but tighter limits are warranted for materials used in products intended for immunocompromised populations or children.

5. Heavy Metals / Elemental Impurities USP <232> defines oral Permitted Daily Exposure (PDE) limits for elemental impurities. Cadmium, for example, has a PDE of 3 µg/day for oral route. Lead is 5 µg/day. Given typical serving sizes and expected daily intake, you can back-calculate maximum acceptable concentration limits for each element in your raw material. Many supplier COAs report heavy metals as a single “total” number or test only against non-binding limits — that’s not adequate for USP <232>/<233> compliance. Specify each element individually.

6. Pesticide Residues For botanical materials, this is non-negotiable. Reference the applicable MRL table — EU MRL regulations (EC 396/2005 and its amendments) offer the most comprehensive list and are widely used as a global benchmark even outside European markets. For organic-certified materials, include verification of organic certification status as a document requirement, not a substitute for testing.

7. Adulterants and Ingredient-Specific Impurities This is the parameter that most generic specification templates omit entirely. Adulterants are specific to the ingredient and the economic incentive. For weight-loss botanicals, include testing for undeclared pharmaceutical substances (sibutramine, phenolphthalein). For protein powders, include testing for non-protein nitrogen compounds (melamine is the most historically significant, but the list is longer). For botanical extracts with documented adulteration history, specify which adulterant markers must be absent. FDA’s Dietary Supplement Label Database and USP’s Dietary Supplements Compendium are useful starting points for known adulterant profiles.

---

Setting Acceptance Criteria That Hold Up Under an Audit

The most common specification problem we see isn’t a missing parameter — it’s acceptance criteria that were copied verbatim from a supplier’s COA and have never been independently verified.

That approach creates a circular validation problem. The supplier says their material contains ≥2.5% rosmarinic acid because their HPLC says so. You accept the material because their COA says ≥2.5%. Your specification says ≥2.5% because that’s what their COA says. Nothing in that chain provides any independent assurance that the number is real, that the method is valid, or that 2.5% is the right requirement for your intended use.

Setting defensible acceptance criteria means:

• Starting from the compendial monograph, where one exists. If USP specifies ≥0.5% hypericin by HPLC for Hypericum perforatum dry extract, that’s your floor.
• Calculating product-level requirements backwards from label claims. If your finished product label claims 600 mg ashwagandha per serving with ≥5% withanolides, your raw material specification for the ashwagandha extract needs to account for blend-down ratios, excipient dilution, and manufacturing losses. What concentration in the raw material actually delivers 30 mg withanolides per serving?
• Tightening microbial and contaminant limits for materials in high-sensitivity finished products. Finished product microbial limits are set at USP <2023> product categories — but if your raw material accounts for 60% of your blend by weight, it needs to come in well under the finished product limit to leave you headroom.
• Documenting the rationale for every limit. “Per supplier specification” is not a rationale. “Derived from USP <232> PDE for cadmium at maximum daily serving of 2 capsules, 600 mg per capsule” is a rationale.

---

Linking Every Parameter to Your Analytical Testing Laboratory

A specification is only as useful as the testing program attached to it. For each parameter on your spec sheet, someone needs to be responsible for generating the data — either your in-house QC team or a contracted analytical testing laboratory. Both are legitimate; most manufacturers use a hybrid.

A few things that make this work in practice:

Define the method at the specification level. “Heavy metals by ICP-MS, elements per USP <232>” tells a third-party analytical testing laboratory exactly what you need. “Heavy metals, NMT 20 ppm” tells them almost nothing.

Clarify which tests are in-scope for which lots. Identity testing under 21 CFR 111 happens every lot. Full compendial panel testing might happen on first qualification of a new supplier and then annually, with abbreviated incoming testing per lot. Your specification should include a testing frequency table or reference your material testing SOP. If your analytical testing laboratory doesn’t know whether a test is required every lot or skip-lot, you’ll get inconsistency.

Make sure your lab’s scope of accreditation covers your methods. ISO/IEC 17025 accreditation means the lab’s specific, listed methods have been independently assessed for competence. If the accreditation scope doesn’t list your specific matrix (e.g., botanical powders by HPLC), ask for the validation data supporting method extension.

Plan turnaround time into your incoming inspection process. If microbiology results take 5–7 days and your specification requires micro testing every lot, your incoming hold area needs to accommodate that cycle time. Releasing materials before results are in is a cGMP deviation — one that’s avoidable with logistics planning.

---

Four Specification Mistakes That Routinely Cause Problems

After reviewing hundreds of raw material specification packages across categories, certain mistakes come up again and again:

Accepting “consistent with supplier’s specifications” as an acceptance criterion. This appears on more specification sheets than you’d expect. It delegates your quality decision entirely to the supplier, which is exactly what cGMP regulations say you cannot do.

No revision history or effective dates. A specification without a version number and effective date is ambiguous under audit. Which specification was in effect when lot 2024-XYZ was received? If you can’t answer that, the batch record is incomplete.

Specifications that list parameters but not methods. “Arsenic, NMT 1 ppm” is a criterion. Without specifying the method (ICP-MS? ICP-OES? colorimetric?), two different labs will potentially produce different results against the same criterion, and you have no way to adjudicate.

Setting all raw material specifications to the same template values regardless of ingredient. A vitamin C powder and a powdered botanical extract don’t share the same risk profile. Microbial limits appropriate for an ascorbic acid with known low bioburden are not appropriate for a milled root powder that passed through soil-contact growing conditions. Specifications should be ingredient-specific.

---

Where to Start

If you’re staring at a library of 40 raw material specifications that you know need work, pick your three highest-volume dietary ingredients first — the ones that represent the largest share of finished product weight. Build those specifications from scratch: USP monographs first, then your product-level requirements, then a testing frequency table. Document your limit rationale. Review them with your contract analytical testing laboratory to make sure every test is within their validated scope.

Then work through the rest systematically, prioritizing by volume and historical non-conformance rate. Ingredients with prior OOS events should move up the list.

A well-built specification sheet doesn’t eliminate supplier risk. But it does mean that when something goes wrong — and eventually, something will — you have a documented, defensible basis for your acceptance decision and a clear chain of evidence showing you exercised due diligence. In a regulatory environment where FDA is increasingly scrutinizing raw material controls under 21 CFR 111, that chain of evidence isn’t just good practice. It’s what separates a warning letter from a closeout.

---

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

Talk to our team about raw material testing Contact us

Related from our network

• How ISO 17025-Accredited Testing Protects Your Supplement Supply Chain — Qalitex Laboratories covers the accreditation standards and scope requirements that US supplement brands should demand from their testing partners.
• Understanding USP and Compendial Testing for Finished Supplements — A deeper look at compendial method requirements and how finished product testing ties back to raw material specifications.