Two phosphatidylserine suppliers can send you passing COAs for twelve consecutive batches and still present very different risks. One runs a centered, stable process that would rarely produce an out-of-spec result even if its controls loosened. The other drifts close to the specification limit every quarter and passes only because its lab reports against generous rounding rules.
Capability analysis is the tool that separates these two. It asks not "did this batch pass" but "how much margin does the process have against the specification, and is that margin stable over time." This article gives US and EU buyers a practical review method: what data to request, which indices to read, what thresholds to demand, and where the review fits in the annual supplier evaluation. For non-GMO and European programmes, our sunflower phosphatidylserine page covers the specification and documentation set buyers typically request.
Why a Passed COA Is Not Proof of a Capable Process
A COA is a point measurement: one sample from one lot, tested once, against a specification. A capable process is a distribution: hundreds of lots whose assay, peroxide value and moisture results cluster tightly and centered, with the specification limits far out in the tails. Under US dietary supplement GMP (21 CFR Part 111), supplier qualification rests on verification of specifications — but your risk as brand owner depends on how often those specifications could realistically be missed, not just whether this lot passed.
The distinction is commercial as well as technical. A process running at Cpk 1.33 — the common minimum for key attributes — produces roughly 63 defective parts per million against a two-sided tolerance when stable. A process at Cpk 1.00 produces about 1,350 ppm. Twelve conforming COAs from that second supplier tell you little about lot thirteen, because the process sits on a knife edge where any small shift — a new enzyme batch, a drier rebuild, a seasonal raw material change — moves it over the line.
The Capability Data to Request from a PS Supplier
Most capable suppliers can produce this pack within days. Put it in the qualification file and refresh it annually:
- Trend data: the last 20 to 30 lots, showing individual results for the attributes you care about — typically PS assay, peroxide value, acid value, moisture or water activity, and particle size where your formulation depends on it.
- Control charts for the same attributes, with limits calculated by the supplier's quality system, not drawn by hand for the audit.
- Capability indices: Cp and Cpk (within-subgroup) and Pp and Ppk (overall), calculated against your agreed specification, not the supplier's internal release limits.
- A statement of process stability over the period, with notes on any special causes and their corrective actions.
If a supplier replies that capability statistics are "not applicable" to enzyme-derived lipids, or offers only a histogram, treat the response itself as a finding. For continuous production of the same grade, capability data is standard practice in ingredient manufacturing.
Reading Cp, Cpk and Ppk Against Real PS Specifications
The indices are simple once tied to real limits. Cp measures spread only — how wide the process distribution is relative to the tolerance band. Cpk adds centering — how far the mean sits from the nearer limit. Ppk recalculates both using overall long-term variation rather than within-subgroup variation.
Keep it concrete with PS examples. For a 50% PS powder specified at 50.0–57.0% assay, a mean of 53.4% with a standard deviation of 0.6% gives a Cpk near 1.9 — comfortable. The same mean with a deviation of 1.6% gives roughly 0.75: the process regularly lands in the low 50s, and a lot below 50.0 is a matter of time. For peroxide value limited to 5.0 meq/kg, a mean of 1.2 with a deviation of 0.5 gives a Cpk around 2.5; a mean of 3.8 with the same spread gives about 0.8 — passing lots, but a supplier oxidizing its way toward the limit, which threatens shelf life as much as release.
Two reading rules protect you. First, insist that Cpk be calculated against your specification — a supplier can raise Cpk by narrowing internal release limits or retesting until a lot passes, so ask whether the data is release testing or a periodic process study, and whether retests are included. Second, compare Cpk with Ppk: a large gap (say Cpk 1.5, Ppk 1.1) means within-subgroup performance is good but the process drifts between subgroups — often a batch-to-batch raw material or operator effect worth probing at audit.
What Control Charts Reveal That COAs Do Not
Ask for the charts, then look for patterns rather than points outside limits. Six or more points steadily rising in peroxide value across lots suggests an oxidation driver — storage temperature, nitrogen headspace practice, holding time — long before any limit is breached. The same lot position repeatedly flagged and adjusted (say, the first lot after maintenance) tells you the process is being steered back on target, not controlled. Control limits tighter than the natural process variation usually mean the limits were drawn where management wanted them, not where the process runs.
During an audit, ask the line operator, not only the quality manager, what happens when a chart signals a special cause. A capable system reacts with a defined reaction plan and records the action; a weak one "looks into it."
