Formulation Science14 Oct 20265 min read

How to choose analytical methods at each stage of cosmetic formulation development

Your lab owns an HPLC, so every question becomes an HPLC question, while the emulsion keeps thinning and nobody has run a flow curve. Choosing methods by the instruments on hand is the most common selection error. Start with the formulation question instead.

By CSL Analytical and formulation teams, Cosmetic Science Lab · Reviewed by Dr. Premalatha, Ph.D., Life Science, Regulatory Consultant

Evidence: Rico et al., Cosmetics 2024 reviewSeries: Part 2 of 2

Cosmetic development bench: a gloved hand dosing cream with a pipette, a texture-analyser probe pressing into a cream sample, and finished jars and pump bottles

Part 1What each analytical method measures
Part 2How to choose the right method at each stage (this article)

Quick answer

Trace every test from a formulation question, through a method and measurement, to a specific decision.
Stages 1 to 3 (intake, prototype, stability) change the formula. Stages 4 and 5 confirm it.
Use entry-level methods in-house and step up to advanced methods only when a decision needs them [1].
Plan regulatory data during development so the EU safety file grows with the formula [2], [3].

01Start with the question, not the instrument

A 2024 review groups cosmetic analytics into five categories and stresses that the right method depends on the product matrix [1]. Figure 1 maps four common problems to the method, measurement and decision each one points to. A weakening emulsion calls for rheology. A degrading active calls for HPLC or LC-MS. A shifting fragrance calls for GC-MS or an electronic nose, and drifting colour calls for colorimetry [1], [6].

Four rows linking a formulation question to a method, a measurement and a decision: emulsion weakening to rheology to modify the emulsifier or polymer; active degrading to HPLC or LC-MS to reformulate or change process or pack; fragrance changing to GC-MS or electronic nose to check oxidation or pack interaction; colour drifting to colorimetry to adjust stability, pack or light protection. — **Figure 1.** Each formulation question points to one method, one measurement and one decision.

The one-line test

Before booking any test, write down the decision it will inform. If no decision changes with the result, the test is reporting, not formulation.

02Which analytical method at which stage

**Table 1.** Which method to use at each development stage
Stage	Question	Primary methods	Output
1. Raw-material intake	Is this lot what the supplier claims?	HPLC, LC-MS/MS, GC-MS	Identity, purity, impurities [1]
2. Prototype design	Will this emulsifier and polymer system hold?	Rheology, interfacial tension, droplet sizing	Flow curve, G′/G″ balance [1], [7]
3. Stability screening	Which prototype survives stress?	Centrifugation, zeta potential, colorimetry	Separation rate, colour shift [4], [6]
4. Scale-up	Does the plant batch match the lab?	Rheology, HPLC assay, e-nose	Batch match to reference
5. Release	Is the batch safe and compliant?	Qualified assays, real-time stability	Certificate of analysis [2], [3]

For emulsion stability, stages 2 and 3 matter most: droplet size and viscosity set early decide how a product ages [6], [7]. Part 1 explains what each of these methods measures.

03Sample preparation for complex cosmetic matrices

A cream is not a clean solution. Oils, pigments and polymers interfere with measurement, so preparation is part of choosing the method [1], [6]. Common routes before chromatography include [1]:

Emulsion breaking: ethanol or methanol, sometimes with low pH or heat, before analysing UV filters.
Solid-phase extraction (SPE): pulls target compounds out of the matrix.
Microextraction: ultrasound-assisted or dispersive methods for preservatives such as parabens.

How to fix it: validate the method for each new matrix. A method built for a serum rarely transfers to a sunscreen unchanged, and zeta potential dilution can alter the structure being measured [1].

04A cost-performance framework for lab methods

The review compares methods on cost, specificity, ease of use and precision [1]. LC-MS/MS is highly specific but scores lower on cost and ease of use; HPLC with UV or DAD detection suits routine industrial work [1].

**Table 2.** Entry and advanced testing options, and when to step up
Need	Entry option	Advanced option	Step up when
Ingredient assay	HPLC-UV or DAD	LC-MS/MS	Unknown impurities or trace levels [1]
Fragrance control	Electronic nose	GC-MS with olfactometry	Off-notes must be identified [1]
Flow and texture	Viscometer	Rotational rheometer	Structure or elasticity matters [7]
Stability screening	Microscopy checks	Centrifugation, zeta potential	Many prototypes to rank [6]
Colour	Visual panel	Spectrophotometric colorimetry	Colour needs a numeric spec [1]

Two panels: the entry tier (routine, in-house) lists HPLC-UV or DAD, electronic nose, viscometer, microscopy checks and visual colour panel; the advanced tier (when a decision needs it) lists LC-MS/MS, GC-MS with olfactometry, rotational rheometer, analytical centrifugation and zeta potential, and spectrophotometric colorimetry. — **Figure 2.** Entry-tier and advanced-tier methods, and the point at which a formulation question justifies stepping up.

