# Which Formula Did We Actually Make? Managing Cosmetic Formula Versions > Keep formula revisions, development samples, test evidence, and manufacturing instructions connected without rewriting earlier product history. Canonical URL: https://cosmeticaffairs.com/blogs/manufacturing/cosmetic-formula-version-management/ Author: [Cosmetic Affairs Editorial](https://cosmeticaffairs.com/authors/editorial/) Published: 2026-10-17T00:00:00-04:00 Modified: 2026-10-17T00:00:00-04:00 Collection: Manufacturing Topics: Formulation, Manufacturing, Documentation ![Three cream prototypes sit beside a notebook and a laptop while a gloved formulator records an observation.](https://cosmeticaffairs.com/blogs/manufacturing/cosmetic-formula-version-management/cover.png) AI-generated illustration; not a photograph of actual laboratory work. A formula can pass through several drafts before a team chooses it for production. The difficulty starts when a sample label, test report and manufacturing ticket all use the same product name but refer to different compositions. A useful version-control process makes those differences visible. It lets a developer explain what changed and lets production identify the instructions that apply to a particular batch. It also preserves the basis for earlier decisions, so new work does not quietly rewrite the product's history. ## Define what your version number identifies Start by defining the content covered by a formula version. This might include material identities, percentages, phases and defined processing instructions. Write down your boundary so everyone applies the same rule when a field changes. FDA's longstanding cosmetic self-inspection checklist treats formulations, processing instructions, transfer and filling instructions, and in-process controls as written manufacturing information to maintain.[^1] This is a useful reminder that composition is only part of the production instruction. An internal formula code should remain recognisable through development, with each saved version clearly identified. Add a status that explains its intended use, such as development, under review or available for a specified manufacturing purpose. Define these terms within your own process rather than assuming a software label has the same meaning for every team. Also identify related specifications, methods and documents. A formula version number should not be expected to describe the revision state of every connected record. A manufacturing ticket needs enough detail to establish which applicable instructions and acceptance criteria were used. ## Make each change understandable Record the reason for creating a new version in language another formulator can understand. "Updated" provides little help. A note such as "replaced the selected emollient following a sourcing review" identifies both the change and the decision that prompted it. A comparison should make changes in material identity, amount and processing instructions easy to review. Pay attention to changes that look small on screen. Replacing a trade material can involve more than changing its name, so retain the relevant supplier and composition information for assessment. Decide who needs to evaluate the proposed change and what questions they must answer. Depending on the change, this could include suitability for manufacture, test coverage, safety assessment, ingredient labelling or purchasing. The review should establish what further work is justified and who will complete it. Keep rejected approaches identifiable too. A previous version may explain why a particular route was abandoned. Recording that reasoning helps future development begin with the available evidence instead of rediscovering an unresolved problem. ## Connect evidence to the version evaluated Give each development sample a reference that leads to the formula version actually prepared. Carry that reference into the test request and report. Where a sample was made differently from the planned instructions, preserve the difference within its record. When reviewing evidence for a later version, distinguish the original result from the assessment of its relevance. A test on version 2 remains a test on version 2. If the team considers it informative for version 3, record the reasoning and the limits of that conclusion. Avoid simply changing the formula reference on the original report. The same discipline helps with sensory reviews and customer feedback. Record which sample was evaluated and which attributes were accepted or remained open. An email approving "the latest cream" is easier to interpret when it points to a particular sample and version. This approach does not prescribe a new study for every edit. It creates a clear place to document the assessment of what changed and which evidence still applies. ## Illustrative case: the pilot and the revised thickener level Imagine a fictional moisturiser, F-214. Version 3 is selected following a texture review. Before a pilot run, a developer prepares version 4 with a different thickener level. The pilot ticket still names version 3, while the recorded quantities correspond to version 4. The team should preserve the original ticket and document the discrepancy. It needs to establish what was actually prepared using the available weighing and processing evidence, then identify the pilot sample accordingly. The texture observations from the earlier sample still belong to version 3. A clear decision record can explain which pilot findings are available for version 4, which questions remain open and what must happen before its next use. Correcting the record should reveal the mismatch and its resolution, so a later reader can follow the decision. ## Make version control usable during production When comparing software, check whether the team can compare formula versions and create a new draft from an earlier version while retaining its history. Also establish how specification changes are recorded: the formula version and the applicable specification need separate attention. Before a batch starts, make the chosen formula version visible on the instructions people use. Decide how the relevant specification and method revisions will be recorded, especially where related information can be updated independently. Check how printed or exported tickets remain connected to the approved manufacturing information after later changes. Give production a clear route for reporting a mismatch or requesting an adjustment. The process should explain how a proposed change is assessed and how the actual execution is recorded. It should also prevent an informal development edit from being mistaken for a manufacturing instruction merely because it is the newest item on screen. ## Check a handover from development to manufacture Use one formula with several development versions to check whether your process can answer six practical questions: - What does this version number identify, and which related documents have separate revisions? - Why was this version created, and which earlier version did it develop from? - Which samples and test results belong to it? - What decisions established its intended use, and what remains unresolved? - Which version and supporting instructions did the selected batch actually use? - Can the team retrieve an earlier record without replacing its evidence with current information? Ask a developer and a production colleague to work through the same example. Resolve differences in their answers by improving the references, status definitions or handover instructions. Then use the next batch to confirm that both people can identify the same formula, evidence and production record. ## Sources [^1]: U.S. Food and Drug Administration. [Good Manufacturing Practice (GMP) Guidelines/Inspection Checklist for Cosmetics](https://www.fda.gov/cosmetics/cosmetics-guidance-documents/good-manufacturing-practice-gmp-guidelinesinspection-checklist-cosmetics). Relevant sections of the legacy self-inspection checklist.