The Future of Cosmetic Formulating

The Future of Cosmetic Formulating

The central message is that cosmetic science is global, but successful products need to be adapted to local markets. Shared technologies and AI can speed development and reduce costs, but they cannot replace understanding local suppliers, regulations, consumers, and product expectations.

Industry strategy and operating trade-offs

• The cosmetics industry shifted from local production to professionalized manufacturing, global distribution, and corporate consolidation. Scale can improve quality systems, expand research, lower costs, and widen product access.
• Consolidation also encourages standardization: fewer suppliers, shared manufacturing systems, and common formula platforms. This simplifies procurement and operations, but can make products less distinctive or less satisfying to some customers.
• INCI names provide a shared ingredient vocabulary, not a guarantee of equivalence. Materials with the same INCI name can differ by supplier in composition, purity, specifications, or performance. Supplier changes may therefore require renewed testing.
• Mergers and acquisitions can be operationally complex, even when companies use similar ingredients. Differences in suppliers, specifications, ingredient codes, and production equipment can obstruct integration and require substantial retesting.

Product-development playbook: global platform, local adaptation

A useful approach is to start with a scalable product platform, then adapt it to the markets and audiences it serves. The presentation identifies three main sources of local variation:

1. Ingredient availability and sourcing: Local materials may differ in quality or performance, while importing ingredients can increase costs. Whenever practical, develop formulas around reliable local supply.
2. Regulation and claims: Ingredient rules and product classifications vary across markets. Claims can change a product’s regulatory category, so teams need to check both whether a formula can be sold and whether its intended claims are permitted.
3. Consumers and use conditions: Climate, water, hair and skin characteristics, bathing habits, fragrance preferences, and expectations of texture or performance vary. A formula can be chemically stable and legally compliant yet still fail because people dislike how it feels or smells.

Before taking a product into a new market, ask:

• Are the ingredients genuinely equivalent and reliably available there?
• Can the product be manufactured consistently with local supply?
• Are the formula, label claims, and marketing story acceptable in that market?
• Do local consumers prefer the product’s fragrance, texture, and performance?

The goal is not to create a completely new formula for every country. It is to build enough flexibility into the product platform to make informed local changes.

Consumer insight and positioning

• Consumer perception is part of product performance. In a halo effect, a preferred scent, name, or package can influence how people rate lather, rinsing, or other attributes—even when the underlying formula is unchanged.
• Blind testing alone may miss the effect of packaging and branding. Test products with the presentation consumers will actually encounter.
• Fragrance preferences are culturally and personally specific. Test scents with the intended market rather than assuming headquarters’ or the development team’s preferences will generalize.
• Formulation changes can help one consumer segment while disappointing another. The speaker recommends that smaller brands consider serving a clearly defined niche rather than trying to satisfy everyone; broad-market targeting is more suited to large players.

Examples discussed

• Pantene and Herbal Essences: The speaker described a strategy in which a shared or standardized shampoo-and-conditioner base supported different brands, with packaging, fragrance, and storytelling helping distinguish their positions and appeal to different consumers.
• Alberto Culver and St. Ives: Integrating the businesses was not as simple as combining similar-looking formulas. Supplier differences, internal specifications, coding systems, and equipment meant materials could not simply be swapped based on INCI names. The team repeated testing, including stability and claims-related testing, while seeking procurement savings.
• TRESemmé’s expansion: A formula successful in North America did not transfer unchanged to South America. Ingredient differences affected conditioning and viscosity, while consumer fragrance preferences also differed. The brand adapted rather than relying on a universal formula.
• Hair conditioner testing: A lighter conditioner addressed complaints that a product felt too heavy for some users, but reduced satisfaction for others. This illustrates how one “fix” can create a need for distinct formulas or segments.
• Johnson’s baby shampoo and V5: The speaker used these examples to show how regulatory differences or consumer concerns about ingredients can trigger reformulation. The business lesson was to anticipate regulatory, supply, and reputational risks rather than assume a technically acceptable formula will remain acceptable to consumers.

AI and the changing role of formulators

AI is best used as a research and productivity assistant, not as an authority on technical decisions. The presentation identifies three practical uses:

• Generating formulation starting points or comparing ingredient approaches.
• Explaining unfamiliar technologies and summarizing research.
• Organizing experimental results and drafting technical documentation.

A formulator still needs to check AI output against chemistry, supplier data, regulations, and laboratory testing. AI can make mistakes, including confidently overstating ingredient levels; polished answers should not be treated as verified answers.

In a speaker-described conditioner-formula comparison, Gemini, ChatGPT, and Claude all identified the product type and produced plausible starting formulas. They tended to overestimate ingredients included at very low “declared” levels. Claude had an approximately 9% overall margin of error in this particular test and performed best among the three systems tested. This is a single example, not a general performance guarantee.

The recommended career and workflow approach is to learn AI tools while maintaining formulation expertise: use them to reduce routine work and spend more time on testing, interpretation, consumer fit, and decisions that require human judgment.

Metrics and targets mentioned

• TRESemmé: The speaker recalled the brand growing from about $20 million to $100 million during their time at Alberto Culver; no exact timeline was given.
• AI formula comparison: Claude’s reported error was approximately 9% in the specific conditioner test.
• A claimed 98% “no hallucination” rate was mentioned as a claim made by companies, but the speaker cautioned that AI systems still fabricate information and require review.
• No formal business targets or deadlines were specified. The speaker said they might update an AI seminar early the following year.

Presenter and source

• Presenter: The webinar speaker; not named in the subtitles.
• Source: Chemists Corner webinar, The Future of Cosmetic Formulating.