Tiers are relative, not price quotes. Most brands run entry methods in-house and use a partner lab such as our testing team when a decision needs advanced data. That keeps routine checks fast without giving up depth when it counts.

05Regulatory analytics: what the EU safety file needs

EU law requires a Product Information File and a Cosmetic Product Safety Report for every product [2]. Much of that report depends on analytical data, and manufacturing must follow good manufacturing practice (Article 8), which ISO 22716 is used to demonstrate [2], [3].

**Table 3.** EU requirements and the analytical evidence behind them
Requirement	Source	Analytical evidence
Composition	Reg. 1223/2009, Annex I [2]	Chromatographic assay of actives and preservatives
Physical/chemical properties and stability	Reg. 1223/2009, Annex I [2]	Rheology, droplet size, colour and stability data [4], [5]
Fragrance allergen labelling	Reg. 1223/2009, Annex III, as amended by Reg. (EU) 2023/1545 [2], [8]	GC-MS quantification of regulated allergens [1]
Controlled production	Article 8; ISO 22716 [3]	Qualified methods, batch records, lab controls

Allergen list is growing

Regulation (EU) 2023/1545 adds new fragrance allergens to the labelling list. It applies to products newly placed on the market from 31 July 2026 and to all products on the market from 31 July 2028, so GC-MS allergen methods need updating now.

Planned during development, this data builds the safety file alongside the formula. Retrofitted after a regulatory query, it costs time, samples and often a reformulation. Our regulatory team plans the dossier from the first prototype.

06Five common mistakes in selecting analytical methods

Testing only at release: the formula is fixed, so a failure means costly rework.
Defaulting to the most advanced instrument: LC-MS/MS adds cost when HPLC-UV can answer the question [1].
Trusting accelerated data alone: stress tests rank prototypes, but real-time monitoring is advisable to confirm shelf life [4], [5].
Reusing one method across products: each matrix needs its own preparation and validation [1].
Measuring without a decision: a test that changes nothing in the formula, process or pack is reporting, not development.

07Inside CSL's integrated analytical workflow

Complete characterisation needs nano, micro and macro data together [1]. CSL matches each method to the question it answers, not to the instrument that happens to be free. Results are reviewed at each stage gate, so a weak prototype is dropped before it reaches scale-up.

Upstream
Intake chromatography and interfacial screening shape the formula.
Midstream
Rheology and accelerated stability rank prototypes.
Downstream
Colorimetry, fragrance analytics and qualified QC methods confirm batches.

08Questions readers ask

Should a small brand buy LC-MS/MS?

Rarely. Most routine assays run on HPLC-UV; send advanced work to a partner lab when a decision needs it.

Which tests does an EU safety report need?

Data on composition, physical and chemical characteristics, stability, microbiological quality and impurities, supported by reliable analytical evidence, under Annex I of Regulation (EC) No 1223/2009.

When should stability testing start?

At prototype stage. Accelerated screening ranks candidates early, then real-time studies confirm shelf life.

How do I validate a method for a new product?

Re-check sample preparation, accuracy and recovery in the new matrix before relying on results, even if the method worked on a similar product.

Not sure which tests your product needs?

Tell us the product and the decision you need to make. We plan the method set by stage and cost tier, and build the safety-file data as we go.

Testing services Talk to the CSL team

09References

Rico F, Mazabel A, Egurrola G, Pulido J, Barrios N, Marquez R, García J. Meta-analysis and analytical methods in cosmetics formulation: a review. Cosmetics, 2024; 11(1): 1. doi:10.3390/cosmetics11010001
Regulation (EC) No 1223/2009 of the European Parliament and of the Council of 30 November 2009 on cosmetic products (consolidated). EUR-Lex
ISO 22716:2007 Cosmetics — Good Manufacturing Practices (GMP) — Guidelines on Good Manufacturing Practices. ISO
Colipa (now Cosmetics Europe) and CTFA. Guidelines on stability testing of cosmetic products. March 2004. Cosmetics Europe PDF
ISO/TR 18811:2018 Cosmetics — Guidelines on the stability testing of cosmetic products. ISO
McClements DJ. Critical review of techniques and methodologies for characterization of emulsion stability. Critical Reviews in Food Science and Nutrition, 2007; 47(7): 611–649. PubMed
Tadros TF. Rheology of Dispersions: Principles and Applications. Weinheim: Wiley-VCH, 2010. Wiley-VCH
Commission Regulation (EU) 2023/1545 of 26 July 2023 amending Regulation (EC) No 1223/2009 as regards labelling of fragrance allergens in cosmetic products. EUR-Lex

CSL

CSL Analytical and formulation teamsCosmetic Science Lab · Reviewed by Dr. Premalatha, Ph.D., Life Science, Regulatory Consultant · Last reviewed 14 Oct 2026Analytical chemistry and characterisation

Update log