Products & Formulas – Master Data, Control, and Release

This topic is part of the SG Systems Global regulatory & operations glossary.

Updated October 2025 • Specifications, Recipes, BOMs • QMS, MES, LIMS, WMS

Products & Formulas are the canonical definition of what is made and how it is made. “Product” captures identity, labeling claims, and specification limits; “Formula” (or recipe) captures materials, quantities, tolerances, sequencing, and required equipment. In regulated manufacturing, these are not tribal knowledge or spreadsheets—they are controlled objects under Document Control with effective dates, change histories, and traceable approvals. Execution must reference the same truth in the MES and the resulting eBMR, warehouse movement must observe it in WMS, and laboratory decisions must compare to it in LIMS. If the spec in LIMS, the instructions in MES, and the BOM in ERP/WMS are not the same object/version, you have a compliance gap waiting to become a customer complaint.

“If it isn’t controlled, versioned, and enforced at runtime, it’s not a formula—it’s a rumor.”

1) Product Identity—Specifications, Labels, and Codes

Product identity binds the commercial name to specification limits, label claims, packaging configuration, and regulatory identifiers. For pharmaceuticals, an NDC may apply; for finished goods, global identification typically uses GS1 GTIN. Label artwork and claims fall under Labeling Control, and line-side scans verify correctness using Label Verification. The identity record references the active specification so that every release decision and certificate aligns to what’s printed and promised.

2) Formula/Recipe—From BOM to Executable Reality

The starting point is a BOM, but formulas go further: they define target quantities, tolerances, units, rounding rules, substitutions, preparation steps, environmental prerequisites, and hold times. Critical adds use controlled Batch Weighing and gravimetric capture to remove arithmetic games. The sequence is explicit, not implied; interlocks ensure downstream steps cannot proceed until identity, quantity, and status checks pass. The exact object executed by the operator must be the one approved—anything else defeats traceability.

3) Master Records—MMR/MBR and Executed BMR/eBMR

Process instructions live in a Master Manufacturing Record (MMR) or MBR. Execution generates a BMR or its electronic counterpart, the eBMR. Version discipline is non-negotiable: only effective versions are selectable; obsolete versions are withdrawn and blocked. Reviewers must be able to see the lineage—who authored, who approved, when it became effective, and which batches used which version—without leaving the record system.

4) Specifications—What “Good” Must Look Like

Specifications define identity, potency/assay, impurities, microbiological limits, physical attributes, packaging acceptance, storage, and label elements. Laboratory results are compared in LIMS against the current spec with harmonized significant figures and validated calculations under CSV. Trends that move atypically within limits are OOT and should trigger proactive action; failures are OOS and must launch controlled investigations. The point is simple: compare raw or correctly rounded values to the spec as written—no masking via display precision and no spreadsheet detours.

5) Equipment Fitness—Only In-Status Assets May Execute

Formulas call out specific equipment classes or IDs, and the system must block the use of out-of-status assets. This is enforced with Asset Calibration Status checks at start-of-use and before any data are accepted. When instruments generate data—balances, spectrometers, or HPLC—the identity of the instrument, user, run ID, and timestamps must travel with the result. All transformations are captured via audit trails under Part 11/Annex 11. “Use it because it’s there” is not a policy; suitability is proven every time the asset is used.

6) Warehouse Controls—Right Item, Right Version

Materials are controlled in WMS with Directed Picking, Bin Location Management, and shelf-life rules like FIFO, FEFO, and Expiration Control. Kitting aligns to the formula, with Kitting enforcing identity and quantity, and movement histories preserving Batch-to-Bin and Batch Genealogy. If a quarantined or expired lot can be picked, you don’t have agility—you have a recall in slow motion.

7) Identification & Traceability—Make Genealogy a Query

Finished goods often carry a GS1 GTIN and internal lot codes. Labels encode item, lot, and expiry, verified online via Label Verification. Every movement and transformation should be reconstructable within minutes: which lots fed which batch, which bins they touched, which equipment executed which steps, which operators and signatures authorized progression, and what the final disposition was. That story lives in the eBMR and WMS histories, not in a hero’s memory or a lost spreadsheet.

8) Execution in MES—Interlocks, Not After-the-Fact Postings

The MES must enforce the formula in real time: correct item, correct lot, correct quantity, correct equipment, correct order, correct limits. Line Clearance prevents carryover; reason-for-change prompts capture the “why” behind deviations; Automated Batch Records write the evidence as events occur. The system should block unsafe progress by design rather than narrate failures after the fact. If the next step can proceed with the wrong lot or out-of-tolerance add, the design is broken, not the operator.

9) Laboratory Proof—Methods, Rounding, and Results

Lab evidence closes the loop. Results captured in LIMS carry method versions, instrument status, and analyst identity. Where chromatographic methods apply (HPLC), preserve raw signals and integration rules; don’t let downstream rounding hide failures. Harmonize rounding across acquisition, calculation, and spec comparison under CSV. Watch OOT signals and address drift before it becomes OOS; when OOS occurs, tie investigations and dispositions back to the product and version so corrective actions actually modify the master data that created the risk.