Capability Thresholds and the Annual Supplier Review
Set thresholds in the quality agreement so the annual review has teeth. For key attributes — assay, peroxide value, moisture — require Cpk of at least 1.33 sustained, investigation below that, and a corrective action plan below 1.0. For safety-critical attributes such as heavy metals or microbiology, capability indices are less meaningful because good control makes failures rare; require validated control points and periodic verification testing instead. For new suppliers and new grades, ask for a capability study across at least 20 consecutive lots before volume ramps to full allocation.
When a threshold is missed, grade the response: a documented investigation and improvement plan for a single miss; interim third-party testing of every lot while capability recovers; re-qualification or dual sourcing if the process cannot hold. What you should not accept is a silent specification change — a supplier "helpfully" widening the assay range from 50.0–57.0% to 48.0–57.0% lifts the Cpk on paper without improving the product you bought.
Close the loop in the scorecard: current Cpk and Ppk per key attribute, trend over the review period, stability statement, open investigations, and any specification changes. Weight capability alongside on-time delivery, complaints and audit results, so commercial teams see that a cheap supplier with a marginal process is more expensive than it looks — one failed lot in a booked OEM production slot costs more than a quarter's unit-price saving. The review is also the moment to align specifications with reality: if the process naturally centers at 53.4%, keep the spec at 50.0–57.0% and let the margin work for you.
FAQ
Q: What Cpk value should an ingredient supplier achieve?
A: For key attributes such as assay, peroxide value or moisture, Cpk of at least 1.33 is the widely used minimum, with 1.67 preferred where margin is cheap. Below 1.0 means the process regularly produces near-limit lots and needs a corrective action plan.
Q: Is a passed COA enough to qualify a phosphatidylserine supplier?
A: No. A COA confirms one lot met specification. Capability data — 20 to 30 lots of trend results, control charts and Cpk/Ppk values — shows whether future lots will. Use both: COA for release, capability for qualification and annual review.
Q: What is the difference between Cpk and Ppk?
A: Cpk uses within-subgroup variation and measures short-term capability; Ppk uses overall variation and reflects long-term performance. A large gap between them indicates the process drifts between subgroups, even if both values look acceptable individually.
Q: What data should I request to prove process capability?
A: 20–30 lots of individual results for key attributes, the corresponding control charts with system-calculated limits, Cp/Cpk and Pp/Ppk against your agreed specification, and a statement of statistical control with any special-cause investigations.
Q: Can a supplier improve Cpk by changing the specification?
A: Only on paper, which is why capability must be calculated against the specification in your quality agreement. Watch for widened ranges or relaxed internal limits; genuine specification changes should go through formal change control, not appear in the next COA template.
Conclusion
Capability review turns supplier quality from a historical record into a forecast. A pack of trend data, real control charts, and Cpk and Ppk values against your own specification takes a competent supplier days to produce and reveals more about next year's supply risk than two years of passing COAs. Set thresholds in the quality agreement, add capability to the annual scorecard, and grade the response when thresholds slip. Buyers who do this stop discovering capability problems from failed lots — and start seeing them in the charts, several lots early, while there is still time to act. When you are ready to move from checklist to quotation, contact our team with the grade, source and document set described above.
Sources
- U.S. Food and Drug Administration — 21 CFR Part 111, Current Good Manufacturing Practice for dietary supplements: https://www.ecfr.gov/current/title-21/chapter-I/subchapter-B/part-111
- U.S. Food and Drug Administration — Guidance for Industry: Process Validation: General Principles and Practices: https://www.fda.gov/regulatory-information/search-fda-guidance-documents/process-validation-general-principles-and-practices
- NIST/SEMATECH e-Handbook of Statistical Methods — Control Charts: https://www.itl.nist.gov/div898/handbook/pmc/section3/pmc32.htm
- ISO 9001 Quality Management Systems overview: https://www.iso.org/iso-9001-quality-management.html
Buyer fit
Who this guide is for
This guide is written for the buyer profiles below. Open your profile for the matching grades, documents and next steps, or tell us your requirements directly.
Recommended next steps
- Review the Phosphatidylserine product page.
- Compare Soy PS and Sunflower PS.
- Check manufacturing proof and Quality & R&D.
Contact sales for product documents
Share source preference, application, country, and annual quantity.
Contact Sales