10) Data Integrity—Trustworthy Evidence or Nothing

Evidence must be attributable, legible, contemporaneous, original, and accurate—ALCOA in practice, not posters. Unique credentials and e-sign meanings live under Part 11/Annex 11, with computer-generated audit trails capturing who/what/when/why. Store controlled records under Document Control and enforce Data Retention & Archival so re-inspection years later can reconstruct decisions without archaeology. “Pictures in a share drive” is not retention; it’s a finder’s fee for auditors.

11) Governance—Change Control that Actually Changes Behavior

Any change to specs, labels, formulas, or calculation rules routes through cross-functional Change Control and operational MOC, backed by GAMP 5 and CSV. Risk assessment decides test depth, training updates ensure adoption, and effectiveness checks confirm that deviations drop or capability tightens. If a spec change doesn’t cut complaint rates or rework, you changed paper, not reality.

12) Release—The Gate That Matters

QA posts disposition via Lot Release; only then can WMS enable pick/pack for shipment and the organization post Finished-Goods Release. If shipments leave before disposition, that is an integration failure, not “speed.” Tie release back to the exact spec and formula version used so any future investigation can trace claims to evidence without guesswork. Where needed, include the product spec on the certificate package; the version tells the whole story.

13) Performance & Capability—Shrink Variation and Prove It

Define critical quality attributes and process parameters, then measure with MSA-vetted systems. Monitor stability with SPC and maintain life-cycle oversight via CPV. Roll results into the APR; when trends degrade, open targeted actions through CAPA or MOC. Capability must be demonstrated where it matters—at the combination of product, site, equipment set, and shift that actually produces risk.

14) How This Fits with V5 by SG Systems Global

V5 Solution Overview. The V5 platform treats products and formulas as first-class, versioned objects. Configuration is controlled, evidence is attributable, and cross-module interlocks—identity, asset status, signatures, and label verification—are testable and reportable. This means the same master that defines the product also drives execution, lab comparison, and warehouse movement without re-keyed copies or side spreadsheets.

V5 MES. In V5 MES, the effective MMR/MBR drives each step. Operators are guided through sequenced adds with gravimetric capture, in-step checks against equipment fit-for-use, and fail-intent tests wired as interlocks. Line-clearance prompts appear where the risk is real; Automated Batch Records write the eBMR contemporaneously, binding actions, users, device signals, and limits to the exact version of the formula executed.

V5 QMS. Within V5 QMS, Document Control governs specifications, labels, and master recipes; Change Control and MOC route proposed updates with risk, validation impact, and training. Audit-ready packages show the lineage from requirement to approval to execution. Deviations, CAPA, and internal audits align to the same product/formula objects, so corrective actions actually change the thing that operators and labs see.

V5 WMS. V5 WMS enforces status, identity, and shelf-life at receiving, kitting, and shipment. Directed picking respects QA disposition; bin/location and FEFO rules keep wrong or expired materials out of the kit; label verification at pack/ship prevents artwork or code mismatches. The genealogy view stitches material movements to batches and finished goods without manual reconciliation.

Bottom line: V5 turns “Products & Formulas” from static documentation into living controls. One governed truth flows through MES, LIMS, and WMS; the same interlocks that protect production tomorrow are the ones you can prove you tested yesterday.

15) Common Failure Patterns (and the Antidotes)

• Spec drift vs. capability. Antidote: trend with SPC/CPV; update masters through Change Control with effectiveness checks.
• Shadow spreadsheets for formula math. Antidote: implement validated calculations under CSV; execute via MES, not Excel.
• Out-of-status equipment usage. Antidote: enforce Asset Calibration Status blocks; reject data when suitability is missing.
• Wrong item/lot additions. Antidote: Directed Picking + scan-based Label Verification with bin rules.
• Evidence that can’t be reconstructed. Antidote: immutable audit trails, rigorous Data Retention & Archival, and complete eBMR.
• Label mismatches at ship. Antidote: strict Labeling Control and pack-line Label Verification tied to item/lot/expiry rules.
• Unaligned masters across systems. Antidote: make MES, LIMS, and WMS reference the same governed objects and versions; audit regularly via Internal Audit.

FAQ

Q1. What’s the difference between a product spec and a formula?
The specification defines acceptance criteria and labeling; the formula defines materials, quantities, tolerances, and steps. Both are controlled under Document Control and executed/verified via MES and LIMS.

Q2. Where should rounding rules live?
In validated calculation objects governed by CSV, applied consistently in MES and LIMS so comparisons to spec are defensible.

Q3. How do we prevent wrong-material additions?
Use Directed Picking, Label Verification, bin/location rules, and MES interlocks; only Released lots per Hold/Release are pickable.

Q4. How are formulas validated?
Through process and system validation under GAMP 5/CSV, plus ongoing monitoring via CPV and trending with SPC.

Q5. Which identifiers matter for finished goods?
Use GS1 GTIN where applicable, control label artwork under Labeling Control, verify on-line with Label Verification, and maintain traceable histories in the eBMR and WMS.